Fourier transform ion cyclotron resonance mass spectrometry investigation of gas-phase ions

MISSING IMAGE

Material Information

Title:
Fourier transform ion cyclotron resonance mass spectrometry investigation of gas-phase ions
Physical Description:
xvii, 150 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kage, David, 1969-
Publication Date:

Subjects

Subjects / Keywords:
Ion cyclotron resonance spectrometry   ( lcsh )
Mass spectrometry   ( lcsh )
Polycyclic aromatic compounds   ( lcsh )
Chemistry thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1999.
Bibliography:
Includes bibliographical references (leaves 137-149).
Statement of Responsibility:
by David Kage.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 021562503
oclc - 43607620
System ID:
AA00018855:00001

Full Text










FOURIER TRANSFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY
INVESTIGATION OF GAS-PHASE IONS












By

DAVID KAGE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1999


































This dissertation is dedicated to the memory of my grandmother, Lila M. Brown,
who would have laughed at all of this gobbledegook.













ACKNOWLEDGMENTS


There are many individuals whom I would like to thank that have contributed to my

educational experience. Firstly, I would like to thank all of the previous and present

members of the group. It was by working side-by-side with them, that most of this work was

accomplished. The knowledge shared by all and passed through the ranks was invaluable. A

big thank you goes out to Dr. Clifford H. Watson for always being there to answer basic

questions. His knowledge of electronics, instrumentation, and trouble-shooting was a very

big plus to have in the laboratory. I would like to thank Professor John R. Eyler for having

the patience to let me stick around and try my best.

My appreciation goes out to Professor Laszlo Prokai for the knowledge and assistance

he gave me during the cyclodextrin project. Professor Martin Vala was very helpful on the

polycyclic aromatic hydrocarbon project in more ways than one. Dr. Jan Szczepanski was

extremely generous in helping me get the polycyclic aromatic hydrocarbon project off and

going. His assistance with laser-oriented questions was always appreciated. I would like to

thank the remaining members of my committee (Professors William Weltner, Robert

Hanhrahan, and Lisa McElwee-White) for always having an open door if I needed to seek

scientific advice or if I just wanted to discuss current events.








Often overlooked is the work performed by people behind the scenes. The work

reported here is definitely no exception to this. Without the skilled talents of the machine

shop and electronics shop, most of this work would not have been possible. The guys in the

machine shop were always there to explain certain aspects of a design, fix parts when they

broke, or help me redesign a critical piece on an instrument. The last members of this

behind-the-scenes-group was Mrs. Lori Clark and all of the staff in the business office. Even

when I would show up on a Friday at four o'clock with an emergency, they would still have a

smile and the time to point me in the right direction for the money.

I can't put into words how much I appreciate my family. I would like to tell my

father, Jerry Kage, that I love him and that he can stop worrying about me because I am

finally finished. I want to thank my mother, Marcia Kage, for always believing in me and for

always putting up with me and all of my bitching for these past few years. I would like to

thank my sisters, Dawn Hopping and Brenda Kage, for always being there when big brother

needed to talk to someone. Finally, I would like to thank A. M. H. for giving me a reason to

complete this venture. Her love, friendship, and patience through this endeavor were

unparalleled.














TABLE OF CONTENTS

page

A CKN O W LED G M EN TS............................................................................................................. iii

LIST O F TABLES....................................................................................................................... viii

LIST O F FIG U RES........................................................................................................................ ix

A BSTRA CT.................................................................................................................................xvi

CHAPTERS


1. HISTORICAL INCEPTION AND FUNDAMENTAL PRINCIPLES
OF FOURIER TRANSFORM ION CYCLOTRON RESONANCE
M A SSSPECTRO M ETRY ..................................................................................... 1

Introduction............................................................................................................. 1
H historical O verview ................................................................................................ 2
The Basic Apparatus............................................................................................... 5
M otion of Trapped Ions........................................................................................... 8
CyclotronM otion........................................................................................ 8
TrappingM otion....................................................................................... 10
M agnetronM otion..................................................................................... 11
Experim entalProcedure........................................................................................ 12
IonForm ation............................................................................................ 13
IonExcitation............................................................................................ 14
Im pulseExcitation......................................................................... 17
ChirpExcitation............................................................................ 17
SW IFTExcitation......................................................................... 19
lonD etection............................................................................................. 19
Conclusion............................................................................................................. 23








2. GAS PHASE BINDING ENERGIES OF SELECTED HOST:
GUESTCOMPLEXES......................................................................................... 24

Introduction..................................................................................................... 24
CyclodextrinBackground ....................................................................... ........... 24
Electrospraylonization .................... ...................................................................31
Production ofCharged Droplets .................................... ............. ......... 33
ChargedDroplet Shrinkage ...................................................................35
M echanism of Gas-Phase Ion Production .......................... ....................36
Collision-InducedDissociation ........................................................................... .. 38
Experimental ..................................................................... ................................... 46
Results...................................................................................................................49
a t-CD:Tryptophan ................................. ....................... ...........................50
a-CD:Proline ............................................................................................. 52
ca-CD:Lysine .............................................................................................. 52
3-CD:Tryptophan...................................................................................... 52
O-CD:Histidine.......................................................................................... 53
Discussion ............................................................................................................. 53


3. MOTIVATION FOR INVESTIGATING POLYCYCLIC AROMATIC
HYDROCARBONSOFASTROPHYSICALIMPORTANCE........................... 71

Introduction ........................................................................................................... 71
Background ................................................................................. .......................... 71
RelatedStudies........................................................... ...................................... 76
CurrentEfforts....................................................................... .......................... 78


4. PHOTODISSOCIATION AND ION-MOLECULE REACTIONS
OFFLUORENECATIONS.................................................................................. 81

Introduction.................................................................................... .......................81
Background........................................................................................................... 81
Experimental ......................................................................................................... 82
Results and Discussion .......................................................................................... 90
Photodissociationvs. Irradiation Time....................................................90
Atomic vs. M olecular Hydrogen Loss ....................................................... 99
lon-M oleculeReactions.......................................................................... 104








5. PHOTODISSOCIATION AND ION-MOLECULE REACTIONS
OF ACENAPHTHYLENE, DIPHENYLACETYLENE, AND
N A PH TH A LEN ECA TION S............................................................................. 110

Introduction......................................................................................................... 110
A cenaphthylene................................................................................................... 110
D iphenylacetylene............................................................................................... 115
N aphthalene......................................................................................................... 126


6. CON CLU D IN G REM A R K S............................................................................. 132

Binding Energies for CD:Amino Acid Complexes............................................. 132
Fluorene............................................................................................................... 133
A cenaphthylene................................................................................................... 135
D iphenylacetylene............................................................................................... 135
N aphthalene......................................................................................................... 135

LITERA TU RE CITED ................................................................................................................ 137

BIO G RA PH ICA L SKETCH ....................................................................................................... 150














LIST OF TABLES


Table pag

1. A brief listing of selected physicochemical properties of the three most common
cyclodextrin molecules. (Adapted from reference 101).................................................... 30

2. The twenty essential amino acids along with their appropriate symbols and
m asses................................................................................................................................51

3. Conditions for studying the various [CD:amino acid]H+ complexes............................... 55

4. Observed fragmentation channels and efficiencies for selected PAH cations using
the ion-trap detector. (Reproduced from reference 197)............................................ 78

5. A list of the twenty-four PAHs examined by Ekem et al. placed within the
appropriate fragmentation category................................................................................... 79













LIST OF FIGURES


Figure pag

1. Schematic representation of a typical cubic trapped analyzer cell commonly used
in FT-ICR MS. The three pairs of parallel electrodes and their orientation with
with respect to the magnetic field are depicted................................................................... 7

2. Origin of ion cyclotron motion. The path of an ion moving in the plane of the
is bent into a circular orbit by the inward-directed Lorentz magnetic force
produced by a magnetic field directed perpendicular to the plane of the paper.
(Taken from reference 6)..................................................................................................... 9

3. Schematic diagram of the natural motions of an ion trapped by a uniform magnetic
and static electric field: oc (cyclotron), (BT (trapping), and Wom (magnetron).
The magnetron motion is circular about a guiding center that follows a contour
of constant electric potential. (Adapted from reference 21)............................................. 12

4. A general experimental pulse sequence that illustrates the four fundamental
steps needed in order to obtain a mass spectrum using FT-ICR MS................................. 13

5. Incoherent ion cyclotron orbital motion (left) is converted to coherent (and therefore
detectable) motion (right) by the application of an oscillating voltage to the
excitation plates. Ions which are in resonance with the excitation frequency
gain kinetic energy and spiral outward from the center of the cell into a larger
cyclotron orbit (right). (Taken from reference 6)............................................................. 15

6. A Fourier excitation waveform and excitation spectrum for impulse (a) and chirp
(b) excitation. (Taken from reference 7).......................................................................... 18

7. An illustration of the principles of SWIFT excitation (a) and a SWIFT excitation
depicting selective ejection of unwanted ions (b). (Taken from reference 7).................. 20

8. A rotating monopole description of signal generation. Positive ions approach one
plate, attracting electrons. As the ions continue moving in a circle, they
approach the other plate and attract electrons. Thus, the ion motion induces a
small AC (sine wave) current, an image current, in the detection plates.
(Taken from reference 7)................................................................................................... 21








9. Overall depiction of an FT-ICR mass spectrometer. The upper diagram depicts
the excitation of the ion packet by an externally applied alternating rf field. The
lower picture shows the detection of the image current that is produced by the
coherently orbiting ion packet on the two opposing detection plates to produce a
time-domain signal. The time-domain signal is then converted to a voltage,
digitized, and Fourier-transformed to yield a frequency-domain spectrum which
is then converted to a mass spectrum. (Adapted from reference 54)............................... 22

10. Compounds 1-3 are the chemical structures of the three most common
cyclodextrins: a-, P-, and y-cyclodextrin, respectively.................................................... 26

11. Portion of a cyclodextrin molecule showing the glucose units connected through
glycosidic ca-1,4 linkages ............................................................................................27

12. Representation of the three most common cyclodextrins (a-, P-, and y-cyclodextrin)
along with approximate dimensions and cavity volumes.................................................. 28

13. A simple depiction of the processes that occur in electrospray mass spectrometry.
(Adapted from reference 115)........................................................................................... 34

14. Schematic representation of the ion evaporation model based on methanol as the
solvent. The parent droplet that is created at the spray tip undergoes uneven
fission as time passes. The depiction demonstrates how the parent droplet
shrinks (losing about 2% of its mass) and loses charge (approximately 15%) as it
produces daughter droplets while drifting towards the counter electrode
(Adapted from reference 115)........................................................................................... 37

15. Resulting mass spectra following five stages of CID of FeS10'. (a) Isolation of Fe'
following laser desorption and collisional cooling with argon and S8. (b) Reaction
ofFe' with Ss. (c) Isolation of FeS10+. (d) CID ofFeS10+. (e) Isolation of FeS8+.
(f) CID of FeSs. (g) Isolation of FeS6'. (h) CID of FeS6. (i) Isolation of FeS4.
(j) CID of FeS4>. (k) Isolation of FeS2. (1) CID of FeS2+. (Taken from
reference 139)...................................................................................................................41

16. Fourier transform ion cyclotron resonance mass spectrometer used to determine the
gas-phase binding energies of cyclodextrin:amino acid complexes. The instrument
employed a shielded 4.7 T magnet, an external electrospray ionization source,
and possessed three stages of differential pumping to achieve analyzer cell
pressures on the order of 5.0 x 109 Torr........................................................................... 47








17. Typical pulse sequence used for the CID studies. HD is the Hexapole Dump, Q
is the Quench pulse, IG is the Ion Generation pulse, MS is MS/MS Coarse
Selection, IA is the Ion Activation pulse, E is the Excitation pulse, and D is the
D election pulse ..................................................................................................................49

18. Mass spectrum of the isolated [a-CD:Trp]H' at m/z 1177. The unlabeled peaks
demonstrate the inefficient ejection of unwanted ions during the isolation of the
parent ion ........................................................................................................................... 56

19. The CID mass spectrum of [a-CD:Trp]H' showing the free protonated tryptophan
at m /z 205..........................................................................................................................57

20. Percent fragmentation versus ion center-of-mass energy for [a-CD:Trp]H'.
Extrapolation of this line to zero yields a threshold binding energy of 1.32 eV ............... 58

21. Mass spectrum of the isolated [a-CD:Pro]H+ at m/z 1088............................................... 59

22. The CID mass spectrum of [at-CD:Pro]H' showing the free protonated proline
at m /z 11l6..........................................................................................................................60

23. Percent fragmentation versus ion center-of-mass energy for [ca-CD:Pro]H.
Extrapolation of this line to zero yields a threshold binding energy of 1.21 eV ............... 61

24. Mass spectrum of the isolated [a-CD:Lys]H' at m/z 1119............................................... 62

25. The CID mass spectrum of [ot-CD:Lys]H' showing the free protonated lysine
at m /z 147..........................................................................................................................63

26. Percent fragmentation versus ion center-of-mass energy for [ca-CD:Lys]H'.
Extrapolation of this line to zero yields a threshold binding energy of 0.71 eV ............... 64

27. Mass spectrum of the isolated [p-CD:Trp]H at m/z 1339............................................... 65

28. The CID mass spectrum of [P-CD:Trp]H. The free protonated tryptophan at m/z
147 can be seen in the expanded portion of the spectrum................................................. 66

29. Percent fragmentation versus ion center-of-mass energy for [P-CD:Trp]H+.
Extrapolation of this line to zero yields a threshold binding energy of 0.58 eV ............... 67

30. Mass spectrum of the isolated [I-CD:His]H at m/z 1290............................................... 68

31. The CID mass spectrum of [p-CD:His]H' showing the free protonated histidine
at m /z 156 ..........................................................................................................................69

xi








32. Percent fragmentation versus ion center-of-mass energy for [1-CD:His]H'.
Extrapolation of this line to zero yields a threshold binding energy of 0.73 eV ............... 70

33. Chemical structure and numbering system of the fluorene molecule (hydrogen
atoms have been omitted from the structure).................................................................... 82

34. Results of DFT calculations2" on the fluorene cation outlining the possible fragmen-
tation pathways. The energies were calculated at the B3LYP/4-31G level of theory......83

35. Schematic representation of the 2 T instrument used to study the photodissociation
of PAHs. (A) 2 T superconducting magnet, (B) ionization gauge, (C) inlet leak
valves, (D) sample tubes, (E) gate valve, (F) oil diffusion pump, (G) solids probe
port, (H) irradiation window, (I) connections for the analyzer cell and electron
gun, (J) vacuum cham ber..................................................................................................85

36. Dimensions of the stainless steel (a) trapping tubes and (b) excitation and detection
p lates..................................................................................................................................86

37. Analyzer cell that was used with the 2 T FT-ICR mass spectrometer to study
PAHs. Depicted in the drawing are (a) the two stainless steel trapping tubes, (b)
stainless steel tube cut into four equal segments used for the excitation and
detection plates, and (c) virgin electrical grade Teflon (TFE) rings used to
electrically isolate the different segments of the cell. The overall length of the
cell is 12.7915" and the diameter of the rings is 3.250".................................................... 87

38. Depiction of the entire analyzer cell assembly used throughout the PAH studies.
The analyzer cell was confined between four stainless steel rods that were held
together by two stainless steel rings. The entire assembly was attached to a
flange that contained the electrical feedthroughs that supplied voltages to the El
gun, trapping plates, and excitation plates........................................................................ 88

39. A typical pulse sequence that illustrates the essential steps in obtaining a mass
spectrum for the photodissociation studies of the fluorene cation.................................... 89

40. A representative pulse sequence used to study the fragmentation as a function of
irradiation time for the fluorene cation. The variable in these experiments was
the length of the irradiation pulse (USER A). An ejection pulse was placed on the
ion at m/z 167 (which was due to the presence of a carbon- 13 atom) in order not to
complicate the spectra unnecessarily. This was done in order to ensure that
the photodissociation products were derived from the parent ion at m/z 166 and
not from the ion at m /z 167............................................................................................... 91








41. Plot of fragmentation as a function of irradiation time from 0 to 5000 ms for the
fluorene cation........................................................... ............................................ 92

42. Expanded portion of Figure 41 showing fragmentation as a function of irradiation
tim e from 0 to 1000 m s..................................................................................................... 93

43. A typical mass spectrum of the fluorene cation. The daughter ions at m/z 163-165
are a result of the El process.............................................................................................. 94

44. Mass spectrum of the fluorene cation after irradiation of 200 ms. Note that after
only 200 ms the daughter ion at m/z 165 now dominates the mass spectrum................... 95

45. Mass spectrum of the fluorene cation after irradiation for 500 ms. Note that after
500 ms the dominant ion in the spectrum is the daughter ion at m/z 163......................... 96

46. Mass spectrum of the fluorene cation after irradiation for 5000 ms. Note that longer
irradiation times lead to new ions as a result of ion-molecule reactions........................... 98

47. Pulse sequence used to determine if daughter ions of fluorene ions were formed by
loss of atomic hydrogens or molecular hydrogens. In this experiment, the
isolated parent ion at m/z 166 was exposed to the lamp for 1000 ms............................. 100

48. Pulse sequence used to determine if daughter ions of fluorene ions were formed by
atomic hydrogen or molecular hydrogen loss. In this experiment, an ejection pulse
was placed on the ion at m/z 164 during the irradiation event........................................ 101

49. Mass spectrum of the fluorene cation after an irradiation of 1000 ms. In this
experiment, the ejection pulse on the ion at m/z 164 was turned off, therefore, all
the daughter ions from [M-H]' to [M-5H] are visible in the spectrum.......................... 102

50. Mass spectrum of the fluorene cation after an irradiation of 1000 ms. In this
experiment, the ejection pulse on the ion at m/z 164 was turned on. The absence
of the ion at m/z 163 proves that the hydrogens are being lost as atomic hydrogens
and not as m olecular hydrogens...................................................................................... 103

51. A typical mass spectrum obtained for the fluorene cation after longer irradiation
times. This particular experiment had an irradiation time of 4000 ms. At 4000 ms,
the parent ion has completely disappeared, leaving [M-3H]' as the dominant
fragment ion. Also visible are ions (m/z 226-324) that result from ion-molecule
reactions of the neutral parent and fragment ions........................................................... 105








52. A mass spectrum with identical conditions as the previous spectrum with on change;
the lamp was turned off for this experiment (the 4000 ms was simply a delay time
before detection). Note the absence of any ions between m/z 226-324 suggesting
that these species are due to ion-molecule reactions initiated by the lamp..................... 106

53. A mass spectrum with identical conditions as in Figure 51 but with an irradiation
time of only 500 ms. This spectrum is dominated by [M-3H] with only a hint of
any ion-molecule reactions occurring. This spectrum thus demonstrates that ion-
molecule reactions are not important at shorter irradiation times
(< 500 m s).......................................................................................................................107

54. Structure (a) results from a reaction between the neutral fluorene molecules and
the fluorene fragment ions. The resulting ion of m/z 328 can further lose
hydrogen atoms to form (b) which has a m/z of 324....................................................... 109

55. Chemical structure and numbering system for the acenaphthylene molecule (the
hydrogen atoms have been omitted from the structure).................................................. 111

56 A typical mass spectrum of the acenaphthylene cation after a 500 ms delay was
placed before the detection pulse. The spectrum essentially depicts only the
parent ion at m /z 152.......................................................................................................112

57. A mass spectrum of the acenaphthylene cation after a 60,000 ms delay time. As a
result of the increased delay time, ion-molecule reactions are occurring that
generate the ion near m /z 304.......................................................................................... 113

58. Possible mechanism for the formation of the ion at m/z 304 resulting from an ion-
molecule reaction between a neutral acenaphthylene and an acenaphthylene cation...... 114

59. Mass spectrum of the acenaphthylene cation after irradiation for 60,000 ms by a
xenon arc lamp. The spectrum depicts the generation of a new ion at m/z 228
which likely forms as a result of the photodissociation of the ion at m/z 304................ 116

60. Proposed scheme for the ion observed at m/z 228. This ion is a
photodissociation product derived from the ion at m/z 304............................................ 117

61. Chemical structure of the diphenylacetylene molecule.................................................. 118

62. A typical mass spectrum of the diphenylacetylene cation with a 500 ms delay
placed before the detection event. At short delay times the mass spectrum is
dominated by the parent ion at m/z 178.......................................................................... 119








63. A mass spectrum of the diphenylacetylene cation with a 5000 ms delay placed
before the detect event..................................................................................................... 120

64. Plot of percent abundance vs. delay time (lamp off) for the diphenylacetylene
cation. At longer delay times, ion-molecule reactions resulted in a decrease in the
abundance of the parent ion at m/z 178 along with an increase in the abundance
of the product ion at m /z 356.......................................................................................... 121

65. Two possible structures for the ion at m/z 356 that results from an ion-molecule
reaction involving a neutral diphenylacetylene and a diphenylacetylene cation............. 122

66. A typical mass spectrum of the diphenylacetylene cation after an irradiation of
500 ms from a xenon arc lamp. The mass spectrum is dominated by the parent
ion at m/z 178. The spectrum also depicts a significant abundance of the
daughter ion at m /z 152................................................................................................... 124

67. A mass spectrum of the diphenylacetylene cation after an irradiation of 5000 ms.
The spectrum depicts the formation of several new peaks which are most likely
photofragments resulting from the photodissociation of the ion at m/z 356 ................... 125

68. Two possible routes for naphthalene cation photodestruction. (Taken from
reference 198).................................................................................................................. 127

69. A typical mass spectrum of the naphthalene cation. The spectrum illustrates the
tendency of the naphthalene cation to dissociate completely. The ions at m/z 102
and 76 result from the El process itself.......................................................................... 128

70. A mass spectrum of the naphthalene cation after an irradiation pulse of 5000 ms.
Note the formation of several new ions at masses above 128 ......................................... 129

71. Possible structures for the ions observed at m/z 202 and 250 resulting from
irradiation of the naphthalene cation............................................................................... 130














Abstract of Dissertation Presented to the Graduate
School of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy

FOURIER TRANSFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY
INVESTIGATION OF GAS-PHASE IONS

By

David Kage

December 1999

Chairman: Dr. John R. Eyler
Major Department: Chemistry

Fourier transform ion cyclotron mass spectrometry (FT-ICR MS) has received

considerable attention for its ability to make mass measurements with a combination of

resolution and accuracy that is higher than any other mass spectrometer. It can be used to

obtain high-resolution mass spectra from ions generated by practically every known

ionization method, to perform tandem mass spectrometric measurements, and to examine ion

chemistry and photochemistry. Its versatility follows from the fact that it is an ion trapping

instrument. The instrument mass analyzes and detects ions using methods which are unique

among mass spectrometers.

FT-ICR MS was used to measure the binding energies of cyclodextrin:amino acid

complexes in the gas-phase. Cyclodextrins are cyclic oligosaccharides that form truncated,

cone-shaped molecules. The most common are referred to as a-, 3-, and








y-cyclodextrin, which contain 6, 7, and 8 glucose units in the ring, respectively. The cavity

that is formed by these molecules is hydrophobic, which lends to their ability as a "host"

molecule for the study of host-guest chemistry. Collision-induced dissociation was used to

measure the binding energies between the cyclodextrin host molecules and amino acid guest

ions.

FT-ICR MS was also used to study the photodissociation of polycyclic aromatic

hydrocarbon cations that are of interstellar importance. It has become a widely accepted

notion that PAH cations are the carriers of the diffuse interstellar bands that have been

measured for years by astronomers. The unknown link is to find exactly what PAH cations

are responsible for the bands. Experiments have been performed to answer this question.

Two major themes are repeated for the fluorene, acenaphthylene, diphenylacetylene, and

naphthalene cations: monitoring the generation of new ions formed as a function of trapping

time in the analyzer cell, and analyzing the products that are generated after irradiation from

a xenon arc lamp. Results from the first series of experiments reveal that new ions are being

formed which result from ion-molecule reactions between neutral parents and parent ions.

The second set of experiments show that the ions generated from the ion-molecule reactions

are fragmenting into smaller ions as a result of the irradiation from the xenon arc lamp.


xvii













CHAPTER 1
HISTORICAL INCEPTION AND FUNDAMENTAL PRINCIPLES
OF FOURIER TRANSFORM ION CYCLOTRON
RESONANCE MASS SPECTROMETRY


Introduction

Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) has,

in recent years, developed into a very powerful analytical technique after years of

promising anticipation. Since 1985, this technique has been the subject of four journal

special issues,1"4 three books,5"7 and more than 60 review articles.8 As of 1998 there were

over 235 FT-ICR mass spectrometers located in several countries around the world that

are being used to solve problems by a wide variety of scientists from analytical and

physical chemists in academic settings to drug discovery scientists in the pharmaceutical

community. In a relatively short time, FT-ICR MS has established itself as a powerful

mass spectrometric technique that combines the advantages of ultra-high mass resolution

and mass accuracy, is capable of utilizing a wide variety of ionization techniques, and can

use a wide range of methods for structure characterization of the primary sample ions.

With future improvements in magnet technology and with cheaper and more powerful

computers, this technique will surely become the method of choice for mass spectrometry

in the future.










Historical Overview

Fourier transform ion cyclotron resonance mass spectrometry is a mass spectrom-

etric technique, whose beginnings can be traced back to conventional ICR mass

spectrometry and Fourier transform nuclear magnetic resonance (FT-NMR) spectroscopy.

The fundamental principles underlying ICR mass spectrometry were explained in 1930 by

Ernest 0. Lawrence9 who invented the cyclotron particle accelerator (he was later

awarded the Nobel Prize in physics in 1939 for this development). In 1932 Lawrence

first demonstrated that a charged particle moving perpendicular to a uniform magnetic

field is constrained to a circular orbit.10 An orbit in which the angular frequency of the

particle's motion is independent of the particle's orbital radius is characterized by the so-

called cyclotron equation



co=qB/m (1)



where co is the angular frequency, q is the charge on the particle, B is the magnetic field

strength, and m is the mass of the particle. Lawrence demonstrated that the cyclotron

motion of a particle could be excited to a larger orbital radius by applying a transverse

alternating electric field whose frequency matched the cyclotron frequency of the particle.

This was a very significant discovery in that he demonstrated that a particle could be

excited to a very large kinetic energy by use of small electric fields. The cyclotron

particle accelerator has been a tremendous research tool used in the field of nuclear

physics.










The first use in an analytical sense of the mass selective characteristics of the

cyclotron motion of ions was the development of the Omegatron at the National Bureau

of Standards by Sommer et al. in 1949."1"113 Their instrument used the frequency-selective

cyclotron acceleration of ions by a radio frequency (RF) field, into an electrometer

collector that detected the current produced by the ions that were ejected from a cell.

Their instrument was designed to achieve the common objectives of early mass

spectrometer development: high mass resolving power and high abundance sensitivity.

They obtained the objectives, but due to stringent stability requirements for the

electronics and for the need of a very high vacuum, the instrument was never

commercially produced as a general purpose mass spectrometer. It did find service as an

affordable analyzer for leak detection.

The modem-day instrument can trace its direct ancestry back to the ICR

spectrometer that was constructed in the mid-1960s in a collaborative effort between John

Baldeschwieler's laboratory at Stanford University and a group of scientists at Varian

Associates led by Peter Llewellyn. 4 The ICR technique soon became widely recognized

as a preferred tool in the novel field of gas-phase ion chemistry with several instruments

installed for basic research. These early instruments did have their share of disadvan-

tages, primarily slow scan speeds and low mass resolution. With these limitations, it is

obvious why the 1970s did not see the advent of a commercially available instrument.

Even so, the limited number of instruments opened up new areas for researchers; they

provided an avenue to study ion-molecule chemistry, ion thermochemistry, and ion

spectroscopy unavailable to them previously.








4

Possibly the most important period of evolution for the technique was initiated by

professors Alan Marshall and Melvin Comisarow when they began to apply FT methods

(which were universally accepted in the field of NMR spectroscopy) to the handling of

ICR data. But before they could be successful, several key technological problems had to

be mastered. The major setbacks included: 1) lack of a suitable means of storing a range

of ions of widely varying mass-to-charge ratio (m/z) simultaneously for the length of time

needed to perform a mass measurement by the FT method, 2) lack of a feasible method to

allow resonant excitation of all ions over a broad m/z range (broadband excitation), 3) an

acceptable method to detect all ions simultaneously, and 4) lack of a fast digitizer with

sufficient memory and bit resolution available at that time. The full advantages of FT-

ICR MS would not be appreciated until these major obstacles were overcome.

Mclver solved the problem of ion storage with the development of the trapped ion

cell in 1970.15 Comisarow was able to solve the problem of broadband excitation by

applying a rapid frequency sweep of a few tens of volts amplitude. This would be

capable of exciting a wide mass range of ions, but in a much shorter time (as compared to

a slow frequency sweep with an amplitude of a few millivolts used to excite one ion at a

time over a long period). To detect all the ions simultaneously, Marshall and Comisarow

decided to measure the image current induced in a set of nearby electrodes by the packet

of ions undergoing cyclotron motion. With most of the technological obstacles solved,

Marshall and Comisarow demonstrated the advantages of the FT mode of operation of

ICR mass spectrometry in 1973."6"18 The advantages of speed, high resolution, and

effective computer data processing that accompanied the advent of FT techniques made








5

the instrument much more attractive as an analytical MS tool. After a few developmental

years, Nicolet Instruments (now Finnigan) manufactured a commercial FT-ICR mass

spectrometer in 1981. Today, three companies market and sell FT-ICR mass spectrom-

eters with the number of instruments used for studying the chemistry of gas-phase ions

rising to over 235.

The Basic Apparatus

An FT-ICR instrument is a mass spectrometer, in other words, an instrument that

observes the abundance of ions, resolved according to their masses. Most mass

spectrometers operate on the basis of spatially separating the ions through a mass-

dependent feature of their motion in a series of magnetic and/or electric fields while

collecting the ions of different masses separately onto a detector. The FT-ICR approach

is quite different since the ions are observed without separation or collection. The

method is facilitated by using the absorption and emission of RF energy at the ion's

characteristic (mass-dependent) cyclotron frequency as the ion undergoes cyclotron

motion in a strong magnetic field. This fundamental detection principle (resonant

absorption and emission of energy at a characteristic frequency) places this technique into

the company of other resonance RF spectroscopies like NMR, electron paramagnetic

resonance (EPR), microwave, and nuclear quadrupole resonance.

All FT-ICR instruments have four main components in common. These are the

need for a strong magnet, an analyzer cell, an ultra-high vacuum system, and a sophisti-

cated data system. Only a short description of each will be given since each component

in its own right could fill a chapter. The magnet can be either a permanent magnet, an










electromagnet, or more commonly, a superconducting magnet. The performance of the

FT-ICR instrument improves as the magnetic field strength increases (discussed later).

Superconducting magnets have field strengths that commonly range from 3 to 9.4 tesla.

As magnet technology improves so does the field strength and with this comes an

increase in the performance of FT-ICR MS.

The second component is the analyzer cell. This is the heart of the instrument

where ions are stored, mass analyzed, and detected. Several analyzer cell designs have

been developed with specific tasks in mind, but the first, and possibly the most common

design, is the cubic cell. It is composed of six plates arranged in the shape of a cube.

The cell is situated in the heart of the magnetic field with one opposing pair of plates

orthogonal and two pair of plates parallel to the magnetic field. The plates that are

perpendicular to the magnetic field are referred to as the trapping plates. The two

remaining pairs of plates are used to excite and detect the ions. Figure 1 depicts a basic

cubic analyzer cell used in FT-ICR MS. A recent review presents the relative advantages

of several analyzer cell designs.19

The third feature is the need for an ultra-high vacuum system. The performance

of the FT-ICR instrument is more sensitive to pressure than other mass spectrometers.

An ultra-high vacuum (pressures on the order of O1-l9-10lO Torr) is required to achieve

ultra-high resolution. To achieve these extremely low pressures, cryogenic or turbo-

molecular pumps (backed by mechanical pumps) are typically preferred rather than oil

diffusion pumps.









xyz
~B
x 'Z








Figure 1. Schematic representation of a typical cubic trapped analyzer cell commonly
used in FT-ICR MS. The three pairs of parallel electrodes and their orientation with
respect to the magnetic field are depicted.


The final feature is the need for a very sophisticated data system. Some of the
major components of the data station are a frequency synthesizer, delay pulse generator,
broadband RF amplifier and pre-amplifier, a fast transient digitizer, and a powerful
computer to coordinate all of the electronic devices during the acquisition of data, as well
as to process and analyze the data. The FT-ICR technique has benefitted tremendously
from the rapid growth and development of the semiconductor industry and will continue
to benefit as new technological breakthroughs are made.










Motion of Trapped Ions

Cyclotron Motion

Ion cyclotron resonance spectrometers are based on the principle of cyclotron

motion, by which the orbital movement of charged particles in an applied magnetic field

can be described. How these charged particles are produced will be discussed later. In a

strong magnetic field, a charged particle will experience an inwardly directed force

known as the Lorentz force ifthat charged particle has some velocity component that is

perpendicular to the direction of the field. In the absence of an electric field, the

expression for the calculation of the Lorentz force experienced by an ion is given by



FL = qv x B (2)



where q is the charge of the ion, v is the ion's velocity, and B is the magnetic field

strength. The cross product in eq 2 indicates that only those velocity components perpen-

dicular to the magnetic field contribute to the Lorentz force. Figure 2 illustrates how the

Lorentz force acts perpendicular to both the velocity and the magnetic field, resulting in a

circular orbit of the charged particle. The centrifugal force for an object undergoing

circular motion is given by



F, = mv/r (3)








where m is the mass of the particle, v is the velocity of the particle, and r is the distance of
the object from the center of rotation.







B0

\ )





Figure 2. Origin of ion cyclotron motion. The path of an ion moving in the plane of the
paper is bent into a circular orbit by the inward-directed Lorentz magnetic force produced
by a magnetic field directed perpendicular to the plane of the paper. (Taken from
reference 6)



When the centrifugal force is equal to the Lorentz force, the ion achieves a stable
circular orbit and Eqs 2 and 3 can be equated (in a simple treatment of the theory of ion
motion) as demonstrated by


mv/r = qB


The quantity vir in Eq 4 is equal to the angular frequency, co, which is the number of










radians swept out by the ion per unit time. By substitution of this expression into Eq 4

and rearranging the terms, the well-known cyclotron equation is obtained



Co. = qB/m (5)



Angular frequency (radians per second) can be converted to linear frequency (cycles per

second) by dividing by 2t. Finally, the celebrated cyclotron equation expressed in terms

of SI units is given by



vc = qB/27m (6)



Eq 6 demonstrates that a group of ions with a given m/z always exhibit cyclotron motion

at the same frequency v, (for a given value of B). This result is one of the reasons that

FT-ICR is capable of measuring spectra with ultra-high mass resolving power.

Additionally, the m/z of an ion is determined from its cyclotron frequency, and since

frequency is the most accurately measured physical quantity,20 FT-ICR provides a very

accurate method for m/z determination.

Trapping Motion

Cyclotron motion is one of three natural motions an ion possesses as a result of

being trapped by static magnetic and electric fields. A static magnetic field applied along

the z-axis effectively confines ions in the x- and y-axes according to the cyclotron motion

just described. However, ions are still free to escape in the z-axis, parallel to the










magnetic field. In order to mass analyze ions they must be trapped. Trapping of ions is

accomplished by the use of a Penning ion trap, or as it is more commonly known in the

FT-ICR MS community, an analyzer cell (described in the previous section). Trapping is

generally accomplished by applying a small (~1 volt) electrostatic potential (same polarity

as the ions of interest) to each of the two trapping electrodes (trapping plates). The ion

trapping frequency has been previously derived.6 The result is given by the expression



oT = (2aqVT/ma) (7)



where a is a constant that depends on the cell geometry, q is the charge of the ion, VT is

the trapping voltage, m is the mass of the ion, and a is a characteristic trap dimension,

which in the case of a cubic cell is the length of one side. In general, the trapping

frequency is much smaller than the ICR orbital frequency (see Figure 3).

Magnetron Motion

The third natural motion is the "magnetron" motion which results from the

relatively mass-independent precession of an ion along a path of constant electrostatic

potential. Magnetron motion arises in a natural way as one of two solutions to the

equations of (transverse) motion of an ion in static electric and magnetic fields. Although

this motion can be excited, either intentionally or as an unintentional consequence of

cyclotron excitation, it can be ignored in most FT-ICR applications.















Trapping Motion--


Cyclotron Motion "


Magnetron Motion- .





Figure 3. Schematic diagram of the natural motions of an ion trapped by a uniform
magnetic and static electric field: Go (cyclotron), oTr (trapping), and om (magnetron). The
magnetron motion is circular about a guiding center that follows a contour of constant
electric potential. (Adapted from reference 21)




Experimental Procedure

The FT-ICR mass spectrometer operates in a very different fashion than most

other types of mass spectrometers. With this technique, the principal functions of ion-

ization, mass analysis, and ion detection occur in the same space (the analyzer cell) but

are spread out in time, whereas with quadrupole and magnetic sector mass spectrometers,

these events occur simultaneously and continuously in different parts of the mass spec-

trometer. The basic series of events that occur in FT-ICR MS are referred to as a pulse

sequence and consists of four events: quench, ion formation, ion excitation, and ion

detection. This sequence of experimental events is depicted in Figure 4. The quench










event is used to empty the analyzer cell of any ions that may be present from a previous

experiment. This is accomplished simply by applying antisymmetric voltages to the

trapping plates. Under these conditions, ions are axially ejected (along the z-axis) from

the cell in less than 1 ms (+10 and -10 V applied to the trapping plates).



BASIC PULSE SEQUENCE



Quench Ionize Excite Detect







time

Figure 4. A general experimental pulse sequence that illustrates the four fundamental
steps needed in order to obtain a mass spectrum using FT-ICR MS.




Ion Formation

Samples to be analyzed can be either a solid, liquid, or a gas. There are several

possible techniques to get these samples into the gas. Solids probes are used to introduce

solids of sufficient vapor pressure into the vacuum chamber, while leak valves and/or

pulsed valves are used for liquids and gases. Once the samples are in the vacuum

chamber, FT-ICR MS detects ions, therefore the samples need to be ionized. There are

many different techniques that are used to ionize a sample depending on the specific








14

needs of the mass spectrometrist and the sample that is to be analyzed. These include (in

no particular order) electron impact ionization (El)22, chemical ionization (CI)23, laser

desorption (LD)2427, plasma desorption (PD)28'29, electrospray ionization (ESI)30'34, fast

atom bombardment (FAB)35, secondary ion mass spectrometry (SIMS)36, field desorption

(FD)37"40, and matrix-assisted laser desorption (MALDI).41"43 Two of these ionization

techniques (El and ESI) will be discussed in more detail in later chapters. Appropriate

references are cited for a more detailed discussion of each of the remaining techniques.

Ion Excitation

The ICR orbital motion that results from the magnetic and electric fields does not

by itself generate an observable electrical signal. At its instant of formation the phase of

each ion's orbital motion is random. In other words, an ion may start its cyclotron motion

at any point along the circle depicted in the left diagram of Figure 5. Thus, any charge

induced in either of the two opposed detector plates will be balanced, on the average, by

an equal and opposite charge induced by an ion whose phase is 180 different.

To circumvent the aforementioned problem of incoherency, the first step in FT-

ICR MS detection is to excite the ions. There are several excellent references that discuss

the fundamental process of excitation in FT-ICR.4446 In order to create a signal on the

detector plates, an ion packet whose cyclotron orbits are initially centered on the z-axis

must be made spatially coherent by moving the ion packet off-center. This is accom-

plished by applying an oscillating resonant phase-coherent electric field excitation that

accelerates the ions of interest into larger cyclotron orbits. An ion's orbital radius

























Figure 5. Incoherent ion cyclotron orbital motion (left) is converted to coherent (and
therefore detectable) motion (right) by the application of an oscillating voltage to the
excitation plates. Ions which are in resonance with the excitation frequency gain kinetic
energy and spiral outward from the center of the cell into a larger cyclotron orbit (right).
(Taken from reference 6)




can be determined from the following expression


r = 1/qB (2mkT)'


where q is the charge on the ion, B is the applied magnetic field strength, m is the mass of

the ion, k is the Boltzmann constant, arid T is the temperature of the ion. The previous

equation can readily be derived from Eqs 4 and 9 which relates the translational energy of

an ion to its temperature


kT=mvmxy2/2










As shown in Eq 8, the cyclotron radius of an ion can be increased by increasing its tem-

perature (its kinetic energy). In addition to increasing an ion's cyclotron radius, excita-

tion simultaneously achieves spatial coherency. The RF electric field component rotating

in the same sense (in resonance with) as the ion of interest will push that ion continuously

forward in its orbit. Thus, ions can be excited to detectable ICR orbital radii by a

relatively small RF electric field. Therefore, all ions of a given m/z range can be excited

to the same ICR orbital radius, by application of an RF electric field whose magnitude is

constant with frequency.

The goal for excitation is usually to excite all of the ions in the mass spectral

range of interest to the same cyclotron orbital radius to produce a flat spectrum without

mass discrimination. Several methods of ion excitation have been developed that achieve

these conditions. The effects of an excitation waveform can be evaluated by displaying

the excitation spectrum. This is obtained by performing an FT on the time-domain

excitation waveform. The excitation spectrum portrays the amount of excitation at any

frequency or mass, from which parameters such as ion radius, ion excitation energy for

MS/MS, and overall evenness of the excitation can be observed. The three most common

types of excitations are impulse excitation,6'47'48 chirp excitation,17'45'49 and stored wave-

form inverse Fourier transform (SWIFT) excitations.50 All three of these excitation

methods have been discussed in detail elsewhere and only a brief description of each will

be given.










Impulse excitation

An ideal delta function pulse (infinite amplitude, zero width) has a flat excitation

spectrum and should excite all of the ions equally. In real world approaches to this idea a

pulse of finite width and height is used (see Figure 6a). The shape of the pulse is not very

important. The mass range of ions that are excited extends from infinite mass to a lower

limit mass whose angular cyclotron frequency is of the order of %Oc,. = l/t, where t is the

pulse width. Mclver et al.47 have discussed the quantitative aspects of impulse excitation

and have shown that it is useful with a pulse amplifier delivering peak pulse amplitudes

of the order of 1 kV.48

Chirp excitation

The most commonly used excitation waveform is the "chirp", an RF pulse whose

frequency sweeps rapidly over the range from the lowest to the highest (or highest to

lowest) frequencies desired in the spectrum (see Figure 6b). A fairly flat excitation

spectrum results if the frequency sweep traverses a constant number of Hertz per second.

The mathematics of the FT with a chirp are slightly complicated, but were formulated

long ago (see reference 6). The advantages of chirp excitation are its rather simple

implementation and the ease with which a wide mass range can be excited without the

need for large RF amplitudes and expensive amplifiers. Disadvantages are the somewhat

nonuniform excitation of the ions, which becomes pronounced for ions near the edge of

the swept frequency range (depicted in Figure 6b).

















E -----


time


FT


frequency


Time
FFT


;P Power
Spectrum


Frequency


Figure 6. A Fourier excitation waveform and excitation spectrum for impulse (a) and
chirp (b) excitation. (Taken from reference 7)


~l/t










SWIFT excitation

In 1985, Marshall and coworkers introduced the SWIFT excitation method, which

is the most satisfactory approach to achieving complete control over the excitation char-

acteristics to date.50"52 It is based on the fact that the excitation that is actually

experienced by each ion, and therefore its final radius, is proportional to the amplitude of

the excitation spectrum at its frequency. Recalling that the excitation spectrum is

produced via an FT of the excitation waveform, a desired excitation waveform can be

produced via an inverse FT of the excitation spectrum. The SWIFT technique specifies

the excitation spectrum actually desired for the excitation waveform. This is ordinarily a

square shape that extends the desired frequency range, as in Figure 7a. It can, however,

just as well be a complicated shape with gaps, changing amplitudes, and other features as

depicted in Figure 7b. The specified excitation spectrum is inverse Fourier-transformed

to give the time-domain excitation waveform, as depicted in the right side of Figures 7a

and 7b. If carried out precisely, this gives an excitation waveform that will excite each

ion to exactly its preselected cyclotron radius.

Ion Detection

Detection in FT-ICR MS is based upon the principle of electric induction,

whereby a current flows through a circuit in response to an accumulation of charge. The

current will always flow in such a way that seeks to minimize the charge buildup. All

ions of the same m/z are excited coherently and undergo cyclotron motion as a packet.

As the orbiting ion packet passes the cell's electrodes (the detection plates), the coherent

orbiting ion packet attracts electrons to first one and then the other of the two detection




















STransform
-"4P


frequency


time


---1 ejection limit


G
0.
2o


S Transform


frequency


time


Figure 7. An illustration of the principles of SWIFT excitation (a) and a SWIFT
excitation depicting selective ejection of unwanted ions (b). (Taken from reference 7)







plates through external circuitry (see Figure 8). This alternating current is referred to as


the image current.53 The periodic cyclotron motion of the ions produces a sinusoidal


image signal which can be amplified, digitized, and stored for processing by a computer.

The frequency of the detected sinusoid is nearly equal to the frequency of the cyclotron


motion of the ions; it is exactly equal to the difference between the cyclotron and


magnetron frequencies.


u
0o
0.
tE











Electrons




X/. /r~. /


PlatT


Electrons
Figure 8. A rotating monopole description of signal generation. Positive ions approach
one plate, attracting electrons. As the ions continue moving in a circle, they approach the
other plate and attract electrons. Thus, the ion motion induces a small AC (sine wave)
current, an image current, in the detection plates. (Taken from reference 7)


Image current detection provides unique capabilities for FT-ICR MS. All other
mass spectrometers detect ions by destructive collisions with an electron multiplier.
Image current detection is non-destructive; the ions remain in the analyzer cell after the
detection process has been completed. Since ion detection is nondestructive, ions can be
repeatedly detected many times which improves the signal-to-noise ratio (S/N) since the
ion signal increases as the square root of the number of detection events. Finally, the
image current is converted to a voltage, amplified, digitized, and Fourier transformed to







22

yield a frequency spectrum that contains complete information about frequencies and

abundances of all ions trapped in the cell. Finally, a mass spectrum can then be produced

by converting frequency into mass (see Eq 6). Because frequency can be measured

precisely, the mass of an ion can be determined to one part in 109 or better.


i Ii--- U--
B. Magnitude


1'
n.e


Figure 9. Overall depiction of an FT-ICR mass spectrometer. The upper diagram depicts
the excitation of the ion packet by an externally applied alternating RF field. The lower
picture shows the detection of the image current that is produced by the coherently
orbiting ion packet on the two opposing detection plates to produce a time-domain signal.
The time-domain signal is then converted to a voltage, digitized, and Fourier-transformed
to yield a frequency-domain spectrum which is then converted to a mass spectrum.
(Adapted from reference 54)










Conclusion

The purpose of this chapter was to provide the reader with a general summary of

the FT-ICR technique, from its roots in the 1930s to the modem day instrument. Even

after its relatively short existence (roughly some 25 years), FT-ICR MS has proven itself

to be the method of choice for many researchers who are interested in the unique qualities

this technique has to offer. The mass resolution and mass accuracy achieved by FT-ICR

mass spectrometers is much higher than any other type of mass spectrometer. The tech-

nology has come to a point where even pharmaceutical and biotechnology companies are

now employing the use of FT-ICR mass spectrometers to assist them in their everyday

analysis of samples. With future advancement of higher magnetic fields (at the National

High Magnetic Field Laboratory and Battelle Pacific Northwest National Laboratory) as

well as more powerful computers, improvements in not only mass resolution and mass

accuracy but also in mass range, will undoubtedly make this technique the choice of mass

spectrometrists in the future.














CHAPTER 2
GAS PHASE BINDING ENERGIES OF SELECTED
HOST:GUEST COMPLEXES


Introduction

Gas phase binding energies of a series of amino acids trapped within the cavity of

cyclodextrin molecules were measured using FT-ICR MS. To begin the chapter, a brief

background regarding cyclodextrins will be offered. This is followed by a description of

the experimental procedures that were applied; namely electrospray ionization and

collision-induced dissociation. Finally, a discussion regarding the results obtained from

these experiments along with a few comments on possible future experiments will be

presented.

Cyclodextrin Background

Cyclodextrins (CDs) encompass a family of cyclic oligosaccharides that are

produced from the enzymatic degradation of starch. The first reference to a substance

that was later proven to be a CD was that of Villiers55 in 1891. Villiers successfully

isolated a white crystalline compound after digesting starch with the enzyme Bacillus

amylobacter. It would be over ten years before a detailed report was published by

Schardinger56 that characterized the preparation and isolation of CDs. Schardinger was

investigating strains of bacteria that were thought to be responsible for certain types of

food poisoning that were occurring near the turn of the century. After digesting the starch










with such a microorganism, he was able to isolate small amounts of two different

crystalline compounds which appeared to be identical with the "cellulosines" reported by

Villiers a decade earlier. Schardinger named the isolated microbe Bacillus macerans.57'5

It was determined much later that, during the preparation process, the starch helix is

hydrolyzed, and its ends are joined together through a-1,4 linkages.59'6 The enzymes that

are used in the digestion of starch are not specific as to the site of hydrolysis resulting in a

product that exhibits a number of different cyclic and linear dextrins.

Investigations into the chemistry of CDs have increased for several decades.

Literature regarding structures, properties, and applications of CDs have been the subject

of several books,61"67 a number of review articles,6884 more than 800 patents, and countless

papers in the years up to 1992. Their physical and chemical properties contribute to the

broad interests from different scientific disciplines. CDs are the first and probably the

most important examples of relatively simple organic compounds which exhibit complex

formation with other organic molecules. They are excellent models of enzymes which led

to their use as catalysts (for both enzymatic and nonenzymatic reactions), and they are

natural products that are readily available to most researchers.

The three most common CDs possess six, seven, and eight glucose units and are

referred to as a-, 3-, and y-cyclodextrin, respectively. CDs that contain fewer than six

glucose units are too strained to exist85 whereas those that have more than eight are very

soluble and difficult to isolate (though they have been identified by column chromatog-

raphy).86 The chemical structures of the three most common CDs are depicted in Figure

10 while Figure 11 shows the glucose units in the relatively undistorted C, chair




















































Figure 10. Compounds 1-3 are the chemical structures of the three most common
cyclodextrins: a-, 13-, and y-cyclodextrin, respectively.













0 OH
HO1
46 5 2 Ho 46 0
3 3
OH
OH HO OH

0- _..


Figure 11. Portion of a cyclodextrin molecule showing the glucose units connected
through glycosidic a-1,4 linkages.



conformation as well as the a-1,4 linkages. This arrangement allows the CD to maintain

an overall shape of a ring, or more accurately a conical cylinder, which is often described

as a doughnut or wreath-shaped truncated cone. The wider side of the cone is created by

the secondary 2- and 3-hydroxyl groups while the narrower side is created by the primary

6-hydroxyl group (see Figure 10). The number of glucose units in the ring governs the

overall dimensions of the cavity, as depicted in Figure 12. The cavity is lined with the

hydrogen atoms and the glycosidic oxygen bridges. The nonbonding electron pairs of the

glycosidic oxygen bridges are directed toward the interior of the cavity providing high

electron density along with Lewis base characteristics. As a result of this arrangement of

functional groups in the CD molecules, the cavity is relatively hydrophobic (compared to

water) while the external faces are hydrophilic. Moreover, a ring of hydrogen bonds is

also formed intramolecularly between the 2-hydroxyl and the 3-hydroxyl groups of

adjacent glucose units. This hydrogen bonding ring gives the CD a remarkably rigid

structure.










0.57 nm


I0 .79nm
0.79 nm


ac-cyclodextrin (C36H60030), cavity volume 0.202 nm3

0.78 nm


SI -0.79 nm


J3-cyclodextrin (C42H70035), cavity volume 0.377nm3


0.95 nm


0.79 nm
Inm


y-cyclodextrin (C48Hgo0040), cavity volume 0.560 nm3




Figure 12. Representation of the three most common cyclodextrins (a-, P3-, and y-
cyclodextrin) along with approximate dimensions and cavity volumes.








29

It is the cavity that generates the attraction of many disciplines to the chemistry of

CDs. As a result of the polar exterior and the relatively nonpolar interior, these com-

pounds have been studied as "host" molecules for the inclusion of "guest" molecules

which are capable of entering the cavity. The unique properties of the CD cavity explain

some of the unusual features of these molecules; thus, they form inclusion complexes

rather unspecifically with a wide variety of guest molecules. The only obvious require-

ment for the guest molecule is that it must fit into the cavity, even if only partially. Based

on this fact, it is not surprising to find that organometallics,87 amino acids,8890

peptides,88'9 1-93 aromatic molecules,94 drugs,95'96 explosives,97 and metal ions98 are

included, just to name a few in a long list of potential host species.

The evolution of host-guest chemistry started in 1967 with the discovery of crown

ethers by Pedersen.99", The term "host-guest chemistry" has been used to designate a

variety of processes occurring in a number of research fields, such as organic, analytical,

biological, pharmaceutical, and organometallic chemistry, and involving molecules and

ions of different structures, dimensions, and properties. Restricting the definition of host-

guest chemistry by considering the common elements that these disciplines possess is

possible. In general, host-guest interactions involve the establishment of multiple non-

covalent bonds between a large and geometrically concave organic molecule (the host)

and a simpler organic or inorganic molecule or ion (the guest). Guest molecules or ions

can be fully entrapped within the cavity or partially trapped as is the case with larger

species such as peptides and proteins. Table 1 lists cavity dimensions as well as other

properties of interest for the three most common CD molecules.








30

Table 1. A brief listing of selected physicochemical properties of the three most common
cyclodextrin molecules. (Adapted from reference 101)


property
no. glucose units
empirical formula (anhydrous)
mol. wt. (anhydrous)
cavity length, A
cavity diameter, A (approx.)
(XD, deg.
heat capacity (anhydrous solid), J mol' K'
heat capacity (infinite dil'n.), J mol' K"'
pK. (25 C)
AH0 (ionization), kcal mol"'
AS0 (ionization), cal mol' K-'
solubility (water, 25 C), mol L"
AH0 (solution), kcal mol'
ASO (solution), cal mol' K-'
'Mole fraction standard state.


6
C36H60030
972.85
7.9
-5.7
+150.5
1153
1431
12.33
8.36
-28.3
0.1211
7.67
13.8a


cyclodextrin

7
C42H7o035
1134.99
7.9
-7.8
+162.0
1342
1783
12.20
9.98
-22.4
0.0163
8.31
11.7"


In aqueous solution, the slightly apolar CD cavity is occupied by water molecules,

which is energetically unfavorable, and therefore, the cavity can be readily substituted by

appropriate guest molecules which are less polar than water molecules. The dissolved

CD is the host molecule with the driving force of complex formation being the substi-

tution of the high-enthalpy water molecules by an appropriate guest molecule. Most

frequently, the host:guest ratio is 1:1; this is the basis of "molecular encapsulation." A

1:1 ratio is the most common case; however, 2:1, 1:2, 2:2, or even more complicated

associations and higher order equilibria exist, almost always simultaneously. How these

CD molecules are promoted into the gas-phase for analysis by mass spectrometry will be

presented next.


8
C48H80040
1297.14
7.9
-9.5
+177.4
1568
2070
12.08
11.22
-17.6
0.168
7.73
14.7a










Electrospray Ionization

Recent advances in the biological sciences have generated a tremendous demand

for the characterization of large biopolymers including peptides, proteins, oligonucleo-

tides, and oligosaccharides. There has been tremendous pressure on the analytical

chemistry community to keep up with these advances. A few of the daily challenges that

are required of modem instrumentation are the need for rapid determination of molecular

weight, purity, sequence, and site and nature of modifications. The biological sciences

have greatly benefitted in recent years from improvements that have been made in mass

spectrometry. More than ever, mass spectrometry has been called upon for the inves-

tigation of biopolymers because it does not suffer from certain limitations of the classical

techniques102 such as gel electrophoresis or Edman degradation. The single event that

finally demonstrated the usefulness of mass spectrometry for the biological sciences was

the development of a new ionization technique. Historically, conventional mass spec-

trometric methods could only be used on low molecular weight, volatile compounds.

Larger species simply could not be promoted into the gas phase without undesirable

degradation and/or fragmentation. Since the inception of electrospray ionization (ESI),

the upper limit to molecular weight of proteins and biopolymers which can be studied has

continued to increase to well over 100,000 kDa.

The capability to impart multiple charges upon an analyte molecule is the

principal feature of ESI that distinguishes it from other ionization techniques. These

highly charged molecular ions, which normally exhibit little or no fragmentation, are

reduced to a m/z range where conventional mass spectrometers routinely operate. The










combination of this fact along with the advantages of FT-ICR MS discussed previously,

makes ESI FT-ICR MS quite possibly the most powerful analytical technique for studying

very large proteins, biomolecules, or biopolymers. An example of the power of this

coupling of techniques has been reported by Smith et al. '03 They recently reported the

mass spectrum of a protein, bovine serum albumin (MW 66,430 Da), which showed ions

with a charge distribution of+30 to +50 that corresponded to m/z values ranging from

2214 to 1329, respectively. Electrospray ionization-mass spectrometry has been used to

study a wide range of systems including proteins and glycoproteins,114 nucleotides

(including DNA, RNA, and oligonucleotides),1'05 fullerenes,16 synthetic polymers,1'07 and

inorganic transition metal complexes.108

Yamashita and Fenn33'19 were the first to demonstrate electrospray mass spec-

trometry (ESMS) which was based upon the pioneering work of Dole et al.30 The

dramatic impact of this new technique was slow to be realized by most in the scientific

community.110""2 After several years of proven success and universal acceptance of the

ESMS technique, much attention has now turned to understanding the mechanism of gas-

phase ions production from solution phase analytes. There are at least three essential

steps to consider: creation of charged droplets from dissolved electrolytes; solvent

evaporation that leads to charged droplet shrinkage followed by repeated droplet disinte-

grations (fissions); and the mechanism of gas-phase ion production from small, highly

charged droplets. A brief description of the three processes is presented below.










Production of Charged Droplets

The ES events are depicted in Figure 13. A voltage of 2-3 kV is applied to a

small metal capillary (typically 0.2 mm o.d. and 0.1 mm i.d.) which is normally located

from 1 to 3 cm from a larger planar counter electrode. The counter electrode will have an

opening that leads to the mass spectrometric sampling system in most ESMS applications

(for example, this opening allows ions to be transferred to the analyzer cell in FT-ICR

MS). Due to the size of the capillary tip, the electric field in the atmosphere surrounding

the tip is very high (E = 106 V m'). When the capillary of radius rr is located at a

distance d from the planar counter electrode, the magnitude of E, for a given potential Vr

is given by113"14



Ec = 2 V/r ln(4d/r.) (10)



This equation provides the field at the capillary tip in the absence of solution. The

electric field is proportional to the applied potential while the dominant geometric par-

ameter is the capillary radius.

Typical solutions used in ESMS will consist of a dipolar solvent in which electro-

lytes are at least fairly soluble. Methanol (or methanol/water) is generally the solvent of

choice with small amounts of acetic acid added as the source of electrolytes. For

optimum operation of ESMS, low electrolyte concentrations (10.-l104 M range) are

required. The choice of solvent, solvent mixtures, solvent mixture ratios, and













-1-3 cm







Oxidation Reduction












Figure 13. A simple depiction of the processes that occur in electrospray mass







system of interest and which would need to be characterized for optimum efficiency.
G 0 + ++























The applied electric field partially penetrates the liquid at the capillary tip. If the

capillary is the positive electrode, the negative ions in the liquid will migrate toward the

electrode while the positive ions migrate toward the liquid surface until the imposed field

inside the liquid is essentially removed by this charge redistribution. Negative ions can
also be generated for further investigation if the capillary is the negative electrode. The
High voltage
powerr supply




Figure 13. A simple depiction of the processes that occur in electrospray mass
spectrometry. (Adapted from reference 115)




concentrations are, of course, parameters that can be different depending on the specific

system of interest and which would need to be characterized for optimum efficiency.

The applied electric field partially penetrates the liquid at the capillary tip. If the

capillary is the positive electrode, the negative ions in the liquid will migrate toward the

electrode while the positive ions migrate toward the liquid surface until the imposed field

inside the liquid is essentially removed by this charge redistribution. Negative ions can

also be generated for further investigation if the capillary is the negative electrode. The










accumulation of positive charge at the liquid surface tends to destabilize the liquid

surface since the positive ions are repelled down field but cannot escape from the liquid.

The surface is drawn out such that a liquid cone forms which has been referred to as a

Taylor cone"6 in honor of Sir Geoffry Taylor.

Eventually, as the electric field is increased above a certain value, the cone

becomes unstable, resulting in a liquid filament with a diameter of a few micrometers

(with a surface enriched with positive ions) being emitted from the cone tip. Separate

droplets are formed downstream as the liquid filament becomes unstable with a continual

increase of the electric field. The droplet surfaces are enriched with positive ions for

which there are no negative counterions. The length of the unbroken liquid filament will

decrease if the electric field is increased.

Charged Droplet Shrinkage

Since the initial size and the number of charges on the droplets depend on the

spray conditions, it will be convenient for the next discussion to consider droplets that are

formed by low flow rates (-5 [tL min"') and concentrations that are less than 103 M.

Droplets formed by these conditions are considered to be monodisperse since they are

small and have a narrow distribution of sizes. It has been shown that the size distribution

peaks at a radius of about 1.5 pm while possessing a charge on the order of 10-4 C, which

corresponds to approximately 50,000 singly charged ions."7"'8

The conditions that determine when the charge, Q, becomes sufficient to overcome

the surface tension, y, that holds the droplet together are given by the Rayleigh equation"9








36

Q2R = 646oYR3R (11)



where e. is the permittivity of vacuum and RR is the Rayleigh radius.

Larger droplets (within the micrometer range or larger) maintain their charge and

do not emit gas-phase ions.8120,'121 After the droplets have decreased in size to near the

Rayleigh limit, they become unstable and begin to divide (undergo fission) into smaller

droplets as seen in Figure 14. Studies have shown that the droplets do not produce

offspring droplets of equal size and charge.118'120'121 Furthermore, it was observed that the

droplets tend to vibrate alternately from prolate to oblate shapes. These vibrations cause

disruptions where the droplet releases a tail of much smaller offspring droplets. The

offspring droplets take about 15% of the original charge and nearly 2% of the original

mass with them when they are emitted. The radius of the offspring droplets is about one-

tenth that of the parent droplets.181122'123 The total time for this sequence of events is in

the hundreds of microseconds as seen from Figure 14 (calculated using methanol).

Mechanism of Gas-Phase Ion Production

Over the years, two different mechanisms have been proposed to account for the

formation of gas-phase ions from the charged droplets. Dole et al.30 devised the first

mechanism which involved the formation of extremely small droplets (R = 1 nm) that

feature only one ion. Their mechanism allows gas-phase ions to evolve directly from

these extremely small, solvent-evaporated droplets. How these extremely small droplets

were formed or whether the process should include selectivity that may favor the

formation of gas-phase ions A' relative to B' was not addressed in their proposal.

















A' N= 51250
R = 0.945
At = 462 gis


0 N = 43560
R = 0.848
0 ~74H j ^s
N= 43560 -----_____
R = 0.939
N = 37026
N+ R = 0.761
R 34 OO 0000.....
S0.09 20droplets N=37026 OP
R=0.844 70g s 0
+
N=326 0000.....
R 0.08
SN= 31472
R 0.756 N=278
+ R = 0.03
R278 0000 Do 0
R = 0.07 r
0 .07 39 ps N
N=-236
0 R = 0.03
N=2 +
R = 0.003






Figure 14. Schematic representation of the ion evaporation model based on methanol as
the solvent. The parent droplet that is created at the spray tip undergoes uneven fission as
time passes. The depiction demonstrates how the parent droplet shrinks (losing about 2%
of its mass) and loses charge (approximately 15%) as it produces daughter droplets while
drifting toward the counter electrode. (Adapted from reference 115)






The second mechanism, proposed by Iribarne and Thomson, 124,125 assumed that


ion evaporation resulted from very small and highly charged droplets. Normally, the


droplets have a radius of about 8 nm and roughly 70 elementary charges124'125 when ion


emission becomes competitive with Rayleigh fission. At this point, the droplet releases


gas-phase ions rather than undergo fission to produce yet smaller droplets. As the










number of charges decrease, emission is still possible as a result of a decrease in the

radius of the droplets by solvent evaporation. Thus, the Iribarne mechanism does not

require the production of extremely small droplets that contain only one ion (as in Dole's

theory). Iribarne emission can occur even when the droplet contains other solutes such as

charge-paired electrolytes. At present, it is not possible to state with certainty which

theory fits better with the available evidence.

Collision-Induced Dissociation

The goal of this project was to ascertain the gas-phase binding energies of a series

of amino acids that were trapped within the cavity of CD molecules. The previous dis-

cussion pertained to the generation of gas-phase ions; next a short discussion on how the

binding energies were determined will be presented.

Several different techniques have been developed for ion structure determination,

but collision-induced dissociation (CID) remains one of the most useful and widely

implemented mass spectrometric techniques, especially when employed in an MS/MS

technique for complex mixture analysis. 126"128 Basically, this technique consists of

isolating an ion of a specific mass, accelerating the chosen ion, and allowing it to pass

through a collision gas. Upon collision, some of the ion's kinetic energy is converted into

internal energy. This allows for a faster redistribution of energy throughout the normal

modes allowing for higher energy fragmentations to occur. Normally, high kinetic

energies (3-30 keV) are required to observe CID using mass-analyzed ion kinetic energy

spectrometry (MIKES) in reverse-geometry mass spectrometers.126128 But Yost and

Enke129 were able to show that a low-energy (10-200 eV) CID process was possible with








39

high efficiency by using a triple-quadrupole mass spectrometer. This low energy pathway

is readily accessible with an ICR spectrometer using the double-resonance technique to

irradiate a given ion at its cyclotron frequency in order to accelerate it. The amount of

kinetic energy transferred to the ion is limited to being less than that required to totally

eject the ion from the cell, typically 10-1000 eV. As long as a collision gas is used at a

sufficiently high pressure (-10'5 Torr), dissociation may be observed instead of the

ejection of the ion from the analyzer cell. Thus, their quadrupole results suggested that

CID should be feasible in an FT-ICR mass spectrometer. In fact, CID was reported in the

literature using conventional ICR mass spectrometers well before the quadrupole results

appeared"30'33 but received little attention and remained essentially a curiosity. A

possible reason for the lack of interest was due to the cumbersome nature of the

experimental procedure for a conventional ICR mass spectrometer.

The infinite parallel plate capacitor approximation134"35 (given in Eq 12) has

commonly been used to calculate the translational energy imparted to an ion during the

excitation stage of the FT-ICR CID process



E1o. = q2V2?l8md (12)



where q is the charge of the ion, V is the amplitude of the RF excitation pulse, t is the RF

pulse width, m is the mass of the ion, and d is the distance between the excitation plates

of the analyzer cell. However, since the actual analyzer cell in many cases is a cubic cell,

and therefore not an infinite parallel plate capacitor, the electric fields and translational








40

energies will undoubtedly be less than those values predicted for an infinite parallel plate

capacitor. In fact, calculations and ion motion simulations'36 have demonstrated that

excitation of ions located at the center of a cubic analyzer cell reached a radius that was

only 72% of the theoretical radius that was calculated using an infinite parallel plate

capacitor approximation. The actual excitation is even less since an ion's translational

energy is proportional to the square of its radius. Therefore, actual excitation is only 0.52

of that predicted by Eq 12. It is extremely important to keep the ion excitation time to a

minimum. The important thing is that collisions between the ions and the collision gas

occur after translational excitation, otherwise, uncertainties in the amount of energy

actually imparted during the excitation process will result.

Some of the most dramatic and promising MS/MS applications in FT-ICR MS

involve sequential dissociations in which successive fragmentation of the parent ion into

smaller and smaller fragments is followed by a series of excitation/observation steps on

the successive fragments.'37"38 An excellent illustration of MS" analysis was

demonstrated by Freiser and Gord,139' where five CID steps were used, along with

selective ejection, to proceed from FeS O0 down to Fe (see Figure 15). Carrying such

multiple MS/MS observations to four or five steps is basically equivalent to a multisector

or multi-quadrupole MS/MS experiment using a long (and impractical) series of sectors

or quadrupoles. This experiment illustrated the use of an FT-ICR MS as a series of

temporally separated mass analyzers.

The basic CID process can be thought of in terms of two consecutive steps that

occur on well-separated time scales. The first is a rapid step (-10'-10-"4 s) in which a

















Fe' FeS,

FeS7

C FeS,



FeS FS


5 0 i 26 6 i 46 56 10 l 200 250 300 350 0
m/z
CFeSo FS








FeSFoe

F*Sa*
SO 100 ISO 200 250 300 350 400 100 SO 200 250 300 3SO 400









Figure 15. Resulting mass spectra following five stages of CID of FeS10. (a) Isolation of Fe' following laser desorption and
(C) (8








FScollisional cooling with argon and S. (b) Reaction ofFe with S. (c) Isolation of FeS (d) CID of FeS0. (e) Isolation of FeS.
FeS,
Fe510

FeS4/





50 100 ISO 200 350 300 3S0 400 50 100 ISO 200 250 300 350 400

Figure 15. Resulting mass spectra following five stages of CID of FeS10. (a) Isolation of Fe+ following laser desorption and
collisional cooling with argon and S8. (b) Reaction of Fe~ with Sg. (c) Isolation of FeS i0*. (d) ID of FeS,0. (e) Isolation of FeS8g.
(f) CID of FeS8. (g) Isolation of FeS6. (h) CID of FeS6. (i) Isolation of FeS4. (j) CID of FeS4t. (k) Isolation of FeS2. (1) CID of
FeS2. (Taken from reference 139)




















FeS*


FeS,+
FeS,+





......... ISO 200 2.. 300 350 400
50 ,oo ,So 200 250 .o +o


FeS4+


FeS j-




--m... w -. .o aal, .S,...,4,0.-


SO 100 ISO 0 260 0 300 S3O 400 Si t0 ISO 260 iO 300 30O 400


Figure 15. Continued.


1(e)


I.)



U


FeS,+


FeS6,


so50 100 ISO150 200 250 300 350 400
m/z
(g)

FeS6











FeS4"















FeS4+


FeS+


- 0~
0-



0


50 100 150 200 25O 300 350 400


m/z

















ONOr~ 11


5sO IO o 2d0 2SO 310 3O 400


(1) Fe+


FeS2+


FeS3+





50~ 00 1 0 0 300 'T350 0


Figure 15. Continued.


La


(k)

FeS +













,L.,ln. l..J i.L


FeS4+










small amount of the initial translational energy of the accelerated ion is changed into

internal energy of both the ion and target molecule (the target molecule also acquires

translational energy). The next step in this process is the dissociation of the energized

(and typically isolated) ion. The yield of product ions after collisional dissociation

depends on the probability of unimolecular decomposition of the precursor ion after

excitation. To explain the rates of such reactions, quasi-equilibrium theory (QET) has

been used.40"144 In simple terms, the theory states that unimolecular decomposition

reactions depend upon the random distribution of the internal energy of the ion among all

the vibrational modes of that ion. In other words, the rate of decomposition is related to

the probability of a given vibrational mode or modes acquiring enough energy to rupture

bonds.145 Since there are 3N 6 vibrational modes in a nonlinear ion that contains N

atoms, the number of vibrational modes will be in direct proportion to the molecular mass

for a given class of compounds. Because the random distribution of internal energy

among the vibrational modes is required by QET, the average energy per mode must

decrease with increasing molecular mass. Because the decrease is related to the inverse

of the mass of the ion, the fragment ion yield should decrease similarly beyond some

threshold.

In the early days, low-energy CID mass spectra were observed in quadrupole

reaction chambers of triple quadrupole or hybrid sector-quadrupole mass spectrometers,

but in recent years low-energy CID has been performed more and more with FT-ICR

mass spectrometers. Regardless of the choice of instrument configuration, an obvious










requirement is the presence of a collision cell that can be pressurized with a suitable

target gas.

Collisions that occur at large energies (keVs) are assumed to result in excitation of

electronic internal modes (and to a lesser extent rotational-vibrational modes); collisions

that occur at lower energies (<100 eV) will no longer result in efficient transfer of trans-

lational energy to electronic internal modes. A typical interaction time of a selected ion

of mass 200 and a translational energy of 30 eV with a target molecule over a few

angstroms is on the order of 10"13 s. This is longer than the time needed for internal

electronic excitation and thus, the probability of such excitation is reduced with respect to

excitation by high-energy (keV) CID.146

The interaction time of ca. 10"13 s is comparable to the reciprocal of typical

vibrational frequencies. Under these conditions, the collisions are nonadiabatic and the

interaction is described as having an impulsive character that can effectively induce

energy transfer.46 The subsequent transfer of translational to vibrational energy is

believed to occur by internuclear momentum transfer. 147

In the so-called binary or spectator model, the selected ion and the target gas

engage as essentially structureless elastic spheres. Momentum is transferred between the

two bodies which leads to rotational-vibrational excitation of the ion, as well as the gas,

along with a shift in momentum in the center of mass of each. If the ion is much larger

than the collision gas, the gas will only interact with a small portion of the selected ion.

The maximum amount of energy (center-of-mass kinetic energy, Eom) accessible for










internal excitation is given by Eq 13148



Ecom, = Elabmt/mp + m(mp/piM)] (13)



where mp is the projectile mass, mpi is the impact portion of the projectile, and m, is the

target mass. The elastic limit is thought to be reached when mp = mpi.

With low-energy collisions, the composition of the target gas plays a much more

important role than it does with high-energy collisions. The reason for this is that a

different excitation mechanism vibrationall excitation) is at work. Furthermore, a larger

portion of the maximum available energy is converted into internal energy of the target

ion. Bursey and co-workers.49'.5. demonstrated that heavier targets are preferred over

lighter targets because they provide a larger Ecom. They found that collisions using helium

transferred very small amounts of energy when compared with collisions using nitrogen,

argon, or krypton.'49"'52 Specifically, for every volt change in Elab of the selected ion at a

given pressure, there was an increase of 0.04, 0.25, and 0.32 eV, in maximum possible

energy transferred when using helium, nitrogen, and argon, respectively.'52

Experimental

The experiments were conducted on a Bruker CMS 47X FT-ICR mass spec-

trometer (Bruker Daltonics, Billerica, MA) incorporating a shielded 4.7 T supercon-

ducting magnet (Magnex Scientific Limited, Abingdon, England) and a modified external

electrospray ionization source (Analytica of Branford, Inc., Branford, MA). Figure 16

depicts the instrument used throughout this work. The commercially-sold glass capillary

















Turbo-


I ~ 2 lT 1 LA M I Needle





400 Ls 800 L Syringe
., .,,. l -I | *^ _ PL2 M echanical
| EV2/DEV2 I 1 F-^ DPL2 PumpP
4.7 T Magnet Vat Fly



400 L/s 800 Lis 10
Cryo Cryo
Pumps Pump








Figure 16. Fourier transform ion cyclotron resonance mass spectrometer used to determine the gas-phase binding energies of
cyclodextrin:amino acid complexes. The instrument employed a shielded 4.7 T magnet, an external electrospray ionization source,
and possessed three stages of differential pumping to achieve analyzer cell pressures on the order of 5.0 x 10-9 Torr.










which utilized heated N2 for desolvation was replaced in favor of a heated metal (brass)

capillary (designed and built in-house)153 which produced a more stable ion current. The

capillary utilized a cartridge heater and under normal operating conditions was heated to

between 100-130C to assist in generating desolvated gas phase ions.

The CDs (a-, P3-, and y-) and amino acids were provided by Dr. Lazslo Prokai and

used without further purification. Typical solution concentrations were 10'4 M for the

CDs and 10-3 M for the amino acids. The samples were sprayed from a water/methanol

(50/50) solvent with a small amount of acetic acid added to provide charge. The

solutions were introduced to the electrospray needle with the aid of a 74900 Series

syringe pump (Cole-Parmer Instrument Company) normally operating at 60 tl hr'. The

electrospray needle potential was normally maintained near +3500 V while the capillary

was effectively kept at ground. The pressure in the external ion source was maintained at

1 xl 06 Torr via pumping by an 800 L s' cryopump (Edwards High Vacuum International,

West Sussex, England). After entering the external ion source, the ions were guided to

the analyzer region of the mass spectrometer by a series of electrostatic ion optics which

were optimized for efficient ion transfer. Pressures in the analyzer region were typically

maintained at 2xl0.9 Torr by additional pumping from two 400 L s' cryopumps. Immedi-

ately before entering the analyzer cell, the ions were given a "sidekick" to minimize any

z-axis loss.'54 Once in the cell, the ions were trapped using trapping potentials of+1.0 V

and +1.4 V on the two opposed trapping plates.

Inside the analyzer cell, the ions were isolated using an RF notch ejection pulse

(see Figure 17 for a typical pulse sequence). Following a 3 s cooling delay to allow for











HD Q IG MS IA E D









Figure 17. Typical pulse sequence used for the CID studies. HD is the Hexapole Dump,
Q is the Quench pulse, IG is the Ion Generation pulse, MS is MS/MS Coarse Selection,
IA is the Ion Activation pulse, E is the Excitation pulse, and D is the Detection pulse.




thermalization of the ions, the collision gas was introduced through either a piezoelectric

pulsed valve (50 ms pulse) or a leak valve (Varian) to perform CID of the parent ions

(cell pressure 1 x 10-7 Torr). A 100 us RF activation was used to translationally excite the

ions, which were then allowed to undergo collisions and fragment during a subsequent

250 ms reaction delay. The resulting reactant and product ions were detected via

frequency-chirp excitation. Broadband detection, covering a mass range of 50 to 2500

amu, was utilized in these experiments. During detection, 64 spectra (64k data sets) were

normally acquired and signal averaged in order to increase the S/N.

Results

The aim of this project was to determine the gas-phase binding energies between

the three common CDs and the twenty essential amino acids, with thoughts of increasing

the size of the guest molecule to include peptides and small proteins. The goals of this

project were only partially realized. Results for a-CD with tryptophan, proline, and










lysine, and P-CD with tryptophan and histidine are presented below. A reference to the

twenty essential amino acids is offered in Table 2 below.

a-CD:Tryptophan

The experimental parameters used to study the [a-CD:Trp]H' system (as well as

the other systems) are summarized in Table 3 (found at the end of the chapter). The

tryptophan side chain possesses a nitrogen-containing ring as well as a phenyl ring and is

therefore classified as an aromatic amino acid. The cavity diameter of a-CD has been

reported to be between 4.7-5.7 A; therefore, the encapsulation and complexation of the

side chain occurs easily by a number of interactions (hydrogen bonding, electrostatic,

hydrophobic, and/or van der Waals). Figure 18 shows the isolated complex at m/z 1177.

A typical CID mass spectrum depicting the uncomplexed [Trp]H' (m/z 205) as well as

other unidentified fragments is given in Figure 19. Fragmentation of this ion was

measured as a function of the ion center-of-mass kinetic energy (an indirect measure of

the amount of energy imparted to the ion in the collision). The variable throughout the

experiment was the amplitude of the RF excitation pulse, which is to say, the amount of

energy imparted in the collision. Attenuation of the RF excitation pulse varied from 11 to

17 dB for the [aot-CD:Trp]H' system (which corresponded to 240.2 to 89.8 Vp.p applied the

analyzer cell plates). The threshold binding energy was determined by extrapolating the

linear portion of the graph to zero (see Figure 20). The threshold binding energy for [ca-

CD:Trp]H was determined to be 1.32 eV.











Table 2. The twenty essential amino acids along with their appropriate symbols and
masses.


H2N CH COOH
I

R

General Amino Acid Structural Unit With Distinctive R Group


Name R group Symbols Monoisotopic Mass Average Mass

Nonpolar, aliphatic R groups

Glycine -H Gly, G 57.02146 57.0520
Alanine -CH3 Ala, A 71.03711 71.0788
Proline -(CH2)- Pro, P 97.05276 97.1167
Valine -CH(CH3)2 Val, V 99.06841 99.1326
Leucine -CH2CH(CH3)2 Leu, L 113.08406 113.1595
Isoleucine -CH(CH3)C2H5 Ile, I 113.1595 113.1595

Aromatic R groups

Phenylalanine -CH2C6H5 Phe, F 147.06841 147.1766
Tyrosine -CH2C6H4OH Tyr, Y 163.06333 163.1760
Tryptophan -CH2(C2H2N)C6H4 Trp, W 186.07931 186.2133

Polar, uncharged R groups

Serine -CH2OH Ser, S 87.03203 87.0782
Threonine -CH(CH3)OH Thr, T 101.04768 101.04768
Cysteine -CH2SH Cys, C 103.00919 103.1448
Asparagine -CH2CONH2 Asn, N 114.04293 114.1039
Glutamine -(CH2)2CONH2 Gin, Q 128.05858 128.1308
Methionine -(CH2)2SCH3 Met, M 131.04049 131.1986

Positively charged R groups

Lysine -(CH2)4NH2 Lys, K 128.09496 128.1742
Histidine -CH2(C3H3N2) His, H 137.05891 137.1412
Arginine -(CH2)3NH-C(N2H3) Arg, R 156.10111 156.1876

Negatively charged R groups

Aspartate -CH2COOH Asp, D 115.02694 115.0886


rl1*Y P 110 nA)%Q 110 1 1 q;


.hl 1tamatff


-Mi_ "O nU


rI; ii


19o t A- go


1nQ llqq<










a-CD:Proline

Proline has an aliphatic ring and is, therefore, classified as a nonpolar amino acid.

Figure 21 is the mass spectrum of the isolated [ct-CD:Pro]H at m/z 1088. A typical CID

mass spectrum depicting the free protonated proline (m/z 116) as well as the parent ion is

presented in Figure 22. Attenuation of the RF excitation pulse used for the [a-CD:Pro]H'

system ranged from 13 to 19 dB (169.7 to 65.2 Vp.p). Figure 23 highlights the linear

portion of the graph which resulted in a threshold binding energy of 1.21 eV when

extrapolated to zero.

x-CD:Lysine

The side chain of lysine possesses an amine group and lysine is thus classified as a

positively charged amino acid. The isolated parent ion (m/z 1119) can be seen in the

mass spectrum shown in Figure 24. Figure 25 is a representative CID mass spectrum of

[a-CD:Lys]H showing the free protonated lysine (m/z 147) as well as the parent ion.

Attenuation of the RF excitation pulse used for [a-CD:Lys]H varied from 12 to 30 dB

(204.9 to 17.83 Vp.p). Figure 26 depicts the linear portion of the graph which resulted in a

threshold binding energy of 0.71 eV when extrapolated to zero.

P-CD:Tryptophan

The cavity of the P-CD is larger than that of the a-CD; therefore, the encapsul-

ation of the amino acids should be facilitated. The isolated [0-CD:Trp]H at m/z 1339

can be seen in the mass spectrum depicted in Figure 27. A typical CID mass spectrum is

illustrated in Figure 28. The free protonated tryptophan (m/z 205) can be seen in the

expanded region of the spectrum. Attenuation of the RF excitation pulse for








53

[P-CD:Trp]H' varied from 11 to 20 dB (240.2 to 57.1 Vp.p). Figure 29 depicts the linear

portion of the graph which resulted in a threshold binding energy of 0.58 eV when

extrapolated to zero.

P-CD:Histidine

Histidine is classified as a positively charged amino acid since the side chain

contains a ring with two nitrogen atoms. Figure 30 shows the mass spectrum depicting

the isolated parent ion at m/z 1290. Attenuation of the RF excitation pulse for [P-

CD:His]H ranged from 12 to 21 dB (204.9 to 50.7 Vp.p). A CID mass spectrum showing

the free protonated histidine (m/z 156) as well as the parent ion is given in Figure 31.

The threshold binding energy was determined to be 0.73 eV from the graph in Figure 32.

Discussion

Unfortunately, a great deal of information cannot be inferred from these data due

to the limited number of systems. A comparison can be made between the complexation

of tryptophan to a-CD and P-CD. The intermolecular forces that bind the tryptophan to

the interior of the CD would presumably be stronger for the at-CD case due to the smaller

diameter. With a smaller cavity diameter (closer proximity of atoms), the a-CD would

bind more tightly to the tryptophan than the P-CD would. This is reflected in the two

threshold binding energies that were measured; 1.32 and 0.58 eV for the ac-CD and p-CD

systems, respectively. It appears that twice the energy was necessary upon collision to

eject the tryptophan from the a-CD cavity than the P-CD cavity.








54

The threshold binding energy between a-CD and lysine, tryptophan, and proline

can be compared since the cavity size in each case remains the same. Given the size of

the a-CD cavity the binding strength should be the greatest for tryptophan and the

weakest for lysine. Since the side chain of lysine is simply a straight chain (weakest

intermolecular forces), the lysine would not be encapsulated as tightly as with the other

two amino acids. As for the tryptophan and proline case, they both possess a ring and,

therefore, would be of comparable size. However, the tryptophan would extend deeper

into the cavity. In addition, the intermolecular forces holding the tryptophan complex

together would be stronger due to the increased mass over the proline complex (all other

intermolecular forces taken to be equal). The calculated binding energies did, in fact,

follow this trend yielding 1.32, 1.21, and 0.71 eV for a-CD with tryptophan, proline, and

lysine, respectively.










Table 3. Conditions for studying the various [CD:amino acid]H' complexes.


n-CflPrn n,..rr')1 v~


R-_CfTrn R-.'-li4


[CD] (M)
[amino acid] (M)
L1 (V)
L2 (V)
L3 (V)
L4 (V)
L5 (V)
capillary temperature (C)
needle voltage (V)
flow rate (tl hr')
collision gas
collision gas pressure (Torr)
RF amplitude (Vp.p)
RF pulse width (ps)
Bindin2 Eneryv (threshold. eV)


8.20 x 10-4
3.59 x 10-3
143.9
23.5
-3.3
-48.6
103.8
132
3397
60
Kr
-5.0 x 10-"
240.2-89.8
50
1.32


8.20 x 10-4
8.13 x 10-3
82.3
26.3
-1.6
-49.9
102.1
132
3406
60
Kr
-5.0 x 10-8
169.7-65.2
50
1.21


2.06 x 10-5
2.33 x 10-3
136.6
27.9
-22.8
-55.3
106.9
132
3403
60
Kr
-5.0 x 10-8
204.9-17.83
50
0.71


2.06 x 10-5
5.43 x 10-3
155.4
27.1
6.5
-34.8
85.9
130
3400
60
Kr
-5.0 x 10-8
240.2-122.2
50
0.58


t-Cfl-Trn


2.06 x 10-'
4.25 x 10-3
142.3
18.1
25.2
-56.1
103.7
134
3410
60
Kr
-5.0 x 10-
204.9-50.7
50
0.73


rf-("T)oTrn ry-('TI*Prn rv-('TIaT
-Lba S &M 2 hz


132















1177.5103



80

U


1 60




40-




20





200 400 600 800 1000 200 10 1400
m/z


Figure 18. Mass spectrum of the isolated [a-CD:Trp]H at m/z 1177. The unlabeled peaks demonstrate the inefficient ejection of
unwanted ions during the isolation of the parent ion.



















1177.5786


205.0912
. I


200


400


800


1000


Is
II --


1200


Figure 19. The CID mass spectrum of [a-CD:Trp]H' showing the free protonated tryptophan at m/z 205.


1400


! J


.11 *




















4.0-


3.5--


"1 _


0
1X
2.5

2.0




1.5
M),,


F


4-


4-


L


I.U


0.5


I I I


1.0 1.5 2.0
Ion Center-of-Mass Kinetic Energy (eV)


3.0


Figure 20. Percent fragmentation versus ion center-of-mass energy for [ao-CD:Trp]H'. Extrapolation of this line to zero yields a
threshold binding energy of 1.32 eV.
















1088.6634


I. -, T.... z--- lr1A*A


200


400


600


800
m/z


1000


Figure 21. Mass spectrum of the isolated [a-CD:Pro]H' at m/z 1088.


* A


1200


1400












100-
1088.6375


80-
U

< 60-
c 116.0714
I /
ci
40-



20-




200 400 600 800 1000 1200
m/z


Figure 22. The CID mass spectrum of [ct-CD:Pro]H* showing the free protonated proline at m/z 116.


1400
1400
















30 -T-


0




15



S 10


i II I [ I I i Ii I
o.0 0.5 1.0 1.5 2.0 2.5 3.
Ion Center-of-Mass Kinetic Energy (eV)


Figure 23. Percent fragmentation versus ion center-of-mass energy for [ca-CD:Pro]H+. Extrapolation of this line to zero yields a
threshold binding energy of 1.21 eV.







































200


400


600


800


Figure 24. Mass spectrum of the isolated [a-CD:Lys]H' at m/z 1119.


1119.4756


1000


12 00o


1400













1119.4769


147.1086
1


200


400


600


800


1000


1200


Figure 25. The CID mass spectrum of [a-CD:Lys]H showing the free protonated lysine at m/z 147.


1400


.i ~ ~~I ..." '


I























84--
0 /









2 --
0







0 0.5 1 1.5 2 2.5 3
Ion Center-of-Mass Kinetic Energy (eV)





Figure 26. Percent fragmentation versus ion center-of-mass energy for [a-CD:Lys]H'. Extrapolation of this line to zero yields a
threshold binding energy of 0.71 eV.











































200


400


800


Figure 27. Mass spectrum of the isolated [I3-CD:Trp]H at m/z 1339.


1339.475


1000


1200


1400


. . .ll . . .
















1339.365



80
205.0974
4)
u
o-

I 60-
<4
>



40-



200 205 210
20-




&I I Am qI A O i s o
200 400 600 800 1000 1200 1400
m/z




Figure 28. The CID mass spectrum of [p-CD:Trp]H. The free protonated tryptophan at m/z 147 can be seen in the expanded portion
of the spectrum.

















































I I I I I


I*I I I
0 1 2 3
Ion Center-of-Mass Kinetic Energy (eV)


Figure 29. Percent fragmentation versus ion center-of-mass energy for [P-CD:Trp]H'. Extrapolation of this line to zero yields a
threshold binding energy of 0.58 eV.


10-r


8-

C
'U

6-


pro
06





2-
0 4
ul


I !


V


















100'
1290.5442



80
u
u



60
OS



40-




20- 1312.8553
/



-(L4
200 400 600 800 1000 1200 1400
m/z





Figure 30. Mass spectrum of the isolated [0-CD:His]H' at m/z 1290.
















156.1006


1290.5645


200 400 600 800 1000 1200 1400


Figure 31. The CID mass spectrum of [P3-CD:His]H+ showing the free protonated histidine at m/z 156.



















100 -


40+


20 +


.1 I I I I I


0.0 0.5


I I I i
1.0 1.5 2.0 2.5
Ion Center-of-Mass Kinetic Energy (eV)


3.0 3.5
3.0 3.5


Figure 32. Percent fragmentation versus ion center-of-mass energy for [PI-CD:His]H. Extrapolation of this line to zero yields a
threshold binding energy of 0.73 eV.














CHAPTER 3
MOTIVATION FOR INVESTIGATING POLYCYCLIC AROMATIC
HYDROCARBONS OF ASTROPHYSICAL IMPORTANCE

Introduction

The work presented in chapters four and five was undertaken with the purpose of

studying the products of photodissociation from a series of polycyclic aromatic hydro-

carbon cations that have potential interest within the astrophysical community. The goal

was to find laboratory analogues of the species that are possible carriers of the diffuse

interstellar bands that have been observed in specific regions of space. An appropriate

background discussion will be offered in the present chapter to show the importance of

these studies. Following the background discussion, a brief review will provide previous

experimental results along with the current direction of our research.

Background

Polycyclic aromatic hydrocarbons (PAHs), more simply known as polyarenes,

embody an extraordinarily large and diverse class of organic molecules that are generated

from fused benzene rings. The major sources of PAHs on this planet are crude oil, coal,

and oil shale. The fuels produced from these fossil sources constitute the primary source

of energy for the industrial nations of the world, and the petrochemicals produced from

these raw materials are the basis of the synthetic fibers and plastics industries. Major

considerations of this family of molecule have been their prominent roles as










environmental toxins,155 exemplified by the discovery in the mid-1930s of the carcin-

ogenic properties of benzo[a]pyrene and other polyarenes. This discovery was an

important landmark in biomedical science since it was the first indication of disease

caused not by a microorganism, but by a relatively simple organic molecule. Since this

discovery, researchers have found significant levels of these environmental contaminants

in the air we breathe, the food we eat, and the water we drink; most of which can be

attributed to the burning of fossil fuels as we have continued to exist as an industrialized

nation.

During the past two decades, PAHs have again received a considerable amount of

interest, but this time it was the astrophysical community that found PAHs to be

intriguing molecules. During this time period, researchers have presented insurmountable

evidence that PAHs are also important members of the interstellar medium (ISM), quite

possibly being the third most abundant detected molecules behind H2 and CO.'56 Yet, the

likely presence of PAH molecules and ions beyond the earth's atmosphere in space is still

somewhat of a novelty for many chemists. Even though they have not been unequiv-

ocally identified, aromatic and PAH molecules and ions are now generally accepted by

the astrophysical community to be present in interstellar and circumstellar environments.

Some have even argued for the presence of nonplanar PAHs and of hollow cages of

carbon atoms known as fullerene molecules.

The hypothesis that PAHs are present in interstellar and circumstellar environ-

ments already has a substantial historical record.157-162 Discussions dealing with the

presence of PAHs in interstellar environments began with the visionary suggestion,163 in










1956, that related carbonaceous species are responsible for visible diffuse absorption

bands. This fact crystallized with the discovery64 that some astronomical objects emit a

broad infrared emission band which peaks at 3050 cm' as well as other unique emission

features 165,166 which peak in the region between 1610 and 890 cm'. Astronomical objects

which emit these features include regions associated with individual stars such as H-II

regions and reflection nebulae as well as interstellar clouds like the IR Cirrus, both in our

own and other galaxies.'67170 Soon after their discovery, these infrared emission features

were attributed to infrared fluorescence from molecular-sized emitters excited by the

absorption of ultraviolet and visible photons.'7172 The idea that the fluorescence

originated from vibrations of chemical groups attached to aromatic constituents of

amorphous carbon particles'73 led to the proposal that individual PAH molecules are

responsible for the infrared emission due to their stability against photodissociation and

the resemblance of laboratory infrared fluorescence data of such species to the

astrophysical spectra.'74""76 PAH molecules have also been proposed as carriers of visible

diffuse interstellar bands.156177'178 Presently, the PAHs responsible for the infrared

features are thought to be more abundant (-17% of the cosmic carbon) than all of the

other known gaseous interstellar organic molecules combined.'8'

Current models of interstellar and circumstellar chemistry have emphasized planar

PAHs with arrangements of hexagonal rings. These rings are more or less compact

("catacondensed") with the general formula C6pH6p such as coronene (C24H,2) or

elongated polyacenes with the general formula C4,+2H2,+4 such as naphthalene (CIoH8),

anthracene (C 4H10), or tetracene (C18H,2). 57 PAHs with loose arrangements of hexagonal










rings, such as those bound by single carbon-carbon bonds, have largely been neglected.

In addition, non-hydrocarbon aromatic molecules have also been excluded. Neutral and

positively-charged fused-ring molecules such as pyrene, coronene, and ovaline, either

completely or partially hydrogenated, have been invoked to account for both the observed

broad IR emission features in nebulae74'176 and the observed diffuse interstellar

absorption bands.77'79 Recent detection of additional sharp emission features'80 has led

to the proposal that much simpler linear fused-ring molecules such as naphthalene,

anthracene, and tetracene are responsible for the infrared emission and that benzene may

also be present in these environments.'8' For example, anthracene has been suggested as

the most abundant of these linear polyacenes in the Orion Ridge."8'

Diffuse interstellar bands (DIBs) are ubiquitous absorption features in astro-

nomical spectra. They absorb from approximately 4,400 A into the near infrared.

Identifying the carriers of the DIBs has become a classic spectroscopic problem of the

20th century. Since their original discovery in 1922 by Heger, 82 these bands have

challenged spectroscopists, astronomers, and physicists, and their origin remains the

longest standing unsolved problem in all of spectroscopy.'83 During this time, so many

suggestions have been made, experiments carried out, and theories proposed that it would

be impossible to review them in this work. The experimental challenge was succinctly

stated by Johnson'84 almost thirty years ago, "...one not only has to match 25 diffuse

interstellar lines as far as wavelengths are concerned, but also as far as intensity. In

addition, the... interstellar line widths vary from 40 A to 1 A... and are invariant to 0.1

A..." Since that time, the number of DIBs has grown to nearly 160 and their relative








75

intensities have been shown to vary from one line-of-sight to another. In addition, some

DIBs seem to associate in loosely connected families."1857 Wavelength invariance has

been taken to indicate that the carriers cannot reside in or on dust particles. This is

because particle size, shape, and composition influence peak position and profile, and it is

difficult to imagine that the grains along all lines-of-sight are exactly the same.

The criteria which must be met for a particular material to be considered for

acceptance as a DIB carrier are that its visible and near-infrared absorption features match

the known DIBs in wavelength, bandwidth, and relative intensities while not possessing

additional features that are absent in the interstellar spectra. Of the aforementioned 160

DIBs, a carrier for two has almost certainly been identified; the DIBs at 9577 A and

9632 A closely agree with expected spectral features due to the C60 fullerene cation.'88

Current theories from leading researchers in the field believe that the carriers are PAH

cations or, to account for the discrepancies in peak intensities, mixtures of PAH cations

and neutrals.

Since PAHs are believed to be ubiquitous and abundant in the interstellar

medium175"176 and are stable against ultraviolet photodissociation, they are attractive

candidates for the DIB carriers. Moreover, a large fraction of the PAHs are expected to

be ionized in the interstellar medium,156,176-178 and thus, absorb lower-energy photons

(mainly in the visible and near-infrared regions of the spectrum) than their neutral

precursors. "'89










Related Studies

The only available data on PAHs for many years have come from absorption

spectra of neutral and ionized PAHs suspended in perturbing media (solid phase.89"191 or

solution192) or from gas-phase photoelectron spectra which do not provide information on

all the possible optical transitions.'93 Allamandola was one of the first investigators to

systematically measure the spectroscopic properties of neutral and ionized PAHs in the

ultraviolet, visible, and near-infrared range (1,800-9,000 A) under conditions relevant to

astrophysical environments. This was accomplished by studying the isolated species in

the least-perturbing solid medium known, neon matrices at 4.2 K. Neon generally

produces shifts in vibronic positions of only a few tenths of a percent with respect to the

gas phase.'94 Since their first report on naphthalene (the smallest PAH) in 1991, the

Allamandola laboratory has published several articles dealing with matrix-isolated PAH

and PAH cations. But, as expressed by Allamandola, even at 4.2 K in a neon matrix,

there still exist matrix effects that are absent in the interstellar medium. The ideal

situation would be to study these systems in the gas phase where perturbations from a

matrix are absent.

Boissel and co-workers195 reported the results of a study using a FT-ICR MS with

the Penning trap placed in an ultrahigh vacuum cell attached to a closed cycle helium

cryostat. The temperature of the parts ranged from 12 K to 30 K, ensuring a very low rate

of ion-neutral collisions (confirmed by the long trapping times of up to ten minutes). Ions

were produced by laser ablation of a solid PAH pellet located near one of the trapping

plates. The trapped ions were irradiated by a focused cw xenon arc lamp that had a








77

computer-controlled mechanical shutter to allow light into the cell. The length of irradi-

ation (in other words the amount of energy imparted to the ions) was varied throughout

the experiments. Mass spectra recorded the fragmentation after irradiation of the anthra-

cene, anthracene-d,0, and pyrene cations. The photodissociation pathways were found to

be the loss of C2H2, C2D2, and H2 (or two hydrogen atoms) for the three previously listed

ions, respectively.

Two more papers appeared in the literature in late 1997 that pertained to PAHs

and astrophysical implications. Ekem and co-workers196 used FT-ICR MS to study the

coronene and naphtho[2,3-a]pyrene cations. Upon irradiation from a xenon arc lamp,

each of the two ions was found to dehydrogenate completely to leave the C24' bare carbon

cluster cation. In the second paper, Wang et al.'97 examined the CID mass spectra of a

series of PAH cations using a modified ion trap. Ions of interest were the naphthalene,

acenaphthylene, acenaphthene, anthracene, phenanthrene, pyrene, coronene, and the

corannulene cation. Mass spectra were recorded after the selected PAH cation was

allowed to undergo collisions with argon buffer gas at a pressure of 1-3 x 104 Torr.

Several different fragment ions were observed and are reported in Table 4.

Ekern et al.'98 extended their earlier study by examining the photostability of a

series of twenty-four PAH cations and one fullerene, C60'. Electron impact and laser

desorption were used to generate the ions which were then trapped and mass-analyzed by

an FT-ICR mass spectrometer incorporating a 3 T magnet. The trapped ions were

subjected to irradiation from a xenon arc lamp and the fragmentation products were

recorded. From this series of PAH cations, it was discovered that the observed fragmen-

tation patterns fell into one of four categories: photostable, loss of hydrogen atoms only,











Table 4. Observed fragmentation channels and efficiencies for selected PAH cations
using the ion-trap detector. (Reproduced from reference 197)


PAH+ 'Relative abundance of fragment ions bEf(%)
(M-H)* (M-2H) (M-3H) (M-4HW (M-2C.2H) (M-2C.3H)(M-2C.4HY(M-4C.2Hr
naphthalene 100 95 0 0 10 0 0 1 90
acenaphthylene 100 60 0 0 20 0 0 0 40
acenaphthene 100 30 0 0 5 0 0 0 75
phenanthrene 5 100 0 10 35 15 20 0 75
anthracene 100 65 0 10 50 20 30 0 65
pyrene 40 100 20 60 0 0 0 0 65
coronene 35 100 0 0 25 10 15 0 75
corannulene 40 100 10 100 0 0 0 0 50

' The relative uncertainty in the relative abundance is about fifteen percent.
b The fragmentation efficiency, Ef, is the fraction of the product ion intensity following CID expressed as a
percentage. It is the ratio of the sum of the daughter ion intensities to the total ion intensity of the daughter
ions and the undissociated parent ion detected after CID.





loss of hydrogen and carbon atoms, and photodestroyed. Their results are condensed in

Table 5. No real correlation was found to explain why certain PAH cations fragment one

way while a very similar cation will fragment in a different fashion (e.g. naphthalene vs.

anthracene).

Current Efforts

Our efforts to study the photodissociation of PAHs of astronomical importance

began where the work of Ekem et al.198 left off. Our initial plans were to continue with

their work by examining wavelength-dependent photodissociation of the same ions. In

the previous work, each of the twenty-four ions was generated, trapped, and irradiated

with the full spectrum from a xenon arc lamp for 500 ms. After the 500 ms irradiation










Table 5. A list of the twenty-four PAHs examined by Ekemrn et al. placed within the
appropriate fragmentation category.


Photostable Loss of H atoms) Loss of H and C atom(s) Photodestroyed
acenaphthylene fluoranthene phenanthrene naphthalene
azulene pyrene chrysene decacyclene
biphenylene perylene anthracene
fullerene, C6o benzo[ghilperylene diphenylacetylene
acenaphthene benz[a]anthracene
coronene benzo[k]fluoranthene
triphenylene benzo[a]pyrene
fluorene tetracene
nlaphtho[2.3-a]pvrene dibenzanthracene


period the resulting mass spectrum was recorded. We believed a closer study of the

dissociation was warranted for these systems. We generated ions with internal El and

trapped them in an open-ended analyzer cell. The trapped ions were mass analyzed by a

FT-ICR mass spectrometer incorporating a 2 T magnet. Several modifications were

made, including the design and construction of a much larger cell to allow more light to

enter. A computer-controlled mechanical shutter was also built in-house. The source of

irradiation was an argon ion laser used to pump a dye laser. After several months of

designing, constructing, and troubleshooting, the studies were ready to begin. At first, we

did not observe any photodissociation for the PAHs of interest. We then decided to work

with a different molecule that would easily dissociate, p-bromochlorobenzene. After

some success with this system and after becoming familiar with the experimental setup,

we switched back to the PAHs. Once again, we had no success. It was decided that we

simply were not accessing the appropriate wavelengths for photodissociation with our








80

setup. Thus, we decided to take a step back, and use the same xenon arc lamp that Ekern

et al. 9 used. Our thought was to repeat the experiments while using in-line filters so that

we could narrow down the appropriate wavelengths in order to purchase the correct laser

dyes. While performing the filter studies, I decided to change the experiment slightly

after reading the paper published by Boissel and co-workers. Instead of opening the

shutter each time for 500 ms, I started to extend this time. The results of this variation

were (as Boissel had witnessed) that at longer irradiation times (up to 60 s) the mass

spectra drastically changed from what Ekem et al. 198 had reported. Since that time, the

photodissociation of the PAHs has been studied by varying the length of irradiation from

the xenon arc lamp. The next two chapters will report the findings for the fluorene cation

as well as the naphthalene, acenaphthylene, and the diphenylacetylene cations.














CHAPTER 4
PHOTODISSOCIATION AND ION-MOLECULE
REACTIONS OF FLUORENE CATIONS

Introduction

Photodissociation and ion-molecule reactions of fluorene cations using Fourier

transform ion cyclotron resonance mass spectrometry are discussed in this chapter. First,

a brief background of relevant theoretical work will be presented. Next, a review of the

experimental setup used to study fluorene will be offered, including photodissociation

studies using a xenon arc lamp and ion-molecule reactions without the lamp. A mention

of plausible structures for some of the products formed from both lamp-on and lamp-off

experiments will be provided. The chapter will end with concluding remarks and a

mention of future experiments that need to be carried out.

Background

Fluorene (C13Ho10, see Figure 33) falls into the category of nonalternant polyarenes.

A nonaltemant polyarene is defined as a PAH molecule that comprises one or more rings

that are other than fused six-membered benzenoid rings.99

Previous work performed198 on the fluorene cation revealed that upon irradiation,

up to five hydrogen atoms were lost from the parent ion. Density functional theory

calculations have been carried out at the B3LYP/4-31G level of theory to determine the

most likely positions for the hydrogen atom losses.2" The results of the calculations











7 >2


6 3
5 4



Figure 33. Chemical structure and numbering system of the fluorene molecule (hydrogen
atoms have been omitted from the structure).


reveal that the first hydrogen loss is from the sp3 carbon atom at the number nine position.

The remaining four hydrogen atoms are lost in a sequential order around one of the

aromatic rings. These calculations were employed to determine the energies of the

appropriate cations which might be formed in the photodissociation experiment with no

consideration given to the potential energy barriers that exist on the pathway from one

cation to the next. Figure 34 depicts a flow diagram of the energy pathways leading from

the parent ion at a m/z of 166 to the daughter ion at m/z 161.

Experimental

The underlying principles of FT-ICR MS have been presented earlier. Mass

spectra were acquired on a home-built FT-ICR mass spectrometer incorporating an

IonSpec data station (IonSpec, Corp., Irvine, CA) along with a 2 Tesla superconducting

magnet (depicted in Figure 35). The analyzer cell used throughout these studies was a

cylindrical cell built in-house (see Figures 36-38). Fluorene samples were placed on the

end of a solids probe and introduced directly into the vacuum chamber through a small

gate valve located on the vacuum chamber. The vacuum port which supported the probe





















H aT / '
H H


H **H
H HH H


3.84 eVI -2H





H H

H H


2.94 eV
S-H


H

H I I H

H H H
H H H


3.62 eV H *
-2H H H


H H
H H


4.62 eVI -H






to^'
H H
H H

4.94 eVi -H



H H


H DH
H H


5.07eV H --
-2H H H


432 eV H
4.32 eVH H


-^H


4. 81 eV

-2H









4.83 eV
-2H1





-H
464eV


4.97 eVI -H

+





H H
H


H
4.36 eV ]H

+- +





H H


Figure 34. Results of DFT calculations2 on the fluorene cation outlining the possible fragmentation pathways. The energies were
calculated at the B3LYP/4-31G level of theory.




Full Text
xml version 1.0 standalone yes
Volume_Errors
Errors
PageID P173
ErrorID 4
P176
4
P179
4
P191
4
P215
1007
ErrorText orientation error: 0071-0086
P299
4
P305
1008
orientation error: 0101-0104
P323
1009
orientation error: 0107-0112
P344
4
P350
1010
orientation error: 0116-0119
P365
4
P368
4
P371
4
P386
4
P389
4
P392
4
P398
4
P401
4
P407
4
P410
4
P413
4
P422
4
P425
4
P434
4
P437
4
P440
4