PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU In An Affirmative Action/Equal Opportunity Institution
cMMomHJ

 

ANALYTICAL APPLICATIONS OF ION/MOLECULE REACTIONS USING
A TRIPLE QUADRUPOLE MASS SPECTROMETER

By

Timothy Gordon Heath

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
, for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1990

IE

3)

./ 4(2) -- :5

ABSTRACT

ANALYTICAL APPLICATIONS OF ION/MOLECULE REACTIONS USING
A TRIPLE QUADRUPOLE MASS SPECTROMETER

By
Timothy Gordon Heath

Tandem mass spectrometry (MS/MS) has been a valuable tool in the
study of gas-phase ion chemistry. Triple quadrupole mass spectrometers
have proven to be versatile instruments for MS/MS studies, in both
fundamental and analytical applications. The second quadrupole collision
chamber (Q2) provides a region where low-energy ion/molecule interactions
may probe structural features of gaseous ions. Evaluation of the ionic
products detected following the second stage of mass analysis provides
insight into the gas-phase chemistry occurring within Q2, and may lead to
innovative analytical applications based on selective chemistry.

Utilization of a triple quadrupole mass spectrometer for exploring
ion/molecule reactions has resulted in novel analytical applications. Studies
of ion/molecule reactions involving aryl cations led to applications for
detection of aromatic ions which contain a vacant charged site on the ring.
The reactivity of C7H7+ isomers with selected neutral reagents has been
investigated and, as predicted by the thermochemistry, only the tolyl cation
undergoes ring addition reactions with nucleophilic reagents while benzyl
and tropylium cations do not.

Ion/molecule reactions are not limited to positive ions. A reaction
involving molecular anions of abscisic acid methyl ester (ABA-Me) and
molecular oxygen was elucidated. Oxygen-activated fragmentation results in

the formation of structurally diagnostic ions. This ion/molecule reaction

approach was used to analyze ABA-Me in 18O--labeling studies, and the
results provided insight into the biosynthetic pathway of abscisic acid.

Not all the studies involve reactions carried out in the collision cell.
Ion/molecule reactions may occur in the ion source, with ionic products then
being subjected to collision-induced dissociation (CID). This approach was
used extensively in the study of reactions involving protonated carbonyl
compounds and alcohols. Comparison of the CID daughter ion mass spectra
derived from known ion structures with those of unknown product ions
suggests alkylation occurs on the carbonyl oxygen. In addition, the inherent
selectivity of conventional MS/MS using CID is demonstrated in the analysis
of dexamethasone. Finally, ion/molecule reactions of protonated molecules
with hexamethyldisilazane were studied. Whereas, the gas-phase silylation
reaction was demonstrated, its selectivity for protonated molecules remains

unknown.

To my darling Melissa.

ACKNOWLEDGMENTS

I would like to thank my research advisor, Dr. J. Throck Watson for his
support and encouragement throughout my graduate school career. I would
also like to thank the rest of my guidance committee, especially Dr. John
Allison, who continually emphasized the importance of evaluating the
experimental mass spectral results in terms of the thermochemistry of the
reaction.

There are many people to whom I am grateful for having the
opportunity to work with, and whose assistance contributed to my successful
completion of the degree requirements. Thanks to Dr. J .A.D. Zeevaart, Doug
Gage, Chris Rock, Greg Dolnikowski, and Kathleen Kayganich for their
efforts in enjoyable collaborative projects. I would also like to thank Curt
Heine, Paul O'Connor, Ron Lopshire, Mark Cole, Mike Davenport, Mel Micke
along with the rest of the crew at the Michigan State Mass Spectrometry
Facility, for their assistance in my research efforts.

I am also grateful for the love, encouragement, and support I received
from my family, especially from my parents, while I was pursuing my degree.
Also, I would like to thank fellow 'Hags', Kurt Kooyer and Kevin Heath, for
their tremendous friendship. I have many fond memories of; spring training
drills in March, softball with "The Goats" (and SBC), "Prairie Home..."
barbecues in the summer evenings, football in September, and the annual
UM-MSU battle in October. And to Kevin and Greg Heath, I will never forget
the two-day, thousand-mile, four-state expedition to see (in 24 hours) ball
games at Municipal, Three-Rivers and Riverfront Stadiums.

I am very grateful to my wife, Melissa. I feel blessed to have married
someone as wonderful as you. Thanks for your continued love, support, and
encouragement throughout these past few years while I have been a student.
At last, I think I am done.

And finally, I would like to thank my heavenly father for his grace and
kindness shown to me. I echo the psalmist's words, "Praise the Lord, 0 my
soul, and forget not all his benefits. " (Psalm 103:2).

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................

LIST OF FIGURES ................................................................................

CHAPTER I: INTRODUCTION AND OBJECTIVES ............................ 1

A. Introduction ......................................................................... 1

B. Triple quadrupole mass spectrometry ...................................... 2

1. Instrumentation ............................................................. 2

a) Operating principles of quadrupoles ............................ 3

b) Ionization modes ....................................................... 7

c) Detection .................................................................. 8

2. MS/MS scan modes with TQMS ........................................ 9

a) Daughter ion scan ..................................................... 9

b) Parent ion scan ......................................................... 11

c) Functional Relationship scan ..................................... 11

d) Selected reaction monitoring (SRM) ............................. 12

3. Methods of ion activation .................................................. 12

a) Collisional activation (CA) ......................................... 14

b) Surface induced dissociation (SID) .............................. 16

c) Photodissociation ....................................................... 17

d) Ion/molecule reactions .............................................. 18

4. Analytical utility of TQMS ................................................ Z)

5. Limitations to dissociation (CID, SID) methods .................. 21

vi

C. Ion/molecule reactions in a TQMS .......................................... 23

1. Advantages of using TQMS for studying
ion/ molecule reactions .................................................... 23

2. Disadvantages of using a TQMS for studying

ion / molecule reactions .................................................... 24
D. Literature review of ion/molecule Reactions in TQMS ............... %
1. Fundamental studies ...................................................... %
a) Ion/molecule reactions of C2H5O+
isomers with NH3 ..................................................... 25
b) Endothermic proton transfer reactions ........................ 28
c) Reaction of benzoyl ions with ammonia ........................ 28

(1) Formation of ammonium ions in reaction of

protonated carbonyls with NH3 ................................... a)
2. Analytical applications .................................................... E
a) Reaction of CgH3+ ions with acetylene .......................... 29

b) Reactions of CzH5O+ isomers with
reagent molecules ..................................................... 3)
c) Reactions of cations with hydrocarbons ........................ 3)
(1) Reaction of protonated esters with ammonia ................ 31

e) Protonated natural products reacting

with ethyl vinyl ether ................................................. 31
0 Reaction of protonated hexachlorobiphenyl

molecules with ammonia ........................................... 32
g) Discrimination of 1,2-cyclopentanediol isomers ............ 32

h) Reactions in a double quadrupole mass spectrometer.... 33
i) Quantitative analysis using reactions in ammonia ........ 33

5) Reactions of protonated trichothecenes with ammonia... 34

k) Reaction of CH3CO+ with l-methylcyclopentene ............ 34

vii

1) Reactions of tetrachlorodibenzo-p- dioxins with 02 ......... 34

m) Tetraethylsilane molecular ion reacting with air ......... 35
11) Charge exchange and proton transfer reactions ........... 35
0) Reactions of selected ions with GC effluent ................... {B

E. Low-energy ion/molecule reactions in a four-sector instrument.. 37

F. Research objectives ................................................................ 40

CHAPTER II: ION/MOLECULE REACTIONS OF ARYL CATIONS ..... 42

A. Introduction ......................................................................... 42
B. Survey of ion/molecule reactions involving

the phenyl cation ................................................................... 44

1. Experimental ................................................................. 44

2. Reaction with alcohols ..................................................... 47

a) Methanol .................................................................. 47

b) Ethanol .................................................................... 54

3. Reaction with amines ...................................................... 57

a) Ammonia ................................................................. 57

b) Methyl amine ............................................................ 62

4. Reaction with CH3CN ...................................................... 65

5. Reaction with ethers ........................................................ 65

a) Dimethyl ether .......................................................... 65

b) Diethyl ether ............................................................. 67

c) Vinyl methyl ether ..................................................... 70

6. Summary of reactions ..................................................... 72

C. Screening for aromatics ......................................................... 73

viii

D. Ion/molecule reaction of [C7H7t] isomers for selective

detection of the tolyl cation ...................................................... 81
1. Introduction ................................................................... 81
2. Experimental ................................................................. 84
a) Instrumentation ....................................................... 84
b) Generation of 'pure' isomers ...................................... 85
c) GC/MS/MS analyses .................................................. 86
3. Results and discussion .................................................... 86
a) Thermochemistry ..................................................... 86
b) Experimental results ................................................. 88

0) Analytical utility of the ion/molecule reaction approach. 92

d) Effect of higher gas pressure ...................................... 100
e) Reaction with ammonia ............................................. 103
f) Conclusion ................................................................ 105
E. Conclusion ........................................................................... 106

CHAPTER III: ION/MOLECULE REACTIONS OF ABSCISIC ACID
METHYL ESTER WITH MOLECULAR OXYGEN IN

ELECTRON CAPTURE NEGATIVE IONIZATION ...... 107

A. Introduction ......................................................................... 107
1. Abscisic acid .................................................................. 107

2. Mass spectrometry of ABA ............................................... 109

3. Reactions with oxygen in ECNI-MS ................................... 113

B. Experimental ........................................................................ 114
1. Instrumentation ............................................................. 114

2. Materials ....................................................................... 115

C. Results and discussion .......................................................... 116
1. MS /MS with argon as collision gas .................................... 116

2. MS / MS with 02 as collision gas. ........................................ 118

3. MS / MS analyses of isotopically-labeled compounds ............. 123

4. Analysis of structurally-similar compounds............ .......... 133
a) ABA metabolites ....................................................... 133
b) Analysis of a and B ionone .......................................... 135
c) Analysis of 2,4-hexadiene ........................................... 141
5. Wall-catalyzed reactions .................................................. 141
D. Application .......................................................................... 144
E. Conclusion ........................................................................... 149
CHAPTER IV: COLLABORATIVE RESEARCH PROJECTS ................. 150
A. Introduction ......................................................................... 150
B. ECNI-MS/MS analysis of oxidized dexamethasone .................... 152
1. Introduction ................................................................... 152
2. Experimental ................................................................. 154 '
a) Instrumentation ....................................................... 154
b) Collision energy and pressure studies .......................... 157
c) Selected reaction monitoring studies ............................ 158
d) Methods ................................................................... 158
3. Results and discussion .................................................... 159
a) Optimization of MS/MS parameters ............................. 159
b) SRM vs. SIM ............................................................. 163
C. Study of the gas-phase reaction between protonated
acetaldehyde and methanol .................................................... 168
1. Introduction ................................................................... 168
2. Experimental ................................................................. 171
3. Results and discussion .................................................... 171
a) Formation of the product ion with m/z 59 ..................... 171
b) Reactions of acetaldehyde and dimethyl ether ............... 185

c) Reaction of benzaldehyde and dimethyl ether ................ 187

(1) Ion trapping ............................................................. 189

4. Summary ....................................................................... 19!

D. Reactions of protonated alcohols and ethyl acetate ..................... 198
1. Experimental ................................................................. 198

2. Results and discussion .................................................... 199

E. Ion/molecule reactions of protonated molecules

hexamethyldisilazane ........................................................... m8

1. Experimental ................................................................. 210

2. Results and discussion .................................................... 210
CHAPTER V: SUMMARY AND FUTURE WORK ............................... 215
LIST OF REFERENCES .................................................................... 220

xi

Table 1 -1
Table 1 -2
Table 1 -3

Table 1-4

Table 2-1

Table 2-2

Table 2-3

Table 2-4

Table 2-5

Table 2-6
Table 2-7

Table 2-8

Table 4-1

Table 4-2

LIST OF TABLES

Function of the quadrupoles in various TQMS scan modes....10
Methods of ion activation in a TQMS ...... 13
Low-energy ion/molecule reaction collision processes ........... 19

Reaction enthalpies for dissociation of C2H50+ isomers in
collisions with N2 and NH3 ................................................. 27

Calculated enthalpies for reactions of the phenyl cation
with alcohols .................................................................... 56

Reaction enthalpies for the phenyl cation reacting
with ammonia .................................................................. 59

Reaction enthalpies for the phenyl cation reacting

with diethyl ether .............................................................. 70
Structural candidates of the product ions formed in the
reaction of the phenyl cation and vinyl methyl ether .............. 70
Components in mixture used to obtain chromatograms
shown in Figures 2-17 and 2-18 ........................................... 77
Bond strength in kJ/mol ..................................................... 87

Components in test mixture separated by GC to obtain
chromatogram shown in Figure 2-24 ................................... 94

Reaction enthalpies forming C3H9+ and methanol from
C7H7+ and dimethyl ether ................................................. 101

Procedure used to implement inject, trap and pulse on the
TSQ-7O triple quadrupole mass spectrometer ....................... 193

Protonated molecules which were reacted with HMDS in the
collision cell of the TQMS .................................................. 213

xii

Figure 1-1

Figure 1-2

Figure 1-3

Figure 2-1

Figure 2-2

Figure 2-3

Figure 2-4

Figure 2-5

Figure 2-6

Figure 2-7

LIST OF FIGURES

Schematic diagram of a triple quadrupole mass
spectrometer (TQMS) .......................................................... 4

The a-q stability diagram indicating region which corresponds
to stable ion trajectories; a relates to the RF-only field and q
relates to the DC field .......................................................... 6

Relative abundances of various ions in reaction of protonated
leu—enkephalin with ammonia as function of ion kinetic
energy. (Adapted from reference 100 ) ................................ 39

Schematic diagram of the vacuum system of the Finnigan
TSQ-70 triple quadrupole mass spectrometer (Adapted
from reference 17) ............................................................. 46

Product ion mass spectra of the phenyl cation and methanol
obtained with the A) Extrel TQMS and B) Finnigan TSQ-70....48

Ion/molecule reaction mechanism to form the molecular
ion of phenol from the reaction of the phenyl
cation and methanol ......................................................... 49

CID daughter ion mass spectrum of reaction product ion of
m/z 109 formed when benzene and methanol were
simultaneously introduced into the ion source. Collision
energy was 30 eVLab at single collision conditions ................ 51

CID daughter ion mass spectra of the protonated forms of A)
anisole, B) o-cresol and C) p-cresol. Collision energy was 30
eVLab at single collision conditions ..................................... 52

Product ion mass spectrum of the phenyl cation reacting
with ethanol at a pressure of 1.5 mtorr and with a collision
energy of 1 eVLab .............................................................. 55

Product ion mass spectrum of the phenyl cation and

ammonia. Collision pressure ~ 2 mtorr and collision
energy is 2 eVLab .............................................................. 5 8

xiii

Figure 2-8

Figure 2-9

Figure 2-1 0

Figure 2-11

Figure 2-1 2

Figure 2-13

Figure 2-14

Figure 2-15

Figure 2-16

Figure 2-1 7

Figure 2-1 8

Energy-resolved curve of the reaction products generated
from reaction of the phenyl cation and ammonia.
Collision pressure is 1 mtorr ................................ . ............. 61

Product ion mass spectrum of the phenyl cation and methyl
amine at collision energy of A) 1 eVLab and B) 20 eVLab.
Methyl amine pressure is 1 .5 mtorr .................................... 63

Product ion mass spectrum of (M-H)+ of napthalene with
methyl amine at a pressure of 1.5 mtorr and collision
energy of 1 eVLab .............................................................. 64

Product ion mass spectrum of (M-H)+ of napthalene with
acetonitrile at a pressure of 2 mtorr and a collision energy of
1 eVLab ............................................................................ 66

Product ion mass spectrum of the phenyl cation and
dimethyl ether at a pressure of 1 mtorr and collision
energy of 1 eVLab .............................................................. 68

Product ion mass spectrum of the phenyl cation and
diethyl ether at collision energy of 2 eVLab and pressure
of 1 .2 mtorr ...................................................................... 69

Product ion mass spectrum of the phenyl cation and vinyl
methyl ether at collision energy of 2 eVLab and a pressure
of 1.2 mtorr ...................................................................... 71

Product ion mass spectra of the (M-Cl)+ ion of 1 ,2,4,5 tetra-
chlorobenzene with DME at 1 mtorr. Parent ion is A)

C5H235Cla+ (m/z 179) B) C(5H237C135C12+ (m/z 181) .................. 74

Product ion mass spectrum of the (M-H)+ ion of pyridine (m/z
78) with dimethyl ether at a pressure of 2 mtorr and collision
energy of 1 eVLab .............................................................. 76

A) Reconstructed mass chromatogram of m/z 77 and B) total
ion current chromatogram, following EI of mixture
introduced through GC ..................................................... 78

A) Total ion current chromatogram, following E1 of mixture
introduced through GC, and B) total ion current
chromatogram from neutral gain experiment (+31) with DME
in Q2 at a pressure of 1.1 mtorr and a collision

energy of 1 eVLab .............................................................. 80

xiv

Figure 2-1 9

Figure 2-20

Figure 2-21

Figure 2-22

Figure 2-23

Figure 2-24

Figure 2-25

Figure 2-26

Figure 2-27

Figure 2-28

Figure 3—1
Figure 3-2

Structures and heats of formation for C7H7+ isomers. Heats of
formation taken from references 108 and 125 ....................... 82

CA mass spectrum of A) pure tropylium and B) pure benzyl
ions. Accelerating voltage = 8 kV. (Adapted from reference
1 1 7) ................................................................................. 83

Product ion mass spectrum of the tolyl cation with DME at a
pressure of 1 mtorr and collision energy of 1 eVLab ............... 90

Proposed reaction mechanism for the formation of the product
ion of m/z 122 from the reaction of the tolyl cation and
dimethyl ether .................................................................. 91

CID daughter ion mass spectrum of A) ion/molecule reaction
product of mass 122 and B) molecular ion of authentic 3-
methyl anisole. Collision energy was 25 eVLab at single
collision conditions ........................................................... 93

Reconstructed mass chromatogram of A) m/z 91 and B) m/z
122 when mixture of components identified in Table 2-7 was
injected into the GC and the ion of mass 91 chosen to react with
dimethyl ether in the second quadrupole at a pressure of 1 _
mtorr and collision energy of 2 eVLab .................................. 96

CID daughter ion mass spectrum of mass 91 generated by 70
eV E1 of A) 2-bromotoluene and B) 3-nitrotoluene. Collision
energy was 25 eVLab and argon was the collision gas at 0.3
mtorr .............................................................................. 97

Ratio of 1122 / 1022411) as a function of electron energy during
electron impact of 3-nitrotoluene and transmission of ion
current at m/z 91 into Q2 for reaction with dimethyl ether at

indicated pressures .......................................................... 99

Product ion mass spectrum of A) tolyl cation, B) benzyl cation,
and C) tropylium cation in reaction with DME at a pressure
of ~ 3 mtorr and collision energy of 1.5 eVLab ...................... 102

Product ion mass spectrum of the 3-toly1 cation and
ammonia at a pressure of 1.5 mtorr and a collision
energy of 2 eVLab ............................................................. 104

Structure of abscisic acid methyl ester (ABA-Me) ................ 108

ECNI mass spectrum of ABA-Me obtained under conditions in
which: (A) no effort was made to purge air from the CI lines
and (B) the CI lines were carefully purged free from
contamination with air .................................................... 111

Figure 3-3

Figure 3-4

Figure 3-5

Figure 3-6

Figure 3-7

Figure 3-8

Figure 3-9

Figure 3-10

Figure 3-11

Figure 3-12

Figure 3-13

Figure 3-14

Figure 3-15

Structures of fragment ions of ABA-Me as proposed by
Netting et al. in reference 136 ........................................... 112

Product ion mass spectra (ELab=2 eV) of the molecular anion
(m/z 278) of ABA-Me obtained with a TQMS following (A) CID
with Ar and (B) CID and/or ion/molecule reactions with
molecular oxygen ............................................................ 117

Energy-resolved curves in collisions of the molecular anion of
ABA-Me with molecular oxygen at a pressure of 1 mtorr. (A)
Ions with m/z 260 and 245, and (B) ions with m/z 141

and 152 .......................................................................... 119

Product ion mass spectrum of the molecular anion of ABA-Me
(m/z 278) following ion/molecule reactions and/or CID with
1302 in Q2 at a collision energy of 2 eVLab ........................... 121

The daughter ion mass spectrum following CID with argon at
10 eVLab collision energy of the adduct ion of m/z 310 formed
during ion/molecule reactions of the molecular anion of ABA-
Me and 02 in the ion source .............................................. 122

Product ion mass spectrum of the molecular anion of
ABA-ethyl ester (m/z 292) and molecular oxygen at a pressure
of 1 mtorr and a collision energy of 1 eVLab ........................ 125

Ion/molecule reaction mechanism accounting forthe
formation of the ion of m/z 141 following the reaction of the
molecular anion of ABA-Me and 02 .................................. 126

Ion/molecule reaction mechanism accounting for the
product ion of m/z 179 ..................................................... 128

Mechanism for the formation of the ions of m/z 152 and 136
following the ion/molecule reaction of M'- of ABA-Me

and 02 ........................................................................... 130
Fragmentation of M'- of ABA-Me (m/z 278) to yield ions with
m/z 260 and 245 ............................................................... 132
Structure of ABA metabolites ........................................... 134
Formation of ion with m/z 168 from the ion/molecule reaction
of the molecular anion of phaseic acid-Me and 02 ............... 136
Structures of the isomeric forms of ionone (MW=192) .......... 138

xvi

Figure 3-16

Figure 3-1 7

Figure 3-18

Figure 3-19

Figure 3-20

Figure 3-21

figure 3-22

Figure 4-1

Figure 4-2
Figure 4-3

Figure 4-4

Figure 4-5

ECN I mass spectrum of (A) a-ionone and (B) B-ionone with
ammonia as the buffering reagent .................................... 139

ECNI mass spectrum of (A) a-ionone and (B) B-ionone with
methane as the buffering reagent ..................................... 140

Product ion mass spectrum of (M-H)‘ of trans,trans-2,4-
hexadiene (MW=82) with molecular oxygen at a pressure
of 1.5 mtorr and a collision energy of 1 eVLab ...................... 142

Ion/molecule reactions of (M-H)' of trans,trans-2,4-hexadiene
in reaction with molecular oxygen to form enolate anions.
(Adapted from reference 151 ) ...................................... .....143

Mass chromatograms at m/z 141 (A), m/z 165 (B), m/z 260 (C),
and m/z 278 (D) reconstructed from mass spectra obtained
during ECN I with sample introduction via the GC inlet ...... 145

ECNI mass spectrum of a mixture of isotopically labeled
ABA-Me ......................................................................... 1 47
Product ion mass spectra of mono-labeled ABA-Me (m/z 280)
with molecular oxygen at a collision energy of 5 eVLab ........ 148
Conversion of dexamethasone to its 11 ,17-keto analogue by
chemical oxidation .......................................................... 153
ECNI mass spectrum of 11,17-keto-dexamethasone ............. 155
CID daughter ion mass spectrum of the molecular anion

(m/z 330) of 11,17-keto-dexamethasone ............................... 156
(A) Reconstructed TIC from CID of the M'- of oxidized
dexamethasone as a function of collision energy at different
collision gas (argon) pressures in Q2. (B) Magnitude of
daughter ion (m/z 310) current as a function of collision

energy at three pressures of argon in Q2 ............................ 160
Comparison of selected ion current profiles obtained by

GC/ECNI with (A) selected ion monitoring (top two panels for
oxidized dexamethasone; bottom two panels for internal
standard) and (B) selected reaction monitoring (top panel for
oxidized dexamethasone; bottom panel for internal

standard) ....................................................................... 1 64

Figure 4-6

Figure 4-7

Figure 4-8

Figure 4-9

Figure 4-10

Figure 4-11
Figure 4-12

Figure 4-13

Calibration curves for the determination of dexamethasone in
plasma by (A) GC/ECNI/MS with selected ion monitoring, (B)
GC/ECNI/MS/MS with selected reaction monitoring, and (C)
DIP/ECNI/MS/MS with selected reaction monitoring. The
quantities introduced into the ion source represent a range of
4-200 pg of dexamethasone and 700 pg of internal

standard ........................................................................ 167

Product ion mass spectrum of ions produced during
ion/molecule reactions of protonated acetaldehyde with
methanol in Q2 at a pressure of 2 mtorr. Those peaks marked
by an asterisk represent known product ions in the
ion/molecule reaction between protonated methanol
(generated by proton transfer in this case) and methanol....169

C-alkylation or O-alkylation for the formation of the
product ion with m/z 59 ................................................... 1 70

Reaction product mass spectrum of ions produced during
ion/molecule reactions following E1 of methanol and
acetaldehyde in a high pressure ion source of the TSQ-70.
Those peaks marked by an asterisk represent known product
ions of methanol self-CI. Those peaks marked with a #
represent known product ions of acetaldehyde self-CI ......... 173

Daughter ion mass spectra obtained during CID in Q2 of a
TQMS of a parent ion of mass 59 corresponding to: (A) product
ion from high pressure ion source containing methanol and
acetaldehyde; (B) protonated methyl vinyl ether (self-CI at
lower pressure); (C) protonated methyl vinyl ether (self-CI at
higher pressure); (D) protonated acetone; (E) protonated allyl
alcohol; (F) fragment ion from diethyl ether following EI; and
(G) fragment ion from acetaldehyde dimethylacetal

following EI .................................................................... 174
Candidate structures for product ion with m/z 59 ................ 176
Structure of m/z 59 formed by homolytic cleavage

of the molecular ion of acetaldehyde dimethylacetal ............ 177
Proposed mechanism for the formation of the ion with m/z 59

from the collision-complex intermediate of m/z 77 following
the ion/molecule reaction of protonated acetaldehyde and
methanol ....................................................................... 1 79

xviii

Figure 4-14

Figure 4-15

Figure 4-1 6

Figure 4-17

Figure 4-18

Figure 4-19

Figure 4-20

Figure 4-21

Figure 4-22

Figure 4-23

(A) CID (ELab= 4 eV) daughter ion mass spectrum of the
adduct ion of m/z 77 formed by ion/molecule reactions
occurring after methanol and acetaldehyde were placed in the
high pressure ion source. (B) Daughter ion mass spectrum
of m/z 79 formed when CH313OH and acetaldehyde were
placed in the ion source ................................................... 181

Proposed mechanism for the formation of the ion of m/z 59
from the intermediate of m/z 91 following the ion/molecule
reactions occurring in a mixture of methanol and

acetaldehyde ................................................................... 1 83
CID (ELab= 2 eV) daughter ion mass spectrum of m/z 93
formed by ion/molecule reaction occurring after acetaldehyde
and 18O-methanol were introduced into the ion source ........ 184
Partial reaction product mass spectrum of ions produced
during ion/molecule reactions following E1 of acetaldehyde
and dimethyl ether in a high pressure ion source ............... 186
Partial reaction product mass spectrum of ions produced

during ion/molecule reactions following E1 of benzaldehyde
and dimethyl ether in a high pressure ion source. Peaks
produced from self-CI of dimethyl ether were subtracted ..... 188

Proposed mechanism for the formation of the ion of m/z 121
following ion/molecule reactions in mixture of benzaldehyde
and dimethyl ether .......................................................... 190

Implementation of inject, trap and pulse methodology on
the TSQ-70 ..................................................................... 192

Product ion mass spectrum obtained when the inject, trap and
pulse method was implemented on the TSQ-70. The collision
cell, containing methanol vapors,was filled for 50 msec with
protonated acetaldehyde. Immediately after L23 was gated,
the ions were extracted by L31 ........................................... 195

Product ion mass spectrum of protonated acetaldehyde with
methanol using the inject, trap and pulse method
implemented on the TSQ-70. The collision cell, containing
methanol vapors,was filled for 50 msec with protonated
acetaldehyde. Ions were extracted after a containment

period of 5 msec .............................................................. 196

Acetal ion formed in reaction of protonated propanol
and ethyl acetate ............................................................. 200

xix

Figure 4-24

Figure 4-25

Figure 4-26

Figure 4-27

Figure 4-28

Figure 4-29

Figure 4-30

Ion/molecule reaction of protonated ethyl acetate with ethyl
acetate to form the product ion of m/z 131 ........................... 201

CID daughter ion mass spectrum of the ion of m/z 134 formed
when propanol and CZH3COOC2H5 are introduced into the
ion source ...................................................................... 203

CID daughter ion mass spectrum of the product ion of m/z
145 formed when ethyl acetate and butanol are introduced
into the ion source ........................................................... 204

CID daughter ion mass spectrum of the ion of m/z 122 formed
when CH3COOszH5 and ethanol are introduced into the ion

source ............................................................................ 206

General scheme showing the elimination of an alcohol or
olefin molecule from the acetal ion formed in ion/molecule
reactions of alcohols and ethyl acetate .............................. 207.

Proposed reaction mechanism of protonated amines with
HMDS ........................................................................... 209

Product ion mass spectrum following ion/molecule reaction of .
protonated acetone (m/z 59) with HMDS at a pressure of 3
mtorr and a collision energy of 1 eVLab .............................. 212

XX

CHAPTER I
INTRODUCTION AND OBJECTIVES

A. Introduction

Mass spectrometry has provided the capability for studying
ion/molecule chemistry in a solvent-free environment. Tandem mass
spectrometry (MS/MS) has extended this capability, enabling further
exploration of gas-phase ion/molecule phenomena. The primary focus of the
research described in this dissertation involves the investigation of gas-phase
ion/molecule reactions using a triple quadrupole mass spectrometer (TQMS).
The principal intent has been to develop novel analytical methodology which
utilizes gas-phase ion/molecule reactions of a selected mass-analyzed reactant
ion with a reagent gas which is introduced into the collision cell of a TQMS
for a tandem MS analysis.

There are numerous avenues by which this goal of developing the
analytical utility of ion/molecule reactions may be realized. Some of the
experimental results could have obvious analytical implications, as in the
study of an ion/molecule reaction which leads to development of an MS/MS
assay for a biologically important compound. However, the analytical
implications of other experimental results may be more subtle. Perhaps
studying ion/molecule reactions in a TQMS will reveal alternative methods
which would assist in the elucidation of ion/molecule reaction mechanisms, or
in the probing of ionic or neutral structures. Various applications are

discussed in this dissertation and all serve to demonstrate the continued

2
flexibility and diversity of a triple quadrupole mass spectrometer for studying
gas-phase ion/molecule chemistry.

The objective of this first chapter is to: (i) introduce the
instrumentation used in this research, that being the triple quadrupole mass
spectrometer, (ii) discuss the different modes of ion activation in an analysis
by MS/MS using TQMS and, (iii) lay out the rationale for pursuing the
investigation of ion/molecule reactions with a TQMS. Towards the end of this
chapter, the effort extended by other researchers in studying ion/molecule
reactions with triple quadrupole mass spectrometry, will be reviewed. This
section provides an overview of the published successes, which serves to
demonstrate the analytical value of a TQMS in studying ion/molecule
reactions and the types of experiments accessed by a TQMS. Finally this
chapter will conclude with a guideline to the research objectives.

B. Triple quadrupole mass spectrometry
1. Instrumentation

The ability to mass-select an ion of a specific mass-to-charge ratio (m/z)
in the first stage of analysis, for introduction into a spatially-separated
reaction chamber, which is followed by a second stage of mass analysis,
provides a powerful tool for studying gas-phase ion chemistry. This technique
is referred to as tandem mass spectrometry, mass spectrometry/mass
spectrometry, or just MS/MS (1,2). With the development of the triple
quadrupole mass spectrometer by Yost and Enke (3-5), a new and relatively
straightforward approach emerged for studying low energy (1-200 eVLab)

ion/molecule interactions.

3

A schematic of a triple quadrupole mass spectrometer is shown in
Figure 1-1. Following ionization in the ion source, the ions are focussed and
passed into the first mass analyzer, Q1. After the first stage of mass analysis,
the ions continue their flight path into the second quadrupole, Q2, which does
not function as a mass analyzer, but as a reaction (or collision) chamber and
an ion transmission device. Introduction of a collision gas in this region (Q2)
provides the mass-selected reactant ion with a reagent molecule for
interaction. The reaction products generated in Q2 may then pass to the
second mass analyzer in a TQMS, which is the third quadrupole, Q3. With
detection of the ionic products, the researcher may infer the reaction
processes that are occurring in the second quadrupole reaction chamber.
Currently, triple quadrupole mass spectrometers provide the capability of
analyzing gaseous ions up to 4000 mass units, although resolution of the ions
suffers beyond 2000 mass units.

a) Operating principles of quadrupoles

The core of the triple quadrupole instrument is the three sets of
quadrupole rods. The triple quadrupole mass spectrometer consists of two
quadrupole mass analyzers, Q1 and Q3, and one set of rods, Q2, which acts as
a reaction chamber, and an ion transmission device. With the proper
combination of direct-current (DC) and radio-frequency (RF) potentials
applied to the rods, the four quadrupole rods act as a mass filter. Ideal
performance is obtained with rods having hyperbolic geometry (6-8), however,

cylindrical rods are most often used and provide satisfactory results.

.6295 u3o8oeuooam 29: 325.655 32.5 a no Bahama. Omumaoaom 3A 933..“

momafiommm mom: goon ” A

.8anme mmma 988m ” m@
.8389? 56330 ... N G
sunbeam mama swarm u H 3

so 8330

 

I? 72 .m<.m .5 .5
”GOmn—QNEOH
A 1H , A
- - I. QOHH—hv
l. m
_.....E©|.. .6. __

 

 

 
 

J.

53033955
can—om

5
The function of the first and third quadrupole rod sets is to mass-
analyze the ions. As the ions travel through the central longitudinal space
between the rods, their motion in the X-Y plane is influenced by the combined
DC and RF fields and is described by the Mathieu equation with ax and qx

parameters defined as follows:

ax = -ay = 4zeU / ma) 2r02
qx = -qy = 22eV / ma) 2r02

where m/z represents the mass-to—charge ratio of the ionic species, r0 is the
distance from the center of the quadrupolar field to the closest surface of any
rod, to is the alternating RF frequency, U is the magnitude of the DC field,
and V is the amplitude of the RF field. Figure 1-2 represents the a-q stability
diagram, and the tip of this region, which is the bound area where the mass I
scan line intersects the stability diagram, represents coordinates Of those ions
which will have stable trajectories through the quadrupole field. The slope of
the mass scan line is defined by the ratio 2UN. Altering the slope of the
mass scan line changes the mass-resolution of the ions. Typically, in TQMS
experiments, the voltages selected by the researcher provide unit mass
resolution. There are certain applications, however, where it may be
desirable to either increase or decrease the mass resolution of the ions. In
any case, if the voltage applied to the rods is swept while maintaining a
constant slope of the mass scan line, the mass of ions with stable trajectories
within the field changes, thus one is able to obtain a mass spectrum with
uniform mass resolution over the mass range.
In contrast to the mass filtering capabilities of Q1 and Q3 is the RF-

only second quadrupole. There is no DC potential applied to the second

0.2 —

_ Mass
Scan Line

0.1 —

 

 

Figure 1-2: The a-q stability diagram indicating region which correspOnds
to stable ion trajectories; a relates to the RF-only field and q relates to the DC
field.

7
quadrupole, which implies theoretically that ions of all m/z ratios will have
stable trajectories within the field, thus RF-only quadrupoles have commonly
been referred to as total ion transmission devices. However, as suggested by
Miller and Denton (9), the transmission efficiency for ions of all m/z values is
not 100% in an RF-only quadrupole. Not only is the transmission of ions
through a quadrupole influenced by the stability of the ions as defined by the
Mathieu stability diagram, but by (i) the acceptance aperture of the
quadrupole and (ii) the spatial focusing conditions that arise due to the
trajectory of the ions through the quadrupole (9). Therefore, mass
discrimination effects are apparent, which implies that RF-only quadrupoles
are not strictly total ion transmission chambers, that is, ions of different m/z
ratios are transmitted through RF-only quadrupole rod assemblies with
different degrees of efficiency. It should also be noted that most of the
experimental work reported in this dissertation was performed on a Finnigan
TSQ-7O triple quadrupole mass spectrometer. This TQMS instrument is
unique in that the axis of the second quadrupole collision cell is non-linear,
which helps minimize the detection of chemical noise arising from fast
neutrals in fast atom bombardment (FAB) experiments. Recently, the
transmission characteristics of non-linear RF quadrupole collision cells were
investigated (10). The authors of this study concluded that the transmission
of the ions in a non-linear device was not measurably different than the

transmission of ions in a linear quadrupole collision cell (10).

b) Ionization modes
Several ionization modes are available with triple quadrupole mass
spectrometers. These include but are not limited to electron impact (EI),

chemical ionization (both positive and negative CI) (11), fast atom

8
bombardment (FAB) (12,13), and atmospheric pressure ionization (API)
techniques (14-16). In addition, triple quadrupole instruments are well
suited to interfacing with a gas chromatograph (GO) or liquid chromatograph
(LC). In many MS/MS experiments, it is desirable to use a 'soft' ionization
technique, such as CI or FAB, where most of the ion current is concentrated
at one particular m/z value, usually representing the protonated species,
which reflects the molecular weight of the compound of interest while
maintaining the structural integrity of the molecule. Structural information
supplied by fragmentation of the ion can then be obtained by subjecting the
ionic species to an analysis by MS/MS.

c) Detection

An electron multiplier is often used as the detector in triple quadrupole
mass spectrometry. Ions strike the electron multiplier, causing electrons to
be ejected. These electrons then strike the walls of the multiplier creating a
cascade of electrons which finally results in a measurable current at the base
of the multiplier. The typical gain of an electron multiplier is 105 (17). With
recent advances in the ionization efficiency of large biomolecules, post-
acceleration conversion dynodes are more frequently being used in the
detection scheme providing more uniform detection of ions within a large
mass range. The conversion dynode is located before the electron multiplier.
Ions strike this dynode which is biased with a large positive or negative
potential depending on the polarity of the detected ions. After the ions strike
this conversion dynode, charged particles are ejected and accelerated, with
energy in the kV range, towards the electron multiplier.

2. MS / MS scan modes with TQMS

There are four different MS/MS scan modes available with a triple
quadrupole mass spectrometer, the daughter ion scan, parent ion scan,
functional relationship scan and the selected reaction monitoring (SRM)
mode (18,19). Recently, details of the vaIious scan modes possible for a MS"
system have been described (18,19). The TQMS may also be operated as a
single stage quadrupole mass spectrometer by scanning only one of the two
mass analyzers while the other two quadrupole rod assemblies in the RF-only
mode, function as ion transmission devices. The role of the rods in the

different scan modes is illustrated in Table 1-1.

a) Daughter ion scan

The most familiar and commonly used MS/MS scan mode is the
daughter ion scan. In this mode, the first mass analyzer, Q1, is set to pass
ion current of a particular mass-to-charge (m/z) ratio which is referred to as
the parent ion or reactant ion. Parent ions chosen by Q1 interact with the
collision gas in Q2 generating product ions, referred to as daughter ions when
the process is collision-induced dissociation (CID). These fragment ions
formed in the collision cell enter the third quadrupole which is scanned to
obtain a daughter ion mass spectrum. Throughout this dissertation, the term
daughter ion mass spectrum will be used when the collision gas
introduced is inert, such as the case with argon. However, if a reagent gas is
introduced, the term product ion mass spectrum will be used. It should be
noted that use of a reagent gas does not preclude the formation of ions formed

by CID.

Q1
Scan
RF-only
Fixed
RF-only
Fixed

Scan

Fixed

Q2
RF-only
RF-only
RF-only
RF-only
RF-only

RF-only

RF-only

RF-only

MS: single stage analysis
MS/MS=tandem analysis,

Q3
RF-only
Scan
RF-only
Fixed

Scan

Fixed

Scan

Fixed

Table 1-1: Function of the quadrupoles in various TQMS scan modes.

Mode
MS
MS
SIM (MS)
SIM (MS)

Daughter
Ion Scan
(MS/MS)

Parent Ion
Scan (MS/MS)

Functional

Relationship
Scan (MS/MS)

SRM (MS/MS)

SIM: selected ion monitoring,
SRM: selected reaction monitoring,

11

b) Parent ion scan

A parent ion mass spectrum is obtained if Q3 is set to pass a selected
daughter (or product) ion, while the first mass analyzer, Q1, is scanned
sequentially passing parent ions with different mass into Q2. This resulting
spectrum will indicate all those ions generated in the source which upon
reaction with collision gas in Q2, yield a specific m/z value as chosen by the

second mass analyzer, Q3.

c) Functional relationship scan

In a functional relationship scan, both Q1 and Q3 are scanned with a
mass offset. The most familiar functional relationship scan is a neutral loss
or neutral gain scan, where the mass offset remains constant. Neutral loss
scans are used when examining dissociation reactions in the collision cell. A
neutral gain scan may be utilized if a reactive collision gas is introduced into
the collision cell for associative ion/molecule reactions. In a neutral loss/gain
experiment, ions will be detected provided that the parent ion eliminates (or
gains) a neutral moiety prior to entering the second stage of mass analysis;
the mass lost or gained being equal to the constant mass offset of Q1 and Q3.

A second example of a functional relationship scan is in the search for
proton-bound dimers; an anticipated fragment would be the protonated
monomer (18). To search for symmetric proton-bound dimers, a scan mode
such that m2+ = (m1+ + 1)/2 could be implemented, where m1+ is the parent
mass and my‘ is the daughter mass (18). In this case the mass offset is not
constant. Functional relationship scans could get quite complex, but at times

may be useful.

12

(1) Selected reaction monitoring (SRM)

Selected reaction monitoring refers to the MS/MS mode in which Q1
and Q3 are set to pass ions of only one m/z value. Neither of the mass
analyzers is scanned, a feature which provides greater sensitivity at the
expense of selectivity. Selected reaction monitoring (SRM) is analogous to a
selected ion monitoring (SIM) experiment which is employed with a single

mass analyzer, and is most often used in the targeted detection of a low-level

analyte.

3. Methods of ion activation

The heart of the MS/MS technique is the ion activation step. This
section summarizes the more prominent ion activation methods which have
been used in tandem MS. These include: collisional activation (CA) (20) with
a collision gas resulting in either charge permutation reactions, collisionally-
induced dissociation (CID), or associative ion/molecule reactions;
photodissociation, and surface-induced dissociation (SID). Activation Of the
parent ion by collision with a gas was the first activation method to be used
for analytical MS/MS analyses, and currently is the most popular method (1).
Table 1-2 lists the various activation methods used in tandem MS, and the

following section discusses in more detail those methods most prominent in

TQMS studies.

13

Table 1-2: Methods of ion activation in a TQMS.

A. Collisional Activation With a Gas:

M‘ + N ------- > M+ + N + 29'

W + N ------- > M2++ N + e'

M+ + N ------- > M + N+

M+ + N -------> M“ + N ------- >M2+ + M3 + N

M+ + N ------- > MN” -------- > M4+ + M5

B. Surface Induced Dissociation:

M+ + S -------> M+. --------'"> M6+ + M7
C. Photodissociaton:

M4. + hv ----..-> M'I'. ---------- > M8+ 4' M9

N = neutral collision gas
M“ = ion with excess internal energy.

MN+" = adduct ion with excess internal energy.

Charge Inversion (1)
Charge Stripping (2),
Charge Exchange (3)

Collision Induced (4)
Dissociation

Ion/Molecule (5)
Reaction

8 = solid surface
hv= photon

14

a) Collisional activation (CA)

Efficient transfer of the parent ion's translational kinetic energy to
internal vibrational energy can occur when an ion collides with a neutral gas
with kinetic energy in the eV range. This first step, termed collisional
activation (CA), produces an ion with excess vibrational energy. The
maximum energy available for conversion from kinetic to internal energy

distributed within the incident ion, is described by the expression:
ECM = ELab x Mg/(Mp-i-Mg) Equation 1-1

where ELab is the potential difference between the ion source and Q2 which
determines the kinetic energy of the parent ion, Mg is the mass of the target
gas, and Mp is the mass of the parent ion. The important variable is the
collision energy in the center of mass system, ECM, as this represents the
energy available for conversion to internal energy within the ion (21). The
energy in the laboratory frame may be varied by changing the DC offset
potential applied to the second quadrupole rod assembly, which is in addition
to the ramping DC voltage. Following collisional activation, the excited ion
depicted as M” may either be collisionally stabilized by third-body collisions
or if the collisionally-excited ion has sufficient internal energy, it can undergo
unimolecular decomposition forming fragment ions. This latter process is
termed collisionally induced dissociation (CID).

The mathematical expression of Equation 1-1 suggests a massive
target gas is preferred as a collision gas for CID experiments because there is
more efficient conversion of kinetic energy to internal energy relative to that

in a lighter target. However, the use of heavy mono or polyatomic target

15

gases has not always been found to be advantageous in low-energy
collisionally induced dissociation studies (22). With polyatomic targets, for
example, significant amounts of kinetic energy may be transferred to internal
modes of the target gas (22) robbing the parent ion of this available kinetic
energy. In CID studies the ionization potential of a target gas also needs to
be considered so that charge exchange does not effectively compete with CH).
The expression for conversion of kinetic energy to internal energy is ideal,
and does not exactly indicate the conversion efficiency. Studies have shown
that at very low ion kinetic energies, the fraction of the theoretical maximum
energy ECM transferred into internal energy is high, but levels Off as the
collision energy (ELab) is increased (23).

Not only does the nature of the target gas affect the collisional
activation results (22,24), but the pressure of the target gas influences the '
average internal energy deposited into the parent ion (23,25,26). In a study
by Cooks' group (23), they reported that in the collision of (C2H5)4Si+e ions
with argon in Q2 of a triple quadrupole instrument, the calculated amount of
internal energy deposited in this parent ion was 3.5 times greater for roughly
15 collisions with argon than for a single collision with argon at collision
energy of 28 eVLab for both experiments. Multiple collisions increases the
energy deposited in a stepwise fashion. The use of multiple collisions,
although increasing the average internal energy deposition, may lead to
complications in characterizing an ion structure as (i) both parent and
fragment ions may undergo a collision, thereby complicating the CID
daughter ion mass spectrum, and (ii) isomerization may occur between
collisions (23,27 -29). Furthermore, although multiple collisions increases the
average energy imparted, these conditions broaden the range of internal

energy deposited in the incident ion. The energetics of fragmentation

16
processes in CID is dependant on the nature of the collision gas and the

pressure of the collision gas, and these factors must be considered, especially

in fundamental studies.

b) Surface induced dissociation (SID)

In recent years, research pioneered primarily by Cooks' group at
Purdue University has demonstrated the feasibility of dissociating a parent
ion via its collision with a surface, usually made of stainless steel (30-32). In
triple quadrupole surface induced dissociation (SID) studies (32), a stainless
steel target was placed at a fixed 90 degree scattering angle following the
second quadrupole. The parent ion was selected by Q1, passed through the
lens system and the Q2 focussing chamber, and collided with the target
surface. Following activation of the ion due to collision with the surface, the
extracted ions were mass-analyzed by Q3. The collision energy was changed
by maintaining the ion source at ground while applying a potential of -10 to -
100 eV to the steel target. If collision with the surface supplied sufficient
internal energy for the ion, the activated ion may undergo unimolecular
dissociation, hence the term SID. Recently, in-line SID was demonstrated in
a triple quadrupole mass spectrometer (33). The advantages to the SID
method include: (i) improved vacuum as no collision gas is introduced, (ii)
narrow and very high internal energy deposition into the parent ion, and (iii)
highly reproducible spectra. Interestingly, with the SID technique,
ion/molecule reactions occur, usually for radical ions, even at collision
energies up to 50 eV (30-32). The most common reactions observed include
reduction by addition of hydrogen to the parent ion. This phenomenon is
attributed to hydrocarbon impurities which are adsorbed on the metal

surface, so that the ion collides with an organic molecule rather than a clean

17
stainless steel surface. Greater vacuum requirements are necessary to
ensure a clean surface, free of any adsorbant. SID techniques appear
promising in that greater internal energy is deposited into the parent ion
than in CA methods with a target gas, although it is not known the extent to

which this conversion of kinetic energy to internal energy takes place.

c) Photodissociation

Excitation of an ion by photon absorption is another means of
activating an ion and promoting dissociation. One of the great advantages is
that, in principle, photoexcitation can impart a wide range of energies to the
ion, assuming the ion is able to absorb the light. Of all the ion activation
methods, photodissociation is most capable of depositing a well defined
energy into the ion. Where photodissociation suffers compared to collisional
activation is that the cross sections are on the order of 10'2 A2, whereas
typical CID cross sections are 10-100 A2 (1). Mass spectrometers with the
capability to trap ions are best suited for photodissociation experiments as
ions may be irradiated by light for varying lengths of time. Most experiments
have been performed in an ion cyclotron resonance (ICR) instrument or in an
ion trap mass spectrometer, but photodissociation studies have been
performed on fast ion beams, and a recent review addresses the analytical
utility of this technique (34). A tandem quadrupole Fourier transform
instrument has also been built for studying laser photodissociation (35,36).
Interestingly enough, the first triple quadrupole mass spectrometer was
constructed with the intent of studying photodissociation of organic ions with
the RF-only quadrupole (37).

18

d) Ion/molecule reactions

All of the previously described activation methods result in
fragmentation of the parent ion. In contrast, ion activation may be promoted
by exploiting the reactivity of the parent ion with a neutral reagent gas
leading to associative ion/molecule reactions. Associative ion/molecule
reactions do not imply, however, that the product ions are always of higher
mass than the reactant ion. Associative ion/molecule reactions produce
product ions which are attributed to the ion/molecule chemistry that occurs
with the interaction, not just the unimolecular decomposition process
following CA. Ion/molecule reactions most often have been studied in ICR
instruments (38,39), flowing afterglow instruments (40), or drift tubes (41-43)
with most of the attention in these studies directed towards determining the
kinetics of ion/molecule reactions, reaction cross-sections, and gas-phase
basicities. The study of ion/molecule reactions with a triple quadrupole mass
spectrometer has remained largely underutilized.

Ion/molecule reactions proceed via the initial formation of a collision-
complex. If an ion collides with a reagent gas, a collision-complex may form.
This collision-complex intermediate may then undergo one of four processes
indicated in Table 1-3. The complex may: (i) fall apart back to the original
reactants, (ii) undergo collisions with a third body, thereby stabilizing the
intermediate, (iii) proceed to yield reaction products through one or more
reaction channels, or (iv) emit a photon in a radiative emission process (44).

The formation of a collision-complex may occur following the
ion/molecule interaction in the second quadrupole of a TQMS, and under
kinetically and thermodynamically favorable conditions, reaction products, as
reflected by (iii) in Table 1-3, can be detected. Generally, only exothermic or

19

Table 1-3: Low-energy ion/molecule reaction collision processes.
(i) M1+ + R ----- > M1R+I ----- > M1+ + R
(ii) M14“ + R -----> M1R+" ---+R----> M1R+ Collisional Stabilization

(iii) M1+ + R ----- > M1R+" ----- > M? + M3 Ion/molecule reactions

(iv) M1+ + R ----- > M1R+* ----- > M1R+ + hv Radiative Emission
where ho is a photon

20
thermoneutral ion/molecule reaction products are observed from ion/molecule
reactions occurring in the collision cell of a TQMS, although by increasing the
collision energy of the reactant ion, it is possible to deposit more internal
energy into the collision complex thereby driving the formation of
endothermic reactions. However, increasing the kinetic energy of the
reactant ion decreases the lifetime of the collision-complex, consequently the
lifetime of the collision complex may become too short to form reaction
products. Increasing the collision energy decreases the cross-section of the
reaction (45), thereby decreasing the likelihood of detecting product ions.
Finally, raising the collision energy increases the abundance of products
arising from collisionally-induced dissociation which complicates the product
ion mass spectrum. For these reasons, the collision energy is usually very

low (<5 eVLab) for studying ion/molecule reactions in the reaction chamber of '

a TQMS.

4. Analytical utility of TQMS

Triple quadrupole mass spectrometry has proved to be a versatile
analytical instrument capable of selectively detecting a targeted compound to
levels of 10'12 to 10'15 moles. Some of the recent applications of TQMS
include: screening for trace levels of environmental pollutants and
contaminants (46-49), detection of carcinogenic compounds in foods (50), drug
metabolism studies for structural characterization and identification of
unknowns (51-55), quantification of drugs in biological matrices (56-60),
analysis and sequencing of peptides and small proteins (61-65), and on-line
characterization of products produced by electrochemical oxidation/reduction

reactions of uric acid (66) and purines (67). Advances in the analysis of large

21

biomolecules has been proceeding rapidly. Recent studies have shown that
for some ionization methods (thermospray and electrospray techniques, in
particular) multiply charged ions are formed (14,63,68). This provides two
significant advantages to analysis by triple quadrupole techniques, the first
being that the mass range of the quadrupole analyzer is extended. For
example, a quadrupole mass analyzer with a range of 2000 u, will
theoretically be able to detect ions with eight charges up to 16000 u. The
second advantage offered by multiply charged ions is that the effective
collision energy in the laboratory frame of reference (Econ), is equal to the
number of charges present on the ion multiplied by the collision energy in the
laboratory frame of reference (ELab), EColl = ELab x number of charges. This

multiplication of the kinetic energy should result in greater amounts of
internal energy being deposited into the parent ion. Finally, triple
quadrupole mass spectrometry continues to be useful for CID studies of

ion/molecule reactions that occur in the ion source (69-71).

5. Limitations to dissociation (CID, SID) methods

Despite the many successes of triple quadrupole mass spectrometry
with conventional ion activation techniques, collisional activation resulting in
fragmentation does not always offer solutions to analytical problems. With
the advent of desorption ionization, and the API techniques, larger and larger
biomolecules can be introduced into the gas phase and ionized, often
producing a protonated molecular species. Minimal structural information is
obtained so tandem MS analysis is required. Coupling these ionization
methods with a quadrupole mass spectrometer generally is preferred.

However, with larger ionic species, there are more internal vibrational modes

22
for which the internal energy is distributed following CA and with low-energy
triple quadrupole methods, this may result in insufficient energy being
available for bond fragmentation. This is especially apparent in collisional
activation with a gas, whereas in SID, larger relative amounts of internal
energy are deposited. Additional work is required before one may assess the
utility of SID in fragmenting larger ions.

Poor fragmentation efficiencies is not only limited to larger ions. In
certain cases, ions of low mass (<500 u) may not fragment very readily due to
the high stability of the parent ion. In these cases, low collision energy CID
does not impart sufficient energy for bond cleavage. Highly cyclic aromatic
compounds are a class of compounds for which this holds true. In addition,
dissociation methods may not always provide the specificity necessary to
distinguish between isomeric species. Reaction channels by which two
isomers fragment may proceed through an identical intermediate, thereby
eliminating variances in the CID daughter ion mass spectrum of the two
isomers. Finally, there may not always be diagnostic fragment ions from
dissociation methods which would supply the necessary structural or
compositional information of the analyte. These limitations point to the need
of alternative activation methods other than dissociation techniques. The
approach being investigated and reported on in the bulk of this dissertation
involves the use of low-energy reactive collisions. These reactions have
occurred in the second quadrupole, although for some applications it is useful
to perform the ion/molecule chemistry in the ion source, thereby generating

the ionic species for conventional CID MS/MS analysis.

23

C. Ion/molecule reactions in a TQMS

The triple quadrupole mass spectrometer has some unique features
which make it an attractive tool for the study of gas-phase ion/molecule
reactions. Ion/molecule reactions may be studied in an ion source, and the
reaction products may be analyzed by collisional activation. An alternative
approach for using a TQMS for studying ion/molecule reactions is to select an
individual parent ion by Q1 and introduce it into the collision chamber filled
with the reagent gas. Provided that the kinetics and thermodynamics are
favorable, interaction of the ion and neutral molecule may result in the
formation of reaction products. The ion/molecule reaction conditions,
including target gas pressure and collision energy, may influence which

reaction pathway is available.

1. Advantages of using TQMS for studying ion/ molecule reactions

There are several features of a TQMS which are advantageous for the
study of ion/molecule reactions. First, quadrupole mass spectrometers are
versatile in that a wide range of ionization techniques is available for use,
including those which involve relatively high source pressure, such as FAB,
CI and API. This feature makes quadrupoles amenable to GO and LC
interfacing. The quadrupoles are able to tolerate relatively high pressures
without significant degradation of performance. Second, with a tandem
quadrupole instrument, both the reactant ion and product ions may be mass-
selected with unit resolution. Third, the reactant ion and neutral reagent gas
are separated spatially which likely eliminates mixing of the reactants prior

24
to reaction in the collision cell. Finally, the researcher has some control over
the energetics of the reaction as the collision energy may be adjusted. This
enables ion/molecule reactions to be examined in crude energy-resolved
studies. Also, by raising the collision energy, CID products may be observed.
On one hand, this may complicate the product ion spectrum, but it does
enable one to observe both reactive and dissociative collisions within the

same experiment.

2. Disadvantages of using a TQMS for studying ion/ molecule reactions

There are some undesirable characteristics of a triple quadrupole mass
spectrometer that need to be considered when studying ion/molecule
reactions. First, it is difficult to determine exactly the collision energy in
current triple quadrupole mass spectrometers. The potential difference
between the ion source and the second quadrupole provides most of the
kinetic energy to the reactant ion, but additional kinetic energy arises from
other sources. The radio frequency voltage applied to the quadrupole rods
imparts an undetermined amount of radial kinetic energy to ions within the
field. In addition, the voltages applied to the set of focusing lenses located
just prior to and after the second quadrupole will affect the kinetic energy of
the ion. This makes accurate thermochemical measurements difficult due to
the uncertainty of the kinetic energy of the reactant ion. However, the focus
of this research is not in determining thermochemical measurements, so this
limitation should not interfere with the intent of this research project.

As discussed previously, one advantage of the triple quadrupole mass
spectrometer is the capability to examine both reactive collision and

dissociation product ions. However, this may complicate the product ion

25

mass spectrum and it is necessary to determine reaction channels accessed by
each reactant ion. One way that this dilemma may be overcome is by
comparing interaction of the reactant ion with reagent gas to the interaction
of the ion with an inert gas under identical collision conditions. Comparison
of the daughter ion mass spectrum with the product ion mass spectrum
should be useful in determining the competitive processes that are occurring
in the ion/molecule interaction.

A final disadvantage with studying ion/molecule reactions in the
TQMS is that it is difficult to control the residence time in the collision
chamber. If there is insufficient residence time, it would be expected that the
yield of reaction products may be low for slower reactions. There has been
some successes with ion trapping techniques in a TQMS which effectively
increase the residence time of the ions in Q2 (72). The ions are trapped in the ‘
reaction chamber by applying a positive potential to the Q2 exit lens so that
the ions are confined in the collision cell. Following this trapping, the lenses
are pulsed with a large negative potential which serves to extract the ions.
This method enhances the signal of reaction product ions generated in the
collision because (i) the residence time was increased so that more product
ions were formed from slower ion/molecule reactions and (ii) product ions

that had low kinetic energy were extracted from the second quadrupole.

D. Literature review of ion/molecule reactions in TQMS

There have been a limited number of reports in the literature where
ion/molecule reactions were investigated in the second quadrupole and the
following section summarizes this work. Although the distinction is

somewhat arbitrary, this section has been divided into two subsections: i)

26
fundamental studies and ii) analytical applications of ion/molecule reactions
in a TQMS.

1. Fundamental studies

a) Ion/molecule reactions of C2H5O+ isomers with NH3 (73)

Two isomers of C2H5O+ composition were reacted with NH3 and N2 at
single collision conditions and the energy-resolved results were discussed in
relation to the thermochemistry of the reactions. Thresholds for the
formation of CH3+ in CID studies with N2 as the target gas were determined
from the energy-resolved plots to be 3.8 eV and 6.5 eV for CH3OCH2+ and
CH3CHOH+, respectively. These values correlated to thresholds predicted by
the thermochemical calculations of the reaction enthalpies. Thresholds were
then experimentally determined for the formation of the same ion, in this
case CH3+, in the ion/molecule reaction with NH3. It was found that the
reaction threshold, which is the energy at which product ions of m/z 15 were
detected, with NH3 as the target gas was 3.0 eV and 4.2 eV for CH3OCH2+
and CH3CHOH+, respectively. The lower observed threshold energy in the
ion/molecule reaction with ammonia was attributed to reaction-induced
fragmentation (RIF). The lower threshold values obtained experimentally are
in agreement with calculated enthalpies of reaction. The formation of stable
neutral NH2CH20H leads to the shift of the CH3+ threshold, as shown in
Table 1-4.

The authors illustrated the use of ion/molecule reactions for inducing
the formation of product ions at a lower energy than was required for

unimolecular decomposition occurring with CID. Similar values for the

27

Table 1-4: Reaction enthalpies for dissociation of CzH5O+ isomers in
collision with N2 and NH3.

CID CH3-O=CH2+ ---> CI'13+ 4» H200

AHf 6.7eV 11.4 eV -1.2eV AHr= 3.5 eV
observed threshhold = 3.8 eV

CID CH3-CH=OH+ ----> CH3+ + HCOH

AHf 6.1eV 11.4eV 1.2eV AH,=6.5 eV
observed threshhold = 6.5 eV

 

RIF CH3-O=CH2+ + NH3 ----> CH3,+ + NH20H20H

AHf 6.7eV -0.5eV 11.4eV -2.1eV AHr=3.1 eV
observed threshhold = 3.5 eV

RIF CH3-CH=OH+ + NH3 ----> CH3+ + NH2CH20H

AHf 6.1 eV -0.5eV 11.4eV -2.1eV AH,=3.7 eV
observed threshhold = 4.2 eV

28
experimental thresholds for reactions and calculated reaction enthalpies lent

support to the suggested reaction mechanisms.

b) Endothermic proton transfer reactions (74)
A triple quadrupole was used to determine the proton affinities of the

conjugate bases [C5H51- of the following acids: benzene molecular ion, 2,4-

hexadiyne molecular ion, and 1,5-hexadiyne molecular ion. The C5H6+ ion
was reacted with dimethylamine, ammonia, methanol, and water under
single collision conditions. Proton affinities for [06H5]- radicals were

established from the kinetic energy at which the onset of proton transfer to
less basic species occurs within the second quadrupole rod assembly. The
values obtained by this method were compared with proton affinities obtained
using bracketing methods, involving exothermic proton transfer reactions. In
the TQMS used for this study, the lateral motion of ions in the second
quadrupole did not appreciably influence the kinetic energy of the parent ion.
A second report by Bursey's group discussed the difficulties that need to be
addressed when measuring onset potentials for proton transfer reactions in a

triple quadrupole (75).

c) Reaction of benzoyl ions with ammonia (76)

In this study, the reaction of the C6H5O+ ion with ammonia was
investigated. The authors were interested in determining the affect of
multiple collision conditions on the determination of thermochemical
information. Onset potentials, determined experimentally, for endothermic
processes under multiple collision conditions were not in agreement with
calculated enthalpies and the authors concluded that information on the

energetics of reactions could not be obtained when the parent ion undergoes

29
more than one collision. However, they stated that multiple collision
conditions may be useful analytically for maximizing the formation of product

ions.

d) Formation of ammonium ions in reaction of protonated carbonyls
with NH3 (77)

Ammonia was introduced into the collision cell, and the dependence of
the protonation of neutral ammonia on the axial kinetic energy of protonated
carbonyl reactant ions was investigated. Evaluation of the resultant energy
curves led to the postulation that formation of protonated ammonia may
occur via two non-competitive reaction channels. The ammonium ion, NH4+,
may be generated by either a very low-energy direct proton exchange

reaction, or by fragmentation of an initially formed adduct ion.

2. Analytical applications

a) Reaction of C3H3+ ions with acetylene (78)

The first demonstration that reactive collisions in the center
quadrupole can be analytically useful was demonstrated by the reaction of
C3H3+ ions with acetylene (Csz). These C3H3+ ions were generated by
charge exchange of propargyl halides (C3H3X where X=Cl, Br or I). ICR
experiments indicated that the linear C3H3+ ion was reactive with acetylene,
forming a C5H5+ ion, whereas the cyclic structure was not. The experiments
on the triple quadrupole corroborated those on the ICR indicating that the
reactive linear C3H3+ structure is generated from C3H3I, and that the

structure of C3H3+ ions may be probed by ion/molecule reactions with

acetylene.

30

b) Reactions of CzH5O+ isomers with reagent molecules (79)

In this work, reactive collisions with either benzene or 1,3, butadiene
distinguished two C2H5O+ isomers, the 1-hydroxyethyl cation and the
methoxymethyl cation. These two isomers can be distinguished by
conventional CID methods, but the distinction is based on differences in
relative intensities of fragment ion peaks. In ion/molecule reactions of these
isomers with benzene, both isomeric ions transferred a proton to benzene
resulting in a peak at m/z 79 representing protonated benzene. However,
only the methoxymethyl cation reacted with benzene to give a peak at m/z 91

which probably represents an ion with the benzyl structure.

c) Reactions of cations with hydrocarbons (80-82)

This series of papers described research in which, small cations,
including CH3+, CH4+-, and C2H4+-, were reacted with hydrocarbons.
Reaction pathways were elucidated by selecting the particular parent ion and
reacting it with the neutral reagent gas in the collision cell of a TQMS. Much
of the work was done to confirm earlier experiments performed with an ICR
instrument or in high pressure CI studies. They demonstrated that the triple
quadrupole was an elegant instrument by which ion/molecule reactions could

be studied.

31

d) Reaction of protonated esters with ammonia (83)

The authors in this study investigated the formation of adduct ions
when a series of protonated esters was reacted with ammonia. They proposed
that the loss of H20 from the protonated ester increased when ammonia was
the collision gas rather than nitrogen due to reaction-induced fragmentation
(RIF). Elimination of water from the initially formed protonated ester-
ammonia collision complex yields stable neutral products, which effectively
decreases the reaction enthalpy of the dehydration reaction. The structure of
this collision complex adduct ion was discussed, and they proposed that it was
a covalently bound tetrahedral structure rather than a hydrogen bound
structure. Formation of an adduct ion was not observed when the molecular

ion of the esters was reacted with ammonia.

e) Protonated natural products reacting with ethyl vinyl ether (84)

This report describes ion/molecule reactions in ethyl vinyl ether of
protonated natural products, including quabalactones, psorospermins, and
diterpene dilactone compounds. The reaction channels accessed by these
protonated compounds were explained by proton affinity differences,
delocalization of the charge within the compound, along with structural
differences amongst the natural products. In some compounds, ion/molecule
reactions resulted in product ion peaks which were larger than any of the
CID daughter ion peaks, implying that the TQMS is well suited for sensitive
analytical techniques employing ion/molecule reactions.

32
0 Reaction of protonated hexachlorobiphenyl molecules with
ammonia (85)

The ion/molecule reaction between six hexachlorobiphenyl protonated
molecules and ammonia was investigated in the TQMS. The product ion
(M+NH3-HCl)+- was detected in all six cases, yet under uniform conditions,
the abundance of this reaction induced fragment ion varied by at least an
order of magnitude depending on which compound was being analyzed,

thereby offering a method for distinguishing these isomers.

g) Discrimination of 1 ,2-cyclopentanediol isomers (86,87)

The cis and trans isomers of 1 ,2-cyclopentanediols were distinguished
by ion/molecule reactions with ammonia (86). These isomers could be
differentiated by isobutane positive CI or by comparison of the CID daughter
ion mass spectrum of the protonated molecule, yet the distinction was based
on the differences of the relative abundances of some fragment ions.
However, due to proton affinity differences of the isomers, in collisions with
ammonia in a TQMS, the protonated trans isomer transferred the proton to
ammonia yielding a peak at m/z 18, whereas proton transfer with the cis
isomer was minimal. There also was a significant abundance of the
(M+NH4)+ adduct formed in the reaction of the trans isomer with ammonia,
whereas no adduct was observed with the protonated cis isomer.

In contrast to this work just described, the authors reported an
alternative approach for discriminating the stereoisomers of 1,2-
cyclopentanediols (87). Rather than distinguish the isomers in their
protonated form, a method was developed by which the neutral isomeric
compounds could be differentiated. In this case, 1,2-cyclopentanediols were

individually introduced into the second quadrupole reaction chamber. When

33
[Si(CH3)3]+ and 1,2-cyclopentanediol were reacted, a (M+Si(CH3)3)+ adduct
was formed for both isomers, but there were significant differences in the
subsequent decomposition behavior of the two isomeric adducts. The cis
isomer favored decomposition of the (M+Si(CH3)3)+ adduct to the hydrated
trimethylsilyl ion [Si(CH3)3OH2]+ ion with m/z 91. The formation of the
hydrated trimethylsilyl ion from the trans isomer adduct is endothermic, and

a definite ion kinetic energ threshold was observed.

h) Reactions in a double quadrupole mass spectrometer (88)

Although this work was not done in a triple quadrupole instrument, it
deserves mention as it demonstrated the capability of studying ion/molecule
reactions when two quadrupole mass analyzers were utilized. In this double
quadrupole instrument, a collision-cell was located between the two
quadrupole mass analyzers. Within this region, reagent gases were
introduced, and Glish demonstrated that ion/molecule reactions could supply
structural information for ionic species. Isomers of CzH5O+ were studied by
reaction with 2-propanol. Fragment ions of n-hexane reacting with n-hexane

itself were also investigated in the double quadrupole instrument.

i) Quantitative analysis using reactions in ammonia (89)

Ion/molecule reactions with NH3 were used for developing a selected
reaction monitoring (SRM) method in a GC/MS/MS analysis. The targeted
analytes were 2-methoxyethanol and chlorobenzene in methanol and in urine.
A fi'agment ion from electron impact of 2-methoxyethanol, and the molecular
ion from chlorobenzene were chosen to react with ammonia forming an

addition product ion. The authors demonstrated an increase in sensitivity

34
and selectivity compared to the conventional SRM analysis in which a CID
daughter ion fragment was monitored.

j) Reactions of protonated trichothecenes with ammonia (90,91)

Characterization of trichothecenes, Fusarium mycotoxins produced by
a variety of fungi and found in food, grain, and feed, was done utilizing a
TQMS and reactive collisions with ammonia. Reactions between the
protonated molecule with ammonia were highly specific for structural
features in this class of compounds. Good detection limits (20-300 pg) and
good reproducibility was obtained (relative standard deviation, 5-10%) by

means of reactive collisions with ammonia.

k) Reaction of CH3CO+ with 1-methylcyclopentene (92)

The authors were interested in the functional group interaction in the
fragmentation of protonated 2,7-octanedione. Over the course of this elegant
study, reactive collisions of CH3CO+ with 1-methylcyclopentene and
deuterated analogues of 1-methylcyclopentene, were performed in the
collision cell. The study of these reactions was used to lend further support to
the reaction mechanisms proposed for the fragmentation of 2,7-octanedione.
The use of ion/molecule reactions in the TQMS complemented their CID
fragmentation studies.

1) Reactions of tetrachlorodibenzo-p- dioxins with 02 (93,94)

An isomer-specific determination method of tetrachlorodibenzo-p-

dioxins (TCDD) utilizing reactions of the molecular anion with 02 was
developed by Kostiainen. Low-energy collisions of the molecular anion with

molecular oxygen result in an oxidative ether cleavage reaction yielding an

35

abundant phenoxide ion at (M-19)' which is formed by the addition of 02 and
elimination of OCl-. Variations in the abundance of this ion was dependent
on the specific isomer, and these variations were used to distinguish them.
When the isotope ion (M+2)‘ is selected, the isotope ratio pattern of the
product ions provides information concerning the chlorine distribution in the
rings. A second paper (94) dealt with the effect of collision gas pressure and
collision energy on this reaction. They concluded that it was desirable to use
high pressures of oxygen (3-6 mtorr) and a low collision energy (0.1-7
eVLab). Here at Michigan State, a similar approach has been used
successfully by Ron Lopshire for sensitive detection of polychlorinated
biphenyls (PCBs). Molecular anions of PCBs are reacted with oxygen in the
collision cell yielding the (M-1 9)’ anion.

m) Tetraethylsilane molecular ion reacting with air (95)

Reaction of a selected parent ion, either the tetraethylsilane molecular
ion with m/z 144 or the molecular ion of argon, was used to probe the 'purity'
of air. If the tetraethylsilane molecular ion reacted with dry air, only CID
product ions were observed. If the air present in the collision cell contained
moisture, products were detected which were explained by ion/molecule
reactions with water. Similarly, when Ar+- was chosen to react with air
which had been intentionally contaminated with organics, charge exchange
product ions were observed in the MS/MS spectrum, provided that the

neutral molecules had a recombination energy lower than that of argon.

n) Charge exchange and proton transfer reactions (96)
This paper reports on the advantages of first mass-selecting a reactant

ion, which then is used to ionize sample molecules in the center quadrupole.

36

It was demonstrated that control of the energetics of a specific charge-
exchange or proton-transfer reaction could be achieved with mass selection of
the reactant ions, whereas the energetics of these processes were not as well
defined when they were done in the ion source due to interfering reactions.
For these studies, n-butyl benzene was used as the energy required to form
the ions with m/z 91 and m/z 92 is well established and reflects the internal
energy imparted during ionization of the molecule. Different reactant ions
were chosen for charge exchange (CE) experiments with n-butylbenzene, and
through comparison of the relative abundances of the peaks at m/z 91 and
m/z 92, it was possible to assess the energy imparted by CE. The results were
in good agreement with values predicted theoretically.

0) Reactions of selected ions with G0 effluent (97)

Perhaps the most novel study involving ion/molecule reactions in a
triple quadrupole mass spectrometer is this study from Yost's group. The
effluent from a GC column was introduced into the collision cell, so that
reactions of a selected reactive ion with components in a mixture could be
investigated. Charge exchange reactions and chemical ionization reactions
were studied. This approach allowed the control of the energetics involved in
the ionization of the eluting GC components. By alternating the reactant ion,
different internal energies could be imparted to the neutral reagent enabling
both structural information and molecular weight information to be obtained
during the same chromatogram. A search of El library spectra using the
charge exchange spectrum was found to be generally reliable, with CH3+
charge exchange spectra yielding the most reliable match. This method is

being evaluated for selectively detecting aromatics in jet fuels. Detection

37
limits for benzophenone with this method were achieved at the low picogram

level.

E. Low-energy ion/molecule reactions in a four sector instrument
(98-101)

Recently, a series of papers has been published which describes
endothermic ion/molecule reactions that occur between protonated molecules
and ammonia; of particular interest is the reaction of protonated leu-
enkephalin and ammonia. A second report describes reactions of N,N
dimethyl myristamide and monomethyl amine (102). Remote site
fragmentation is observed at very low center-of—mass energies (102).
Although these experiments were not done with a TQMS, this type of
experiment is well suited for a TQMS, and the results are interesting and
deserve mention. These experiments were done in a hybrid instrument of
BEQQ geometry and in a HX110/HX110 four sector (EBEB configuration)
tandem mass spectrometer equipped with a collision cell on which large
potentials could be applied. This allows the study of very low beam collision
energy ion/molecule reactions. The authors report on the increase in
fragment ion abundances of protonated leucine enkephalin (m/z 556) when
ammonia is the target gas compared to a neutral gas such as helium, for
collision energies on the order of 5-10 eVLab (98-100). This is attributed to
reaction induced fragmentation, and results in an overall fragmentation
efficiency that is even greater than high-energy CID. For their experiments,
the ion beam was attenuated up to 99% in some cases, yet they correlate their
results with predicted energetics of the endothermic reaction and report

energy-resolved curves. As discussed earlier in this chapter, thermochemical

38
information cannot be obtained under multiple collision conditions, which
surely are occurring with these beam attenuations. In spite of what is
perceived as certain flaws and that according to one of the investigators,
difficult to reproduce (103), these experiments are intriguing, and
demonstrate the power of increasing the fragmentation efficiency for large
ionized molecules by utilizing ion/molecule reactions. These studies led to the
development of a new technique, called neutralization chemical reionization
mass spectrometry (101) where deprotonation of the reactant ion, followed by
reprotonation, occurs in the collision cell.

This report described above merits some additional attention. As
mentioned, it seems as if a TQMS is more suited to these types of
experiments, so a preliminary attempt was made exploring this reaction
using the TQMS. Protonated leu-enkephalin was produced in the ion source
by FAB, and the ion (m/z 556) was passed into the collision cell containing
ammonia. A target gas pressure was chosen which resulted in 80%
attenuation of the primary beam at a collision energy of 10 eVLab. In the
TQMS study, the (M+H+NH3)+ adduct was most abundant at 1-2 eVLab
collision energy. Beyond 2 eVLab, no adduct ion was observed in the TQMS.
This is in contrast to their study where 6 eVLab collision energy provided
maximum formation of the (M+H+NH;3)+ adduct as seen Figure 1-3. They
also report that fragment ion abundances for ions with m/z 425, 397 and 279
reached a maximum at a slightly higher collision energy, and then leveled off.
They believe that forming the (M+H+NH3)+ adduct, via an endothermic
ion/molecule reaction, provides internal energy for this adduct ion which
induces fragmentation. In preliminary studies done in our laboratory using
the TQMS, this fragmentation behavior was not observed. Additional

investigations are required before one may assess the utility of this reaction

39

 

 

 

 

 

 

 

12
——a— NH4 (x10)
0" —9—- (M+H+NH3) (x10)
—-o—— 279
a ——6— 397
8- I + 425
5’ 1
.3 I
g 6-1 . c
.5 l
C)
p
'43
.9. 4-
'4
2 - I J
I, e ’ e
0 “ I Y Y I ' I 1
0 S 10 15 20 25 30
E(lab) eV

Figure 1-3: Relative abundances of various ions in reaction of protonated
leu-enkephalin with ammonia as function of ion kinetic energy. (Adapted

from reference 100).

40
when utilizing a TQMS, and this may be a fruitful area of research

considering the interest in peptide sequencing at Michigan State University.

F. Research objectives

A primary goal of this investigative research is to use a triple
quadrupole mass spectrometer for developing novel applications of gas-phase
ion/molecule reactions. Manifestation of this goal may be realized by a
variety of diverse applications. The following guidelines were used as an aid
in directing this research to the overall objective:

(i) Apply reactions which have been previously characterized, either in
a TQMS or another instrument, and seek to develop methodology which
exploits the unique intermolecular activity of specific ionized functional '
groups present in a molecule. The chemistry of the ionized species with a
reagent molecule may provide selective detection of a class of compounds.

(ii) Commit to understanding the mass spectrometric results in terms
of the chemistry that is occurring in the gas-phase. This may lead to
intensive study of a specific ion/molecule reaction, perhaps in an attempt to
unravel mechanisms of ion formation or fragmentation. Where possible, the
reaction will be evaluated in terms its thermochemistry. At times,
calculating the reaction enthalpies may prove valuable in directing the future
course of the research.

(iii) Being observant to the chemistry that is occurring in the ion
source of the mass spectrometer may prompt ideas for ion/molecule reactive
collisions which could occur in the second quadrupole collision cell. This

would allow a specific reaction to be examined in an environment free of any

41
interfering reaction, and may help in understanding the chemistry that is
happening within the ion source.

(iv) Realizing that ion/molecule reactions are not a panacea for all
analytical problems involving mass spectrometry, but instead just a
complement to the more conventional MS/MS techniques. Some of the work
reported in this dissertation does not involve ion/molecule reactions, but new
applications for the traditional collisionally induced dissociation methodology.

Mass spectrometry is a tool used for studying isolated ions in the gas-
phase. Understanding the chemistry of these ions in their interaction with
neutral molecules is critical in the evolution of analytical methodology. This
dissertation is the culmination of the effort to further demonstrate the
capabilities of a triple quadrupole mass spectrometer in studying
ion/molecule reactions, and reveals some novel applications for this unique

instrument.

CHAPTER II
ION/MOLECULE REACTIONS OF ARYL CATION S

A. Introduction

Ion/molecule reactions involving aromatic cations, especially those ions
which contain the six-membered benzene ring, have been widely studied in
the gas-phase (104-106). The impetus behind the research that is described
in this chapter arises from the previous effort put forth by Dr. Gregory
Dolnikowski while he was a graduate student at Michigan State University
(106). In his ion/molecule reaction studies utilizing a TQMS, he'
demonstrated that aryl cations which are even-electron ions with a vacant
charged site on an aromatic ring, react with methanol to form a hydroxylated
product ion. In the case of the CaH5+ phenyl cation with m/z 77, reaction
with methanol vapor was shown to produce a product ion with m/z 94. This
product ion of m/z 94 was shown to have the structure of the molecular ion of
phenol as established by their strikingly similar CID daughter ion mass
spectra (71 ). Aryl cations were also found to be reactive with ammonia (106).
Once again using the phenyl cation for illustrative purposes, the product ion
with m/z 93 was obtained in the reaction with ammonia. This ion was shown
to be identical in structure to the molecular ion of aniline based on the
similarity of their CID daughter ion mass spectra. Ammonia has also been
used as a reagent gas in the collision cell of a TQMS to probe isomeric
XCsH4+ ions for X: OCH3, CH3, F, and ON (107). In these studies, a
nucleophilic substitution reaction occurred forming the molecular ion of

derivatives of aniline.

42

43

The phenyl cation was the model aryl cation used in these earlier
studies, and it proved to be very reactive in the gas-phase. This chapter
describes the continued investigation of ion/molecule reactions involving aryl
cations. For most reactions described in this chapter, the phenyl cation or the
(M-H)+ aryl cation generated from napthalene, is the model aryl cation which
serves as the reactant ion.

It should be noted that all of the research performed by Dr.
Dolnikowski was done on an Extranuclear (now Extrel) triple quadrupole
mass spectrometer which was located in the NIH Michigan State University
Mass Spectrometry Facility. Following his departure, a new triple
quadrupole mass spectrometer was purchased, that being a Finnigan TSQ-70.
Some of the initial experiments performed on the newly acquired Finnigan
TQMS were done in the attempt to reproduce the experiments performed on
the Extrel TQMS. As mentioned in chapter one, the Finnigan TQMS
instrument contains a non-linear second quadrupole rod assembly, and all
three sets of quadrupole rods are physically much smaller than the
quadrupole rods in the Extrel instrument. Therefore, it was necessary to
attempt initially experiments on the Finnigan TSQ-70 that had been
characterized previously with the Extrel TQMS, so the feasibility of
performing ion/molecule reactions in the TSQ-70 could be determined. These
investigations led to the further study and elucidation of the hydroxylation
ion/molecule reaction between the phenyl cation and methanol, establishing
in more detail the reaction mechanism. Furthermore, additional experiments
of aryl cations with different reagent gases were performed.

This chapter is divided into three broad sections following this
introductory section. In the first section, the chemistry of the reaction of aryl

cations, predominantly the ion/molecule reaction of the phenyl cation with

44
various reagent gases is described. The intent was not to perform an
exhaustive study for each reagent gas investigated, but rather to examine
reactions with some neutral reagent candidates and comment on the
ion/molecule reaction products in terms of the computed thermochemistry.
Unless otherwise indicated, all of the reaction enthalpies reported for the
ion/molecule reactions, were calculated based on ionic and neutral heats of
formation obtained from reference 108. The latter sections of this chapter
describe two analytical applications for an ion/molecule reaction with
dimethyl ether which appears to be selective for aryl cations. One application
demonstrates the potential for using neutral gain experiments in screening
for aromatic compounds present in mixtures. The second application
described in this chapter involves probing the structure of C7H7+ isomers

utilizing low-energy ion/molecule reactions in the center quadrupole of a

TQMS.

B. Survey of ion/molecule reactions involving the phenyl cation
1 . Experimental

The experiments described in this chapter were performed on the
Finnigan TSQ-70 triple quadrupole mass spectrometer. The phenyl cation
was generated by electron impact of benzene which was introduced directly
into the ion source via a variable leak. A direct inlet system was constructed
by the machine shop which fitted onto the flange mounted for use with the
GC transfer line. This set-up enabled the CI reagent gas lines to be utilized
simultaneously which provides the capability of mixing neutral vapor

molecules in the ion source. Solid samples analyzed were dissolved in a

45
suitable solvent and introduced into the ion source via the direct insertion
probe. When the ion/molecule reactions were performed in the collision cell,
the reagent vapors were introduced through the CID gas lines.

For liquid reagents, the sample was introduced into a glass bulb which
was constructed at the Department of Chemistry's glass blowing shop. These
bulbs were fitted with 1/8" stainless steel tubes which enabled them to be
connected directly to the gas lines via SwagelokTM connections. The bulb also
was fitted with a rubber stopper which permitted an efficient vacuum seal to
be created when the bulb was exposed to the pumping system of the TSQ-70.
After connecting the empty glass bulb to the collision gas inlet line of the
TSQ-70, the gas line valves were opened which served to purge the bulb prior
to the introduction of the liquid reagent. Once the bulb was purged, the
valves were closed and the liquid was introduced via an injection by a syringe '
through the rubber stopper. Provided that there was sufficient vapor
pressure, the gaseous molecules diffused into the second quadrupole collision
cell (or into the ion source) and the pressure could be regulated by the needle
valves connected to the gas lines. The pressures in the collision cell and in
the ion source were measured by a convectron gauge, and are not indicative of
the actual pressure. An ion gauge is located towards the rear of the manifold
region of the TSQ-70, and the pressure readings indicated by the ion gauge
generally remain stable throughout the duration of an experiment. The
pressure reading indicated by the convectron gauge fluctuates often, even
while the ion gauge reading remains steady. For all of the experiments
reported in this chapter, the temperature of the ion source was maintained at
1 50°C, while the manifold region was held at 70°C. Figure 2-1 is a
representation of the vacuum system of the Finnigan TSQ-70.

46

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SOURCE nnNIFOLo SOURCE COLL CELL FOREPUHP nnNIFoLo
Cclsrus Cclsrus mrllrtorr millitorr millxtorr tor-r
F259 F259 Fresco Fresco reeec rE-B
~2ee »2ee ~reee rreee ~reee :-€-4
ise use “100 nee —rreo L-E—s
use P190 ~10 ~10 r-re r'E-G
—Ii
— so |~ so ~E-7
- e I — e LE-e
119 66 .se E-7
. .ee E-4
Solenord Operated .99 5-5
Valve\ Convectron
1 PressureGauge
CI ens AJ
Rotary-Vane ‘
Pump ‘— ' \ \
CID ens ° :4 Vent ”mes
I DHnRounRE w ON vacuun CHECKS Pessso I Dav surrcu
I Usorruarzs w on I Deux PouER ON I Devpnss JUNPER
I Usorrunae PROTECT l Deux PROTECTED U Inuro VENT as rain.
anvE- Calibration T
_ urbomolecuterPump
Compound Vial

 

 

Figure 2-1: Schematic diagram of the vacuum system of the Finnigan TSQ-
70 triple quadrupole mass spectrometer. (Adapted from reference 17).

47
2. Reaction with alcohols

a) Methanol

The first ion/molecule reaction investigated in the collision cell of the
Finnigan TSQ-70 was the reaction of the phenyl cation with m/z 77 and
methanol. The product ion mass spectrum is shown in Figure 2-2, along with
the product ion mass spectrum obtained on the Extrel triple quadrupole prior
to this instrument being transported to the Chemistry Building. In both
cases, the pressure in the collision cell was approximately 1 mtorr and the
collision energy was 2 eVLab. Considering the vast differences in the physical
dimensions of the Q2 collision cells, the product ion mass spectra obtained
with the two TQMS instruments appear remarkably similar. Based on the
results of this initial experiment, it did not appear that the physical
dimensions of the rods in the TSQ-70 would deter the formation of low-energy
ion/molecule reaction products and subsequent detection.

It had been demonstrated by Dolnikowski that the reaction product ion
of m/z 94 was identical in structure to the molecular ion of phenol (71). This
determination was made by comparing the CID daughter ion mass spectrum
of the molecular ion of authentic phenol with the CID daughter ion mass
spectrum of the reaction product ion of m/z 94 generated when benzene and
methanol were simultaneously introduced into the ion source. It seems likely
that the structure of the intermediate collision-complex preceding the
formation of the product ion of m/z 94, is identical to the structure of
protonated anisole as shown in Figure 2-3. However, no experimental
evidence had been obtained on the Extrel which would support this
hypothesis. To verify this premise, methanol and benzene were

simultaneously introduced into the ion source, and the product ion which

48

 

 

 

 

 

 

77
100' Phenyl
A Cation
5° 80-
3‘
og m-
3
.5
Q) 40-
.g + Product Ion
:3 20_ (CH3OH)H 94 .
33 51
0 'F'!"""I' 'fi'inl" I "l' "I "'I""'" :T"'I""I'3"‘I""I""'""I """'l'
30 40 50 60 70 80 90 100 110 120
m/z
77
100‘ B Phenyl
Cation
s 80"
5‘
'8
t: 60“
3
a 40 Product Ion
.2 ' 94
‘63 +
'3 20_ (CH30H)H
33 51
o “J :"C -:.: x": I--- ---- ---- ---- -- -- 2 -- --- ---- Icfi‘---- -n- n" .1- "n T
'I I ' I ' I ' I ' I' ' I I I I I 'I I
30 40 50 60 70 80 90 100 110 120
m/z

Figure 2-2: Product ion mass spectra of the phenyl cation and methanol
obtained with the A) Extrel TQMS and B) Finnigan TSQ-70.

 

 

 

3:233: 3:: :23“. 3:23 23
.3 5323: 23 89¢ 3:23 .3 :2 3383:: 23 8.8m 3 :33232: 5303: moi—83833 ”TN 9:5:

3 ES 9: «E:
q E. 3:
m
9 m 1mm 10% +
$0. + 0 Al 0 ll mOmmO + 0
3:23 .3 3328:2228 :23“. 3:23

:2 3383:: 3238.235

50
appeared with m/z 109 was subjected to CID. The CID daughter ion mass
spectrum of the proposed intermediate ion of m/z 109 is shown in Figure 2-4.
Protonated isomers of cresol along with protonated anisole were all viewed as
possible structural candidates for the intermediate with m/z 109. All of these
ions have the C7H90+ composition. The CID daughter ion mass spectrum of
protonated anisole, along with the CID spectra of the protonated forms of two
cresol isomers, are shown in Figure 2-5. All of the daughter ion mass spectra
show a fragment peak at m/z 94, therefore it appears that all of these
isomeric ions with m/z 109 give rise to the phenol molecular ion upon CID.
The CID daughter ion mass spectrum of protonated anisole, however, is
nearly identical to the CID daughter ion mass spectrum of the product ion
produced from the phenyl cation and methanol reaction (Figure 2-4). The
distinguishing feature is the presence of peaks at m/z 66 and m/z 77 in the '
CID daughter ion mass spectra of both protonated anisole and the reaction
product ion of m/z 109. These peaks at m/z 66 and m/z 77 are absent in the
CID daughter ion mass spectra of the o-and p-isomers of cresol. This clearly
establishes that the molecular ion of phenol is formed via the initially-formed
intermediate collision-complex which has the structure of protonated anisole.
These experiments demonstrate how CID and ion/molecule reaction
experiments complement each other. The isolated reactants of a product ion
may be determined by examining the ion/molecule reaction in the center
quadrupole, while mechanisms of the product ion formation may be verified
by CID studies following the formation of the reaction products in the ion
source. The formation of the product ions in the ion source does not preclude
the possibility of other isomeric product ions being formed with identical m/z
values, however, to generate the product ion in the second quadrupole, it

would be necessary to have a second collision chamber and a third mass

51

Parent Ion
109

8.8:?

Relative Intensity %
3.

77 94

A A A A A A II I
v'lv vvv'vvvvlwvvv'vvwv lvvvv'vvv'l"

- "2- n.--
30 40 50 60 7O 80 90 100 110 120
m/z

9.3

 

l-l 1 ll

 

 

A A I
vv'vv WIVfiv—tvv-v'vv 'v' vvvv'vvvv V‘v I vwv V'vvvvl'

figure 24: CID daughter ion mass spectrum of product ion of m/z 109
formed when benzene and methanol were simultaneously introduced
into the ion source. Collision energy was 30 eV Lab at single collision

conditions.

52

Relative Intensity %

 

109

 

 

#8:???

Relative Intensity %
B

 

'vv 'm

100 110

vvv Iv

120

109

 

 

O
'1!

9f

1

Relative Intensity %
o a a a a

 

94

1-1 .
--'----,

vvv'v

100 110 120

109

 

-1

 

m/z

A A A
- wv vvv vvvv'vvvv vwvv'w vv vvv -v—v'—vvv ‘vv

‘v'

v.. I'

"165" '1'1'o"" 120

Figure 2-5: CID daughter ion mass spectra of the protonated forms
of A) anisole, B) o-cresol and C) p-cresol. Collision energy was

30 eVLab at single collision conditions.

53
analyzer for the subsequent CID. This would require a penta-quadrupole
mass spectrometer which is not available at Michigan State University. An
alternative approach would be to perform an MS/MS/MS experiment in an
ion-trap mass spectrometer.

One of the interesting features of this ion/molecule reaction is that the
initial reactant ion, C6H5+, is an even-electron ion, while the product ion with
m/z 94 contains an odd number of electrons. The even-electron rule, often
applied as a basic principle in organic mass spectrometry, is usually applied
to fragmentation of ions following ionization by EI (109,110). It states that
odd-electron cations may eliminate a radical or an even-electron neutral
species, but even-electron ions generally will not lose a radical to form an odd-
electron cation. In our case, an even-electron ion reacts with methanol to
form protonated anisole, an even-electron intermediate collision-complex.
This intermediate ion (m/z 109) eliminates a radical, which is in violation of
the 'rule’ used for interpreting mass spectra generated by E1. The question of
an even-electron reactant ion forming an odd-electron product ion was posed
by a referee who reviewed a manuscript (71), and additional experimental
work was performed to verify our claims concerning this hydroxylation
reaction. The referee thought that perhaps the reactant ion was not the
phenyl cation, but rather the molecular ion of benzene, which is a much more
intense peak in the EI mass spectrum. If the benzene radical cation, 05H6+-,
is chosen to react with methanol in the collision cell, there is no peak detected
at m/z 94, but rather at m/z 95. This product ion with m/z 95 presumably
represents protonated phenol. Formation of protonated phenol and the
methyl radical from CsHs-+ and methanol is 16.2 kJ/mol exothermic. On the
other hand, the reaction enthalpy for the production of the molecular ion of

phenol along with a methyl radical and hydrogen radical from C6H6-+ and

54

methanol is 311.6 kJ/mol endothermic. If methane were the neutral product
along with the phenol molecular ion in the reaction of 06H6-+ and methanol,

the reaction enthalpy would be 126.7 kJ/mol exothermic. This exothermic
sequence, however, would require extensive hydrogen rearrangement and
does not appear to occur as suggested by the absence of a peak at m/z 94 in
the product ion mass spectrum. Although our experimental results of the
reaction of the even-electron phenyl cation with methanol violate the so called
'even-electron rule', thermochemical data support the experimental claims.
The reaction of the phenyl cation with methanol to form the molecular ion of
phenol and a methyl radical is 54 kJ/mol exothermic. Additional reaction
products may be expected based on the calculated exothermic heats of
reactions, however, these product ions are either absent or present only as
minor peaks in the product ion mass spectrum. This is attributed to
unfavorable kinetic conditions, that is, the residence time within the collision

cell is too short for these reactions to occur.

b) Ethanol

Following the success of the hydroxylation reaction with methanol, the
reaction of the phenyl cation with a higher order alcohol was investigated. It
was anticipated, rather naively, that in similar fashion to the reaction with
methanol, the reaction with ethanol would yield the hydroxylation product
ion with the structure of phenol. However, as shown in Figure 2-6, the
predominant product ion appears as a peak at m/z 95, which represents
protonated phenol. Once again, examination of the reaction enthalpies
provides insight into the chemistry by clearly indicating that the formation of
protonated phenol is favored thermodynamically over the formation of the

molecular ion of phenol. In fact, as shown in Table 2-1 , methanol is unique in

55

 

 

 

 

1CD ' Phenyl cation
77
Q
3 80'
'55
g a)“ Product Ion
.5 95
ff; 40‘ (EtOH)H+
43
an
.3 ms 47
o TT"""‘"'I"?'""-‘I'fi'v'"l'"'"'"l“"'.' :'l‘""""l‘: " ""I"#"‘l"""'"l'
3O 40 5O 60 7O 80 90 100 110 120
m/z

Figure 2-6: Product ion mass spectrum of the phenyl cation reacting with
ethanol at a pressure of 1.5 mtorr and with a collision energy of 1 eVLab'

Table 2-1: Calculated enthalpies for reactions of phenyl cation with
alcohols“

AH,“ (kJ/mol)

C6H5+ + CH3OH ----------> CsH50H+- + -CH3 -54

CeHs+ + CH30H ----------> C6H50H2+ + CH2 +77
C5H5+ + CzH5OH ---------- > CeH50H+- + ~02H5 - 52
C6H5+ + CszOH ---------- > C6H50H2+ + C2H4 - 227
C6H5+ + CaH7OH ---------- > C6H50H+- + -CaH7 -50

C6H5+ + C3H7OH ---------- > C6H50H2+ + C3H5 -2§£
C5H5+ + i- CaH7OH --------- > C6H50H+- + -C3H7 -$
C5H5"’ + i- C3H7OH ---------> C5H50H2+ + CaHs -221

a Ionic and heats of formation found in reference 108.

57

that the thermodynamically-favored reaction product is the molecular ion of
phenol and not the protonated form. For reactions of the phenyl cation with
alcohols which contain more than one carbon, the formation of the molecular
ion of phenol and a radical species is exothermic, however, the elimination of
a neutral olefin along with the formation of protonated phenol, provides
neutral products with increased stability. This significantly increases the
exothermicity of the reaction. During the course of formation of protonated
phenol, hydrogen located on an ethanol carbon atom apparently may undergo
rearrangement, transferring to the oxygen atom prior to elimination of the
olefin. If this transfer can also readily occur for higher order alcohols and
proceed rapidly, the thermochemistry suggests, as indicated in Table 2-1 , that
protonated phenol would be the expected ionic reaction product of the phenyl
cation with alcohols other than methanol.

3. Reaction with amines

a) Ammonia

The reaction of the phenyl cation with ammonia was previously shown
to result in the formation of reaction product ions with m/z 93 and m/z 94
(106). Based on the similarity of the respective CID daughter ion mass
spectra, the ion with m/z 93 was shown to be identical in structure to the
molecular ion of aniline (106). The ion with m/z 94 may either represent
protonated aniline or a non-covalently bound phenyl cation-ammonia adduct.
This reaction was performed in the second quadrupole of the TSQ-7O and the
product ion mass spectrum shown in Figure 2-7 appears similar to that

obtained on the Extrel TQMS. The reactions forming ionic aniline and

protonated aniline are exothermic as indicated in Table 2-2.

 

 

58

Phenyl cation

100‘ 77

80‘

Relative Intenstity (%)
59

 

 

 

 

 

A I A 1 AA
vv'vvvv'vvvvlvvvv'v 'v'vVVV‘vvvv'vv rvvvvl'ffvv'vvva

3O 40" 50W 60 70 80 90 100 110
m/z

Figure 27: Product ion mass spectrum of the phenyl cation and ammonia.
Collision pressure ~ 2 mtorr and collision energy is 2 eVLab'

59

Table 2-2: Reaction enthalpies for the phenyl cation reacting with ammonia.

C6H5+ + NH3 ---------- > C5H5NH2+' + -H Aern = -34 kJ/mol
C5H5+ + NH3 ---------- > C(5HsNH3+ AHl-xn = ~341kJ/mol

60

It is interesting to examine the energy-resolved curve for the reaction
of the phenyl cation with ammonia as shown in Figure 2-8. At collision
energies greater than 5 eVLab, the product ions obtained in the interaction of
the reactant ion of m/z 77 and ammonia are a result of CID. At collision
energies nearer to O eVLab, the abundance of the ion/molecule reaction
products increases, while the CID products decrease in intensity. Of
considerable importance is the maximum abundance of the ion/molecule
reaction product of m/z 93, which represents the formation of a covalently-
bound product ion as compared to the maximum abundance of the CID
product ion of m/z 51 over the range of collision energies examined. The
intensity of the CID peak at m/z 51 representing the elimination of ethyne, at
optimal fragmenting conditions is only one-fourth that of the maximum
abundance of the covalently bound ion/molecule reaction product ion (m/z 93)
formed at low-collision energy. The implication of this observation is
important. Here is a case where the inherent stability of the parent ion
impedes fragmentation in CID, whereas with the ion/molecule reaction, the
formation of more stable products results in an abundant product ion being
detected. The formation of the C4H3+ ion and C2H2 from the CsH5+ phenyl
cation is 318 kJ/mol endothermic. By taking advantage of the exothermic
reactivity of the phenyl cation with ammonia, an ion/molecule reaction
sequence occurs which forms the molecular ion of aniline and -H; the reaction
enthalpy is -34 kJ/mol. As discussed in chapter one, a possible disadvantage
with the ion/molecule reaction approach using a TQMS is the potential
difficulty in detecting product ions formed at low collision energy due to
insufficient kinetic energy. At low collision energies, this can lead to poor

transmission efficiency. However, in this case, the detectability of the low-

61

 

 

INTENSITY

 

 

-2o -15 ‘ -1o -5 o
COLLISION OFFSET (Wm)

Figure 2-8: Energy-resolved curve of the reaction products generated
from reaction of the phenyl cation and ammonia. Collision pressure
is 1 mtorr.

62
energy ion/molecule reaction product ion is greater than the detectability of
CID fragments. These results imply that ion/molecule reactions may be
useful in analytical applications where the formation of a product ion in high
yield is critical for the development of MS/MS methodology which could
provide low detection limits.

b) Methyl amine

Methyl amine was introduced into the collision cell and the phenyl
cation was selected as the reactant ion. The product ion mass spectrum at a
collision energy of l eVLab, obtained when methylamine was present in Q2 at
a pressure of 1.5 mtorr, is shown in Figure 2-9A. The spectrum obtained
under multiple collision conditions, indicates a peak at m/z 93 which again
represents the molecular ion of aniline, and a peak at m/z 32 representing
protonated methyl amine. Protonated methyl amine is formed by
ion/molecule reactions in methyl amine occurring in Q2. If the collision
energy is raised to 20 eVLab, the product ion mass spectrum, as seen in
Figure 2-9B, lacks the peaks at m/z 93 and m/z 32, but the CID peak at m/z
51 is more prominent. The enthalpy of the reaction forming the aniline
molecular ion and the methyl radical is 129.1 kJ/mol exothermic. If the (M-
H)+ of napthalene is generated by EI ionization and chosen to react with
methylamine, once again a product ion at 16 u (corresponding to NH2) higher
than the parent ion is detected in the product ion mass spectrum as shown in

Figure 2-10.

Relative Intensity %

Relative Intensity %

 

63

 

Phenyl cation

 

 

 

 

10)‘ A 77
m1
6) 32 Product ion
40 ‘ 93
m-
0 .' :"clwafx'l moth"; :-'--,-,-,--l-:--:m-rv-3,----'--",--"I-
30 4O 5O 7O 80 90 100 110 120
m/z
Phenyl cation
Im‘ 77
B
m -I
m ..
4) 51
m-
0 'l:"":":l"-‘" f"?! """" l """"" l ""5 :'l""""'l': ":" 'l' '1' "'l""""'I'
30 40 5O 70 80 90 100 110 120
m/z

Figure 29: Product ion mass spectrum of the phenyl cation and methyl
amine at collision energy of A) 1 eV Lab and B) 20 eV Lab . Methyl amine
pressure is 1.5 mtorr.

64

 

 

 

 

 

Parent ion
100' 127
39
>. 33'
'5 Product ion
3 60‘ 143
,5 32
3 40*
23
.9.
a? Z"
158
0 * l'""*""l‘l" WW. w up .-.-:-,----,----, ---. -,! --.----,
20 4O 60 80 100 120 140 169 180

m/z

Figure 2-10: Product ion mass spectrum of the (M-H)+ of
napthalene with methyl amine at a pressure of 1.5 mtorr
and collision energy of 1 eVLab .

65

4. Reaction with CH3CN

In contrast to the previous reagents described in reaction with the
phenyl cation, the computed thermochemistry suggests that a covalently
bound addition product ion would not be formed when the phenyl cation
reacts with acetonitrile. The reaction enthalpy for the formation of the
C(3H5CN+ cation and a methyl radical from the phenyl cation and acetonitrile
is 392 kJ/mol endothermic. The CH3-CN bond strength is 507 kJ/mol, and is
not overcome by the formation of a new bond with a ring carbon in the phenyl
cation. Although this reaction was not examined with the phenyl cation, the
analogous reaction was performed with the (M-H)+ aryl cation of napthalene.
In the studies being described in this chapter, ion/molecule reaction with the
(M-H)+ ion of napthalene always parallels a successful reaction of the reagent
with the phenyl cation. Therefore, reaction of the napthyl cation with
acetonitrile should be indicative of the reactivity of the phenyl cation with
CHgCN. The product ion mass spectrum of the napthyl cation with m/z 127
and acetonitrile is shown in Figure 2-11 . A peak at m/z 168 representing the
non-covalently bound adduct ion is the only product ion detected. There is no
product ion detected which represents covalent bond formation during a ring

addition reaction.

5. Reaction with ethers

a) Dimethyl ether

The product ion mass spectrum of the phenyl cation with dimethyl

ether at a pressure of 1 mtorr and a collision energy of 1eVLab shows a

66

 

 

 

Parent ion

100‘ 127
3°

mu:
.3
2 .
3 so
:3
t—l
g 4)
"3
"" 20* 168
if 77

o ‘ --,?--t'-.‘.:,-‘..|.In"?--- -,-- I ,‘ 4 -,- La"

40 60 80 100 120 140 160 180 200
m/z
Figure 2-11: Product ion mass spectrum of (M-HY of napthalene

with acetonitrile at a pressure of 2 mtorr and a collision energy
0f 1 eVLab.

67
product ion peak at m/z 108 which represents the addition of -OCH3 (Figure
2-12). Calculation of the reaction enthalpy indicates that this reaction
forming the molecular ion of anisole is 73.1 kJ/mol exothermic. Dimethyl
ether has been used as a reagent gas in chemical ionization studies, yet there
were no peaks detected at (M+31)+ in the chemical ionization (CI) mass
spectrum for either benzene or napthalene (111). Perhaps with CI being a
soft ionization method, there is not sufficient internal energy in the
protonated ion to eliminate H2 forming the aryl cation for subsequent
reaction with dimethyl ether. Ion/molecule reactions involving dimethyl

ether will be discussed in greater detail later in this chapter.

b) Diethyl ether

The product ion mass spectrum of the phenyl cation with diethyl ether
at a pressure of 1 .2 mtorr and a collision energy of 2 eVLab is shown in Figure
2-13. Two abundant product ion peaks are observed at m/z 95 and m/z 105.
The product ion with m/z 95 may represent protonated phenol; this reaction is
158 kJ/mol exothermic as seen in Table 2-3. This reaction requires
consecutive hydrogen rearrangements for the elimination of two ethylene
molecules. The second product ion with m/z 105 could represent a CgH9+ ion.
Generation of this product ion requires a collision-complex intermediate
structure which would allow for the formation of a C-C bond and the
elimination of ethanol, an ion/molecule reaction which is 279 kJ/mole

exothermic.

68

Phenyl cation

 

 

 

 

100‘ 77
5° 804
g .
3 60-
3
.5 Product ion
0
.3 40 108
2
“3 20- 4'5

47 51 61
o I .1 II - II '.l I *- j
40 60 80 100 120
m/z

Figure 2-12: Product ion mass spectrum of the phenyl cation and dimethyl
ether at a pressure of 1 mtorr and collision energy of 1 eV Lab .

69

 

 

 

1CD . Phenyl cation
89 77
.2» 80'
'Er'i
§ 60‘
.52
.E’ 404
«OJ
'9' 75
m-
0 '1 "3' l"':"L"l!" ""'I""""'I"L" l'l""""'1‘:"!""l""""'l"""'"I'
30 40 50 60 70 80 90 100 110 120
m/z

Figure 2-13: Product ion mass spectrum of the phenyl cation and diethyl ether
at collision energy of 2 eVLab and pressure of 1.2 mtorr.

70

Table 2-3: Reaction enthalpies for the phenyl cation reacting with diethyl
ether.
C6H5+ + (C2H5)2O ---> CeH5OH2+ + 2C2H4 AHm = -158 kJ/mol
CsH5+ + (C2H5)20 ---> C6H5C+HCH3 + C2H5OH Aern = -279 kJ/mol

0) Vinyl methyl ether

The product ion mass spectrum of the phenyl cation with vinyl methyl
ether (VME) at a pressure of 1.2 mtorr is shown in Figure 2-14. The major
product ion, although not as intense as product ion peaks derived from other
reagents, is observed at m/z 103 which represents the addition of C2H2.
Minor product ion peaks are detected at m/z 91 and m/z 109. Table 2-4 shows

postulated reactions which may account for these product ions.

Table 2-4: Structural candidates of the product ions formed in the reaction of
the phenyl cation and VME

AHm(kJ/mol)
CeH5+ + CH2=CHOCH3 ----> C5H5C+=CH2 + CHgOH na
C6H5+ + CH2=CHOCH3 ----> C5H5CH2+ + HOCH=CH2 - 252
CsH5+ + CH2=CHOCH3 ----> (C(5H5OCH3)H+ + C2H2 - 174

na Heat of formation for CsH5C+=CH2 not found.

71

 

 

 

 

 

1m . Phenyl cation

§ 77

>~. m"

a I > X5
: -

:3 so 103
.g 40'

43

.2

r3 20‘ 91 109

- r *5] 1--- "n.--” ,2 :- ----.---- I: --!-- l I "II-

‘Tv vw'vrvv

3O 4O 50 60 7O 80 90 100 110 120

Figure 2-14: Product ion mass spectrum of the phenyl cation and vinyl
methyl ether at collision energy of 2 eVLab and a pressure of 1.2 mtorr.

72

6. Summary of reactions

A variety of neutral reagents are shown to be reactive with the phenyl
or napthyl cation forming a nucleophilic addition reaction product ion. In the
reactions described, the reaction channels accessed forming covalently bound
addition products were all exothermic. For the smallest homologues, such as
methanol, dimethyl ether, and methyl amine, simple bond cleavage reactions
forming an addition product with the elimination of a radical were the
dominant reactions observed. With more complex molecules, hydrogen
rearrangement reactions took place. For the analytical applications which
were sought for this nucleophilic addition reaction with aryl cations, dimethyl
ether was chosen as the reagent gas. There were two reasons for choosing
dimethyl ether over other neutral candidates. First, dimethyl ether is a gas
at standard temperature and pressure (STP) which makes it easier to handle
for introduction into the collision cell of a TQMS. Second, the product ion
which is formed, representing the addition of 31 u, is an isolated product ion,
that is to say, there is not a non-specific product ion with a mass in close
proximity to the mass of this methoxylated product ion. Ammonia satisfies
the first criterion in that ammonia is a gas at STP, however, not only is there
a product ion representing the addition of NH2, but there is the potential of a
non-specific ion forming at one mass unit greater than the covalently bound
ion. This ion could represent the formation of the non-specific ammonia
adduct ion. For analytical applications, it was desirable to have a 'clean'

region surrounding the peak representing the reaction product ion.

73
C. Screening for aromatics

As discussed above, the phenyl cation reacts with dimethyl ether to form
a methoxylated product ion. Additional experiments were performed to
determine the selectivity of this ion/molecule reaction. If this reaction is
selective for aromatics with a vacant charged site on an aromatic ring, then
perhaps a rapid screening procedure could be developed where aromatics
could be detected in a mixture of compounds. Under EI conditions, it would
be likely that aromatics, regardless of the substituents located on the ring,
would yield ions which satisfy the criteria established for a successful
ion/molecule reaction.

Initially, an experiment was performed which demonstrates that the
reaction is observed even with up to three substituents on the ring, which
might be expected to sterically hinder the ion/molecule reaction. Figure 2-15
indicates that the methoxylation reaction occurs for the (M-Cl)+ ion of 1 ,2,4,5
tetrachlorobenzene. The position of the chlorine atoms on the benzene ring
ensure that following loss of one C1, the charge on the ring will be on a carbon
atom adjacent to a carbon with a chlorine substituent. The 1 ,2,4,5
tetrachlorobenzene was introduced into the ion source via the GC, and the
(M-C1)+ ion was selected for reaction with DME. As shown in Figure 2-15,
when the ion with m/z 179 which represents CeH235C13+, is reacted with
dimethyl ether, a product ion is detected at m/z 210. If the ion of m/z 181 is
selected (C5H235C1237Cl+), the product ion is detected as a peak at m/z 212.
The presence of three C1 substituents on the aromatic ring does not seem to

hinder the reactivity of this €5H2C13+ aryl cation.

74

1(1)‘
39 A 179
§ 33- 35 +
2 06H2 CI3
eo-
2
'5 4°"
3 210
a: z)-

 

 

 

IW'"I'"I

0 I ' LI ' I ' I ' 'I ' l ' ‘ I '
160 170 180 190 200 210 220 230 240 250

m/z

1(1)‘ 1 1
s B 8
>. m“ 37 35 +
3:2 06H2 Cl Cl2
£60-
Q)
.2. 4)“
E 212
s ao-

o A ~va. . . - .

 

 

 

160' U 170‘ q 180 V 199 ' 200 210 220 230 240 250
m/z

Figure 2-15: Product ion mass spectra of the (M-Cl)+ ion of 1,2,4,5 tetra-
chlorobenzene with DME at 1 mtorr. Parent ion is A) C3H235Cl3+ (m/z 179),
B) C6H237Cl35C12+ (m/z 181).

75

Over the course of these studies, a variety of compounds were ionized
and ions present in the mass spectrum were selected for reaction with
dimethyl ether. These compounds included primary, secondary and tertiary
alcohols, esters, hydrocarbons, ketones and additional aromatic compounds.
A product ion peak 31 u greater than the parent ion mass was only detected
for aromatic compounds which contained a vacant charged site on the ring.
The product ion mass spectrum of the (M-H)+ ion of pyridine, an aromatic
without the benzene ring substructure, is shown in Figure 2-16. The
methoxylation product ion peak is seen at m/z 109 following the reaction of
the ion with m/z 78 and DME. The peaks at m/z 93 and m/z 124 represent
the protonated DME dimer, and the adduct ion of DME and the pyridine ion,
respectively.

An experiment was performed which serves to demonstrate the
selectivity of this ion/molecule reaction. A simple mixture, with components
listed in Table 2-5, was injected into the GC interfaced to the TQMS. Three
of the components are aromatic. Figure 2-17 shows the total ion current
chromatogram and the reconstructed mass chromatogram of m/z 77, following
ionization by E1. The phenyl cation with m/z 77 is indicative of aromatic
compounds, however, from this simple mixture, only one of the components
gives a peak at m/z 77. There is not an ion formed during EI which is unique
to these three aromatic compounds. If the mixture were more complex, it
would be even less likely that one ion would characterize the aromatics, yet
not be found in the EI mass spectrum of the non-aromatic components. To
determine which components are aromatic, it would be necessary to evaluate

each individual mass spectrum.

76

 

 

 

 

1(1) ‘
47 Parent ion

3° 3) -
g, 78
"c? m '
*3
H m u
109
*5 m- 45 93
3.} | 124

O Inn": !:'!::ift'annf'fln.-- !‘:- ".4"; --.-,----r.:w,fi-- 1-- --,----l-- n," "1-

40 5O 60 70 80 90 100 110 120 130

m/z

Figure 2-16: Product ion mass spectrum of the (M-HY ion of pyridine
(m/z 78) with dimethyl ether at a pressure of 2 mtorr and collision

energy of 1 eVLab .

77

Table 2-5: Components in mixture used to obtain chromatograms shown in
Figures 2-17 and 2-18.

1) o-xylene

2) 1 ,2 dibromopropane

3) 1,2 dichlorobenzene

4) 1 ,2,4 trichlorobenzene

5) 6—undecanone

78

.00 smacks“. acumen—E 8358 no a wagozom
.EquSmEEAu 26.56 :3 :33 Am 98 R. £8 mo Eahwoeafiouno mmwa wflogamaooom 2 fifia 953%

as: can. 8383

 

 

 

 

 

 

 

 

comm ooK comm oonm oouw comm
b.—.F..._b.bPrL»-un._..+-pn...rh_
1 l 11 O
-8
E E E -9
A:
'8 m
.SW.
9
l E can .SHW
i cf 7 4 o m
P
9
I8 m
-8
new
Rea < .9:

79

In contrast, aromatic compounds would be expected to yield at least one
ion in the EI mass spectrum which would contain a vacant charged site on
the ring. Regardless of the mass of this ion, it would be expected that the aryl
cation would react with dimethyl ether to form a methoxylation product ion.
A neutral gain experiment of +31 will provide detection of only those ions
which are able to undergo an addition of 31 11 following reaction with
dimethyl ether. Figure 2-18 illustrates the total ion current chromatogram
obtained in a +31 neutral gain experiment following injection of this five-
component mixture. Only the three aromatic compounds 1), 3), and 4), are
detected as peaks in the GC trace. As demonstrated, none of the ions in the
E1 spectrum of 6-undecanone or 1,2-dibromopropane will form a
methoxylation product ion 31 mass units higher than the reactant ion.
Electron impact ionization of 6—undecanone offers a slew of ions for reaction
with dimethyl ether, yet none undergo the methoxylation ion/molecule
reaction with dimethyl ether. Detailed structural information is not obtained
with this approach, yet this neutral gain experiment could be used to rapidly

identify aromatics within mixtures.

80

. a3 >o H mo 3.88 .8358 a 38 £38 #4 mo 9:533
a 3 a a E BS £5 :3 “assuage new 358: 89a 833585 snobs =3 33 a 93
.00 smack: @835qu magnum mo Hm ages...“ .EEMSmEoEo 39:3 :3 Eon. 2 ”a?“ 953'.—

AEEV 085. nouaoaom

 

 

 

 

 

 

% 801181311an OAQBIBH

 

U..----.oao.mm1.-h phloemfillwwmww- - - ”swoum ooze ooumo
L . 8

3 . 9.

. 8

- 8
A V 5+ Emu 3.58 Z m r 9:

... . ... . .tc .lcc..p..l._.4.-. .c
.8

E E E .9.

A: .8

- 8

. A8 . 0E. < .8“

 

 

81

D. Ion/molecule reaction of [C7H7+] isomers for selective

detection of the tolyl cation

1 . Introduction

Perhaps the most widely investigated series of ionic isomers in the gas-
phase are those with C7H7+ composition (105). Most of the attention has
been directed towards three of these isomers, the benzyl cation (112), the
tropylium cation (113), and the tolyl cation (114) which are shown in Figure
2-19. In the effort to distinguish these individual species, high energy
collisional activation (CA) mass spectrometry has been utilized with some
success (114-118). There are minor differences in the relative abundances of '
the fragment ions in the m/z 74 to m/z 77 region when the individual ionic
species are generated in the ion source and subjected to high-energy
collisional activation with a target gas. These variations in relative
abundances are indicative of the individual isomers and provide a
'fingerprint' of each ionic species.

Despite the success of the CA methodology, disadvantages exist with
this high-energy CA technique. The primary disadvantage being that the
capacity to distinguish these isomers is dependant on the reproducibility of
specified ratios of fragment ion abundances. The m/z values of the fragment
ion peaks in the daughter ion mass spectra are identical; there are no unique
daughter ion peaks which unequivocally characterize the individual isomers
using the high-energy CA approach. Figure 2-20 indicates the relatively
minor differences in the high-energy CA spectra of the benzyl cation and

tropylium cation.

.mNH 98 wow 805393 89a
5&5 “Bamako“ ma Baum .muofiomm Wm N.0 you 55538 mo 38: can mondfiabm Axum arm

3.3 gm 9% :CEBVEAV
ac A8 A:
Tfiofim Eamon Engage
. O
m

:0 NIw

83

 

 

 

 

A l A

1 1 1
("/1 74 75 76 77

.L.‘

1 1 1 1 '
ml: 74 75 76 77

Figure 2-20: CA mass spectrum of A) pure tropylium and B) pure benzyl
. ions. Accelerating voltage = 8kV. (Adapted from reference 117).

84

An alternative approach for analysis of C7H7+ ions has been to exploit
their unique intermolecular reactivity (1 O5). Ion cyclotron resonance (ICR)
mass spectrometers are usually chosen for these ion/molecule reaction
studies, although some work has been done utilizing high pressure ion
sources. Tropylium ions have been found to be non-reactive in the attempts
to utilize ion/molecule reactions as structural probes (105). However, both
the benzyl and tolyl ions have been shown to be reactive with a variety of
neutral reagent molecules, thereby permitting these ions to be characterized
via an ion/molecule reaction (11,110,119-124).

Heretofore, little effort has been directed towards differentiating these
isomers by low-energy collisions with a reagent gas in a triple quadrupole
mass spectrometer. The CID daughter ion mass spectra of these isomers are
nearly identical in the collision energy regime accessible with a TQMS.
However, based on the results described previously in this chapter, the
reactivity of the ions with neutral reagents was investigated in the attempt to
selectively detect one or more of these isomeric ions. Only the tolyl ion has
the positive charge located on a ring-carbon, therefore, it was expected that

this isomer would react with DME whereas the other isomeric ions would not.

2. Experimental

a) Instrumentation

All experiments were performed on a Finnigan TSQ-7O triple stage
quadrupole mass spectrometer. Following ionization of the aromatic
compounds by electron impact, the structurally 'pure' reactant ion of m/z 91
was selected by Q1 and passed to the collision cell. The reagent gas was

introduced into the Q2 collision cell via stainless steel gas lines at an

85
indicated pressure of 1E-6 to 8E-6 torr measured by a remote ionization
gauge; this corresponds to approximate Q2 pressures of 0.2 mtorr to 1.5
mtorr, as measured by a convectron gauge. Upon reaction with the neutral
vapor, the product ion mass spectrum was obtained by scanning the second
mass analyzer, Q3. This yielded a mass spectrum indicating the m/z for the
reaction products formed in the collision cell. The collision energy was always
1-2 eVLab which optimized the detection of the ion/molecule reaction product
ion. For the studies in which the electron energy was varied, the values are

reported as indicated by the Finnigan TSQ-7O data system.

b) Generation of ‘pure’ isomers:

A. Tropylium cation: ‘Pure’ tropylium ions were generated by 70-eV
electron impact of the tropylium tetrafluoroborate salt (118), introduced into
the mass spectrometer ion source via the direct insertion probe or by electron
impact of toluene at an ion source pressure of ~0.1 torr (114).

B Benzyl cation: The benzyl cation was generated by low electron
energy electron impact ionization of benzyl bromide (118). Benzyl bromide
was introduced into the ion source via a controlled variable leak valve, or via
the gas chromatograph.

C'. Tolyl cation: Tolyl cations were initially generated by methane
chemical ionization (CI) of 3-fluorotoluene (114). However, later in our
investigations, the tolyl cation was generated by low energy E1 of 3-
nitrotoluene (114,118). To generate a steady production of the tolyl ion of m/z
91, the precursor neutral molecules were introduced into the ion source via a

controlled leak valve.

86

c) GC/MS/MS analyses

The Finnigan TSQ-70 instrument was equipped with a Varian 3700
gas chromatograph which over the course of this study was used to introduce
a prepared mixture consisting of substituted aromatics which are known to
produce specific isomers or mixtures of isomeric C7H7+ ions. The mixture of
aromatic compounds (described below), was separated on a DB-5 capillary
column (30m x 0.25mm, J &W Scientific, Inc., Rancho Cordova, CA). Helium
was used as the carrier gas at a flow rate of 1 ml/min. The column was
operated with a splitless injector, and following an injection, the GC oven
temperature was held at 60°C for three minutes, ramped to 90°C at 4

deg/min, and then ramped from 90°C to 180°C at 6 deg/min.

3. Results and discussion

As described earlier in this chapter, selected hydrocarbons which
contain a charged vacant site on an aromatic ring react with various neutral
reagents resulting in an addition product ion. The success of these reactions
for detection of aryl cations led to the examination of the thermodynamic
requirements for selective detection of the tolyl cation among other C7H7+
isomers, and to test the hypotheses experimentally using this low-energy

ion/molecule reaction approach in a TQMS.

a) Thermochemistry
As shown in Figure 2-19 , the tolyl cation is the least stable of the three
isomers under investigation (108,125). The relatively large heat of formation

(1 25) for the tolyl cation contributes to the greater bond strength for [C7H7+-

X] for a variety of ring substituents (Table 2-6). Conversely, there is more

87

. WNH mug wOH wofiammmh A: 630% ficflfléom MG 350: 0E3 Us Ehufiwfl Scum $353030 mSHMGQHum 638 a

 

8m mum 3a a: - £088“
5. mum w; e: - $658“
an among m8 new - Zqu
5 among 82 K - Fx
m9. mum
wmm amoum 34 we - £23m
Sm m5.1.x 84 NS - £608“
in Manx Rm ”2 - mmoux
9% :53. mom 5 - flux
an $63. was SN - :Oux
NR «mommeUum
we. manna Km 3a 2 max
038 3235 3235 32:5
QBQG~02 ~Nha§¢z + 0 ©
mxo mzw
3&3 a 94555

e 38>.“— E 43:22 98mm "TN 038,—.

88

energy generated in formation of a bond with a ring carbon (as with the tolyl
cation) than with the methyl carbon located on the benzene ring (i.e., benzyl).
This thermochemical information aids in our search for a neutral candidate
which may selectively react with the tolyl cation. If the energy released in
the formation of the (C7H7+-X) bond is greater than the R-X bond strength
(D(R-X)) for the neutral reactant, the ion/molecule reaction is exothermic.
Therefore, a neutral reagent is sought for which the R-X bond strength is
greater than D((tropylium)-X) or D((benzyl)-X), yet less than D((tolyl)-X).
Satisfying these criteria would yield a thermodynamically favorable
ion/molecule reaction for the tolyl cation. It has been our experience in
investigating ion/molecule reactions in Q2, that ionic products formed by
ion/molecule reactions are observed only for thermoneutral or exothermic
reactions. In considering neutral reagent candidates, methyl iodide, for ‘
example, would not be expected to react with any of the C7H7+ isomers to

form C7H7I+-, because the reaction would be endothermic in all cases; the

. CH3-I bond strength is greater than the D(C7H7+-I) for the benzyl and tolyl

isomers, and based on the trend for R=H, the tropylium isomer. Upon
examination of Table 2-6, a few neutral species emerge as reactant
candidates which would be selective for the tolyl cation in an low-energy
ion/molecule reaction, including methanol, dimethyl ether and ammonia, all

of which were reactive with the phenyl cation.

b) Experimental results

To test our hypothesis that the tolyl cation will react selectively with
methanol, dimethyl ether, or ammonia, the individual 'pure' isomers with the
tolyl, tropylium or benzyl structure were generated in the ion source, and the

ion with m/z 91 was chosen for reaction with the reagent gas in the collision

89

cell. A product ion mass spectrum for each individual isomer upon reaction
with methanol was obtained, and only the tolyl cation gave a product ion
peak which was detected at m/z 108 representing the addition of -OH. This
same protocol was performed with dimethyl ether as the reagent gas, and the
product ion mass spectrum, corresponding to the reaction of the 3-toly1 cation
with DME at a pressure of 1mtorr, is shown in Figure 2-21. The dominant
ion/molecule reaction product ion is represented by a peak at m/z 122. Minor
peaks are observed at m/z 45 which represents (M-H)+ of dimethyl ether
formed by hydride abstraction, and at m/z 65 which represents a CID
fragment ion (loss of C2H2) from the tolyl cation. There are no product ion
peaks observed when the benzyl cation or tropylium cations, generated from
the appr0priate neutral precursor, are individually selected and reacted with
DME at the same collision conditions. These results are consistent with
predictions from the thermochemistry and suggest that either methanol or
dimethyl ether may be used as a reagent gas for selective detection of the
tolyl cation. In these studies of the ion/molecule reactions of C7H7+ isomers,
DME is the preferred reagent gas, primarily due to the ease with which DME
is introduced (DME is a gas at STP) into the mass spectrometer.

It is postulated that the product ion with m/z 122 formed by the
reaction of the 3-tolyl cation and DME represents an ion structure identical to
that of the molecular ion of 3-methyl anisole, as suggested by the proposed
reaction mechanism shown in Figure 2-22. This reaction is 57 kJ/mol
exothermic. Product ions formed by low energy collisions of the reactant ion
with a neutral reagent in the center quadrupole may be detected only if the
reaction is exothermic or thermoneutral, and if there is sufficient reaction
time. In contrast to the favorable thermochemical conditions for the

methoxylation reaction of the tolyl cation with DME, the reaction enthalpy of

90

h3>m H mo 3.55 5358 was .238 H mo
c.5395 a an MED fits H833 ~33 on... ma 85.38% mama com 835.5 "an.“ arm

 

 

58
OS. omw oow om cm 0? ON C
nod-bu bulb pink-hhhbuhbbhb- nbbnbh-n ub-bh-b-PPbe+bn—nnun-bbb-nb-ban-nnnbb O
mm
mv
ION
IO?
NNH
:3 33.95 :00
now
H@
5E3 38. .. on:

 

 

aouapunqv GAI'JBIGH

91

.pofio 138% can condo 33.. on» we now—ouch
on... 89G «NH £8 mo :3 gunman on... me 5358qu 3... .8.“ Sam—8:88 gunmen vomogoum ”Nufi awn

BEE E- u§I<

5: £8
«fl 3.: a... 8 NE.

mIOmw mIOO\\, 00 IO +
mI _m IO_ +

I0

92

the benzyl cation with DME to form a methoxylated product ion is positive
213 kJ/mol, whereas reaction with tropylium would be expected to be more
endothermic, due to the even higher stability of the tropylium cation.

Evidence for the structure of the reaction product ion with m/z 122 was
obtained by comparing the CID daughter ion mass spectrum of the product
ion with the CID daughter ion mass spectrum of the molecular ion of 3-
methyl anisole. The product ion with m/z 122 was formed when dimethyl
ether was introduced into the ion source simultaneously with 3-nitrotoluene,
the source of the tolyl cation. Under low-energy EI conditions, the initially
formed tolyl cation reacts with DME in the ion source to give the product ion
with m/z 122 which is then subjected to CID. The collision induced
dissociation daughter ion mass spectrum of the reaction product of m/z 122 is
shown along with the CID daughter ion mass spectrum of the molecular ion
of authentic 3-methyl anisole in Figure 2-23. The similarity of these two
daughter ion mass spectra provides good evidence for the structure of the
product ion being identical to that of the molecular ion of 3-methyl anisole, as
shown in Figure 2-22.

C) Analytical utility of the ion/molecule reaction approach

Having established that the ion/molecule reaction of these three C7H7+
isomers with DME is selective for detection of the tolyl cation, the analytical
utility of this reaction was evaluated. A simple mixture was prepared with
components listed in Table 2-7. Each component was present at an
approximate concentration of 20 ng/ul. One microliter of the mixture was
injected into the GC, and following ionization by 70-eV electrons, the ion
current at m/z 91 was selected and passed to Q2 which contained DME at a

pressure of lmtorr. Ifthe ion with m/z 91 had the tolyl structure, then a peak

93

100' A

80-

60 -
x4

 

92

20 - 79 107
9 l

J

Relative Abundance
h
o
‘—

 

122

 

1
'1' v

"'

 

l A
0 'V VVIVVVV‘IVYVVIVVV I'vvv II vmjvv‘rv‘vvvv'vvvv'v vvvrTvnI‘vvv

60 7O 80 90 100 110

100'

m
o
1

O)
O
1

x12

92
107

Relative Abundance
8
‘—

N
o
1

79 9

 

o
I

70 80 90 100 110
m/z

03
O

120

12

 

120

130

2

130

Figure 2-23: CID daughter ion mass spectrum of A) ion/molecule reaction
product of mass 122 and B) molecular ion of authentic 3-methyl anisole.
Collision energy was 25 eVLab at single collision conditions.

Table 2-7: Components in test mixture separated by GC
to obtain chromatogram shown in Figure 2-24.

Tolyl (%) 70 eV E13-
(CA results)

a) o-chlorotoluene 3
b) m-chlorotoluene 24
c) p-chlorotoluene 3

d) benzyl methyl ether NR
e) o-bromotoluene 3

f) p-bromotoluene 3
g) benzyl bromide NR
h) m-iodotoluene 40
i) m-nitrotoluene 57
j) p-nitrotoluene 70

3- results taken from reference 118.
NR, not reported.

95

at m/z 122 was anticipated in the product ion mass spectrum. The results of
this experiment are shown in Figure 2-24 which is a composite of the
reconstructed mass chromatograms at m/z 91 and at m/z 122. Coincidence of
peaks in the two mass chromatograms for h, i,andj suggest that at least

some of those ions with m/z 91 have the tolyl structure; this observation is in
agreement with recently published results as summarized in Table 2-7 (118).
Lack of reactivity for the ions with m/z 91 generated from the other
components to form an ion of m/z 122, suggests that an insignificant portion
of these ions have the tolyl structure under these ionization conditions.

The tolyl isomers which have been investigated include the meta and
para isomers, and their product ion mass spectra in reaction with dimethyl
ether appear identical. The ortho isomer was not investigated, as it is not
possible to generate ‘pure’ tolyl cations by electron impact from o-
nitrotoluene. However, it is not clear whether the three tolyl isomers are
even stable or freely interconvert (125). If the ortho isomer is formed, not
only may it interconvert to meta or para isomers, but rearrangement to the
benzyl ion which is 1.8 eV more stable, via a simple 1,3-H atom transfer,
seems a very likely process (125).

It is interesting to compare the low-energy CID daughter ion mass
spectra of m/z 91 representing two predominantly different ionic structures.
Figure 2-25 shows the CID daughter ion mass spectra of the ion with m/z 91
from 2-bromotoluene and 3-nitrotoluene, respectively. The collisionally
induced daughter ion mass spectra of these two isomeric ions appear nearly
identical at a collision energy of 25 eVLab, yet the difl'erence in reactivity of
these two ions with DME as evident from Figure 2-24 is great and can be

used to distinguish the ionic structures. The ions of m/z 91 were produced

96

 

 

53>». N .3 $.85 822:8 v8 .238 H mo 833.8 m 8 3095356 H883 on...
8 SEE 13383 at? Home 3 :3vo Hm 938 we com 05 v8 DU 0:» 38 cougar: ma? Hum mine 8 358mg
358388 mo 0.888 5:? NNH £8 Am 88 Ha £8 2 no 88803885 mmm8 wouoeumcoomm "vué. 0.53m

@3552 S «8? 8.55qu

 

 

 

 

 

 

 

 

 

 

 

 

003:. OOHNH oouow comm comm oouv
_ . . . _ . . . _ . . . h . _ . _
a 1 4 o
ION
V3
-2 m...“
a
low
NNH N): m
loop
- I o
_ now
-3. Wm
a ma
05.» mm
. Too w
a
u low
a s8
w
c B < IooH

 

97

$9? 81‘?

Relative Abundance

?3

 

 

 

 

Relative Abundance

 

 

 

A
91

91

 

Figure 2-25: CID daughter ion mass spectrum of mass 91 generated by 70
eV E1 of A) 2-bromotoluene and B) 3-nitrotoluene. Collision energy was 25

eVLab and argon was the collision gas at 0.3 mtorr.

98
under identical conditions for the experiments represented in Figure 2-24 and
Figure 2-25.

Another assessment of the selectivity and quantitative nature of the
ion/molecule reaction was conducted under conditions which isomerize C7H7+
ions in a known manner (118). It is known that during dissociative ionization
of 3-nitrotoluene, 'pure' tolyl ions are generated at low electron energy. As
the electron energy is increased, isomerization to the more stable benzyl
structure occurs in the ion source, prior to sampling in the collision cell. With
the straightforward ion/molecule analytical approach, at the very least, a
relative assessment of the tolyl composition present in C7H7+ mixtures can be
established. The ratio of I122/(I122-i-191). where In represents the intensity of
the ion current detected at m/z n, should mirror the actual tolyl composition
of C7H 7+ mixtures. The composition of the C7H7+ mixture from 3-
nitrotoluene as a function of electron energy was investigated by calculating
the 1122/(11224-191) ratio for three different pressures of dimethyl ether, and
the results are depicted in Figure 2-26. Each data point represents the
results of 40-50 averaged scans; this is necessary for good ion counting
statistics as the intensity of the reactant ion peak at m/z 91 is small at low
electron energies. The tolyl content reflected by the 1122/(1122+191) ratio,
decreases as the electron energy is increased from 12 eV to 20 eV. After 20
eV, the tolyl ion content remains relatively constant. These results suggest,
as expected, that the relative amount of tolyl cations in C7H7+ mixtures
generated by EI of 3-nitrotoluene is greater at low electron energy than at

high electron energy.

@8338 638%.: am .85» 3:888 fits con—cam: .Sm «G 38 Hm £8 “a 83.88 :3 mo 823888... 98
@538959m .Ho .889: 80.583 matsc 3.85 88803 We 80388 a ma Smémcgmfi .Ho 38m "@Nfi earn

$8 388m 888on
mm mm vm mm om 3 0H 3 NH 0 H

 

   

 

 

. . . _ . . _ p F L . P t . . 0.0
w .3 w
2 . V
0 Inl-
-v.o R
use :25 No 3 I H
6
use :25 F «3 HS .3 (

msE :SE v; «E «E .
we

 

100

d) Effect of higher gas pressure

As alluded to earlier in this chapter, the enthalpy of the reaction is not
the only factor which determines whether product ions are formed in the
second quadrupole with subsequent detection as peaks in the product ion
mass spectrum. The ion/molecule reaction kinetics also must be considered.
The absence of a peak at m/z 105 in the product ion mass spectra of the
C7H7+ cations with DME is noteworthy. The ion/molecule reaction of the
tolyl cation with DME to yield a methylbenzyl cation with m/z 105 and
neutral methanol, is more favored thermodynamically (Aern=-217 kJ/mol),
than is the methoxylation product ion of m/z 122. However, this peak at m/z
105 is very minor in the product ion mass spectrum as seen in Figure 2-21.
The thermochemistry also reveals, as shown in Table 2-8, that reactions of
benzyl and tropylium cations to form C3H9+ ions are favorable, yet product
ions with m/z 105 are not detected at a pressure of 1mtorr DME. If, however,
the pressure of dimethyl ether is raised to 3 mtorr, the product ion mass
spectrum of all three isomers gives rise to peaks at m/z 105 and at m/z 137 as
shown in Figure 2-27. The product ion peak at m/z 122 is still unique to the
tolyl cation. In all three cases, the product ion of m/z 137 represents the
intermediate collision-complex, (C7H7+-DME)+, which is stabilized at higher
gas pressures due to third-body collisions within Q2. The peak at m/z 105
represents the ion with CgHg+ composition, with suggested structures shown
in Table 2-8. Since higher reagent gas pressure is required, it follows that,
kinetically, the reaction to form the C8H9+ cation with m/z 105 proceeds
slower than the ion/molecule reaction of the tolyl cation and DME to produce
the molecular ion of methyl anisole. Raising the collision pressure increases

the lifetime of the collision-complex of m/z 137, as evident by the detection of

101

Table 2-8: Reaction enthalpies forming CgH9+ and methanol from C7H7+
and dimethyl ether.“

AHrm (kJ/mol)

(l) + CH3OCH3 ---------- > [(l)-H]-CH3+ + CH3OH - 7.6
(2) + CH3OCH3 ---------- > C6H§C+HCH3 + CH3OH - 83.6
(3) + CH3OCH3 ---------- > 3-C6H4(CH3)CH2+ 4' CH3OH - 217

a. Heats of formation found in references 108, 125 and 130.

102

60‘ Parent ion
9 1
45 137

Relative Abundance

 

 

 

I 1105

O "' rvaV'vvvvvvvvv'vvaIIvvv' 'U'V'It "t'vf‘v it" 'VVVVI ‘VV‘Tr‘ I

o 20 4o 60 so "100" 120 340

 

100‘ B 91
80‘
60-

137

40-

O """"V TY'r'Y‘V'hvvvv'v'v' """' vvrvvvvv ‘I'U'VV‘Vv‘V'Ivvv‘Vir'

0 2b 40 so so 100 120 140

Relative Abundance

 

 

 

 

100‘ C 91
80"
60-

40-

20- 137

I I 1 J -

o 20 40 so so "156" 1'26"" 140

Relative Abundance

 

 

 

Figure 2-27: Product ion mass spectrum of A) tolyl cation, B) benzyl cation,
and C) tropylium cation in reaction with DME at a pressure of ~ 3 mtorr
and collision energy of 1.5 evlab.

103
peaks at m/z 137, allowing access to slower reaction channels. Hydrogen
rearrangement is necessary in the sequence forming C3H9+ ions and neutral
methanol; this is a slower reaction than the tolyl-DME methoxylation
reaction, where only simple bond formation and cleavage are necessary to

form the products (Figure 2-22 ).

e) Reaction with ammonia

Based on the experimental results of the reaction of the phenyl cation
with ammonia, and examination of the bond strengths listed in Table 2-6, it
was anticipated that the reaction of the tolyl cation with ammonia would
yield a product ion with m/z 107, which would represent the addition of NH2.
However, as seen in Figure 2-28, this reaction yields a peak at m/z 93.
Neither the benzyl or tropylium cation reacts with ammonia to form a.
product ion with m/z 93. Once again the thermodynamics of the reaction
provides insight into the chemistry that is occurring. Following the formation
of the collision-complex, elimination of a methyl radical is thermodynamically
more favorable than is the elimination of a hydrogen radical. The enthalpy of
the reaction forming the ion with m/z 93 presumably representing the
molecular ion of aniline and the methyl radical is 33 kJ/mol exothermic,
whereas the reaction enthalpy to form the methyl-substituted aniline
molecular ion and H- is only 12 lemol exothermic. It does seem possible,
however, that this ion with m/z 93 may be a distonic radical cation (126,127)
as the reaction of NH3 occurs at the charged carbon, and the methyl radical is
eliminated at a different site along the ring.

In a study by Tabet (1 07), reactions of tolyl cations with ammonia
produced peaks at m/z 93 and m/z 92. It was suggested that the ion of m/z 92
was C6H4NH2+ (107). Ratios of I92 / 193 varied for the different ortho, meta,

104

 

 

 

1 OO- TOlyl cation
91
80-
8
g 60.
2
T, 40.
.2
E
d? 20- 93
65 108
o-vavvvvfijtrrv1-n-I-vv-I'vvrji'fiii"'Vrn" "'r'f.""':'l‘1‘"*‘:‘r
0 20 4O 60 80 100 120 140
m/z

Figure 228: Product ion mass spectrum of the 3—tolyl cation and ammonia at
a pressure of 1.5 mtorr and a collision energy of 2 eVLab'

105
and para isomers. However, it appears from their short report that the
authors assumed that the ion with m/z 91, generated from nitrotoluene, was
pure tolyl, yet this is not a valid assumption for ionization by electron impact
(118), especially in the case of o-tolyl. Also, it is likely that the o-isomer, as
mentioned earlier, could interconvert to either p- or m- isomers or undergo a
rearrangement leading to the benzyl cation (125). Both factors were
neglected, which could lead to erroneous results in the reported intensity
ratios. In the experimental work performed on the TSQ-70 mass

spectrometer, there was no product ion peak detected at m/z 92.

0 Conclusion

A novel approach for selective detection of the tolyl cation among other
C7H7+ ions has been developed based on its unique reactivity with dimethyl
ether. The results recommend the use of triple quadrupole mass
spectrometry for detection and quantification of the tolyl cation in isomeric
C7H7+ mixtures. The search should continue for an ion/molecule reaction
utilizing a TQMS which would be selective for the benzyl cation. For an
ion/molecule reaction to be thermodynamically favorable for the benzyl
cation, but not the tolyl cation, the relative stability of the reaction products
formed from the benzyl cation and the neutral reagent must overcome the
157 lemol difl'erence in the AHf of the tolyl and benzyl reactants, thereby
providing an exothermic reaction for benzyl, yet endothermic for the tolyl
cation. If these reactions can be accomplished in the center quadrupole of a

TQMS, then perhaps complete quantification of C7H7+ mixtures could be

made by this low-energy ion/molecule reaction approach.

106

E. Conclusion

These series of investigations demonstrate the reactivity of aryl ions,
those with a vacant charged site on an aromatic ring, with nucleophilic
reagents. In all cases, the thermochemical calculations correlate with
experimental observations. Although the thermochemistry may suggest
numerous reaction products, the more prominent ionic products observed
required only simple bond-cleavage and subsequent covalent bond formation.
Ionic products which require rearrangement are in some cases detected, but
generally at pressures which are higher than required for the addition
reaction.

Two analytical applications are explored which take advantage of the
this ion/molecule reaction. The first method described offers an approach for
detecting aromatics in mixtures. The presence of aromatics in our
environment poses as a real health concern (46,128,129) and this method
could perhaps allow aromatics to be identified as such in one GC-MS/MS
analysis employing an ion/molecule reaction. The approach takes advantage
of the unique reactivity of a class of ions where certain even-electron aromatic
ions always gain mass in their reaction with dimethyl ether. It would not be
possible to select a neutral loss characteristic of all aromatics. The second
application, more fundamental in nature, demonstrates a novel approach to
selectively detect the tolyl cation amongst other C7H7+ ions. The CID
daughter ion mass spectra of the three C7H7+ isomers using a TQMS are
indistinguishable, yet the reactivity of the tolyl cation with dimethyl ether

provides a way to probe for this cation in mixtures.

CHAPTER III

ION/MOLECULE REACTIONS OF ABSCISIC ACID METHYL ESTER
WITH MOLECULAR OXYGEN IN
ELECTRON CAPTURE NEGATIVE IONIZATION

A. Introduction

1. Abscisic Acid

Abscisic acid (ABA), a plant growth regulator, is ubiquitous in higher
plants and is also produced by certain algae and several phytopathogenic
fungi (131). Abscisic acid has multiple roles during the lifecycle of a plant
and the various functions of ABA are determined developmentally and
environmentally. A recent review thoroughly addresses the physiological and
biochemical responses of this plant hormone (131). Although the structure of
ABA, which is shown as the methyl ester in Figure 3-1, has been known since
1965, much of the details concerning the biosynthesis of this plant growth
regulator remained obscure. Research has focused on two biosynthetic
pathways which may account for ABA production (132). The first of these
postulated pathways is termed the direct pathway as this proposed sequence
involves the direct formation of a 015-precursor derived from farnesyl
pyrophosphate. The second proposed pathway is coined the indirect pathway.
It is postulated in this second pathway that ABA is a breakdown product of
large C4o-carotenoids, following on'dative cleavage reactions. Over the years,

the research group of Dr. Jan Zeevart, a Professor at Michigan State

107

108

 

 

 

Abscisic Acid Methyl Ester (ABA-Me)
M.W. = 278

Figure 3-1: Structure of abscisic acid methyl ester (ABA-Me)

109
University in the MSU—DOE Plant Research Laboratory, has been a pioneer
in this field and have unraveled numerous details concerning ABA
biosynthesis. Mass spectrometry has been a vital tool in their efforts,

providing valuable structural information during the course of their studies

(132-135).
2. Mass spectrometry of ABA

Electron capture negative ionization mass spectrometry (ECNI-MS)
has proven to be a highly sensitive technique for detection of the methyl ester
derivative of abscisic acid (ABA-Me). The biosynthetic pathway of ABA has
been investigated by conducting isotope-labeling studies with 1302 and
H2180, and analyzing the samples by mass spectrometry (132-134). ECNI
mass spectrometry has been used for the determination of both the
enrichment of 1302, and the site of oxygen incorporation in ABA; the latter
was possible by observing mass shifts of fragment ions following ECNI.
ECNI-MS is preferred over positive ion detection in mass spectrometric
analysis of ABA as under ECNI conditions most of the ion current is
concentrated in the molecular anion providing the isotopic enrichment
information. In addition, there are a few structurally-diagnostic fragment
ions which provide the structural information. Finally, the sensitivity using
ECNI-MS for this electrophilic compound is superior to that obtained with
positive ionization techniques.

Structures and mechanisms for the formation of fragment ions of ABA-
Me under ECNI conditions have been suggested by Netting et al. (136).
However, during the course of our studies, the appearance of certain

fragment ion peaks in the mass spectrum, specifically those at m/z 141,

110

representing the side chain fragment, and at m/z 152, representing the ring,
were quite erratic. Figure 3-2 shows two ECNI mass spectra of ABA-Me
taken at different times. In Figure 3-2A, intense peaks are observed at m/z
141 and m/z 152, along with several peaks in the m/z 260 region, while in
Figure 3-2B, only two fragment peaks are observed, those at m/z 260 (loss of
H20) and at m/z 245 (M‘---H - CH30H)‘ (136). The structures proposed by

Netting for the fragment ions, including those 'erratic' ions are shown in
Figure 3-3. After the analysis of many ABA-Me samples, it was realized that
a peak at m/z 310, (M+32)'-, is always observed in conjunction with the peaks
at m/z 141 and m/z 152. This peak at m/z 310, apparently representing an
adduct ion, was observed by Netting et al. when a capillary column was used
for introduction of the sample. The origin of the adduct was said to be
unknown, although aging of the GC column was offered as a possible cause
(136). However, an aging column may be dismissed as the source of the
adduct ion with m/z 310, as this ionic species of m/z 310 was observed not
only when a gas chromatograph (GC) was used for sample introduction, but
also when a direct insertion probe (DIP) was employed. Experiments also
ruled out that formation of the ion of m/z 310 was dependent on the buffering
gas used to thermalize electrons. The ion represented by a peak at m/z 310
appeared when either ammonia or methane was used. We suspected that
formation of this ion was due to the presence of oxygen in the CI gas lines,
since rigorous purging of the inlet lines eliminated the peak at m/z 310.
However, careful purging of the CI lines also eliminated the fragment ion
peaks at m/z 141 and 152. The spectrum, after rigorously purging the gas
lines, is shown in Figure 3-2B. The dramatic differences in the appearance of
these spectra led us to systematically investigate the role of oxygen in the

formation of fragment ions and adduct ions observed in the ECNI mass

100'- A

on
O
J

141

O)
O
l

152

N
O
l

 

Relative Intensity %
a.
O
n

 

o
L

120 140

100- B
so.
60«
404

20-

Relative Intensity %

 

0..
120 140

165

160

160

111

I
'OH C’ OCH3

245

 

278

310
$3

180 200 220 240 260 280 300 320

245

180 200 220 240
m/z

 

278

260 280 300 320

Figure 3-2: ECNI mass spectrum of ABA-Me obtained under conditions
in which: (A) no efl'ort was made to purge air from the CI lines and
(B) the CI lines were carefully purged free from contamination with air.

112

.mmH 8:888 5 .3 8
95.82 .3 6323.5 3 mE-<m< mo 32 Boga mo 89333 ”a.” 93.3,.—

HE E8 N3 EB
m. I o
fooxo
: / 0
I»: 0
own as
new as
mu. 0
mzooxo
/ /

 

113
spectrum of ABA-Me. It was crucial that the the formation of the fragment
ions be elucidated, as the biosynthetic pathway of ABA was inferred from the
mass spectrometric results, specifically, from the mass shift of fragment ions
under ECNI conditions. The structure of the fragment ions must be known,
eliminating the possibility of erroneous conclusions as a result of faulty ionic

structural assignments.

3. Reactions with oxygen in ECNI-MS

Reactions involving oxygen impurities in the ion source have been
reported previously (137,138), with most describing the use of oxygen-
enhanced negative chemical ionization for identification of specific congeners
of aromatic species (49,139-143). Recently, Stemmler and Buchanan ‘
described reactions of polycyclic aromatic hydrocarbons (PAHs) with oxygen-
containing impurities (137,138). Mechanisms for the formation of unusual
product ions were proposed. These studies of PAHs indicated that gas-phase
reactions were not responsible for all the observed product ions, but that
wall-catalyzed reactions can occur, and these processes were enhanced by
adding oxygen and elevating the ion source temperature (137). In many
cases, both ion/molecule reactions and wall-catalyzed reactions with oxygen
yield products such as (M + -O-2-H)' or (M + 20-2.H)', ions whose mass is
higher than that of the molecular anion. However, lower-mass fragment ions
also may result from source reactions following subsequent decomposition of
the initial reaction products (143). As demonstrated in a recent
interlaboratory comparison, rigorous control of the presence of trace gases is
required to minimize interfering reactions in ECNI and thus obtain

reproducible spectra of compounds at levels near the limits of detection (144).

1 14

The use of low-energy collisions in a triple quadrupole mass
spectrometer allowed us to investigate reactions of the negative ions observed
in the mass spectrum of ABA-Me with molecular oxygen. With selection of
the molecular anion by Q1, gas-phase ion/molecule reactions may be studied
in an environment in which the involvement of wall-catalyzed reactions, and
pre-ionization reactions between the analyte and neutral reactive species are
eliminated. Kostianen and Auriola used analogous reactions of M'- of
tetrachlorodibenzo-p-dioxins (TCDDs) with oxygen in the collision cell of a
TQMS for examining isomer-specific oxidative ether cleavage reactions
(93,94). In this chapter, the structure of some of the fragment ions formed
under ECNI conditions and the role played by molecular oxygen in inducing
their formation is examined. Interpretation of the fragmentation patterns of
isotopically-labeled ABA analogues allowed us to propose structures of these
fragments, and possible mechanisms to account for these unusual gas-phase
ion/molecule reactions. Structurally—similar ABA metabolites were also

analyzed, which allowed us to test the proposed mechanisms.

B. Experimental

1. Instrumentation

All experiments were performed on a Finnigan triple stage quadrupole
(TSQ-70) mass spectrometer equipped with a Varian 3400 gas
chromatograph. The samples were introduced through a DB-5 capillary
column (30 m x 0.25 mm, film thickness 0.25 n; J &W Scientific, Inc.) with He
as the carrier gas (flow rate 1 ml/min). The GC oven was programmed from

60-2100C degrees at 50°C/min, and 210-28000 at 12°C/min. The analyzing

1 15

quadrupole was scanned from m/z 50-330 at a scan rate of 0.5 sec/scan. For
samples introduced by the direct insertion probe, the probe was heated from
50-280°C at a rate of 20°C/min. For all the experiments, the ion source
temperature was maintained at 150°C. Methane or ammonia at an indicated
pressure of approximately 0.8 torr was used as the moderating gas, resulting
in the formation of secondary electrons available for capture by the analyte.
The electron energy was 100 eV, and the emission current was 0.3 mA.

Molecular anions were generated by electron capture in the ion source,
mass-selected by the first quadrupole, Q1, and passed to the collision
chamber, Q2. Argon or oxygen (1602 or 18Oz) was introduced into the Q2
collision cell at pressures of 0.7-1 .4 mtorr, as indicated by a convectron gauge,
which provides multiple collisions of the ion with the collision gas.
Interaction of the parent ion with the collision gas results in the formation of
ionic products which were analyzed by scanning the third quadrupole, Q3.
For low-energy reactions of anions with oxygen, the collision energy,
determined by the axial DC-ofi‘set voltage of Q2, was 1-5 eVLab. Higher

collision energies of 10-30 eVLab were used for the collisionally-induced

dissociation (CID) studies, with argon as the target gas.
2. Materials

All of the compounds analyzed were extracted and purified from plant
tissues by known literature methods (132,134). Labeled analogues were
either prepared by known literature procedures (145,146) or prepared
biosynthetically (134). In the latter case, confirmation of the location of the
isotope labels were made by EI mass spectral measurements (147,148). The

deutero methyl ester derivative was prepared with Tri-Deuter-8 TM (Pierce,

1 16
Rockford, IL). The author is grateful to Dr. Jan Zeevart and his graduate
student Chris Rock, for the preparation and provision of the samples
analyzed over the course of this study.

C.Results and dlscusslon

1. MS / MS with argon as collision gas

The preliminary tandem mass spectrometry studies on the molecular
anion of ABA-Me were done with argon introduced into Q2 as the collision
gas. Based on the report of Netting and coworkers (136), in which peaks at
m/z 141, 152, 245, and 260 were observed in the ECNI mass spectrum, it was
anticipated that CID of the molecular anion with m/z 278 would readily yield
these corresponding ions as fragments. However, as shown in the CID
daughter ion mass spectrum represented in Figure 3-4A, only peaks at m/z
245 and 260 were detected. Collision gas pressure and collision energy
adjustments were made in an attempt to increase the formation of the lower-
mass fragments, but these adjustments did not result in a significant
increase of the desired fragment ion peaks. We did observe, however, that at
low energies (<5 eVLab), ion current at m/z 141, on the order of 2% of the base
peak, was detected. At collision energies higher than 5 eV, no ion current at
m/z 141 was detected. This observation was attributed to inadequate purging
of the CID gas lines. Ifthe gas lines were not rigorously purged, air could be
introduced into Q2 as a contaminant in argon. It was thought that perhaps
oxygen present in the collision cell was activating fragmentation at low
collision energy. These observations, along with those stated earlier

concerning oxygen and the co—occurrence of peaks at m/z 310, 141, and 152 in

117

 

 

 

 

 

 

100-
A M"
80. 278
601
40-
5° $0
20
>. '1 245
43
To. 0- W
3 120 140 160 180 200 220 240 260 280 300 320
‘5
3100‘ B 141
'33
ca 80..
'3 M
m -.
60-
278
x25
40-1 Im3
152
20- 179(x5)
136 MI, 245 2a) 310
o ....!. ... ... ..,...... JrLV 4......
120 140 160 "I180 200 220 240 260 280 300 320

m/z

Figure 3-4: Product ion mass spectra (ELab = 2 eV) of the molecular anion
(m/z 278) of ABA-Me obtained with a TQMS following (A) CID with Ar

and (B) CID and/or ion/molecule reactions with molecular oxygen.

118
the mass spectrum suggested that processes other than CID were required to

induce fragmentation of the molecular anion.

2. MS / MS with 02 as collision gas

In order to investigate ion/molecule reactions of ABA-Me and oxygen,
02 was intentionally introduced into Q2 as the collision gas. The low-energy
collisions of M-- of ABA-Me with 02, as shown in the product ion mass
spectrum of Figure 3-4B, results in greatly enhanced peaks at m/z 141 and
152, along with small peaks at m/z 136 and m/z 179 which were not detected
in the conventional ECNI mass spectrum. There also are minor, yet
discernible, peaks at m/z 310 and m/z 293, presumably representing (M+02)'-
and (M+02 - ~OH)‘ species, respectively. Low-energy collisions with oxygen ‘
also promoted the loss of a hydrogen radical as evident by the increase in ion
current at m/z 277. As the collision energy was raised, the lower-mass
fragments, along with the peaks at m/z 310 and m/z 293 disappeared, while
the fragment ion peaks at m/z 260 and 245 became more intense. The
energy-resolved curves for product ions resulting from collision of the
molecular anion of ABA-Me with molecular oxygen is shown in Figure 3-5.
As indicated by Figure 3-5A, ions represented by peaks at m/z 260 and 245
are formed by direct decompositions of the molecular anion as the intensity of
these peaks increases with collision energy, while the product ions
represented by peaks at m/z 141 and 152 are formed as result of low-energy
ion/molecule reactions of the molecular anion with molecular oxygen. The
energy-resolved curves for ions with m/z 136 and 179 are similar to those
shown in Figure 3-5B for m/z 141 and 152. Contrary to what had been
previously suggested by Netting (136), the biologically active 2-cis isomer and

119

 

I/I(total) %

 

 

 

 

Collision Offset (eV)

 

I/I(total) %

 

 

 

 

 

Collision Offset (eV)

Figure 3-5: Energy-resolved curves in collisions of the
molecular anion of ABA-Me with molecular oxygen

at a pressure of 1 mtorr. (A) Ions with m/z 260 and 245,
and (B) ions with m/z 141 and 152.

120
the photoisomerized 2-trans isomer, which is formed in small amounts
during the extraction procedure, have identical fragmentation patterns.

To verify the role of oxygen in the activation process, 1302 was
introduced into Q2 and reacted with the molecular anion of ABA-Me to
produce the product ion mass spectrum as shown in Figure 3-6. A peak at
m/z 314 replaced the peak at m/z 310, confirming that the (M+36)'- peak
represents addition of molecular oxygen (1302) to the M'- of ABA-Me in the
collision cell. Quite unexpectedly, the peak at m/z 141 shifted to m/z 143, and
the peak at m/z 179 shifted to m/z 181. It follows that these product ions
result from initial reaction of the molecular anion with oxygen to form an
intermediate species, which subsequently decomposes with incorporation of
an 130 atom from molecular oxygen. The formation of fragment ions with
m/z 152 and 136 from the molecular anion requires molecular oxygen, but the
peaks at m/z 152 and 136 did not shift when 1302 was introduced into the
collision cell.

It is postulated that the ions with m/z 136, 141, 152, and 179 arise
from the highly unstable (M+32)'- species. Lack of sufficient third-body
collisions to offer stabilization of the Oz adduct of m/z 310 in the collision cell
allows decomposition to give these fragment ions. In contrast, collisional
stabilization can occur in a high-pressure CI source. When oxygen was
intentionally leaked into the ion source with the buffering reagent gas,
formation of the adduct ion of m/z 310 was promoted. A daughter ion scan of
m/z 310 following CID with argon was obtained as shown in Figure 3-7. The
daughter ion mass spectrum of m/z 310 clearly indicates that the ions with
m/z 136, 141, 152, and 179 along with the initial reactant anion M‘-, with m/z
278, arise from the (M+02)'- intermediate adduct species.

121

33>». N no .385
5858 m an a G 5 a Oma fins 80 8:38 28338 mBofiofiiom 9:333
Sum NE: mzémz mo :35» 33838 05 me 838on $9: :3 83.95 um.» 95.»?—

 

 

 

58

8m 8m 8m 8a SN cam 08 of cm: 8. cm.
up.».nun-hb-nnub-n-b-p-npunnnn:upunuu—pppnpnnpn—pp:-n pnnb-nnbbp

t . t _ 0 "d

_ _ — _ 9: a

mm as I

\ -8 n

65:: «9 M.

Ice 9

I

U

'8 m

u

S

-8 u.

.A

a: re: %

 

 

122

.0938 :3 05 3 «0 v5 32-4% .3 :3:m 53332: 33 me
3383.: 6:632:33 93:6 35:3 o5 a): .«o :3 3:33 05 .3 38:0 :33:8

 

5N

 

 

 

 

n3>m 3 a: :03: :33 BO 336:3 8:583 39: :3 839.3 23. 8.” 93.3%
N}:
own com owm 8N ovw omw cam of cow o: cup
pbbphpb-npflnnpnphbfibbuEbbbnflbp-pnppnphi-nuhhhnnptb-.--- n: nun-puppy. o
_ _ H
mfiN _ a a
_ .1
new NS .. cm 1.
8m I.
:5 w A
A a .. ov H
, 3x . W
I 60 9
u
9
u on n...
K
H13 l cow %

123

In an attempt to determine whether the structure of the ion with m/z
310 is a covalent species or a clustering adduct of M'- and 02, the ion of m/z

310, (M+1502)'-, formed in the ion source, was chosen as the parent ion and

introduced into Q2, which contained 1302 at a pressure of 1.8 mtorr. If the
(M+02)‘- ion were a non-covalent clustering species, it would be expected that
at low collision energy (~0.5 eVLab) exchange of 1602 with 1302 would occur

in Q2 resulting in formation of some ions of m/z 314, with subsequent
decomposition to ions of m/z 143 and m/z 152; that is, if some 1802 were
exchanged into the parent ion, one 180 atom would be expected to be retained
in the side-chain fragment ion. However, negligible ion currents at m/z 314
and 143 (less than 1% relative to that at m/z 310 and 141) were detected in
this experiment suggesting that the (M+02)‘° ion is predominantly a

covalently-bound species or, at the least, a tightly-bound adduct complex.

3. MS/ MS analyses of isotopically labeled compounds

The availability of stable isotope-labeled analogues of ABA allowed us
to re-evaluate the structure of the fragment ions of m/z 141 and m/z 152, and
to postulate the structures of ions with m/z 136 and m/z 179. It is evident
from these studies that the peak at m/z 141 represents the side chain portion
of ABA-Me as was originally proposed by Netting et al. (136). However, their
suggested structure and mechanism for its formation, where one of the gem-
dimethyls and one hydrogen from the 5' position transfers to the side chain,
are not correct. In the stable isotope labeling studies, expected shifts in the
mass of this fragment occur with a number of labeled-analogues including

both the deutero methyl-ester derivative and the [2H31-06 analogue (shifts to
m/z 144 from m/z 141). Similarly, the analogue containing 130 atoms in the

124

carbomethoxyl oxygens of the side chain displays a shift from m/z 141 to m/z
143 with one 130 atom, and to m/z 145 with two 180 atoms present. In
addition, analysis of the ethyl-ester analogue shows that the side chain
fragment ion peak is found at the anticipated m/z 155, as evident by Figure 3-
8 which shows the product ion mass spectrum of the molecular anion of ABA-
ethyl ester (MW=292) and 02. However, when the ring positions of ABA-Me
are deuterated ([2H3]-C7', [2H1J-C3', [ZHZJ-C5') or when either one of the ring
oxygens is labeled, the fragment ion peak at m/z 141 does not shift.

The ion with m/z 141 contains three oxygen atoms: two in the
carbomethoxyl group on the side chain and a third from the reaction with 02.
High pressure flow tube studies have shown that ground-state (triplet) .
molecular oxygen reacts with negative ions which contain a conjugated diene
system, involving incorporation of an oxygen atom, forming enolate anions
(149-151). However, these flow tube reactions with oxygen involve even-
electron reactant ions; the reaction we describe with ABA-Me is that of an
odd-electron molecular anion with molecular oxygen.

In order to account for the observed fragmentation and mass shifts for
the ion with m/z 141, the mechanism shown in Figure 3-9 is proposed.
Reaction of triplet oxygen and the radical molecular anion (a) occurs at the
radical site at C5 along the side chain of the molecular anion. Initially, an
unstable peroxy intermediate (b) is formed. The proposed mechanisms
suggests that this adduct with m/z 310 decomposes to transfer an oxygen to
01', forming a neutral radical fragment and the enolate anion of m/z 141 with
structure (c). This product ion of m/z 141, which contains one of the oxygens
from 02, has a highly resonance-stabilized enolate structure, as shown in
Figure 3-9, which resists further fragmentation. In collisions with argon at
an energy of 25 eVLab, the ion with m/z 141 yields a daughter ion represented

125

Sam aid :38 Enamémdw we :3:m 3:532: 33 mo 8:583 33: :3 838.5

. n3>e fl .«o .338 333:8 : c5 33:: H me 8:33: a a: 5998 3:832: an:

own com

van

Sm

Qinxfil;

 

0mm

«$N

com

ovm

a}:
CNN

cow oar

\

sic m3

09

38:3..—

ovw our

3...

N9

 

I
O

I
O
N

l
0
v

10m

l
C
no

% Kirsuequl 911118188

 

Ioo_.

126

.NO «:8 o2-<m< mo coma $3838 on» no assume»

2: wage—Eu d. as): we :3 05 mo sewage 23 you mcflgoooa Engage“: :ouoaou mammoEEoH 5.x 953m

IO

05 ES

2“ H 5; NE
0
w 0.

«8
MIOO 0 mIOO\
I I
O O

 

127
by a minor peak at m/z 109. Examination of the CID daughter ion mass
spectra of labeled analogues reveals that when one 130 atom is located in the
carbomethoxyl group of the side chain, the ion of m/z 143 yields fragments
upon CID with m/z 109 and m/z 111 in a 1:1 ratio. This indicates that the
fragment ion of m/z 109 results from expulsion of methanol (32 u) from one of
the equivalent (prior to methylation) carbomethoxyl oxygen atoms.

Although the ion with m/z 179 is represented by only a minor peak in
the product ion mass spectrum obtained during reaction of M‘- with oxygen
(Figure 3-4B), it is of interest due to incorporation of an oxygen atom from
collision with 02 as demonstrated by a shift in its peak to m/z 181 when 1302
is the reactant gas (Figure 3-6). Mass spectra of structural analogues of ABA-
Me containing 180 also show that the ion of m/z 179 contains both ring
oxygens. The fragmentation sequence of the intermediate ion with m/z 310
decomposing to the ion with m/z 179 proceeds by a two-step process as
represented in Figure 3-10. The conventional negative ion mass spectrum
shown in Figure 3-2A indicates a minor peak is found at m/z 293 which
represents loss of OH from the (M+02)'- of m/z 310; we postulate that the loss
of 'OH involves the distal oxygen of the peroxy moiety and a hydrogen atom
from C5'. The CID daughter ion mass spectrum was obtained for the ion of
m/z 293; it revealed that the ion with m/z 17 9 is a daughter of the ion of m/z
293 as the mechanism of Figure 3-10 suggests. As indicated in the product
ion mass spectrum resulting from reaction of the molecular anion with 1802
(Figure 3-6 ), the peak at m/z 295 corresponds to loss of a single 180 atom
from the intermediate (m/z 314). In this mechanism forming the ion of m/z
179 as shown in Figure 3-10, it is suggested that molecular oxygen reacts
with the charged site at C4, rather than at the radical site (C5), to form once

again, a peroxy intermediate of m/z 310, but with an alternative structure

128

 

.m: a): mo :3 35on 05 .8.“ wcsgooom Swansea" someone." magfioSEoH “3.» 0.33%

 

a at. N\E
«roof «3
+ E; 0 0
ID: I 10¢
mIo I / I z
I :
mam N\E
«E
:0.
05 NE. .0 m8. mum N\E ms.-<m<

10“

129
(b '). Following the formation of intermediate (b'), sequential losses occur
initially to form first the ion of m/z 293 ((1), then an enolate anion with m/z
17 9 (e) as illustrated in Figure 3-10. Two radical species are postulated as
the neutral products in the formation of the ion of m/z 17 9 from the ion of m/z
293 (d) . No ion current is detected at m/z 141 or at m/z 152 upon CID of the
ion of m/z 293 indicating that these product ions do not form via this pathway
involving the intermediate with m/z 293.

Formation of the ionic reaction products of m/z 152 (f) and 136 (g), as
shown in Figure 3-11, also requires activation of the molecular anion with
oxygen. However, these do not retain an oxygen atom from the adduct, as is
evident from the product ion mass spectrum with 1302 shown in Figure 3-6.
Analysis of the isotopically-labeled compounds lend support to the structure
of m/z 152 as was originally proposed by Netting et. al. (136), but formed via
the mechanism shown in Figure 3-11 which indicates the role of oxygen in
activating the fragmentation process. After initial formation of (b), m/z 310,
the peroxy moiety abstracts a hydrogen from the acidic hydroxyl group at Cl '.
Subsequently, the charged site initiates cleavage of the Cl'-C5 bond. The
oxygen-activated product ions of m/z 152 and m/z 136 always co-occur which
suggests sequential decomposition pathways from the common intermediate
of m/z 310 ( b). The ion with m/z 136 has the stable quinoid structure, (g),
which results from the net loss of CH4 (a methyl radical is eliminated from
one of the gem-dimethyls along with a hydrogen radical from the ring) from
the ion of m/z 152. This was confirmed by the CID daughter ion mass
spectrum of the ion of m/z 152. Analogous CID experiments with ring
deuterated ABA-Me ([2H31-C7', [2H1]-C3', [2H21-C5') indicate that the
hydrogen lost is from the ring, most likely the 5' position.

130

.«o e5 m2.<m< a. .2
mo sesame." moi—83823 on» Marga“ m2 can «3 £8 no 23 93 mo gangs.“ .8.“ Sagaooz "2.” 0.33%

E «8

 

131

The molecular anion of ABA-Me was the only ion in the mass spectrum
observed to undergo reaction with molecular oxygen. Other anions, including
the radical anion with m/z 260, which represents the loss of H20 from ABA-
Me, and the even-electron species with m/z 277 and m/z 245, were selected by
Q1 and reacted with molecular oxygen in the collision cell, but no reaction
products were observed. The anion with m/z 277 may not react because it
lacks the radical site necessary for attack by triplet oxygen. The ion with m/z
260, although a radical anion, lacks the 1'-OH which precludes hydrogen
abstraction by the peroxy intermediate, a process that would lead to the
formation of the ion with m/z 152 (see Figure 3—11). Stable isotope studies
indicated, in contrast to what had previously been suggested (136), that the
water loss does not abstract a ring hydrogen (at 05'), but most likely removes
a hydrogen from C5 of the side chain as illustrated in Figure 3-12. MS/MS
studies revealed that in the formation of the dehydration product, the
molecular anion with m/z 284 of ring-deuterated ABA-Me ([2H3]-C7', [2H1]-
C3', [2H21-C5') eliminates H20, rather than H2HO, resulting in a peak at m/z
266. In addition, the loss of H20 allows the charge in this fragment ion (1:)
to be more delocalized which may preclude the reaction of 02 at C5 to form
the ion with m/z 141. Finally, as might be expected, the anion with m/z 245,
which lacks a radical site and the 1'-OH group, also is not reactive with
oxygen. In these studies, the minor quantities of product ions, which were
generated in collision with 02, could be attributed to CID processes, as
similar experiments with argon in the collision cell resulted in peaks at
identical m/z values. This conclusion is supported by the fact that the
abundance of these fragment ions increased as the collision energy was

raised.

132

.mvm can 08 $8 at.» 9.8m Earn 3 35 was 02-422 mo :2 mo domuflaoawgm "a.” charm

new as
E

 

133

Parenthetically, the MS/MS studies of ABA-Me indicated that there
are two possible structures and mechanisms for the formation of the ion of
m/z 245. The first structure being that suggested by Netting et al. (136)
shown in Figure 3-3, which results from a primary loss of -H at C5' followed
by a concerted loss of CH30H from the tertiary hydroxyl and one of the gem-
dimethyl groups. A second structure (i), shown in Figure 3-12, results from
elimination of a methyl radical from the dehydration product formed by the
mechanism described above. The CID daughter ion mass spectrum of the
dehydration fragment ion of m/z 266 from ring-deuterated ABA-Me ([2H3]-
C7', [2H1]-C3', [2H2]-C5') indicated that a methyl radical may be eliminated
resulting in a fragment ion peak at m/z 251. This fragment ion with m/z 251
must contain all the ring deuterium atoms, including [2H1]-C5'. The CID
daughter ion mass spectrum of the ion of m/z 284 for this ring-deuterated
analogue results in a major fragment peak at m/z 250 due to loss of a
deuterium atom and methanol (Figure 3-3), and a minor peak at m/z 251 (i),
which represents the loss of H20 and a methyl radical. The results of these
CID experiments are consistent with the formation of the fragment ion of m/z

245 formed via two separate decomposition pathways.

4. Analysis of structurally-similar compounds

a) ABA metabolites

Several metabolites shown in Figure 3-13 were available for analysis
and analogous fragment ions requiring oxygen-activation also were observed
for these compounds. When the parent ion of abscisic aldehyde (M'- at m/z
248) reacts with oxygen in the collision cell, oxygen-activated fragment ions
result in peaks at m/z 111, 152, and 136. The ion with m/z 111 corresponds to

134

him 3%: 33. 0363... Belugas:

O:
m
10000
IO s

//..

8323308 <m< mo ouaaogum "an.” 0.53m

32-4% saw ~38: e5. 2825

 

 

135
the enolate anion containing the side chain portion of the molecule after
incorporation of an oxygen atom from 02. When 1302 is introduced into Q2,
the peak shifts to m/z 113 indicating its formation is likely from the same
mechanism that produces the ion with m/z 141 in ABA-Me. Ions with m/z
152, 136, and 179 also form during reaction with abscisic aldehyde, and they
are thought to be identical in structure to the ions represented in Figures 3-
10 and Figure 3-11. Likewise, when methylated phaseic acid was the
analyte, the parent ion of m/z 294 yielded oxygen-activated fragments with
m/z 141 and m/z 168, representing the side chain and ring portion of the
methyl ester of phaseic acid, respectively. The structure of the ion with m/z
168 is shown in Figure 3-14. When the molecular anion of trans-dial abscisic
acid methyl ester is reacted with 02, the predicted side-chain fragment ion
peak at m/z 141 is observed. The ring fragment ion of trans-diol ABA-Me is
represented by a minor peak at m/z 153. Apparently loss of -H from the
predicted ring fragment of m/z 154 results in a more stable even-electron
quinoid anion with m/z 153. Reaction of the molecular anion of these

metabolites with oxygen does not appear to be as efficient as that with ABA-

Me. Under identical conditions (02 pressure at 1.4 mtorr and collision energy
‘ at 2 eVLab), the intensity of the oxygen-activated fragment ion peak at m/z
111 from abscisic aldehyde, for example, is on the order of 30 percent of the
base peak, whereas with ABA-Me, the corresponding peak at m/z 141 is the

base peak in the product ion mass spectrum.

b) Analysis of a and B ionone
Over the course of the investigation of the fragmentation of ABA-Me, a

few additional compounds structurally similar to ABA were analyzed by

ECNI, and anions were selected for analysis with 02 in the collision cell. The

136

29.8: new

. NO and 3632» 3393 no .8an

83838 23 me 5338 «goofing—om 23 89a mm: a): 4.33 :3 mo magma—pom

“2.» 953m

wed £8 was a):

 

 

1 37

first two compounds analyzed were a- and [3- ionone, whose structures are
shown in Figure 3-15. Prior to analysis by ECNI, the purity of the
compounds was checked by El, and the mass spectra were in good agreement
with those reported in the literature (109). The ECNI mass spectra for both
a- and B- ionone with ammonia as the buffering reagent gas to generate the
thermal electrons are shown in Figure 3-16. Unlike the case with ABA-Me,
the base peak in the mass spectrum represents (M-H)‘, an even-electron
anion. When this ion was selected for reaction with molecular oxygen, there
were no product ion peaks obtained in the product ion mass spectrum.
Further investigation of these two isomeric compounds did reveal some
interesting results, not related, however, to oxygen-induced fragmentation.
The ECNI mass spectra of oc- and B- ionone with methane as the buffering gas

are shown in Figure 3-17. A striking difference is observed in the mass '
spectra if methane is used as the modifying gas. As seen in Figure 3-17, both
isomers show an unusually abundant MH- ion at m/z 193. In fact, with the B-

isomer, this unusual ion, MH', is represented by the base peak. The mass
spectrum of the a- isomer shows an intense peak at m/z 136 when methane is
used, whereas this peak is minor when ammonia is the buffering gas. These
isomers were introduced by the GC, and B—ionone eluted 20 seconds later
than the a-isomer. Introduction of the sample by GC ensured that the source
conditions were identical. Recently, it was reported that MH' ions were
observed in the methane ECNI mass spectra of chlorprothixene and aromatic
sulfur compounds (152). This MH‘ ion was accounted for by electron capture
followed by H- transfer from the reagent gas (152). This process likely
accounts for what was observed in the ECNI mass spectrometry of a and B

ionone, and is presented here as a curious phenomenon.

138

\

a—Ionone

B- Ionone

Figure 345: Structures of the isomeric forms of ionone (MW=192).

139

100' 191
so-
604

40-

20- 136
173 l

100 120 140 160 180 200 220 240
m/z

 

 

Relative Intensity %
o
l

.0.
o
o

J

191

CD
0
l

O)
O
l

401

 

N
O
I

Relative Intensity %

 

W

100 120 140 160 180 200 220 240
m/z

O
I

Figure 3-16: ECNI mass spectrum of (A) a-ionone and (B) B-ionone with
ammonia as the buffering reagent.

140

100-
A 136

191

m
0
l

O)
O
l

A
O
l

193

 

N
O
l

 

 

Relative Intensity %

O
I

100 120 140 160 180 200 220 240
m/z

100- 193
so—
60-

40"

 

20-

 

Relative Intensity %

100 120 140 160 180 200 220 240
m/z

Figure 3-17: ECNI mass spectrum of (A) a-ionone and (B) B-ionone with
methane as the buffering reagent.

141

c) Analysis of 2,4-hexadiene

A second compound which was investigated in this study was 2,4
hexadiene. The ion/molecule reaction of the molecular anion of ABA-Me and
oxygen is postulated to occur along the side chain which contains a
conjugated diene system. As mentioned above, reactions of even-electron
anions containing a conjugated diene system with oxygen in flow tube
studies, resulted in the formation of enolate anions (149-151). Perhaps
similar reactions could be observed in a TQMS to provide further insight on
the reaction of ABA-Me with oxygen. The ECNI mass spectrum of 2,4-
hexadiene showed the base peak at m/z 81, representing the (M-H)' anion.
Even though this ion is not a radical anion, the product ion mass spectrum of
the (M-H)' of 2,4-hexadiene and 02 at a pressure of 1.5 mtorr was obtained
and is shown in Figure 3-18. Peaks are seen in the product ion mass
spectrum at identical m/z values to those reported in the flowing afterglow
study when 2,4-hexadiene was analyzed (151). It seems likely that these ions
represent enolate anions, with reaction mechanisms shown in Figure 3-19.
These results indicate that it is possible to observe ion/molecule reactions of
even-electron anions with oxygen in the collision cell of a TQMS. Isomeric
carbanions were distinguished by ion/molecule reactions with oxygen in the
flow-tube studies; maybe a TQMS could be utilized in similar fashion for

probing the structure of anions.
5. Wall-catalyzed reactions
One oxygen-related ion that stands in contrast to the gas-phase

ion/molecule reaction products of ABA-Me and 02 is that with m/z 165. This
peak at m/z 165, as seen in Figure 3-2A, appears in the ECNI mass spectrum

142

 

 

 

 

 

 

5, 100— 81
53‘ so-

'53

G

0 50..

2 ’ x5

”:3 404 l V

>

'3 204 43 57 69

'1' 32 “I I I

q’ 0 I . I - _
pr, . . ,l, I , Him... “”1

Figure 3-18: Product ion mass spectrum of (M-H)‘ of trans, trans-2,4-
hexadiene (MW=82) with molecular oxygen at a pressure of 1.5 mtorr and a

collision energy of 1 eVLab.

143

5- 5- 5-
CH3— CH ZLCH :CH: CHIS-CH2 + 02

———> CH3— CH = CH —CH 0 + -O-CH=CH2

m/z 43
——> CHa" OH = CH O' + OHC-CH= CH2
m/z 57
———> CH3CHO + ' O---CH--CH-CH—CH2
m/z 69

Figure 3-19: Ion/molecule reactions of (M-H)' of trans,trans-2,4-hexadiene
in reaction with molecular oxygen to form enolate anions. (Adapted from
reference 151).

144

when oxygen-activated product ions are present, yet it is not observed in any
MS/MS study using oxygen or argon. It appears that this reaction product
may be attributed to wall-catalyzed source reactions (153). Examination of
reconstructed mass chromatograms shown in Figure 3-20 for the ions in the
mass spectrum shown in Figure 3-2A indicates that the ion current at m/z
165 lags in time behind that for other ions, including those ions which are due
to gas-phase ion/molecule reactions with 02. These observations are

consistent with the results of the study by Stemmler et al. where wall-

catalyzed reactions resulted in tailing in the mass chromatogram for the
mass of interest (137). Isotope labeling studies indicate that this fragment
ion contains the ring portion of the ABA molecule, including both ring oxygen

atoms, yet the structure remains undetermined.

D. Application

With the interpretation of the fragmentation pattern of ABA-Me
obtained under ECNI conditions following reactions of the molecular anion
with molecular oxygen, the 18O atoms may be readily assigned to the ring or
side chain of ABA-Me. In labeling studies of ABA with 1302, the conventional
ECNI mass spectrum provides the isotopic enrichment information, while the
MS/MS approach can be used to quantitatively determine isotopic enrichment
at the various positions of 18O-labeled ABA-Me from different plant tissues.
For example, in the MS/MS analysis of single 18O-labeled ABA-Me (m/z 280),
the presence of an ion with m/z 260 indicates that the 18O-label is at the OH-
C1'. If, however, ion current is detected at m/z 262, then the label is located
elsewhere, yet could be determined precisely from the pattern of the peaks
surrounding m/z 141 and 152.

145

100 . .*E+06
so . A m/z 141 4'446
60 -

2-trans isomer
40 -
20 . ] L

5‘

 

 

 

 

 

aw

 

1001 '*E+5
80‘ B m/z 165 7'0“
60*
40‘
20-

 

 

 

1 00 1 '*E+O5

6.836
80 ‘ C m/z 260
60 ‘

40*
20'

Relative Abundance

 

 

 

 

, l *an6
10° D m M" la :21

 

80‘
60‘
40‘
20*

 

 

 

 

 

 

740 750 760 770 780 790 800 810 820 830 84
Scan Number —>

Figure 3-20: Mass chromatograms at m/z 141 (A), m/z 165 (B).
m/z 260 (C), and m/z 278 (D) reconstructed from mass spectra
obtained during ECNI with sample introduction via the GC inlet.

146

The conventional ECNI mass spectrum of ABA-Me extracted from
plant tissue, which was stress-induced and then exposed to an 1802
environment, is shown in Figure 3-21. The need for an MS/MS approach is
obvious as a mixture of single-, double-, and triple-labeled analogues are
present. Provided that there is oxygen present in the ion source, the
diagnostic fragment ions are observed as peaks in the mass spectrum.
However, it would be difficult to determine the specific location of the isotopic
label. Figure 3-22 shows three MS/MS mass spectra of ABA-Me which
contains one 180 atom, but extracted from different plant tissue. In these
product ion mass spectra shown in Figure 3-22, the collision energy was 5 eV,
resulting in less intense peaks at m/z 141 and 152, while increasing the
abundance of the CID peaks at m/z 260 and 245. This figure serves to
illustrate the analytical utility of the ion/molecule reaction approach. The
relative abundance of ion current representing the diagnostic ions is different
in each spectrum, yet in all the spectra the parent ion is monolabeled ABA-
Me. The peak ratios provide information concerning the biosynthesis of ABA
in the respective tissue, pointing out the location of the isotopic label in the
labeled material containing a single 180 atom. This approach was used
extensively in the investigation of ABA with the results indicating that a
universal pathway exists for biosynthesis of ABA in higher plants (134). The
intent of this chapter is not to describe in detail the biosynthesis of ABA, but
rather to demonstrate how this novel ion/molecule reaction approach, once
the role of oxygen activation was elucidated, was useful in probing the

biosynthetic mechanism of ABA.

147

mzéfi 633$ Annoaoaomm mo 9338 a .«o 55.58% mama EOE 3N.» 933%

ES
CNN com omN omN OVN CNN OON 03 cm? 0: ONF

i
l
O

l
O
N

v5 3m
3%

 

l
O
V

l
C
(D

 

mum tom

 

 

% financial “1181821

I
O
O
F

148

 

 

 

 

 

 

 

 

 

 

100- 280
€30 - Apple Fruit
360‘
.5
340-
:g 141
320- 260
143 154 245 J262
0 ‘L A' ' 4'1 IA-l f L ' ' ' r j ' ' fl- I l ' A ' I
100 150 200 250 300
m/z
100- 230
5.80 .. Avocado Fruit
560.
.5
340- 262
as
3 20-1 143 247
53 141 152 245 260
o- -1 4- L- t al.,
100 150 200 250 300
m/z
100- 280
>330 . Avocado Leaves
:3
8
360- 143
.5
3404 262
:3
5' 7
20--
62 152 T L
o T‘fi ‘4. L- 1L-‘- 2 ‘- r‘fi "r -‘ -‘r -
100 150 200 250 300
m/z

Figure 3-22: Product ion mass spectra of mono-labeled ABA-Me (m/z 280)
with molecular oxygen at a collision energy of 5 eVLab .

149

E. Conclusion

The gas-phase reactions between molecular oxygen and odd-electron
molecular anions of the methyl ester derivative of the plant growth regulator
abscisic acid and several of its metabolites are unprecedented in that they
represent the first case where a non-aromatic radical anion reacts with
molecular oxygen, promoting fragmentation. It is postulated that the radical
site has a decisive role in directing the fragmentation process, as this is the
site of initial attack by 02. Mechanisms are suggested to rationalize the
formation of ions generated by the reaction with oxygen. These ion/molecule
reactions have a dramatic effect on the appearance of the mass spectrum, and
thus, could cause problems for the unsuspecting spectroscopist who operates
with an intermittent and minor air leak, and who relies on pattern
recognition for identifying compounds similar in structure to ABA. On the
other hand, ions formed in this process enable the site of oxygen isotope
enrichment to be precisely located within the molecule, an observation which
has provided invaluable insight into the mechanism of biosynthesis of these
compounds, as evidenced by recent publications for which this technique has
played a prominent role (134,135,154). In the case of these particular
compounds, CID-MS/MS with argon does not allow the position of the isotope
label to be determined precisely. Although low-energy ion/molecule reactions
in a TQMS have been used previously to investigate the structure of gas-
phase ions, these have been nearly exclusively in positive ion mode. This
chapter demonstrates the potential analytical utility of oxygen as a reagent
gas for possibly probing the structure of some anions, specifically radical

anions.

CHAPTER IV
COLLABORATIVE RESEARCH PROJECTS

A. Introduction

The opportunity for collaborations with other researchers provided the
author with a great deal of enjoyment. In certain cases, the combined
research efforts resulted in the completion of a project, while in other
instances, preliminary studies demonstrated some interesting chemistry
which may eventually lead to a full-fledged research topic for an incoming
graduate student. This chapter describes four collaborative projects in which
the author played a significant role, and these projects are summarized
below.

The first project involved a study which was undertaken with Kathleen
Kayganich, a former graduate student in the Department of Chemistry under
the direction of Dr. J .T. Watson. Her research involved developing
methodology for the determination of dexamethasone and other synthetic
steroidal drugs in physiological fluids (60). The novel aspect of this re search
involved the unique sample preparation of dexamethasone prior to mass
spectral analysis. Chemical oxidation was used to transform dexamethasone
to a highly electrophilic compound which may be detected by electron capture
negative ionization mass spectrometry (ECNI-MS). A collaborative effort was
made to determine if an MS/MS approach could provide an alternative
method for detection and quantitation of chemically oxidized dexamethasone

in human plasma by electron capture negative ionization (155). Although

150

151
this project did not involve the study of gas-phase ion/molecule reactions, it
was of interest to the author as it involved analytical applications of TQMS.

Unlike the project involving the analysis of dexamethasone, the second
project described was unique in that the genesis of this investigation was
begun at Michigan State University by Gregory Dolnikowski (106). However,
the culmination of this project which involved the study of a gas-phase
ion/molecule reaction between protonated acetaldehyde and methanol,
occurred following his departure. Greg continued his efforts in unraveling
this reaction mechanism while he was completing a post-doctoral assignment
in England. Dr. J .T. Watson studied this ion/molecule reaction on a penta-
quadrupole mass spectrometer while he was on sabbatical leave in France.
Finally, many experiments were carried out by the author at Michigan State,
which were important in determining the ion/molecule reaction mechanism of
this gas-phase ion/molecule reaction (156).

The final two projects which will be described, fall under the heading of
collaborations, as both studies were begun by a different investigator, yet
later, some experiments were performed by the author on the Finnigan TSQ
instrument. Both of these studies can be classified as being in their
preliminary stage, as much more work needs to be accomplished by future
graduate students. The first of these involves the reactions of protonated
esters and alcohols. This study was prompted by the reports of Greg
Dolnikowski describing the condensation reaction of protonated ethyl acetate
and propanol yielding an acetal ion (106). Some additional experiments were
performed mixing ethyl acetate (EtOAc) with other alcohols (Alc) in the ion
source and subjecting the observed product ion representing the

(EtOACtAlCtH - H20)+ ion to CID. The last section describes some

experiments in which protonated molecules were reacted with

152
hexamethyldisilazane (HMDS) in Q2 to determine if silylation would occur in
the gas-phase. Once again, the initial experiments of this type were carried
out by Dr. Watson while he was in France on sabbatical leave. Experiments
were done on the TQMS in the Mass Spectrometry Facility to determine if the
author could reproduce the results obtained in France where a penta-
quadrupole mass spectrometer was used. The results suggest that silylation
may occur in the gas-phase, yet future experiments need to be performed to

determine the selectivity of this silylation reaction for protonated molecules.

B. ECNI-MS/MS analysis of oxidized dexamethasone

1 .Intnoduction

It has been demonstrated previously in our laboratory that
dexamethasone, a synthetic steroid, can be oxidized chemically to a ketonic
steroid structure which can be readily detected by ECNI-MS (60,157). The
structures of dexamethasone and oxidized dexamethasone are shown in
Figure 4-1. The conventional approach for detection of dexamethasone in
biological matrixes has been to prepare the 11 ,17,21-tris-trimethylsilyl ether-
20-enol-trimethylsilyl ether (tetra-TMS) (158). By using this approach,
Kasuya et al. were able to detect an on-column injection of 100 pg of
dexamethasone-tetra-TMS with a signal-to-noise (SIN) of 2.5, from 1mL of
plasma spiked to a concentration of 300 pg/mL (158). On the other hand,
Kayganich et al. were able to detect an injection of 30 pg of oxidized
dexamethasone (S/N=11) from 1 mL of plasma spiked with 209 pg of
dexamethasone and 35 ng of internal standard (60). In addition, extensive

isolation procedures were not required in the sample preparation, as they

153

doBmExo 338on .3 osmofiqm 383%.: mam 3 285508556 no 565250 3% 0.33m

$83K F. P 3.893865 58:: F. F 3.8058865 Ao:9?o~.m.co_cmcaoa-v. 2oz.
{253372033 +.:Eos.83§=.s9 .Fqstfizfioeseéoése
cmcowmcummemn—c .mcommrszmxmos

29: F. F 7.585- a? 29.4. F. r 7359:- 59 ocowszmemo

,. (09.0 .

O

 

 

 

 

 

 

wer
oxiC
spe-
con
a b
up
ta
th

CO

SE

°(

154
were in the Kasuya method. This is attributed to the selectivity of the
oxidation reaction and the ionization by electron capture. The ECNI mass
spectrum of oxidized dexamethasone is shown in Figure 4-2. This spectrum
consists primarily of a peak at m/z 330 representing the molecular anion and
a base peak at m/z 310 representing the (M—HF)'- ion. It was expected that
upon CID, the molecular anion with m/z 330 would eliminate HF yielding a
daughter peak at m/z 310 and this was confirmed experimentally as seen in
the CID daughter ion mass spectrum shown in Figure 43. Few endogenous
compounds contain fluorine, so the loss of 20 u (loss of HF) appeared to be a
selective transition to monitor in a selected reaction monitoring (SRM)
analysis as interfering substances from a plasma matrix are unlikely to
undergo loss of HF. The attempt to further simplify the analysis of oxidized
dexamethasone by using tandem mass spectrometry, and perhaps shorten the
analysis time by introducing the sample by means of a direct insertion probe

(DIP), was investigated and the results are presented in this section.

2. Experimental

a) Instrumentation

All experiments were performed on the Finnigan TSQ-70 triple
quadrupole mass spectrometer which was equipped with a Varian 3400 gas
chromatograph. Samples were introduced into the ion source via a DB-5
capillary column directly interfaced to the mass spectrometer. Helium was
used as the carrier gas at a flow rate of 1 mL/min. The column was operated
with a splitless injector and, following an injection, the GC oven temperature
was ramped fiom 60 or 120° C to 260° C at 40 °C/min and from 260°C to 280

°C at 4 °C/min. For samples which were introduced via the DIP, after

relative intensity

 

155

 

 

 

310
100- _
(M-HF) '
80‘
60-
40. 330
M-.
20- I
0 ..-..:-........,. .-.1.1.-......WJ .L-. n..- .-- L-
280 290 300 310 320 330

an

Figure 4-2: ECNI mass spectrum of 11,17-keto-dexamethasone.

 

 

156

 

 

 

310
100- _.
(M-HF)
a 80‘ 330
g; r l
«2 60- x100 M-
E
g 40-
2 ..
0 280 295
0 1,-.l. WW - ......
280 290 300 31 0 320 330
m/z

Figure 4-3: CID daughter ion mass spectrum of the molecular anion (m/z
330) of 11,17-keto-dexamethasone.

157
insertion of the probe, the temperature was ramped from 35 to 200 °C at a
rate of 70 °C/min and held at 200 °C for 2 minutes. Regardless of the method
used to introduce the sample, (GO or DIP) the ion source conditions were
established to maximize the abundance of the parent ion, M‘-, for analysis by
MS/MS. The molecular anion of oxidized dexamethasone was at the greatest
absolute abundance when ammonia was used as the ECNI modifying gas at a
pressure of about 1 torr. The electron energy was 100 eV, and the primary

electron beam current was 300 “A. The ion source temperature was

maintained at 100°C.

b) Collision energy and pressure studies

The conditions for the MS/MS experiments were optimized by varying
both the collision gas pressure and the collision energy. Conditions for these '
parameters were sought which would provide for the greatest detection of the
daughter ion peak at m/z 310, following CID of the molecular anion. In order
to determine these optimal values, 3.56 ng of oxidized dexamethasone was
introduced via the direct insertion probe (DIP). The probe was held at 40 0C
for 1 minute and then ramped to 200°C at 30 °C/min and held at 200°C for 2
minutes. This temperature program yielded a level plateau region in the
desorption profile over which the total ion current (TIC) was relatively
constant. Over the course of this constant TIC region, repetitive daughter ion
spectra of m/z 330 were obtained at different collision energies. The parent
ion current at m/z 330 was selected by the first quadrupole Q1, and the third
quadrupole, Q3, was scanned from m/z 275 to m/z 345 at 0.1 sec/scan. The
full daughter ion mass spectrum indicated no fragments below m/z 275.

Consecutive daughter ion scans were acquired at collision energies of

1,3,5,12, and 25 eVLab; this cycle was repeated many times during the course

158
of the desorption profile. Five scans at each collision energy, collected during
the period of relatively constant parent ion current in the desorption profile,
were averaged to provide a representative daughter ion mass spectrum.
Replicate experiments were performed for each collision gas pressure. This
protocol permitted a reliable assessment of the fragmentation and collection
efficiency as a function of collision energy at a given pressure of collision gas

in Q2.

c) Selected reaction monitoring studies

For the determination of dexamethasone in plasma by selected reaction
monitoring (SRM), the ion currents corresponding to the transition of m/z 330
to m/z 310 (for oxidized dexamethasone) and m/z 339 to m/z 319 (for the
internal standard, oxidized 13C6,2H3-dexamethasone) were monitored with a
1.0 u window. Dwell times were 50 ms and the electron multiplier was set at

1400 eV.

(1) Methods

The sample preparation based on the oxidation method consisted of
three steps: plasma extraction, chemical oxidation, and removal of excess
oxidation reagent. The author was not involved in this aspect of the research,
and since the intent was not to demonstrate the value of the chemical
oxidation procedure, the details concerning these procedures are not

presented. These sample preparation methods are described in references 60

and 155.

159

3. Results and discussion

The goal of this study was to determine if an assay for oxidized
dexamethasone could be developed utilizing MS/MS. There have been reports
of improvements by the tandem mass spectrometry approach to the detection
of targeted compounds using ECNI with sample introduction by GC (59, 159-
161). Recently, a quantitative assay for leukotriene-B4 (as the
butyldimethylsilyl ether derivative) in synovial fluid was reported (56). The
selectivity of ECNI in combination with MS/MS, allowed quantitation of
lower levels of LTB4 in the matrix than would have been possible using
selected ion monitoring (SIM) with a conventional single-stage mass

spectrometer.

a) Optimization of MS/MS parameters
In order to determine the optimal parameters which would provide
maximum detection of the daughter ion with m/z 310, the experimental

protocol described above was performed. Two essential MS/MS parameters

affecting fragmentation efficiency, ZFi/(ZFi+P), and collection efficiency,
((2F1+P)/Po), where ZFi is the sum of the fragment ion intensity, P is the
remaining parent ion intensity, and P0 is the parent ion intensity without
collision gas, are collision gas pressure and collision energy. The experiments
were designed to dissect out interactive parameters of the CID process that
could be adjusted individually.

First, the ion collection efficiency was assessed as a function of collision
energy. The reconstructed TIC, shown in Figure 4-4A, is presented as a
function of collision energy at argon pressures of 0.4, 0.9, and 1 .3 mtorr in Q2.

When argon was present at these pressures in Q2, the collection efficiency

160

 

 

 

 

a)
25 -
P0 20 "
x 0.4 mtorr Ar
5%. 15 ~ /
8
91°“ 09mtorrAr
.3. .
E ‘/
v 5 ‘
(r3 \ 1.3 mtorr Ar
0 ' l ‘ l ' I ' l v fl
0 5 1o 15 20 25
b) Collision Energy. eV(Lab)
o 15 -
5 A
§ In
E lg
‘66 x 10 ..
E 22
9 g 0.4 mtorr Ar
5 8 /
O
c 2
.2 3 s -
E
“E, g 0.9 mtorrAr
g 1.3 mtorr Ar
n. o I ' T I ' l '
o 5 1o 15 20 25

Collision Energy. eV (Lab)

Figure 4-4: (A) Reconstructed TIC from CID of the M'- of oxidized
dexamethasone as a function of collision energy at different collision gas
(argon) pressures in Q2 . (B) Magnitude of daughter ion (m/z 310) current
as a function of collision energy at three pressures of argon in Q2_

161
decreased at collision energies above 3 eVLab. A similar attenuation of the
TIC at higher collision energy has been noted by other researchers (56). The

poorest collection of ions is observed at the lowest collision energy examined,

1 eVLab. This is attributed to the low kinetic energy of the parent ion. The
highest collection efficiency occurred with a collision energy of 3 eVLab for
each of the pressures examined. As is evident from Figure 4-4A, the
maximum TIC at 3 eVLab was approximately the same for the three
pressures of argon except at 1.3 mtorr, where there was slight attenuation of
the ion beam (about 15% ) due to higher collision gas pressure. N 0
reasonable explanation is offered for the TIC results from experiments with
no collision gas in Q2 being lower than those when Q2 is pressurized at
collision energies below 5 9VLab; the precision of results (+ or - 15% relative
standard deviation) is too poor. A very accurate assessment may be made in '
comparing the TIC at constant pressure, but different energy, as all these
values are obtained over the course of a single experiment and during the
nearly constant desorption region. However, in comparing different
pressures, separate experiments are required. Fluctuations in parent ion
intensity may occur as a result of (i) different desorption profiles, (ii) slight
variations in ion source conditions which may alter the ionization efficiency,
and (iii) error in loading the sample onto the probe tip.

The interdependency of various instrumental parameters involved in
an MS/MS experiment makes it difficult to determine what causes the
decrease in collection efficiency as collision energy is raised. Of the several
variables, greater collision energy increases the fragmentation efficiency, and
these daughter ions may be confined with varying degrees of efficiency in the
RF-only field of Q2 (162,163). Mass-discriminatory effects are generally

pronounced when the daughter ion mass is much lower than the parent ion

162
mass. In this case, however, the most abundant daughter ion of m/z 310 is
94% of the parent mass (m/z 330); therefore losses due to disparities in the
efficiency with which ions are contained in Q2 should be minimal. Alexander
and Boyd (163) have noted that for a hybrid BEQQ instrument, there are
variations in the transmission of ions as a function of collision energy. In
their study, the maximum and minimum transmission of protonated Leu-
enkephalin (MH+, m/z 556), without collision gas present, varied by a factor
of 2 over the collision energy range of 10-30 eVLab (163). Although this study
by Alexander and Boyd was performed on a BEQQ instrument, fluctuations
in the transmission curve for both parent and daughter ions as ELab
increases could account for the apparent losses of ion current in the studies of
oxidized dexamethasone using a TQMS. However, the constant parent ion
flux with Q2 empty (Figure 4-4A), which is quite independent of collision
energy, seems to indicate that the losses observed are not due to the
instrumental dependency reported by Alexander and Boyd. Thus, by default,
target gas pressure factors appear to cause the decrease in collection
efficiency as ELab increases. Scattering of both parent and daughter ions may
lead to lower transmission in the acceptance region of Q3 (22), but the tandem
quadrupole configuration is generally considered to be a good focusing device
for scattered ions, so scattering losses probably are not the major factor.
Finally, with increased pressure and higher collision energies, collision-
induced electron detachment (23) may compete with the CID processes,
resulting in formation of undetectable neutral species. All the factors
described probably contribute to ion losses, but electron detachment is most
likely the predominant factor causing the decrease in collection efficiency as

collision energy is raised.

 

“H

(n

163

The primary function of the optimization studies was to determine the
conditions for which the daughter ion current at m/z 310 is greatest. The
magnitude of fragment ion current at m/z 310 as a function of collision energy
is represented in Figure 4-4B. At the lowest pressure examined, 0.4 mtorr
argon, collision at 25 eVLab provides the greatest amount of (M-HF)'-. At
increased pressures, lower energies were required for maximal formation of
(M-HF)‘-; optimal energy for CID at 0.9 mtorr and 1 .3 mtorr was 5 eVLab, and
the ion current detected at m/z 310 for these conditions is greater than at a
collision energy of 25 eVLab at lower pressure (0.4 mtorr). More effective
transfer of kinetic energy to internal energy occurs with higher collision
frequency (23,162,164); thus, less translational energy is required at higher
gas pressures than at low pressure to induce similar fragmentation.
Secondary dissociation of (M-HF)" also may occur at higher pressures,
primarily due to successive losses of CH3o. This results in an increase in
abundance for ions with m/z 295 and 280 when the collision energy is raised
at argon pressures of 0.9 mtorr and 1.3 mtorr.

As a result of the optimization studies, a pressure of approximately 1.3
mtorr argon in Q2 and a collision energy of 4 eVLab were employed to ensure
both high ion collection efficiency and extensive fragmentation of the parent
ion of m/z 330, resulting in optimal detectability of the daughter ion with m/z

310 during the analysis of plasma samples.

b) SRM vs SIM

The improved selectivity of the GC/SRM analysis over GC/SIM is
shown in Figure 4-5, which is a composite of results from the analysis of a
plasma extract for dexamethasone. Shown in Figure 4-5A are the selected

ion current profiles for M'- and (M-HF)'- of 11,17-keto-dexamethasone (top

1((1‘1

164

3) Selected ion current profiles

        

W» («A-HF)"

‘ 1613-CH:3

 

 

 

 

 

 

 

 

m/z 330 -.
E" M
U)
C
49 M
.E
Q) : m/z 319
e . (M-HF)“
as
E ' “’03- 2H3- internal standard
‘- - A
m/z339 M"
”Cs ' 2 H3- internal standard

1.. 1L

'I'I'I']'IfI'I'I'U'I'U'I'I'U'I'I'I'I‘[V1

9 1o 11 12 13 14 15

time, min—>
b) Selected reaction product ion current

2‘ ; m/z 330 _. m/z 310
“(7) ‘
C ..
S d
s. A
g 1 ml: 339 -> M 319
*5 ..
'65 ‘ 1305- 2H3-internal standard
*- -l

'§""'1'o"‘*'1‘1""Harris?"'1'4 """ 1'5"

time, min —>

Figure 4-5: Comparison of selected ion current profiles obtained by
GC/ECNI with (A) selected ion monitoring (top two panels for oxidized
dexamethasone; bottom two panels for internal standard) and (B) selected
reaction monitoring (top panel for oxidized dexamethasone; bottom panel
for internal standard).

165

two panels) and 11,17 -keto- 1305,2H3-dexamethasone (bottom two panels); the

latter was used as an internal standard during analysis of this plasma
sample. The quantities injected on-column to obtain the data shown in
Figure 4-5A correspond to approximately 20 pg of oxidized dexamethasone
and 0.7 ng of the oxidized isotopically labeled dexamethasone. Shown in
Figure 4-53 are the selected reaction current profiles obtained from another
aliquot of the same sample as analyzed by GC/SRM with CID of the
molecular anions of oxidized dexamethasone (m/z 330), and the isotopically
labeled internal standard (m/z 339). Figure 4-5B demonstrates the inherent
selectivity of MS/MS as all the interfering ion current in GC/MS has been
eliminated by the MS/MS process. The only signal detected is that from the
11,17-keto-dexamethasone and its 163-CH3 epimer (peak at longer retention
time). This 16B-CH3 epimer is formed during the chemical oxidation
procedure. The absence of background makes the determination of peak
height or peak area straightforward and increases the precision of replicate
analyses. It is more difficult to determine the baseline position in
quantitation measurements obtained from the GC/SIM analysis.

The additional selectivity provided by the MS/MS approach, along with
the inherent selectivity of the assay resulting from the chemical oxidation
and ECNI detection, led to the investigation of the use of the direct insertion
probe (DIP) as an alternative method of introducing the sample. Other
researchers have reported that a chromatographic step in the sample
preparation is necessary in methodology based on ECNI/MS/MS when a
general derivatization reagent is used to prepare an electrophilic derivative of .
the sample (160). The chromatographic step is required because the electron
capture cross sections of many matrix components also are enhanced by the

derivatization. The electrophilic matrix components can deplete the

166

concentration of thermal electrons in the ion source (160,165). Without
sufficient chromatographic separation, the low level of analyte cannot
compete with the excess matrix for the pool of thermal electrons in the ion
source. However, as suggested by Figure 4-6 which shows the calibration
curves for the determination of dexamethasone in plasma by GC/SIM,
GC/SRM, and DIP/SRM, the chemical oxidation procedure provides sufficient
selectivity in the conversion of the dexamethasone to the electrophilic
oxidation product without enhancing the electrophilic character of the matrix.
The calibration curves are all similar. Therefore, the results obtained by
DIP/SRM are in good agreement with those obtained when the GC is used to
introduce the sample. The additional selectivity of SRM allows the direct
insertion probe sample inlet to be used for analysis of dexamethasone due to
the selectivity achieved by the chemical oxidation and subsequent ECNI. '
This significantly simplifies and shortens the analysis. Comparison of the
values obtained for the concentration of dexamethasone in plasma were in
good agreement for most of the samples analyzed by the three techniques
(155).

a)

A
O
p-q
m

u—‘r’

\
A
O
'—
m
v

H

I(33o + 310)/I(339 9 319)

O
v

I(330 -> 310)/I(339 b 319)

Figure 4-6: Calibration curves for the determination of dexamethasone in
plasma by (A) GC/ECNI/MS with selected ion monitoring, (B)
selected reaction monitoring,
DIP/ECNI/MS/MS with selected reaction monitoring. The quantities
introduced into the ion source represent a range of 4-200 pg of
dexamethasone and 700 pg of internal standard.

GC/ECNI/MS/MS with

0.8

0.6

0.4

0.2

0.0

167

 

IT'II'IIIT

 

 

11111I1111

 

 

is

10

15

N
0

ng/ml of dexamethasone

 

008 U I T

   

0.6 ‘

GC-SRM

   

0.4

0.2

   

 

 

l

 

o.o' ‘ ‘
0 5 1 0 15 20
ng/ml of dexamethasone

 

0.8 r u w

   

0.6 "

  

DIP-SRM

0.4

0.2

 

0.0

0

4

5

ng/ml of dexamethasone

 

1o

   

L

15

 

20

and (C)

168
C. Study of the gas-phase reaction between protonated acetaldehyde
and methanol

1. Introduction

The study of the ion/molecule reaction between protonated
acetaldehyde and methanol in a TQMS was begun by Greg Dolnikowski
(106). prrotonated acetaldehyde (m/z 45) is selected by Q1 and reacted with
methanol in the collision cell, a product ion mass spectrum is obtained as
shown in Figure 4-7, which shows the presence of a peak at m/z 59. This
experiment was performed on the Extrel TQMS and also on the Finnigan
TSQ-70, and the product ion mass spectra were similar. It is clear that
following the formation of a collision complex with W2 7 7, dehydration occurs
to form the ion with m/z 59. However, it was originally postulated that the
structure of this ion with m/z 59 was that of protonated acetone (106).
Further studies on three different types of mass spectrometers (BEQQ,
TQMS, and a penta-quadrupole instrument) were performed in pursuit of
determining whether C-alklylation (166,167) or the anticipated O-alkylation
(168-172) was involved in the formation of the covalently bonded product ions
(Figure 4-8). The culmination of this odyssey occurred following experiments
performed in England, France, and at Michigan State University.
Comparison of the CID daughter ion mass spectrum of the product ion of m/z
59 formed in the ion source, with known ionic structures at identical mass,
enabled us to determine the structure (156). This chapter describes the role
of the TQMS in determining the ionic structure of the product ion of m/z 59,

and in unraveling the ion/molecule reaction mechanism which accounts for its

formation.

169

 

 

 

 

 

 

. Parent Ion
100 45
80"
.é‘
m -
E), 60 x10
5 f 1
£40.. 33* 59
i3 65*
20'
47* 77 79*
‘I .1
o *f"'sh"'T"" ""l"""" 1'""""l"""':'l""""‘I""""‘l
3 40 50 60 70 80 90 100
m/z

Figure 4-7: Product ion mass spectrum of ions produced during
ion/molecule reactions of protonated acetaldehyde with methanol in Q2 at a
pressure of 2 mtorr. Those peaks marked by an asterisk represent known
product ions in the ion/molecule reaction between protonated methanol
(generated by proton transfer in this case) and methanol.

170

.mm «\8 mo :3 0%. we now—«Sue 2: 3m nofiawafimé .8 :oflflmiwd 3.1 953m

anus
I are
0N: + Vowzw
3.qu :0.\
an 5
ON... + Iflmlo
\0+

 

c2688.

Ucon
Em_m>oo

mm NE.

_H+z + 1095 + :flmzo
o

  
  

35:85
>__mco_m___oo

+1 + zofo + 10.010 II Iofo + 590

__
o
+
I\

__
o

17 1
2. Experimental

The experiments carried out by the author were performed on the
Finnigan TSQ-70 instrument equipped with a Varian 3400 gas
chromatograph. The ion source was maintained at 150°C; the ionizing
current was 200 uA at 70 eV. Ion/molecule reactions between molecules and
ions of methanol and acetaldehyde in the ion source were conducted by
introducing two microliters of both compounds (neat) through the GC inlet,
thereby providing transitory high pressure conditions as the two compounds
coeluted into a CI volume (without CI gas). The other chemicals used
throughout this study were either introduced into the GO, or introduced
through a fixed gas reservoir that was connected to the ion source via a valve
with a variable leak. In all cases, the parent ion of m/z 59 was selected by Q1 '
for CID in the collision cell, with argon at 0.1 mtorr. The CID daughter ion
mass spectra were collected with a collision energy of 25 eVLab. To obtain the
product ion mass spectrum of protonated acetaldehyde and methanol,
methanol vapors were introduced into the collision cell via a controlled

variable leak valve to a pressure of 2 mtorr and protonated acetaldehyde was

introduced at a collision energy of 1 eVLab.

3. Results and discussion

a) Formation of the product ion with m/z 59

The product ion mass spectrum shown in Figure 4-7 was obtained
when protonated acetaldehyde at low kinetic energy reacted with methanol in
the collision cell of the TQMS at a pressure of 2 mtorr. Product ion peaks at
m/z 59 and m/z 77 are observed along with peaks that may be attributed to

172

the proton transfer reaction to methanol (m/z 33), and peaks from subsequent
ion/molecule reactions in methanol (173,187). The product ion with m/z 59 is
not obtained if protonated methanol is allowed to react with acetaldehyde.
This is in contrast to results obtained on the penta-quadrupole instrument
when protonated methanol reacted with acetaldehyde; the ion with m/z 59 '
was observed if ion confinement was used (156). This is attributed to proton
affinity (PA) differences. The PA of acetaldehyde (780.8 kJ/mol) is greater
than that of methanol (761.0 kJ/mol) (108). In the TQMS, the only reaction
that proceeds when protonated methanol interacts with acetaldehyde is a
proton transfer reaction. However, in the penta-quadrupole instrument
using ion confinement which provides more time for reactions, proton
transfer may occur and the protonated acetaldehyde then may react with
methanol vapor which has leaked from the ion source into the collision
region. This sequence of reactions requires more time which is provided by
the ion confinement technique. A second possibility may be that the
formation of a proton-bound complex (m/z 77) generated from protonated
methanol and acetaldehyde may require transfer of the proton within the
collision-complex prior to forming the covalently bound species of m/z 59.
This transfer may be sufficiently slow (173), so that formation of the product
ion (m/z 59) occurs after an extended time, as is available with ion
confinement in the penta-quadrupole instrument.

The structure of the ion/molecule reaction product (m/z 59) was
characterized by CID. The ion of m/z 59 is formed when acetaldehyde and
methanol are mixed in the ion source and ionized by El, as is shown in Figure
4-9. The CID daughter ion mass spectrum of the product ion with m/z 59
formed when methanol and acetaldehyde were simultaneously introduced

into the ion source is shown in'Figure 4-10A. The CID daughter ion mass

100‘ 45#

 

 

 

173

 

 

 

 

 

77
2‘ 80‘
"’2 65*
E so-
0)
.5 89#
g 40-
o: 33*
20- 43f 47* 59 97*
79* 91
0 VA!‘ 5 V '_ I ‘__ - I- l v I- 1' 1' n
1*!t I l v I {ha ]"'vv'~vr~-~
30 4O 50 60 70 80 90 100

Figure 4-9: Reaction product mass spectrum of ions produced during
ion/molecule reactions following E1 of methanol and acetaldehyde in a high
pressure ion source of the TSQ-70. Those peaks marked by an asterisk
represent known product ions of methanol self-CI. Those peaks marked
with a # represent known product ions of acetaldehyde self-CI.

174

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100
2. A 15 590 1 1 Unknown
.. 29 x ‘ Reaction
_j I 31 43 Product Ion
0' f I -- .l. --
I I V WI I I I I I I I
0 5 1o ‘5 20 25 30 5 4o 45 so 55 so 65 i0
100
_4 B 15 59 CH30H=CCH3
: X0.09 and
i 27117343 CHZ=CHE§CH3
0. IV 'T ‘ ' I!I I Iv "1' 1' 'I H
o 5 10 15 2'0 is ab 35 30 45 50 55 so 5 70
100
J C 15 59 .
I 29 x 0-1 CH30H= OCH3
>. J .l31 43
H oqfi't v‘I ' 'V 'CJV‘V—vv I I fiva' I
a, 5 10 15 2'0 is 30 3'5 40 45 so 55 so 65 70
C100
0: " D 31 59
a 4 o
c : 43 "0'2 u +
_ . ll (CH300H3)H
:) o. I v_'V_I' vVV~I~ "fiI' "4 IIV'VI " " "v'
- o 5 10 i5 2'02153'0 35 4045505ss'0mfié5 7'0
100
E J E 41 59
c: J ”'43 (CH2=CHCH20H)l-l'
. 15 3‘
o.--" "5- W L--- H- -: l -- fil -5- ---- n" f--. ---. ----
I I I I I I I I I I I I I
o 5 10 15 20 25 30 35 40 45 50 55 so 65 70
100
j 29 31 59 +
: F 41 x0.07 0115;150:0112
oq""I" 'I'V'I "'I' rI 5"I "'I""I"1!I"' I'VVI'J' 'I""I'fiI
0 5 1o 15 20 25 30 35 40 45 50 55 so 65 70
100
. G 15 ‘59
i 29 X 0-13 CH30H=5CH3
I 31 43
O""'I"" 'Ifi" """""" J3II"fiIV'V'I"!'I""I"r'Ii" I""I'VI'I
o 5 10 15 2'0 2'5 30 35 4o 45 so 55 so 65 70
m/z

Figure 4-10: Daughter ion mass spectra obtained during CID in Q2 of TQMS
of a parent ion of mass 59 corresponding to: (A) product ion from high
pressure ion source containing methanol and acetaldehyde; (B) protonated
methyl vinyl ether (self-CI at lower pressure); (C) protonated methyl vinyl
ether (self-CI at higher pressure); (D) protonated acetone; (E) protonated
allyl alcohol; (F) fragment ion from diethyl ether following EI; and (G)
fragment ion from acetaldehyde dimethylacetal following EI.

175
spectra of several isomeric ions of m/z 59 were obtained with the TSQ-70 in a
comparative effort to identify the structure of the product ion with m/z 59
generated from the ion/molecule reaction of methanol and protonated
acetaldehyde. Isomeric C3H7O+ ion structures have been extensively
investigated (25,174-183). Ifthermochemistry were the determining factor in
the formation of the C3H7O+ product ion structure, then a variety of
candidate structures must be considered as is shown in Figure 4-11. A
composite of low energy CID daughter ion mass spectra of protonated forms
of methyl vinyl ether, acetone, allyl alcohol and fragment ions of diethyl ether
and acetaldehyde dimethyl acetal, is shown in Figure 4-10. Two spectra in
this composite, C and G match the daughter ion mass spectrum of Figure 4-
10A. Figure 4-10G is the CID daughter ion mass spectrum of the fragment
ion of m/z 59 formed by homolytic cleavage in the molecular ion of
acetaldehyde dimethylacetal following ionization by El, with the structure
shown in Figure 4-12. The results provide evidence that the product ion of
m/z 59 from the reaction of protonated acetaldehyde and methanol results
from methylation of the oxygen atom on the acetaldehyde. The structure of
the ion/molecule reaction product ion clearly is identical with this structure
represented in Figure 4-12, as seen by the similar CID daughter ion mass
spectra. Both Figure 4-10B and Figure 4-100 are the CID daughter ion mass
spectra of protonated vinyl methyl ether (VME), however, Figure 10C is
obtained with a higher source pressure of VME. At the higher pressure
conditions, there are sufficient collisions to drive the protonation reaction to
the thermodynamically favored (182,184) C-protonated product (structure
shown in Figure 4-12); at lower pressures, a mixture of O-protonated

(kinetically favored) and C-protonated ions is produced.

176

 

 

 

 

 

* H
A rxnk’J/mol.
H
. *9’ 1
CH30H=OH + 0113011 _. CH300H3 + H20 433.2
+ 1 2'
+ 1
> CH30H20H=OH + H20 732
1
> 0113011265=0112 + H20 -30.2
+ 1,2
t’ (CH2=CHCH20H)H «1» H20 -13.6

 

* 1 Reaction enthalpies calculated from ionic and neutral heats of formation found
in reference 108.

2 P.A. of allyl alcohol, estimated in reference 176, used in calculation of AH rxn.

Figure 4-11: Candidate structures for product ion with m/z 59.

 

177

mzolxonolmzolllm Iol

.300035086 05.300300“ .«0 :2 330038
0:» mo 093003 033080: .3 “008.8.“ an E8 «0 0.883% .NTv 056E

30000130086 093003000

are
3 55

0/2 .oImIo

178

Not only were CID experiments performed on the TQMS, but they
were carried out on BEQQ (in England) and penta-quadrupole instruments.
Although the data from these instruments are not shown, as the experiments
were not carried out by the author, it deserves mention that all of the
daughter ion mass spectra were remarkably similar. In spite of the different
source conditions and different collision conditions, the CID daughter ion
mass spectra were qualitatively in accord, and all the data point to the
structure of m/z 59 from the reaction of protonated acetaldehyde and
methanol being that shown in figure 4-12 which represents O-methylation of
protonated acetaldehyde.

Having established the identity of the ion with m/z 59, we wished to
determine the mechanism or mechanisms for its production in the high
pressure ion source and to determine the structure of the intermediate. A
proposed mechanism is shown in Figure 4-13. As shown by the mechanism, it
is postulated that there are two structures for the ion of m/z 77, each which
are formed at thermal energies in either the ion source or in the quadrupole
reaction chamber. The different location of the charge in protonated
acetaldehyde, the reactant ion, accounts for the formation of different ionic
structures with m/z 77. For the O-charged form of acetaldehyde, the proton
becomes "sandwiched" between the oxygens of the acetaldehyde and the
methanol. Upon CID, this proton-bound "sandwich" form falls apart into
either protonated methanol or protonated acetaldehyde (as shown in pathway
(i) in Figure 4-13). On the other hand, for the C-charged form of protonated
acetaldehyde, the intermediate involves close encounter of the oxygen atom of
methanol with the positively charged carbon in protonated acetaldehyde due
to electrostatic forces, and also close contact between the oxygen of

protonated acetaldehyde and the methyl carbon of methanol. This "close

179

 

4080508 88 0043023000 0308308 mo 85003 03003823 05 mausozom E. E8 m0 080608.808
33800830500 0:0 80¢ an E8 00 8: 08: mo 8030888 05 .8.“ 8883008 00030.5 "35 charm

00005.0an 5 020> ~20 mcficoamotoo 0cm O2 .0 26 020065
fi

 

 

 

E. #5
mm ~>c .
n 9 as 9. cs
Iozouofo Ilv n20100510 . .
+ + Imnxoszo Iofo + Imuxofo
. o
of 20 :0 :0 0.0 5 A v
2: .8 NE
A vs: I v no 56
.255 .255 1 + a.”
k as k 5:. 01:05.10 1m :0
J. I. _._I \a \I A7
£0010 \1 510-010..
\ _ 00760:. \ _
a F +1 0.0
_ i A—~
I
EB. 203650.. 00.08.90
EB. 00.2290 3:952 I
.0an80 000.0. 9.00 E2960 an ~>5 lo lfo
1’ C R E as... o
I. I
.35 .55 I m f0 .7032 +
R ~>t .. \
\/010 f0
£0 + .. I

1M0? I

I ..
£0\+ .. f0

180

encounter" form of the intermediate permits covalent bond formation (as
indicated in the middle of Figure 4-13), thereby producing a protonated
molecular species with m/z 77. This ion of m/z 77 decomposes promptly by
elimination of methanol to form an ion with m/z 45, or rapidly undergoes
hydrogen transfer (as indicated at the right of Figure 4-13) and abruptly
eliminates water to form the product ion with m/z 59. The CID daughter ion
mass spectrum of m/z 77 is shown in Figure 4— 14A, and supports the validity
of the proposed reaction mechanism. At low eV collision energy, some of the
ions with m/z 77 expel water to form a daughter ion with m/z 59. However, at
collision energies above 5 eV, no peak is observed at m/z 59. The "close
encounter" adduct which leads to the covalent bound formation yielding the
ion of m/z 59, may be interrupted before covalent bond formation can be
accomplished, if the collision energy is greater than 5 eV. This kinetic energy
should readily disrupt the electrostatic forces holding two of the four critical
atoms together.

The majority of the ions with m/z 77 exist as the traditional proton-
bound adduct of methanol and acetaldehyde as the dominant CID daughter
ions (Figure 4-14A) are protonated methanol (m/z 33) or protonated
acetaldehyde (m/z 45). Isot0pic labeling studies helped clarify the reaction
mechanism. With the introduction of 18O-methanol and acetaldehyde in the
ion source, the intermediate shifts to m/z 79. The CID daughter ion mass
spectrum of the ion with m/z 79, shown in Figure 4- 14B, indicates that this
collision complex intermediate loses CH3OH to give a peak at m/z 47
representing protonated 18O acetaldehyde (pathway (ii) of Figure 4-13). The
proton-bound adduct form of the intermediate still generates a daughter ion
of m/z 45 upon CID as demonstrated in Figure 4-14B. After hydrogen
transfer within the intermediate, the ion of m/z 79 may eliminate H2130 upon

181

 

 

 

 

 

 

 

 

 

 

100
J I\ 45
g .-
g d 77
.9 x50
.5 -
:5 w 59
m -
a:
J 33 'fi5
c IIII IIII IIII IIII IIII IIII II III IIIIlIIIIIIIIIlI III IIIIWII IIIIIIIIIIII
10152025303540650556067075808590
an
100
. 45
B
.2. .-
§ . 79
g cl
3 .
.5
g -
. 35 J7
c I III—IIIII IIII IIlI—III II III IIfiTIII III III IIIIIIIIII III] II III I IIIII

 

 

"l"' '“l'l
1o 15 20 25 so 35 4o 46 so 55 60 65 7O 75 so 85 90
an

Figure 4-14: (A) CID (ELab = 4 eV) daughter ion mass spectrum of the
adduct ion of m/z 77 formed by ion/molecule reactions occurring after
methanol and acetaldehyde were placed in the high pressure ion source.
(B) Daughter ion mass spectrum of m/z 79 formed when CH3130H and
acetaldehyde were placed in the ion source.

182

CID to give the reaction product ion of m/z 59 (pathway (iii) of Figure 4-13).
The peak at m/z 65 in Figure 4- 14A appears to represent a proton-bound
dimer of methanol due to leakage of methanol into the collision cell from the
ion source. This is supported by the fact that the peak shifts to m/z 69 when
18O-methanol is admitted into the ion source (Figure 4-14B). It is postulated
that a molecular beam of methanol dimers (185,186) from the ion source
enters the collision cell where it captures a proton from the intermediate of
m/z 77 to form a protonated dimer of methanol.

A parent ion scan of m/z 59 was obtained following the introduction of
acetaldehyde and methanol into the ion source, to determine if there were
other intermediates which may account for the formation of the product ion of
m/z 59. The results indicated the ion with m/z 59 arises not only from the
intermediate of m/z 77, but from an intermediate with m/z 91; this second '
reaction sequence is illustrated in Figure 4-15. Ion/molecule reactions in
methanol have been studied previously, and one of the ionic products is
protonated dimethyl ether (173,187 ). Protonated dimethyl ether reacts with
acetaldehyde to give a collision complex of m/z 91, some of which represents a
protonated acetaldehyde dimethyl acetal intermediate. The CID daughter
ion mass spectrum of m/z 91 at low-energy collisions with argon in the TQMS
gives a daughter ion of m/z 59, which is accounted for by pathway (ii) or (iii)
of Figure 4-15. When 180 methanol is introduced into the ion source, the
peak for the intermediate shifts to m/z 93. The CID daughter ion mass
spectrum of m/z 93, shown in Figure 4-16, indicates fragment peaks at m/z 59
and m/z 61 in a 1:1 ratio. This corresponds to loss of CH313OH and CH30H,
respectively, as would be expected statistically from the equivalent structures
in equilibrium through proton transfer, as indicated in Figure 4-15. This
daughter ion spectrum also shows a large peak at m/z 49 which represents

183

00.303380 080 380508 00 0.8058 0 8 80.0.8000 0803000.. 00.00.0880. 0:... 0.8.30.3.
S 08 00 000.008.0008 0:... 80.... mm 08 00 :0. 0:... 00 8050800.. 0:... 00.. 80800008 0000005 "3.0 0.83m

0000500000 E 00_0> ~\E 05000000000 0:0 0 0. .0 0:0 00.00.05
*

0I0I0 "W05

n.

.A .0 NE. 0:0 8 ms 0:05.005

I00I0 Iofo
2.: 90
.80 NE.
.0 as
I
_ I \
0 I00 .040» \
\ _ 00.000;
.. +I

MUI0

0:0

 

 

90 2:

ouI00I0

+
6005.
.. D. NE. 010wa

I

 

 

00:00:00 I
000m 2.2050 ....0

...00000 0500-00.90.

184

 

 

 

 

100‘ 93
80"
2:
'55
c
93 60‘
E 49
3
E 40-
o)
(I:
20‘
59 61
o -2--- ---- ---- --- ---- ---l. !--- 0--- -0“ --:- ---- ---- 5- - ----
I ‘ l ' I ' l ' l ' I ' I ' l
30 40 50 60 70 80 90 100
m/z

Figure 4-16: CID (ELab = 2 eV) daughter ion mass spectrum of m/z 93

formed by ion/molecule reaction occuring after acetaldehyde and 18O
methanol were introduced into the ion source.

185

the protonated 18O-dimethyl ether. This daughter ion spectrum is consistent
with the proposed mechanism. If the ion with m/z 93 only represented a
proton-bound adduct consisting of m/z 59 (formed in source) and 130-
methanol, then the daughter ion spectrum would lack a peak at m/z 61
representing the 18O-labeled product ion. The presence of m/z 59 and m/z 61
is anticipated due to the H+ transfer that can occur prior to the expulsion of
methanol (or 18O-methanol). When [2H4]-methanol is used, the intermediate
of m/z 98 upon CID yields the product ion of m/z 62, which is consistent with
the proposed mechanism.

b) Reactions of acetaldehyde and dimethyl ether

Some further experiments were performed in the continued effort to
verify this mechanism accounting for the formation of the ion with m/z 59.
Dimethyl ether and acetaldehyde were simultaneously introduced into the
ion source at approximately equal pressures totalling ~0.6 mtorr to determine
if the product ion (m/z 59) is detected in the mass spectrum. The mass
spectrum of this mixture is shown in Figure 4-17. The peak at m/z 59 is
observed in this spectrum, and again the mechanism shown in Figure 4-15 is
postulated to account for its formation. The ion with m/z 91 could have two
structures, the first resulting from ion/molecule reactions in dimethyl ether
(188) (m/z 45 cluster to DME), and the second representing the protonated
DME-acetaldehyde species. Before the CID daughter ion mass spectrum of
the ion of m/z 91 was obtained with both DME and acetaldehyde in the ion
source, the daughter ion mass spectrum of the ion with m/z 91 formed with
DME only in the ion source was obtained. There were two fragment peaks
detected, the major peak representing the oxonium ion with m/z 45 (DME-
H)+, and a peak at m/z 61 which represents the elimination of CH20.

 

186

 

 

 

 

 

100-
47
d 45
> 8° 43
a:
U)
5 60-
.E
3
E 40" 91
o
m 93
20‘
o 'I"""""T‘ ' ‘ E'I""""!l5"""" f""":‘l""I'lflrfi'g“f"l
so 40 so 60 70 so 9 100
m/z

Figure 4-17: Partial reaction product spectrum of ions produced during
ion/molecule reactions following E1 of acetaldehyde and dimethyl ether in a

high pressure ion source.

187
Following this experiment, the daughter ion mass spectrum of the ion with
m/z 91 was obtained when DME and acetaldehyde were mixed in the source.
Not only were peaks detected at m/z 45 and m/z 61, but peaks were observed
at m/z 59 and m/z 47. The ion of m/z 59 has the structure shown in Figure 4-
12, and m/z 47 represents protonated dimethyl ether. Based on these
additional experiments it seems likely that the ion/molecule reaction
mechanism postulated in Figure 4-15 accounts for at least some of the

formation of the covalently bound product ion of m/z 59.

c) Reaction of benzaldehyde and dimethyl ether

The results described above indicate that O-alkylation is occurring in
the reaction of protonated acetaldehyde with methanol (and protonated DME
with acetaldehyde). For all the reactions described, the molecular weights of '
the compounds were similar, which can make it difficult to determine the
reactant species in the formation of higher-mass product ions. Product ions
may represent a mixture of ionic structures. A system where the difference
in molecular weights of the reactant species is great may facilitate
determining the reactants involved in formation of product ions as the
possibility of different ionic structures of identical mass is reduced. This led
to the investigation of benzaldehyde and dimethyl ether. Would an O-
alkylated product ion be observed in the ion/molecule reactions occurring in
mixtures of these components? Initially, benzaldehyde and dimethyl ether
were mixed in the ion source, and the reaction product mass spectrum was
obtained. This spectrum (with the self-CI peaks from DME subtracted) is
shown in Figure 4-18. Of interest is the peak at m/z 121 which represents
[(benzaldehyde-DME)H - CH30H]+. The CID daughter ion mass spectrum of

the ion of m/z 153 was obtained to determine if this ion, representing the

 

188

 

 

 

 

 

 

100' 153
107
so-
.3
5 so- 151
s
Q)
.2 _
g 40 105'.
0:
2°" 121
0 ‘l"" :IIVIfififiIlgl‘IT'i'I‘""'"'l"""':'I * l:""l""r"fi1
100 110 120 130 140 150 160 170

m/z

Figure 4-18: Partial reaction product spectrum of ions produced during
ion/molecule reactions following E1 of benzaldehyde and dimethyl ether in
a high pressure ion source. Peaks produced from self-CI of dimethyl ether

were subracted.

189

protonated collision-complex, is the intermediate in the formation of the ion
of m/z 121. Following CID of m/z 153, no ion current was detected at m/z 121,
only at m/z 107 which represents protonated benzaldehyde, formed by
elimination of dimethyl ether. The parent ion mass spectrum of the product
ion of m/z 121 showed that the precursor is the ion with m/z 151. A proposed
reaction mechanism is shown in Figure 4-19. It would be necessary to
perform stable isotope labeling studies to determine the source of the oxygen
and CH3 in the acetal product ion having m/z 121. The mechanism shown in
Figure 4-19 suggests that the methoxy groups prior to the elimination of
methanol are identical, yet this would need to be confirmed experimentally.

d) Ion trapping

Over the course of the studies involving the ion/molecule reaction of
protonated acetaldehyde and methanol, an attempt was made to implement
the ion/trapping technique of Dolnikowski et al. (72,106). Ifa large voltage is
applied to the Q2 exit lens, the lens may function as a gate and trap reactant
ions in the collision chamber in the presence of the neutral reagent. This
provides more time for the ion to react with the neutral vapors in Q2. After
trapping the ions for a designated amount of time (usually milliseconds), the
ions in the collision chamber can be pulsed out by applying a large negative
potential to the Q2 exit lens. This serves to extract the ionic reaction
products from the collision chamber, and may provide greater detectability
for some ion/molecule reaction product ions. This technique was
implemented on the Extrel TQMS, and was called the trap and pulse method
(72,106). If the parent ions are admitted into Q2 for a designated period of
time, and the entrance lens to Q2 is then gated with a large potential so that

no more parent ions may enter, the phrase "inject, trap and pulse" is used.

 

190

.833 3568:. can 3.33:3an .«o 9:558 3

$8388 3:022:22 938:8 HNH Q8 mo :3 2: mo cow-«Snow 23 .8.“ magnum-E coma-98m "3&- charm

HNH a):
I
fob 2: ea
memEmucon 33:39am
Nzoo
.38.
I +
NE NE 0
I 5? \ amp \ mic /:
_ I
«roy- ‘ All 0100 , ‘ All 96/ MUIo
foo
fem J

I

 

191

The trap and pulse method admits parent ions continuously; only the voltage
on the exit lens following Q2 is varied. In the inject, trap and pulse method,
only a discrete packet of ions is admitted into the collision cell (72,106). After
the ions are introduced into the reaction region, the Q2 exit lens performs the
ion containment and ion extraction functions. A similar approach was used
by Beaugrand and coworkers on a penta-quadrupole instrument which
provided access to kinetic and thermodynamic parameters in ion/molecule
reactions (189).

The ion/trapping technique did not prove to be very fruitful on the
TSQ—70. The Finnigan instrument requires that at least one of the two mass-
analyzing quadrupoles be in a scan mode; that is, each quad cannot be set to
pass one fixed mass. This eliminates the possibility of monitoring ion current
in real time as it is always necessary to implement a mass scan to obtain the
data. In the attempt to overcome this difficulty, extremely fast scans were
used over a narrow mass window. The first quadrupole was allowed to scan
over a narrow mass range (0.4 11) which served to introduce the reactant ion
into the reaction chamber, while Q3 passed ion current at one fixed mass
value. A diagram of the instrument is shown in Figure 4-20 which also
summarizes how the inject, trap and pulse technique was attempted on the
TSQ-70. The user output voltage was connected to lens 2-3 to provide better
control of this voltage, and enable it to be adjusted quickly. A procedure was
written as shown in Table 4-1, which would allow adjustment of the user
output voltage and lens 3-1 voltage for altering the potential applied to the
Q2 entrance and exit lenses. The number of scans multiplied by the scan rate
(scan rate is 0.001 sec/scan), will determine the approximate length of time
for the filling of the collision cell. During the filling time, the Q2 exit lens
voltage was 100 volts; no positive ions could leave Q2 at this voltage. After

 

192

.2..de no 333388 035 can nab $8.9: mo :oESqumEEH 5N9 9:53

wcmom m Low «09.68

 

choomw moo. u 9:: cozombxm > 00 F. > 00w ”mac. 5 cowombxm
o u memom #
8:83 6 x See u 235980 > 2: > 2: ”22 s Emacs-coo
u. u 968 t
8:83 E x See u we: use“. > 2: > 2- ”.3595 cozomom 95E All
a mum.—

uofios. 3:5 can 22... .86....

$833 Sod

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$9: :2 839a pox: sous-s : v.0

0-0 NA” 70 m-N N-N TN gm CO_ acmfiump—
Al
Al
A All.
All

A
‘
m0 383 .8520 $23 5

cozomom

 

ssew uo! ionpmd zxeu .10; made:

193

Table 4-1: Procedure used to implement inject, trap and pulse on the TSQ-70
triple quadrupole mass spectrometer.

N=9

PAR N, 44.8,45.2,.001 l Sets parent ion scan, so 03 mass is fixed at m/z N.

REPEAT 80

N=N+1 ;PAR N
U01=-5;L23=-10

REPEAT 50
GO ;STOP
END

L23=100
C=5
REPEAT C
GO
STOP
END

U01 =5

REPEAT 5
GO;STOP
END

END

U01 =0.5;L23=-10

ASTOP

Q1 is scanned from 44.8 to 45.2 u at 0.001 sec/scan
to introduce reactant ion (m/z 45) into 02.

lSelection of fixed product ion (03) mass.
lUser output connected to L31 and will cause L31 to
be +100 V

! Fill reaction chamber for 50 scans of Q1.

! Containment in reaction chamber for 5 scans.

! L23 set to -100 V to extract the ions from 02

! Collect 5 scans of data (ions detected)

l Repeats cycle 80 times so product ion mass
Mass spectrum from m/z 10 to m/z 90 is obtained

! Sets L1 -3 and L3-1 to typical values.

! End of procedure.

194
the entrance gate is 'closed', the number of scans times the scan time defines
the confinement time for the inject, trap, and pulse method. Following the
confinement, the Q2 exit lens is biased with -100 volts, and the ion current at
the fixed m/z value is monitored for 5 scans (0.005 seconds). Generally, ion
current was detected only for the first two scans. This whole procedure is
repeated for the entire product ion mass range of interest.

The procedure was attempted when protonated acetaldehyde was
selected to react with methanol in the collision cell. The reconstructed mass
chromatogram could essentially be interpreted as a mass spectrum, as each
mass scan represents the ion current detected at a fixed mass when Q1 is
scanning rover essentially one m/z value. Most of the scans are obtained
during the filling time or containment time; no ion current is detected during
this time. When the eadt lens to Q2 is pulsed, it is expected that ion current
would be detected for at least one scan or as long as it takes the ions to reach
the detector. The product ion mass spectrum for reaction of protonated
acetaldehyde and methanol for the im'ect, trap and pulse method is shown in
Figure 4-21 for a filling time of 50 milliseconds (50 scans 1 msedscan). After
filling the chamber for 50 milliseconds, the entrance lens was gated so no
more parent ions could enter and, immediately, the exit lens was pulsed to
extract the ions in the collision cell with the data collected for 5 scans. This
was repeated for each mass to obtain the product ion mass spectrum. The
TIC chromatogram shows 'peaks' at m/z 45, m/z 47 and m/z 59 representing
the parent ion, protonated dimethyl ether, and the product ion, respectively.
If the same procedure is followed, but with ion confinement for 5 milliseconds
following admittance of the reactant ion into the collision chamber, the
product ion mass spectrum shown in Figure 4-22 is obtained. This spectrum

shows an increase in the product ions (m/z 47 and m/z 59) as compared to

195

 

.54 .3 “533.558 983 meow 23 63am mm? mug node 3036295 .owmaogflmom
33:39am 5:5 088 on you v25 mg, .9833 68388 mamfiflaoo .28 ”SEES SE. .ouémfi on» no
“6358295 mm? @052: $39 98 93... Joe? 23 5:3 353% 85.58% $9: :3 335.5 3N6 93.3%

806... 00mm 0880 coma OOON son.“ soon Dom

D b”.PhD—bprbbhrt-IDPiblhbb->PbI_b-FF—DFD DFPLIh—DbibblhPhbbbDbIPDDPL-bbtbbLDD-bIbeDDDb D
d

. -1 _

mm
3.

 

mm you

Isa

13,

row
we

 

 

934. . .
+o+ux on food a

 

196

6me m we comma 308:3:8 a San van—ombxo 983 33 .mPEoESmom
wganofizg 5:» 038 on .8,“ 623 mm? .9893 35508 Max:338 £3 nommmzoo use .ouéme mg... no
wgcmEmEEm was» @2308 mafia can nab 58%: on» can? 3:830 833QO 938 :3 835.5 ”umé 0.53m

 

 

 

 

 

 

oom..¢ 00 00mm 8900 oom.~ ooou coma coo." com
P» L... . . r .1.“ ....r_4.‘..‘..4:i _:.:....L. ‘ ... .L»....:.L. .PL.....‘ ...._r:.._.1.r.1_4..¢y G
_ q d _ :_ .. ._ fl 1 _ 4
mm mm
raw
mm mv T0,”.
ram
row
36.0 S. _ . tom: x
mo+mx on

197
when no confinement is used. In both experiments, the methanol pressure in
Q2 was 0.5 mtorr and the collision energy was 1 eV. Although the signal-to-
background appears to have improved when there is ion confinement, the
total current detected is quite low indicating that the procedure does not
contain or extract ions very efficiently. The ion current detected for a steady
state signal (no trapping procedures) is on the order of 1000 times greater
than the signal detected implementing the trap and pulse method. Over the
course of the research, the author did not have difficulty in detecting ionic
products following ion/molecule reactions in Q2, therefore, the trap and pulse
methodology was not pursued any further. Much work needs to be done
before one may accurately assess the feasibility of performing ion-

confinement experiments in the Finnigan TSQ-70.

4. Summary

The results of this study provide direct evidence by CID mass
spectrometry for O-alkylation in the gas phase reaction between protonated
acetaldehyde and methanol (156). A second reaction mechanism was
elucidated involving protonated dimethyl ether and acetaldehyde. Some
additional experiments were performed suggesting that ion/molecule
reactions of dimethyl ether and aldehydes result in O-alkylation of the
aldehyde. Perhaps O-alkylation is a general phenomena for reactions of
aldehydes and ethers in the gas phase, yet further experiments need to be
done to verify this claim. The trap and pulse methodology was difficult to
implement on the TSQ and at this time it is not known whether it would
provide advantages in detection of product ions as compared to the

conventional 'steady state' approach.

198

D. Reactions of protonated alcohols and ethyl acetate

Condensation reactions involving protonated esters and alcohols have
been studied in the gas-phase by other researchers (106,190). It had been
shown by Dolnikowski that the product ion mass spectrum of protonated
ethyl acetate and propanol in the collision cell of a TQMS shows peaks at m/z
131 and m/z 149. These peaks also were observed when protonated propanol
was reacted with ethyl acetate. The ion with m/z 149 represents the proton-
bound adduct of ethyl acetate and propanol, whereas the ion with m/z 131
represents the dehydration product ion. Similar dehydration ion/molecule
reactions involving esters and alcohols also have been reported from studies
with an ICR instrument (190). This brief section describes a few additional
experiments where the product ion, formed via a condensation reaction, was

made in the ion source and selected for CID.

1. Experimental

All experiments were performed on the Finnigan TSQ-70 triple
quadrupole mass spectrometer. A direct inlet was constructed which was
used to introduce vapors of one of the neutral compounds (the alcohol) into
the high pressure CI ion source. Vapor from the second reagent (ethyl
acetate) was introduced via the CI gas lines. This allowed mixture of the
vapors to occur in the ion source for ionization by EI. It also provided some
control over the partial pressures of the individual components. A mass
spectrum was generated which showed peaks representing ion/molecule

reactions involving both reagents. For the CID studies, the ion formed from

199
dehydration of the proton-bound collision complex, was selected for CID with

argon introduced into the collision cell at a pressure of 0.6 mtorr and a

collision energy of 20 eVLab.
2. Results and discussion

The ion/molecule reaction between protonated ethyl acetate and
propanol has been studied previously in an ICR instrument (190) and in a
TQMS (95). In the TQMS study, the CID daughter ion mass spectrum of the
ion with m/z 131, which represents [(Propanola-ethyl acetate)H+ - H20]+,
showed elimination of 60 u (CH30H20H20H). This result led to the
suggestion that the product ion has the structure of an acetal ion as shown in
Figure 4-23, as loss of propanol may be envisioned from this ionic structure
(106).

One of the difficulties in studying this reaction is that ion/molecule
reactions in ethyl acetate produce an ion of m/z 131 with structure shown in
Figure 4-24. This ion is identical in mass to that of the reaction involving
propanol and ethyl acetate, and this fact needs to be considered while
evaluating the CID results. Both ionic structures are likely to be formed in
the ethyl acetate-alcohol mixture. The CID daughter ion mass spectrum of
m/z 131 produced when only ethyl acetate was in the ion source lacks the
fragment peak at m/z 71 which represents loss of (CH3CH20H20H). In the
effort to clarify the ion/molecule reaction mechanism, deuterated analogues of
the ethyl ester, including 2H5-ethyl acetate (022H5COOCH3) and 2H3-ethyl
acetate (C2H5COOC2H3) were reacted with propanol. When the deuterated
analogues of ethyl acetate were introduced into the ion source along with

propanol, in all cases, a peak was detected in the mass spectrum which

 

F1.

200

CHZCHZCH3
+5

||
CHsc—O-CHZCH3

+
C37 ”15 02
acetal ion

Figure 4-23: Acetal ion formed in reaction of protonated
propanol and ethyl acetate.

201

CH3 CHa
\ _ \
OCHZCH3 OCHZCHS

Protonated ethyl acetate (m/z 89)

Ta
OHS + C— CH CH3 . CH3
\0—5 | ——> C=O—C—O-H
l — - OCHZCHa l * IV
OCHZCH3 05‘ °CH20H3
ethyl acetate

loss of ethanol

Figure 4-24: Ion/molecule reaction of protonated ethyl acetate with ethyl
acetate to form the product ion of m/z 131.

202

represented the addition of 43 u (C3H7+) to ethyl acetate. It can only be
presumed that this represents the ion formed in the reaction of protonated
ethyl acetate and propanol, followed by loss of water. The CID daughter ion
mass spectrum of the ion of m/z 134 formed when 2H3-ethyl acetate and
propanol are mixed in the ion source is shown in Figure 4-25. With 2H3-ethyl
acetate, the self-CI peak shifts to m/z 136, eliminating the interference from
the ethyl acetate ion. Once again, the daughter ion spectrum of the
dehydration product ion shows a peak at m/z 76 which is from the loss of
propanol. Based on the structure of the acetal ion (Figure 4-23), it might be
expected that ethanol could be eliminated as well. However, the daughter ion
mass spectrum which lacks a peak representing loss of ethanol suggests that
this process does not take place. Other peaks detected in the CID of m/z 134
include: m/z 92 which represents protonated ethyl acetate formed by the loss '
of the C3H5 olefin, m/z 64 which is protonated acetic acid (with three 2H
atoms), and m/z 43 which is the CgH7+ cation.

Butanol was also simultaneously introduced with ethyl acetate into the
ion source and the anticipated acetal ion was detected as a peak at m/z 145.
The CID daughter ion mass spectrum of the product ion of m/z 145 is shown
in Figure 4-26. The dominant fragmentations include elimination of C4H3
and loss of butanol. Once again, the daughter ion spectrum does not show a
peak which would indicate that the acetal ion may eliminate ethanol.

Finally, ethyl acetate and ethanol were mixed in the ion source, and
the product ion representing the addition of 02H5+ to ethyl acetate with m/z
117 was selected for CID. The acetal product ion can eliminate 02H4 to give
a peak at m/z 89, or ethanol forming an ion with m/z 71. Labeling studies
with CH30 2C 22H5 indicate that the product ion, shifted to m/z 122,

eliminates both 02H4 and C22H4 as seen in the CID spectrum shown in

203

 

 

 

 

 

 

100‘
80 134
£‘
(0
g 601 74
g 92
:2: 40" 43 64
(U
is 20
m:
0---.-"Jfifi pm . ., .- . puma" . n
20 40 60 80 100 120 140 160
m/z

Figure 4-25 : CID daughter ion mass spectrum of the ion of m/z 134 formed
when propanol and C2H3COOCzH5 are introduced into the ion source.

204

100'
71

«Ia-moo
99C?

145

Relative Intensity

m
9

 

 

O
l
1
_
4
4
1
1
q

 

80 100 120 140 :60
m/z

s
31
s

Figure 4—26: CID daughter ion mass spectrum of the product ion of m/z 145
formed when ethyl acetate and butanol are introduced into the ion source.

205
Figure 4-27. The ratio of m/z 94 (loss of CzH4 ) to m/z 90 (loss of 022H4) from
the acetal ion is 2.1 : 1. This can be attributed to an isotope effect favoring H
transfer over 2H transfer (191). In high energy CA studies, Harrison
obtained a ratio of 2.3 : 1 for the losses of C2H4 and szH4 from ion with m/z
122 formed in the reaction of 022H5+ with ethyl acetate (191).

A general scheme for the fragmentation mechanism of the acetal ion is
shown in Figure 4-28. The acetal ion may eliminate the larger alcohol moiety
and the larger olefin. It should be noted that for both primary alcohols and
olefins, the heat of formation for the neutral species decreases with
increasing chain length, so the acetal ion preferentially eliminates the most
stable neutral. For example, the acetal ion of m/z 131 formed in reaction of
protonated propanol and ethyl acetate can eliminate CgHs to form protonated
ethyl acetate (Figure 4-25); it does not eliminate CgH4 to form protonated
propyl acetate. The difference in heat of formation for C3H5 (AHf =20.2
kJ/mol) and C2H4 (AHf = 52.2 kJ/mol) is 32 kJ/mol, whereas the difference is
only 6kJ/mol between protonated ethyl acetate (AHf = 247kJ/mol) and
protonated propyl acetate ((AHf =241 kJ/mol). Therefore, loss of the larger
olefin is thermodynamically favored, and this same trend holds for
elimination of the alcohol moiety. In the reaction involving ethyl acetate and
ethanol, the alkoxy groups within the ion are identical, so that both groups
are included in the fragmentation process.

These CID studies lend support to the structure of the product ion
involving reaction of alcohols and ethyl acetate being an acetal ion. It has
been suggested that this reaction mechanism involves the formation of a
collision-complex, followed by elimination of water. However, these studies
described do not provide definitive information on the reaction mechanism

forming the acetal ion. The results only support the premise that the product

100‘

Amen
c???

Relative Intensity

m
C?

 

206

122

52 94

 

!

 

NC
0

vvvvv‘rfiVTijvvv'vvvv'vvvv‘vv v v: ' W v

40 60 80 100 1 20 140 1 60
m/z

Figure 4-27: CID daughter ion mass spectrum of the ion of m/z 122 formed
when CHgCOOszH5 and ethanol are introduced into the ion source.

 

.Sfioow v3.8 98 £038? mo maoflowou ogoofioEEom E 35.8%
:3 188"“ mg Soc 3322: cum? .8 3:83 :a .«o aoflafifizo m5 masonm 385% 3.856 "warn. earn

207

p+c~1cn_u
.ézcouxo- «:0 Ba ”:0 47:“sz
1% ”o\ 1%.
mm as ”:o
_ m / \
\o /o
nzofo _
o
”:0 fo\
qua—8°83
w+CNI :0
£0
«:0
o
m\ ca 35: :0 7.5.: :0
e “:0 ”:0
:o :o a
n + «:0 :o
C. «\E IO\IO H.m\O/AJ\0\ 0/0\
AI AIIII
__ I _ J acoaomcabaou __
«10.10 zouxo nIo :o

aloe—due lo—

208
ion has the acetal structure. A few product ion mass spectra were obtained
for reaction of protonated propanol and esters in Q2. The esters which where
introduced into the collision cell included ethyl acetate, methyl acetate, and
methyl propionate. In these product ion mass spectra obtained at low
collision energy ( 1 eVLab), peaks were detected representing loss of H20 from
the proton-bound adduct. However, in all cases, a prominent peak was also
detected at m/z 43 (C3H7+) which is a CID fragment ion (loss of H20) from
protonated propanol. At the high Q2 pressures, multiple collisions occur.
Second order reactions involving the C3H7+ ion and the neutral ester may
account for the formation of the acetal product ion. It may be that alkyl
attachment to the carbonyl oxygen of the ester is the reaction occurring and
not a condensation reaction. Furthermore, in CID studies of the various
(ester-alcohol)H+ complexes formed in the ion source, elimination of H20 was
not observed. It is clear from the TQMS studies where the neutral and
ionized reactants were isolated, a product ion corresponding to dehydration of
the original protonated collision-complex is formed, however, the author
believes that many more detailed studies need to be carried out before the

mechanism is unraveled.

E. Ion/molecule reactions of protonated molecules and

hexamethyldisilazane

There are many analytical tests to determine the functional groups
present in an organic molecule in solution (192). Most of them use a specific
reagent such as 2,4-phenylhydrazine for detecting ketones. It would be
useful to extend this strategy to characterize gas phase ionic species.

Preliminary experimental results performed on the penta-quadrupole mass

209
spectrometer in Paris involving ion/molecule reactions of protonated
molecules and hexamethyldisilazane (HMDS) suggested that a silicon-
hydrogen exchange reaction was selective for the amino group as is shown in
Figure 4-29. When protonated amines were reacted with HMDS, a product
ion was observed which represented the addition of 72 u. Silylation was
occurring in the gas-phase. Protonated alcohols and protonated ketones did
not appear to undergo the silicon exchange reaction. These preliminary
experiments were carried out by Dr. J .T. Watson while he was on sabbatical

in France.

RNH3+ + (CH3)3SiNHSi(CH3)3 ----- > RNH2Si(CH3)3+ (CH3)3SiNH2

Figure 4-29: Proposed reaction of protonated amines with HMDS.

In January 1989, Dr. Christian Rolando from Ecole Normale
Superieure, Paris France, visited the Mass Spectrometry Facility. Dr.
Watson and Dr. Rolando were the researchers who initially studied this
ion/molecule reaction on the penta-quadrupole mass spectrometer. While Dr.
Rolando was visiting Michigan State, some experiments were performed to
determine if the results obtained on the penta-quadrupole mass spectrometer
which had ion-confinement capabilities, could be reproduced on the Finnigan
TSQ-70 instrument. A few experiments were carried out where protonated
molecules were chosen for reaction with HMDS, and this section describes

some of the results.

 

21 0
1 . Experimental

All experiments were performed on the Finnigan TSQ-70 mass
spectrometer. Hexamethyldisilazane (HMDS), a liquid at STP, was
introduced into a glass reservoir which was connected to the CID gas lines.
The needle valve was adjusted so that the pressure in the collision cell was
approximately 3 mtorr. Liquid samples which were chosen for reaction with
HMDS were introduced via the glass reservoir directly into the ion source.
The pressure in the ion source was adjusted so that the base peak in the mass
spectrum was the protonated molecule. Solid samples were introduced by the
direct insertion probe. Later in the studies, a few small peptides were placed
on the probe tip in a droplet of glycerol, and ionized by fast atom
bombardment (FAB). In all the experiments the protonated molecule was
selected as the reactant ion and introduced into the collision cell with a
collision energy ranging from 1-4 eVLab. For each reaction, the collision
energy was adjusted to provide optimal detection of a reaction products,
either the silylation product ion or the ion with m/z 162 which represents
protonated HMDS.

2. Results and discussion

The preliminary results obtained in France suggested that the
derivatization reaction involving HMDS was selective for protonated amines.
The first compound chosen for investigation was aniline, with a molecular
weight of 93. When protonated aniline was reacted with HMDS, two major
peaks were detected in the product ion mass spectrum. The first peak was at
m/z 162 which represents protonated HMDS. The second peak was at m/z

 

211
166, which represents the net addition of 72 u. A trimethyl silyl group was
exchanged with a hydrogen atom to form the product ion and (CH3)3SiNH2.
Ion confinement was not necessary for detection of the product ion peaks
representing the silylation ion/molecule reaction.

Next, acetone was introduced into the ion source, and protonated
acetone with m/z 59 was chosen for reaction with HMDS. This reaction was
attempted in France, and the silylation reaction was not observed. However,
on the Finnigan TSQ-70, the reaction was observed to occur quite readily as
can be seen by the product ion mass spectrum shown in Figure 4-30. A few
other protonated ketones, including cyclohexanone and 4-heptanone were
selected to react with HMDS, and all gave a product peak which represented
the net addition of 72 11

Table 4-2 lists the compounds that were investigated and indicates
those for which a successful reaction was observed. In all cases, the intensity
of the product ion was low, yet the distinctive isotopic pattern of silicon
indicated that the peak was not just noise. Although not much is known
about the ion/molecule reaction mechanism, a few points may be made from
the experimental results. Protonated leucine was one of the compounds
analyzed, and it reacted with HMDS to give a product ion with m/z 204.
When 15N -leucine is protonated, the protonated molecule also yields a
product ion representing the addition of 72 u. This result indicates that the
amino nitrogen remains in the product ion; it is not replaced by the nitrogen
from the neutral reagent. Although our results conflict with those obtained
on the penta-quadrupole, indicating that this ion/molecule reaction is not
selective for amines, this reaction is worth pursuing further and suggests
that gas-phase derivatization may be performed in the collision cell of a

TQMS.

 

212

 

 

 

 

x20
100'} 59 l 131

s? 80‘
(O
5
g 60'-
o
.2. 4o-
% 117
03 91

20 I 162

0 v V v all" I l v I v ‘ ' 'A I!I£""':'I

40 60 80 100 120 140 160 180

m/z

Figure 4-30: Product ion mass spectrum following ion/molecule reactions of
protonated acetone (m/z 59) with HMDS at a pressure of 3 mtorr and a
collision energy of 1 eVLab.

213

Table 4-2: Protonated compounds which were reacted with HMDS in the
collision cell of the TQMS.

 

mLLQt

9.9mm! NIL-1+ W+

aniline 94 yes

N,N, dimethyl 122 no ‘
aniline ‘
L-leucine 132 yes

isoleucine 132 yes

[3- alanine 90 ?

L-tryptophan 205 yes l
6-aminohexanol 1 18 yes

glycyl L-isoleucine 189 yes

alanyl leuoylglycine 260 yes

L-tryptophyl L-alanine 276 yes

acetone 59 yes

cyclohexanone 99 yes

4-heptanone 1 15 yes

? Product ion peak would be at m/z 162 which also may represent protonated
HMDS.

 

214

The fact that reaction products are observed for some protonated
molecules with HMDS, suggest that those reactions are exothermic. If excess
energy is generated during the formation of the product ion, this energy may
be redistributed among the bonds within the product ion. Given sufficient
energy, bond fragmentation could take place. It would be worthwhile to
continue the investigation of ion/molecule reactions with
hexamethyldisilazane. A major thrust in mass spectrometry is peptide
sequencing (193). Protonation of peptides occurs at the basic amino terminus.
The results suggest that the protonated peptide may be able to undergo the
derivatization reaction in the gas-phase. Perhaps reaction-induced
fragmentation may occur, thereby provided a novel system for analysis of

peptides.

 

CHAPTER V
SUMMARY AND FUTURE WORK

This dissertation describes gas-phase ion/molecule chemistry studied
by triple quadrupole mass spectrometry. The principal theme is pursuit of
developing analytical methodology based on selective gas-phase ion/molecule
chemistry. Two reactive regions are available with the TQMS, the ion source
and the second quadrupole collision chamber. The collision chamber provides
a region where the ionic and neutral species may react in an isolated
environment. On the other hand, ion/molecule reactions may occur in the ion
source with the ionic products being selected for conventional MS/MS
analysis using CID. Over the course of this research, both methods were
utilized providing complementary information. Throughout this research,
thermochemistry was examined along with experimental results obtained
with the TQMS, in the attempt to clarify the gas-phase chemistry.

The gas-phase chemistry of aryl cations with nucleophilic reagents was
investigated by TQMS. These studies led to the development of new
methodology for distinguishing C7H7+ isomers. Of the three isomers
investigated, the tolyl, benzyl, and tropylium cations, only the tolyl cation
undergoes a ring addition reaction with methanol or dimethyl ether. The
benzyl and tropylium isomers do not undergo this reaction. Examination of
the thermochemistry revealed that in the reaction with dimethyl ether,
methoxylation of C7H7+ is exothermic for the tolyl cation, but endothermic
with benzyl or tropylium. The results suggest that quantitation of the tolyl
cation in C7H7+ mixtures may be possible using this low-energy analytical

approach. Future efforts in this area should focus on continuing to search for
215

216

ion/molecule reactions which may be selective for either the benzyl cation or
tropylium cation. Gas-phase ion/molecule studies using ICR instruments
have previously demonstrated the reactivity of the benzyl cation. In order for
a reagent molecule to be reactive with the benzyl cation, but not with the
tolyl cation, the stability of the reaction products must overcome the 157
kJ/mol difference in the AHf of the tolyl and benzyl reactants. If such a
reaction is found and the reaction enthalpy with the benzyl cation is
exothermic, but endothermic for the tolyl cation, then perhaps this
ion/molecule reaction could be accomplished in the center quadrupole of a
TQMS. This may enable complete quantification of C7H7+ mixtures to be
made by the low-energy ion/molecule reaction approach.

Not only have the studies involving aryl cations led to a method for
selectively detecting the tolyl cation, but the results suggest that methods
could be developed for rapidly screening aromatics in mixtures. The
preliminary results indicate that only aromatic ions with a vacant charged
site on the ring react with dimethyl ether to give the methoxylation product
ion. By utilizing a neutral gain scan mode in a GC/MS/MS analysis, it may
be possible to selectively detect those compounds in a complex mixture which
are aromatic. This was demonstrated for a simple five-component mixture,
but it would be necessary to analyze a more complex mixture to determine
the utility of this approach for rapidly screening aromatics.

The results of other studies indicate that molecular oxygen is involved
in some of the major fragmentation processes of molecular anions of
methylated abscisic acid (ABA-Me). It is proposed that 02 attacks at the
radical site, forming a peroxy anion intermediate. This intermediate anion
readily decomposes yielding predominant fragment ions. Molecular anions of

structurally similar ABA metabolites also were shown to undergo an

217

analogous reaction with molecular oxygen. These results are unprecedented
in that they represent the first case using a TQMS where a non-aromatic
radical anion reacts with molecular oxygen, promoting fragmentation. The
ions formed in this process enable the site of oxygen isotope enrichment to be
precisely located within the molecule, an observation which has provided
invaluable insight into the mechanism of biosynthesis of these compounds.
Electron capture negative ionization mass spectrometry is a rapidly
expanding field, yet not much attention has been devoted to the study of
ion/molecule reactions of anions. Flowing afterglow studies have shown that
anionic conjugated dienes undergo reactions with molecular oxygen to form
enolate anions. The reaction of 2,4 hexadiene with 02 was examined with the
triple quadrupole mass spectrometer, and the ionic products detected are
thought to represent enolate ions. It would be interesting to perform this
experiment with 1302 in the center quadrupole to verify that the product ions
have incorporated an atom from molecular oxygen. It would be worthwhile
for future investigators to engage in research involving reactions of anions
and Oz to determine if oxygen-induced fragmentation has analytical utility.
The location of double bonds in fatty acids, for example, is an important area
of research, and perhaps reactions of oxygen could be used for analytical
purposes with anions which contain a conjugated diene moiety. Also, other
reactions of anions with neutral reagents have been characterized in a
flowing afterglow apparatus, and it would be worthwhile to investigate
similar model systems on a TQMS to determine if the same chemistry is
observed. Some interesting results in this dissertation show that the

buffering gas used in electron capture ionization of a- and B-ionone affects the

appearance of the mass spectrum. These results suggest that ion/molecule

218

reactions or reactions involving radicals may be occurring in the ion source,
and exploring these phenomena in a systematic study would be interesting.

One of the collaborative studies described in this dissertation involves
the MS/MS analysis of dexamethasone. Sufficient selectivity is provided by
the chemical oxidation procedure along with ionization by electron capture to
permit sample introduction into the mass spectrometer via the direct
insertion probe when selected reaction monitoring is employed. Use of the
direct insertion probe rather than the gas chromatograph significantly
simplifies and shortens the analysis of dexamethasone in plasma. Where
applicable, the selected reaction monitoring approach will continue to provide
a simple and rapid alternative to the selected ion monitoring approach for the
analysis of low-level analytes in complex biological matrices.

Other studies include those involving ion/molecule reactions of
carbonyl compounds and alcohols. Evidence for alkylation occurring on the
carbonyl oxygen is presented. A detailed study of the ion/molecule reaction
of protonated acetaldehyde and methanol is described. Although structural
studies of these types did not lead to analytical applications, they
demonstrate the versatility of the TQMS in the determination of ionic
structures. It is important that the chemistry be understood before
attempting to find analytical applications for the chemistry. Further work
needs to be carried out in the study of protonated carbonyl compounds with
alcohols before assessing the analytical utility for such reactions. Perhaps, in
general, protonated carbonyls are methylated on the carbonyl oxygen in
reaction with methanol. This may be useful in developing methodology for
selectively detecting these compounds. Also, these types of studies may be
useful in the study of larger protonated species where the location of the

proton is not known. Once reactions of this sort are characterized with

219

smaller molecules, a potential application for this reaction is in probing the
location of a proton in multifunctional-group compounds.

Finally, some results are described which demonstrate that certain
protonated molecules undergo a silicon-hydrogen exchange reaction with
hexamethyldisilazane (HMDS). These preliminary studies indicate that gas-
phase derivatization forming a silylated product ion can be observed in a
TQMS, but the selectivity of HMDS for protonated molecules remains
unknown. It appears that HMDS reacts with protonated ketones and
protonated amines. This gas-phase silylation reaction may be useful in
peptide sequencing studies, as the site of protonation on a peptide is usually
at the amino terminal group. Formation of the silylated product ion
intermediate may produce excess energy which perhaps could be expended
and redistributed, directing the cleavage of the peptide in a reaction-induced '
fragmentation sequence.

In closing, the data shown in this dissertation give a glimpse of the
variety of analytical applications offered by implementing gas-phase
ion/molecule reactions on a triple quadrupole mass spectrometer. Certainly,
as future researchers are engaged in the study of ion/molecule reactions, this
exciting field will supply the mass spectrometry community with many more
analytical applications based on the unique chemistry ions exhibit in reaction

with selected neutral reagents.

 

LIST OF REFERENCES

 

10.

11.

12.

13.
14.

LIST OF REFERENCES

K. L. Busch, G. L. Glish, S. A. McLuckey, Mass Spectrometry/ Mass
Spectrometry, VCH Publishers: Deerfield, FL, 1988.

F.W. McLafferty, Ed. Tandem Mass Spectrometry Wiley: New York,
1983.

RA. Yost and C.G. Enke, Anal. Chem. , 51, 1251A (1979).

RA. Yost, C.G. Enke, D.C. McGilvery, D. Smith and J .D. Morrison,
Int. J. Mass Spectrom. Ion Phys, 30, 127 (1979).

RA. Yost and C.G. Enke, J. Am. Chem. Soc. 100, 2274 (1978).

PH. Dawson, Ed. Quadrupole Mass Spectrometry and Its
Applications, Elsevier: NY, 1976.

PH. Dawson and Y. Bingqi, Int. J. Mass Spectrom. Ion Processes,
56, 25 (1984).

RE. Miller and MB. Denton, J. of Chem. Educat. 63, 617 (1986).

RE. Miller and MB. Denton, Int. J. Mass Spectrom. Ion Processes,
72, 223 (1986).

J .E.P. Syka and A.E. Schoen, Int. J. Mass Spectrom. Ion Processes,
96, 97 (1990).

AG. Harrison, Chemical Ionization Mass Spectrometry, CRC Press,
Inc.:Boca Raton, Florida, 1983.

H. Mestdagh, N. Morin and C. Rolando, Org. Mass Spectrom. , 21,
321 (1986).

RM. Caprioli, Anal. Chem, 62, 477A (1990).
EC. Huang, T. Wachs, J .J . Conboy and J .D. Henion, Anal. Chem. ,
62, 713A (1990).

220

15.

16.
17.
18.

19.

20.
21.
22.

23.

24.

25.

26.

27.
28.
29.
30.

31.

32.

221

R.D. Smith, C.J. Barinaga and HR. Udseth, J. Phys. Chem. 93,
5019 (1989).

SF. Wong, C.K. Meng and J .B. Fenn, J. Phys. Chem. , 92, 546 (1988).
TSQ-70 Operator's Manual, Finnigan MAT, 1984.

M. Vincenti, J. Schwartz, R.G. Cooks, A.P. Wade and C.G. Enke,
Org. Mass Spectrom. 23, 579 (1988).

J .C. Schwartz, A.P. Wade, C.G. Enke and R.G. Cooks, Anal. Chem. 62,
1809 (1990).

K. Levsen and H. Schwarz, Mass Spec. Reviews, 2, 77 (1983).
PH. Dawson, Int. J. Mass Spectrom. Ion Phys, 50, 287 (1983).

KL. Schey, H.I. Kenttamaa, V.H. Wysocki and R.G. Cooks, Int. J.
Mass Spectrom. Ion Processes, 90, 71 (1989).

V.H. Wysocki, H.I. Kenttamaa and R.G. Cooks, Int. J. Mass
Spectrom. Ion Processes, 75, 181 (1987).

S. Naylor and J .H. Lamb, Rapid Commun. Mass Spectrom. 4, 251
(1990).

J .S. Brodbelt-Lustig and R.G. Cooks, Int. J. Mass Spectrom. Ion
Processes, 86, 253 (1988).

H.I. Kenttamaa and R.G. Cooks, Int. J. Mass Spectrom. Ion
Processes, 64, 79 (1985).

H.I. Kenttamaa and R.G. Cooks, J. Am. Chem. Soc., 107, 1881 (1985).
H.I. Kenttamaa, Org. Mass Spectrom., 20, 703 (1985).
J .L. Holmes, Org. Mass Spectrom., 20, 169 (1985).

MD. A. Mabud, M.J. DeKrey and R.G. Cooks, Int. J. Mass Spectrom.
Ion Processes, 67, 285 (1985).

MD. A. Mabud, T. Ast and R.G. Cooks, Org. Mass Spectrom., 22,
418 (1987).

ME. Bier, J .W. Amy, R.G. Cooks, J .E.P. Syka, P. Ceja and G.
Stafford, Int. J. Mass Spectrom. Ion Processes, 7 7 , 31 (1987).

33.

34.

35.

36.

37.

38.
39.
40.

41.

42.

43.

44.

45.

46.

47.

48.

222
J .H. Callahan, F.L. King, M.M. Ross, V.H. Wysocki, Surface-
Induced Dissociation (SID) In Tandem Quadrupole Mass
Spectrometers; Proceedings of the 38th ASMS Conference on Mass
Spectrometry and Allied Topics, Tucson, Arizona, June 3-8, 1990.

RE. Tecklenburg Jr. and DH. Russell, Mass Spectrom. Reviews, 9,
405 (1990).

DE Hunt, J. Shabanowitz and J .R. Yates III, J. Chem. Soc. Chem.
Commun., 548-550 (1987).

H. Michel, D.F. Hunt, J. Shabanowitz and J. Bennett J. Biol. Chem
263, 1123 (1988).

DC. McGilvery and J .D. Morrison, Int. J. Mass Spectrom. Ion Phys.,
28, 81 (1978).

ML. Gross and DC. Rempel, Science, 226, 261 (1984).
N.M.M. Nibbering, Adv. Phys. Org. Chem. 24, 1 (1988).

E.E. Ferguson, F.C. Fehsenfeld and AL. Schmeltekopf, Adv. At.
Mol. Phys., 5, 1 (1969).

J .M. Van Doren, S. E. Barlow, C.H. DePuy and V.M. Bierbaum, Int.
J. Mass Spectrom. Ion Processes, 81, 85 (1987).

V.M. Bierbaum, G.B. Ellison, S.R. Leone, in M.T. Bowers, (Ed.) Gas
Phase Ion Chemistry Vol. 3 Academic Press, Orlando, 1984, Ch. 17.

KM. Ervin, S. Groner, S.E. Barlow, M.K. Gilles, A.G. Harrison,
V.M. Bierbaum, C.H. DePuy, W.C. Lineberger and GB. Ellison, J.
Am. Chem. Soc., 112, 5750 (1990).

R. Orlando, D.P. Ridge and B. Munson, J. Am. Soc. Mass Spectrom.,
1, 144 (1990).

M. Henchman, Ion Molecule Reactions, Franklin, J .L. Ed. Plenum:
NY, 1972.

TC. Cairns and E.G. Siegmund, Rapid Commun. Mass Spectrom.,
3, 68 (1989).

SN. Ketkar, J .G. Dulak, W.L. Fite, J .D. Buchnar and S.
Dheandhanoo, Anal. Chem., 61, 260 (1989).

K.W.M. Siu, G.J. Gardner and SS. Berman, Anal. Chem., 61, 2320
(1989).

 

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

223
R. Guevremont, R.A. Yost and W.D. J amieson, Biomed. Environ.
Mass Spectrom. 14, 435 (1987).

WC. Brumley, B.J. Canas, G.A. Perfetti, M.M. Mossoba, J .A. Sphon
and RE. Corneliussen, Anal. Chem., 60, 975 (1988).

MS. Lee and RA. Yost, Biomed. Environ. Mass Spectrom. , 15, 193
(1988).

W.M. Draper, F.R. Brown, R. Bethen and M.J. Miille, Biomed.
Environ. Mass Spectrom., 18, 767 (1989).

MC. Pumasia and E. Houghton, Biomed. Environ. Mass Spectrom.,
18, 1030 (1989).

J. Roboz, E. Nieves, J .F. Holland, M. McCamish and C. Smith,
Biomed. Environ. Mass Spectrom., 16, 67 (1988).

WE. Seifert Jr., A. Ballatore and RM. Caprioli, Rapid Commun.
Mass Spectrom., 3, 117 (1989).

M. Dawson, C.M. McGee, P.M. Brooks, J .H. Vine and TR. Watson,
Biomed. Environ. Mass Spectrom., 17, 205 (1988).

H. Schweer, D. Groessl, S. Gund, K. Soeding and H.W. Seyberth,
Prog. Clin. Biochem. Res., 301, 107 (1989).

R. Lorenz, P. Helmer, W. Vedelhover, B. Zimmer and RC. Weber,
Prostaglandins, 38, 157 (1989).

T. Krishnaumurthy and E.W. Sarver, Biomed. Environ. Mass
Spectrom, 15, 13 ( 1988).

KKayganich, J .T. Watson and C. Kilts, J. Ritchie,Biomed. Environ.
Mass Spectrom., 19, 341 (1990).

M. Barber and G.B.N. Green, Rapid Commun. Mass Spectrom., l, 80
(1987).

E.G. Huang and J .D. Henion, J. Am. Soc. Mass Spectrom., 1, 158
(1990).

RD. Smith, J .A. Loo, C.J. Barinaga, C.G. Edmonds and HR.
Udseth, J. Am. Soc. Mass Spectrom., l, 53 (1990).

DP. Hunt, J .R. Yates, J. Shabanowitz, S. Winston and CR. Hauer,
Proc. Natl. Acad. Sci. USA, 83, 6233 (1986).

 

65.

66.
67.

68.

69.

70.

71.

72.

73.
74.
75.

76.

77.

78.

79.
80.

81.

82.

224
DR Hunt, N.Z. Zhu and J. Shabanowitz, Rapid Commun. Mass
Spectrom, 3, 122 (1989).
K.J. Volk, R.A. Yost and A.B.-Toth, Anal. Chem., 61, 1709 (1989).

K.J. Volk, M.S. Lee, R.A. Yost and A.B.-Toth, Anal. Chem., 60, 722
(1988).

RD. Smith, J .A. Loo, C.G. Edmonds, C.J. Barinaga and HR.
Udseth, Anal. Chem., 62, 882 (1990).

H. Mestdagh, N. Morin and C. Rolando, Org. Mass Spectrom., 23,
246 (1988).

NH. Mahle, R.G. Cooks and R.W. Korzeniowski, Anal. Chem. , 55,
2272 (1983).

G.G. Dolnikowski, J. Allison and J .T. Watson, Org. Mass Spectrom. ,
25, 119 (1990).

G.G. Dolnikowski, M.J. Kristo, C.G. Enke and J .T. Watson, Int. J.
Mass Spectrom. Ion Processes, 82, 1 (1988).

M.T. Kinter and M.M. Bursey, J. Am. Chem. Soc., 108, 1797 (1986).
M.T. Kinter and M.M Bursey, Org. Mass Spectrom., 22, 775 (1987).

M.M Bursey, T.R. Blackbum,W.J. Meyerhoffer and WA. Mattson;
Org. Mass Spectrom . 25, 232 (1990).

EL. White, J.C. Tabet and M.M. Bursey, Org. Mass Spectrom., 22,
132(1937x

J .-P. Schmit, S. Beaudet and A. Brisson, Org. Mass Spectrom. , 21,
493 (1986).

DD. Fetterolf, R.A. Yost and J .R. Eyler, Org. Mass Spectrom. , 19,
104 (1984).

J. Jalonen, J. Chem. Soc., Chem. Commun. 872 (1985).

J .H. Batey and J .M. Tedder, J. Chem. Soc. Perkin Trans. II 1283
(1983).

AL. Mitchel and J .M. Tedder, J. Chem. Soc. Perkin Trans. II 667
(1984).

AL. Mitchel and J .M. Tedder, J. Chem. Soc. Perkin Trans. II 1197
(1986).

83.

84.

85.

86.

87.

88.

89.

90.
91.
92.

93.

94.
95.
96.

97.

98.

99.

225

J .P. Schmit, P.H. Dawson and N. Beaulieu, Org. Mass Spectrom., 20,
269 (1985).

RR. Pachuta, H.I. Kenttamaa, R.G. Cooks, T.M. Zennie, C. Ping, C-
j Chang and J .M. Cassady, Org. Mass Spectrom., 23, 413 (1988).

EL. White and M.M. Bursey, Biomed. Environ. Mass Spectrom. , 18,
413 (1989).

W.J. Meyerhoffer and M.M. Bursey, Org. Mass Spectrom. , 24, 169
(1989).

W.J. Meyerhoffer and M.M. Bursey,. Org. Mass Spectrom., 24, 246
(1989).

G.L. Glish, P.H. Hemberger and R.G. Cooks, Analytica Chimica
Acta, 119, 137 (1980).

M.T. Kinter and M.M Bursey, Biomed. Environ. Mass Spectrom., 15,
583 (1988).

R. Kostiainen, Biomed. Environ. Mass Spectrom., 16, 197 (1988).
R. Kostiainen, Biomed. Environ. Mass Spectrom., 18, 116 (1989).

DR. Mueller, B. Domon, W. Blum, W.J. Richter, H. Reiner, R. Keller
and P. Fischer, Org. Mass Spectrom., 24, 157 (1989).

R. Kostiainen and S. Auriola, Rapid Commun. Mass Spectrom., 2, .
135 (1988).

R. Kostiainen and S. Auriola, Org. Mass Spectrom., 25, 255 (1990).
E. Schroder, Org. Mass Spectrom., 24, 205 (1989).

D.W. Berberich, M.E. Hail, J .V. Johnson and R.A. Yost, Int. J. Mass
Spectrom. Ion Processes, 94, 115 (1989).

ME. Hail, D.W. Berberich and R.A. Yost, Anal. Chem., 61, 1874
(1989).

R. Orlando, C. Fenselau and R.J. Cotter, Org. Mass Spectrom., 24,
1033 ( 1989).

R. Orlando, C. Murphy, C. Fenselau, G. Hansen and R.J. Cotter,
Anal. Chem, 62, 125 (1990).

 

100.

101.

102.

103.
104.

105.
106.

107.
108.

109.

110.
111.

112

113.

114.
115.
116.
117.

118.

226
R. Orlando, C. Fenselau and R.J. Cotter, Rapid Commun. Mass
Spectrom, 4, 259 (1990).

R. Orlando, C. Fenselau and R.J. Cotter, J. Am. Chem. Soc., 112,
5747 (1990).

R. Orlando, C. Fenselau and R.J. Cotter, Org. Mass Spectrom., 25,
485 (1990).

R. Orlando, Private Communication, June, 1990.

A. Bjarnason, J .W. Taylor, J .A. Kinsinger, R.B. Cody and DA. Weil,
Anal. Chem. 61, 1889 (1989).

D. Kuck, Mass Spectrom. Reviews, 9, 187 (1990).

G.G. Dolnikowski, PhD Dissertation, Michigan State University
(1987).

J .C. Tabet and C. Guenat, Adv. Mass Spectrom, 1985 Part B, 831.

S.G. Lias, J .E. Barmess, J .F. Liebman, J .L. Holmes, R.D. Levin,
W.G. Mallard, J. Phys. Chem. Ref: Data 1988, 17, Suppl. 1.

F.W. McLafferty, Interpretation of Mass Spectra University
Science Books: Mill Valley, California, 1980.

M. Karni and A. Mandelbaum,Org. Mass Spectrom. 15, 53 (1980).
T. Keough, Anal. Chem., 54, 2540 (1982).

J .-L. M. Abboud, W.J. Hehre and R.W. Taft, J. Am. Chem. Soc., 98,
6072 (1976).

P. N. Rylander, S. Meyerson and H.M Grubb, J. Am. Chem. Soc., 79,
842 (1957).

F. McLafferty and F. Bockhoff, Org. Mass Spectrom., 14, 181 (1979).
F. McLafl‘erty and J .J . Winkler, J. Am. Chem. Soc., 96, 5182 (1974).
PP. Dymerski and F. McLafferty, J. Am. Chem. Soc., 98, 6070 ( 1976).

J .M. Buschek, J .J . Ridal and J .L. Holmes, Org. Mass Spectrom., 23,
543 (1988).

S. Olesik, T. Baer, J .C. Morrow, J .J . Ridal, J. Buschek and J .L.
Holmes, Org. Mass Spectrom., 24, 1008 (1989).

 

1 19.

120.
121.

122.
123.
124.
125.
126.

127.
128.

129.
130.

131.
132.
133.
134.

135.
136.

227
J .-A. A. Jackson, S.G. Lias and P.J. Ausloos, J. Am. Chem. Soc., 99,
515 (1977).
RC. Dunbar, J. Am. Chem. Soc., 97, 1382 (1975).

J .Shen, R.C. Dunbar and GA. Olah, J. Am. Chem. Soc., 96, 6227
(1974).

P. Ausloos, J .A. Jackson and S.G. Lias Int. J. Mass Spectrom. Ion
Phys, 33, 269 (1980).

RD. Wieting, R.H. Staley and J .L. Beauchamp, J. Am. Chem. Soc. ,
96, 7552 (1974).

H.W. Leung, H. Ichikawa, Y.-H. Li and A.G. Harrison, J. Am. Chem.
Soc., 100, 2479 (1978).

T. Baer, J .C. Morrow, J .D. Shao and S. Olesik, J. Am. Chem. Soc. ,
110, 5633 (1988).

V.H. Wysocki and H.I. Kenttamaa, J. Am. Chem. Soc., 112, 5110
(1990).

S. Hammerum, Mass Spectrom Reviews , 7, 123 (1988).

RC. Daugherty, M.J. Whitaker, L.M. Smith, D.L. Stalling and D.W.
Kuehl, Environ. Health Perspec., 36, 103 (1980).

El. Onuska and KA. Terry, Anal. Chem, 57, 801 (1985).

R.G. McLoughlin, J .D. Morrison and J .C. Traeger, Org. Mass
Spectrom., 14, 104 (1979).

J .A.D. Zeevaart and R.A. Creelman, Annu. Rev. Plant Physiol. Plant
Mol. Biol, 39, 439 (1988).

R.A. Creelman, D.A. Gage, J .T. Stults and J .A.D. Zeevart, Plant
Physiol., 85, 726 (1987).

DA. Gage, F. Fong and J .A.D. Zeevart, Plant Physiol., 89, 1039
(1989).

J .A.D. Zeevaart, T.G. Heath and DA. Gage, Plant Physiol., 91, 1594
(1989).

CD. Rock and J .A.D. Zeevaart, Plant Physiol., 93, 915 (1990).

A.G. Netting, B.V. Milborrow, G.T. Vaughan and RD. Lidgard,
Biomed. Environ. Mass Spectrom. 15, 375 (1988).

 

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.
149.

150.

151.

152.

228

E.A. Stemmler and M.V. Buchanan, Org. Mass Spectrom., 24, 94
(1989).

E.A. Stemmler and M.V. Buchanan, Org. Mass Spectrom., 24, 705
(1989).

J .R. Hass, M.D. Friesen and M. Hoffman, Org. Mass Spectrom., 14,
9 (1979).

W.F. Miles, N.P. Gurprasad and GP. Malis, Anal. Chem, 57 , 1133
(1985).

D. Fung, R.K. Boyd, S. Safe and B.G. Chittim, Biomed. Environ.
Mass Spectrom., 12, 247 (1985).

KL. Busch, A. Norstrom, M.M Bursey, J .R. Hass and CA. Nilsson,
Biomed. Environ. Mass Spectrom. 6, 157 (1979).

D.F. Hunt, T.M. Harvey and J .W. Russell, J. Chem. Soc. Chem.
Commun. 151 (1975).

B. Arbogast, W.L. Budde, M. Deinzer, R.G. Dougherty, J.
Eichelberger, R.P. Foltz, C.C. Grimm, R.A. Hites, C. Sakashita and
E. Stemmler, Org. Mass Spectrom., 25, 191 (1990).

B.V. Milborrow, Chem. Commun. 966 (1969).

SJ. Neill, R. Horgan. In: Principles and Practice of Plant Hormone
Analysis, Vol. 1, eds, L. Rivier and A. Crosier, pp. 111-167.
Academic, London (1988).

RT. Gray, R. Mallaby, G. Ryback and VP. Williams, J. Chem. Soc.
Perkins Trans. 2, 919 (1974).

R.A. Creelman and J .A.D. Zeevaart, Plant Physiol. 75, 166 (1984).

A.H. Andrist, C.H. DuPuy and RR. Squires, J. Am. Chem. Soc., 106,
845 (1984).

R.J. Schmitt, V.M. Bierbaum and CH. DePuy, J. Am. Chem. Soc.,
101, 6443 (1979).

V.M. Bierbaum, R.T. Schmitt and CH. DePuy, Environ. Health
Perspect. 36, 119 (1980).

RS. Callery, W.A. Garland and BK. Fukuda,Org. Mass Spectrom. ,
24, 385 (1989).

 

153.

154.
155.

156.

157.

158.

159.

160.
161.

162.

163.

164.
165.
166.

167.

168.

169.

170.

229
L.J. Sears, J .A. Campbell and ER Grimsrud, Biomed. Environ.
Mass Spectrom. 14, 401 (1987).
CD. Rock and J .A.D. Zeevaart, Plant Physiol. 89, S-113 (1989).

K. Kayganich, T.G. Heath and J .T. Watson, J. Am. Soc. Mass
Spectrom. l, 341 (1990).

G. G. Dolnikowski, T.G. Heath, J .T. Watson, J .H. Scrivens and CH.
Rolando J. Am. Soc. Mass Spectrom., in press (1990).

GB. Her and J .T. Watson, Biomed. Environ. Mass Spectrom, 13, 57
(1986).

Y. Kasuya, J .R. Althaus, J .P. Freeman, R.K. Mitchum and J .P.
Skelly, J. Pharm. Sci, 73, 446 (1984).

R.A. Yost, D.D. Fetterolf, J .R. Hass, D. J. Harvan. A. F. Weston,
P.A. Skotnicki and NM. Simon, Anal. Chem, 56, 2223 (1984).

J .V. Johnson, R.Yost and K.F. Faull, Anal. Chem, 56, 1655 ( 1984).

H. Schweer, H. W. Seyberth, C.O. Meese and O. Furst, Biomed. and
Environ. Mass Spectrom., 15, 143 (1988).

PH. Dawson, J .B. French, J .A. Buckley, D.J. Douglass and D.
Simmons, Org. Mass Spectrom., 17, 205 (1982).

A.J. Alexander and BK. Boyd, Int. J. Mass Spectrom. Ion
Processes, 90, 211(1989).

P.H. Dawson, Int. J. Mass Spectrom. Ion Phys., 43, 195 (1982).
D. F. Hunt and F.W. Crow, Anal. Chem, 50, 1781 (1978).

G. Bouchoux, Y. Hoppilliard and R. Houriet, New J. Chem. 11, 225
(1987).

HE. Audier, C. Monteiro and D. Robin, Org. Mass Spectrom. , 24,
353 (1989).

HE. Audier, D. Berthomieu, P. Hudhomme, C. Monteiro and P.
Mourgues, Org. Mass Spectrom., 25, 87 (1990).

J .-P. Morizur, I. Martigny, J. Tortajada and S. Geribaldi, Org. Mass
Spectrom., 25, 89 (1990).

C. Wesdemiotis and F.W. McLafferty, Tetrahedron, 37, 3111 (1981).

 

171.

172.

173.

174.

175.
176.

177.

178.
179.

180.

181.

182.
183.

184.

185.

186.
187.
188.

230
C. Wesdemiotis and F.W. McLafferty, Org. Mass Spectrom., 16, 381
(1981).
A. Maquestiau, C. Jortay, D. Beugnies, R. Flammang, R. Houriet, E.
Rolli and G. Bouchoux, Int. J. Mass Spectrom. Ion Processes, 82, 33
(1988).

J .C. Kleingeld and N.M.M. Nibbering, Org. Mass Spectrom., 17, 136
(1982).

G. Hvistendahl and DH. Williams, J. Am. Chem. Soc., 97 , 3097
(1975).

RD. Bowen and A.G. Harrison, Org. Mass Spectrom., 16, 159 (1981).

RD. Bowen, D.H. Williams, G. Hvistendahl and J .R. Kalman, Org.
Mass Spectrom., 13, 721 (1978).

G. Hvistendah. R.D. Bowen and DH. Williams, J. Chem. Soc. Chem.
Commun. 294 (1976).

F. W. McLafferty and I. Sakai, Org. Mass Spectrum, 7, 971 (1973).

W. Wagner, H. Heimbach and K. Levsen, Int. J. Mass Spectrom. Ion
Phys., 36, 125 (1980).

A.G. Harrison, T. Gaumann and D. Stahl, Org. Mass Spectrom., 18,
517 ( 1983).

S.J.A. Curtis and A.G. Harrison, J. Am. Soc. Mass Spectrom., l,
301 (1990).

R.H. Nobes and L. Radom, Org. Mass Spectrom, 19, 385 (1984).

D.J. McAdoo and CE. Hudson, Int. J. Mass Spectrom Ion
Processes, 88, 133 ( 1989).

G. Bouchoux, F. Djazi, Y. Hoppilliard, R. Houriet and E. Rolli, Org.
Mass Spectrom., 21, 209 (1986).

U. Buck; X. Gu, C. Lauenstein and A. Rudolph, J. Phys. Chem. 92,
5561 (1988).

S. Morgan and A.W. Castleman, J. Phys. Chem. 93, 4544 (1989).
J .M.S. Henis, J. Am. Chem. Soc., 90, 844 (1968).
A. J. Illies, Org. Mass Spectrom., 25, 73 (1990).

189.

190.
191.
192.
193.

C. Beaugrand, D. J aouen, H. szsldagh and C. Rolando, Anal.
Chem, 61, 1447 (1989).

P.W. Tiedemann and J .M. Riveros, J. Am. Chem. Soc., 96, 185 (1974).
A.G. Harrison, Can. J. Chem. 64, 1051 (1986).

T.W.G. Solomons, Organic Chemistry, Wiley: New York, 1980.

K. Beimann, Biomed. Environ. Mass Spectrom., 16, 99 (1988).

"lllllllllllllllllllls