11:3,.
,4... ,.
W fl“

4,144.,
(I upiylel'tf'lt
‘ 1,,4

. . . 4"
‘ M1“
' v: ‘nggfi £2“

1
:4“ WW
$4,141; I. 'm

"'4: ‘4 1
1.4.4,, 1‘“

1
““Mu

\ v
I‘ ._

NW 435
‘4“ 1 ‘1"

4, ‘4‘ 4114, .
414 4 .41.; . -» ,.
4 ‘4“ 1 13¢), 't‘e'g, "if”. -
“3136:" 11!“.9‘w ‘11.”, 1%
.4 4441444454441315‘ 5.4.4411 . «4:44.,

"1,4 [‘3' ' I (I: . . $3434“?
5“ ‘ W11“ 44th
A 1,,"

31:14:, I
5,444.24

~I

4:43;:
222'}?-

u 44..

44.35131

H "W0“;
‘4: .1441

£313“

$413,,»
an“ :‘L‘i‘il‘l

4W

:2
49'5”

‘1

44,444

“4;, -

"i 442?:
”‘79?" é:
xi" is";

15:“;

n

4%?”

‘1 “1“: 3‘1“.
71*“ a...
*m 51:42“ $53344“?

w ,!L w'-
”‘4’; “a.

u

 

4.
.4
1
‘4
M‘U‘l

1
v ( 1 .14!
4.4 , «4, V

m,’ 351““:
[3 #1414"
4.4%; w’fiufia'
‘ 13.1;i‘g‘g‘11u‘gl’x“ ’
(‘R‘ ‘. 21;“:
, ~:31?
;.4

4‘”

h ‘4’. .0 ., 1:43;
4‘1
‘4 "‘4?

_»‘i'v 44“.”:

. :33
1:18,. .1 ' 4;
x 43444
kg '3‘}; 1.3%.; 11?. n

;{.’1'
|-- m

khai‘frw
“44% ,

w
I,

‘u
4::

91x I if."
u . "v‘tzt

'(
‘41.; 3,

WW:

.1441
5144,44:

 

M ‘ t mififi'a‘z
‘ .
\ ‘1
“ I. 1,1. .
. 1;: ‘ .4.
143k ‘ ' ir'fll
WWW} 4

“M .11" ,.
3““

fi‘é?";ma

44.4
‘ ‘41

£133.“
1."‘4

‘02-‘44
.53}.
fiflfiu

1" 4%,,

'11.

‘3 44,711! “i
*112’,‘ 1,

‘12:

n
N

“‘1.

I

3' 45%
“‘fiwfi Hm: '

“211.591. % 554'“ (‘55. k5?“
‘K‘agfiu wg‘é‘: 15,): gig-“)- ’1‘
fl“ " m
"‘6" . 4““,
{‘3‘ I L *
'L :13“ ‘

{‘4‘

' 1,,I. _

’ ""“3'1,i5 1.4“ 4 4

' ' LN;- , ,

£4411 m‘g‘é's‘, 11' ' xégm-‘sfl, 441%.: ,7
1‘ . ‘h'wh Mr;

. 5-4 14%1'1‘3‘3‘ 9|,

.'.~x,Ԥ-:E'. ,_ ,-

. ‘34,
5‘11-

. I“ 'Iui‘i
““ L‘ ‘W
‘ Y

““4.

w
_ .. Sf:
:11:

yas-A, 3&7“.

341% 31°41; W41;
"im .5

1‘ 'y. .
4,55%, ,4,
' 1

4 ’44:;

LL V' 3:1 4'
4. ‘1“ 1741344234
12444.4

.ngt‘?

“w :9 13.41":
“1:14 ,_
“mm: 4

1“ 4:14 4

.1» WSW? :51

N
" 111?;

L1

‘1”. 1
‘1"? ,

m .

. - ‘5").
i 5 r: - '
5 I! .731 51 I i-f‘é. '\ tt“. .3; 'E'

trait again bifitéfié
University

 

W

This is to certify that the

dissertation entitled

COLLISION PROCESSES IN
TRIPLE QUADRUPOLE MASS SPECTROMETRY:
CHAR \CTERIS'I'ICS A1“) APPLICATIONS

presented by

John Anthony Chakel

has been accepted towards fulfillment
of the requirements for

Ph . D Chemistry

degree in

 

 

fléfifi

Major professor

Date December 1, 1986

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-1 2771

 

 

 

MSU

“LIBRARIES
n

v

 

 

RETURNING MATERIALS:
Place in book drop to’_
remove thi¥=checkout from
your record; FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

 

COLLISION PROCESSES IN TRIPLE QUADRUPOLE MASS SPECTROMETRY:
CHARACTERISTICS AND APPLICATIONS

By

John Anthony Chakel

A DISSERTATION

Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1982

ABSTRACT

COLLISION PROCESSES IN TRIPLE QUADRUPOLE MASS SPECTROMETRY:
CHARACTERISTICS AND APPLICATIONS

By
John Anthony Chakel

The characteristics of the collision processes in the
quadrupole collision chamber of a triple quadrupole mass
spectrometer are explored. The results are used to aid in
the analytical application of collisionally activated
dissociation (CAD) and collisionally activated reaction
(CAR) which occur through the ion-molecule interactions in

the quadrupole collision chamber.

Ions formed in the ion source are selected by their
mass-to-charge ratio in the first quadrupole to pass into
the pressurized RF-only quadrupole collision chamber.
Interaction of the low energy (< 200 eV) selected ion with
the target gas present in the collision chamber leads to an
increase of the ion’s internal energy and its subsequent
dissociation or to an ion-molecule reaction with the target
gas. The ionic collision products are mass analyzed by the
third quadrupole and detected. An RF-only quadrupole
exhibits no discrimination of mass-to-charge values and
strongly focuses all scattered ions. The spectra produced

via CAD or CAR in a triple quadrupole mass spectrometer are

I‘ll-l-ulin____

John Anthony Chakel

easy to interpret as unit mass resolution is preserved for
both selection of reactant ions and analysis of ionic

products.

The interaction between the ion and neutral molecule
at low collision energies may occur over several
vibrational periods. This long interaction time permits
association and substitution reactions to occur in addition
to fragmentation and charge exchange reactions. At higher
collision energies (200 eV) charge inversion reactions may
occur. Important variables in the collision process are the
collision energy and the collision gas type. The pressure
of the collision gas affects the number of collisions and
the collision gas species affects the type of products

observed.

An application of CAD to the identification of
compounds present in the acid fraction of the priority
pollutants is presented. The relative merits of different
ionization methods for this analysis are discussed as is

the usefulness of multiple reaction monitoring.

The application of CAR for differentiating among
different hydroxy and halogen containing aromatic
positional isomers is described. Significant differences in
the reaction products of the selected ion with particular
collision gases was shown to provide structural information

and allow differentiation among the chosen isomers.

To my Mother and Father, and to Lynn

ii

ACKNOWLEDGEMENTS

I would like to express my deepest gratitude to Dr.
Christie G. Enke, under whose guidance and insight this
research was performed. I will always value his friendship

and counsel.

I acknowledge my guidance committee at Michigan State
University, including Drs. John Allison, Steven Aust, John
F. Holland and George E. Leroi. Special thanks are
extended to John Allison and George Leroi for their
invaluable insight and friendship. I would also like to
thank Dr. Tom Atkinson for his untiring assistance with

computer graphics on many occasions over the years.

I am also grateful for the opportunity to get to know
many special people during my stay at Michigan State. The
faculty, the staff and the Enke research group members
have all helped to enrich my life and I value their

friendship.

Finally, I acknowledge my family for their love and
support in this endeavor. My parents and my brother Jim
have all helped make this research worthwhile. Most of
all, I thank my wife Lynn for her love and encouragement.

She has helped to make all this possible.

iii

TABLE OF CONTENTS

Page
LIST OF TABLES ......................................... vii
LIST OF FIGURES ........................................ ix
CHAPTER ONE. INTRODUCTION .............................. 1
Organization of the Thesis ................ 6
Historical Perspectives on
Exosource Fragment Formation
in Mass Spectrometry ...................... 8
Collision Processes in Mass
Spectrometry ............................. 11
Overview of TQMS Instrumentation ......... 15
Modes and Utility of Operation ........... 20
CHAPTER TWO. CHARACTERISTICS OF THE COLLISION
PROCESS IN TQMS .......................... 24
Collisions of the Methane
Molecular Ion ............................ 29
Introduction ......................... 29
Experimental ......................... 31
Results and Discussion ............... 32
Dependence of CAD Products
on Collision Chamber
Pressure, RF Voltage and
Target Gas Species ............... 34
Energy Dependence of CAD
Products ......................... 47
Pressure Dependence of CAD
Products ......................... 49
Conclusions .......................... 57

iv

 

CHAPTER THREE.

CHAPTER FOUR.

CHAPTER FIVE

COMPARISON OF DIFFERENT TQMS

INSTRUMENT CONFIGURATIONS ..............

A. Collision of Bromobenzene

Molecular Ion ..........................
Introduction .......................
Experimental .......................
Results and Discussion .............

Conclusions ........................

B. Investigation of Charge Inversion
in the Collision Chamber in Triple

Quadrupole Mass Spectrometry ...........
Introduction .......................
Experimental .......................
Results and Discussion .............

Conclusions ........................

APPLICATION OF COLLISIONALLY
ACTIVATED DISSOCIATIONS FOR COMPONENT

IDENTIFICATION IN A PHENOLIC MIXTURE...
Introduction ...........................
Experimental ...........................
Results and Discussion .................

Conclusions ............................

APPLICATION OF COLLISIONALLY
ACTIVATED REACTIONS FOR
DIFFERENTIATING BETWEEN POSITIONAL

ISOMERS ................................
Introduction ...........................

Experimental ...........................

..61

..61
..61
..63
..65
..83

..85
..85
..87
..89
..97

.100
.100
.103
.107
.124

.126
.126
.134

Results and Discussion .................. 135

Reaction with N2 .................... 136

Reaction with NHs ................... 136

Reaction with D20 ................... 142

Reaction with SF6 ................... 149

Conclusions ............................. 151

LIST OF REFERENCES .................................... 153

vi

 

LIST OF TABLES

Page
Table 2.1 Slopes of the pressure dependence
curves for CAD product formation at
17.5 eV lab collision energy ............... 40

Table 2.2 Cross sections for CAD product
formation at 17.5 eV lab collision
energy ..................................... 42

Table 2.3 Collision process efficiencies as a
function of the RF voltage on the
quadrupole collision chamber ............... 43

Table 2.4 Collision process efficiencies with
argon, nitrogen and helium targets ......... 46

Table 2.5 Collision energy dependence of CAD
fragment ion yield ......................... 48

Table 2.6 Percent total ion current of
components present in nitrogen
collision gas. 70 eV EI .................... 51

Table 2.7 Cross sections for CAD/CAR product
formation at 4.5 eV lab collision
energy ..................................... 58

Table 3.1 Ions resulting from collision of
bromobenzene with nitrogen collision
gas at 25 eV lab collision energy .......... 68

Table 3.2 Effect of decreasing third
quadrupole rod offsets on peak shape ....... 95

Table 3.3 Relative abundances of ions formed
from charge inversion of hippuric
acid at various third quadrupole rod
offsets .................................... 96

Table 3.4 Relative abundances of ions formed
from charge inversion of hippuric
acid at various second quadrupole
pressures ................................. 109

Table 4.1 CAD spectra of 20eV EI generated
ions. Fragment loss from selected
ion (% RA) ................................ 110

Table 4.2

Table 4.3

Table 5.1

CAD spectra of N01 generated ions.
Fragment loss from selected ion

(% RA) .................................... 112

Multiple reaction monitoring

reactions for N01 ......................... 123

Collision activated reactions for

different collision gases ................. 130
viii

 

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

LIST OF FIGURES

Page
Schematic of triple quadrupole mass
spectrometer ................................ 3
Vacuum chamber and quadrupole
arrangement ................................ 18

CAD (N2) spectrum of CH4+ at 17.5 eV
lab collision energy. 6.6 X 10'4 torr N2...33

Efficiency terms for the CAD process
in the quadrupole collision chamber ........ 35

Efficiencies for the collision of
CH4+ with N2 at 17.5 eV lab
collision energy ........................... 36

Pressure dependence of CAD (N2 )
product ions at 17.5 eV lab
collision energy ........................... 38

CAR (N2 ) spectrum of CH4 t at 4 . 5 eV
lab collision energy. 6.6 X 10'4
torr N2 .................................... 50

Efficiencies for the collision of
CH4+ with N2 at 4.5 eV lab collsion
energy ..................................... 52

CAR (N2) spectra of CH4+ (upper) and
CD4+ (lower) at 4.5 eV lab collision
energy. 6.6 X 10“| torr N2 ................. 54

CAD (N2) spectrum of CeHsBr+ at 25 eV
lab collision energy. 3.8 X 10'4
torr N2 .................................... 67

Dependence of bromobenzene CAD (N2)
fragment ion 77+ (CeHst) on second
quadrupole RF setting ...................... 69

Pressure dependence of bromobenzene
parent and CAD (N2) fragment ions at
25 eV lab collision energy ................. 71

Pressure dependence of bromobenzene

CAD (N2) fragment ions 77+, 94+, 51+

and 65+ at 25 eV lab collision

energy ..................................... 73

ix

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

3.

as»

.10

11

.12

Pressure dependence of bromobenzene

CAD (N2) fragment ions 107+, 128+ and

116+ at 25 eV lab collision

energy ..................................... 74

Dependence of 77+ formation on
second quadrupole dc offset ................ 76

Dependence of 77+ formation on

second quadrupole dc offset with
concurrent sweep of interquadrupole

lens and third quadrupole dc

offsets .................................... 77

Dependence of bromobenzene parent
(156+) and CAD (N2) fragments (77*,

51+ and 94*) on collision energy ........... 78
Relative cross sections for
bromobenzene CAD (N2) as a function
of collision energy ........................ 82

Structure and charge inversion
spectrum of hippuric acid.

2.6 X 10'4 torr SF8 collision gas .......... 90
Charge inversion spectrum of

hippuric acid without drawout

potential on interquadrupole lens .......... 92
Charge inversion spectrum of

hippuric acid with drawout

potentional on interquadrupole lens ........ 93
Acid extractable priority pollutants ...... 104

Quadrupole MS/MS technique for
mixture analysis of both positive

and negative ions ......................... 106
20 eV EI spectrum of mixture .............. 115
OH” NCI spectrum of mixture ............... 116

20 eV EI CAD (Ar) spectra of mixture
(a) and reference (b) m/z 142 M+ ion
for p-chloro-m-cresol ..................... 117

20 eV EI CAD (Ar) spectra of mixture
(a) and reference (b) m/z 142 M+ ion
for 2,4,6-trichlorophenol ................. 118

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

NCI CAD (Ar) spectra of mixture (a)
and reference (b) m/z 138 (M-H)’ ion
for o-nitrophenol ......................... 120

N01 CAD (Ar) spectra of mixture (a)
and reference (b) m/z 138 (M-H)‘ ion
for p-nitrophenol ......................... 121

N01 CAD (Ar) spectra of mixture (a)
and reference (b) m/z 183 (M-H)' ion
for 2,4-dinitrophenol ..................... 122

Isomers selected for this study ........... 133

CAD (N2) spectra of IVa and IVb m/z
312 M+ ion (012Ha79Br81Br). 93 V RF
second quadrupole ......................... 137

CAR (NHa) spectra of IVa and IVb m/z
312 M+ ion. 93 V RF second

quadrupole ................................ 138
CAR (NHs ) spectra of IIIa - IIIc

m/z 234 M+ ion (012H931Br). 93 V RF

second quadrupole ......................... 139

CAR (NHa) spectra of la, Ib and lo
m/z 146 M+ ion (CeH435Clz). 67 V RF
second quadrupole ......................... 141

CAR (D20) hydrogen/deuterium
exchange spectra of Ia, Ib and lo
m/z 145 (M-H)' ion. 67 V RF second

quadrupole ................................ 143
CAR (D20) hydrogen/deuterium
exchange spectra of 11a - IIc m/z

169 (M-H)' ion. 67 V RF second

quadrupole ................................ 144
CAR (D20) hydrogen/deuterium

exchange spectra of IVa and IVb m/z
311 (M-H)‘ ion. 160 V RF second
quadrupole ................................ 146

Pressure dependence of CAR (D20)

reaction products for IVb. 3 eV lab
collision energy .......................... 147

xi

 

 

Figure 5.10

Figure 5.11

Collision energy dependence of CAR
(D20) reaction products for IVb.

2.5 X 10'4 torr D20 ....................... 148
CAR (SF6) charge inversion spectra

of 11a - IIc m/z 169 (M-H)' ion.

2.5 X 10“4 torr SFs. 67 V RF second
quadrupole ................................ 150

CHAPTER ONE

INTRODUCTION

In the past few years there has been a rapid increase
in the use of a technique in which two mass spectrometers
are coupled in tandem for chemical analysis. This
technique, called mass spectrometry/mass spectrometry
(MS/MS), performs both separation and identification in

the same instrument resulting in a powerful analytical

 

tool with broad utility similar to that experienced with
GC/MS, LC/MS, GC/IR and other coupled techniques. In
MS/MS, ions generated in the source from the sample may be
selected by the first MS device and allowed to enter a
collision region preceding the second MS device. In this
region the selected ion may undergo unimolecular
decomposition resulting in metastable ions or the ion may
encounter a neutral target molecule and gain sufficient
internal energy from conversion of kinetic energy to
induce dissociation. Ionic products of these reactions are
then analyzed by the second MS device and are subsequently
detected. The presence of the second MS device gives rise
to an extra dimension of information which is extremely
valuable in both complex mixture analysis and structure

elucidation as will be discussed later.

Early applications of MS/MS were performed with

tandem sector instruments, the most prevalent of these

 

being mass-analyzed ion kinetic energy spectrometry, or
MIKES. In MIKES, which utilizes a reversed geometry (B-E)
double focusing MS, ions from the source are accelerated
by a high voltage (3,000-10,000 V) and enter the magnetic
sector where they are analyzed according to their
momentum. The selected ion is then allowed to undergo
unimolecular decomposition or collision induced
dissociation (CID) with a target gas molecule in the
field-free region preceding the electric sector. The
resulting fragment ions then enter the electric sector
where they are energy analyzed and an ion kinetic energy
(IKE) spectrum is recorded. The fragment ions appear in
the IKE spectrum at energies corresponding to (m2/m1)E1,
where m2 and mi are the fragment ion and parent ion
masses, respectively, and E1 is the parent ion kinetic
energy. The collision process in the field-free region
involves a kinetic energy release which produces a spread
of ion velocities. This velocity spread serves to broaden
the peaks obtained by scanning the energy-selective
electric sector in MIKES MS/MS instruments and results in

poorly resolved spectra of daughter ions.

Rather than use a high energy sector-type instrument
for performing MS/MS, our lab has pioneered the
development of a tandem quadrupole instrument in which
three quadrupoles are used. The resulting instrument, a
triple quadrupole mass spectrometer (TQMS) is well suited

for an investigation of the analytical applications of and

the fundamental processes occurring in this form of MS/MS.
Briefly, this instrument consists of an ion source, a
quadrupole mass filter, a radio frequency (RF) only
quadrupole collision chamber, a second quadrupole mass
filter, a conversion dynode and an electron multiplier

detector as shown in Figure 1.1.

 

 

 

 

 

ION QUAD QUAD COLUSION QUAD PARTICLE
SOURCE MASS FILTER CHAMBER MASS FILTER MULTIPLIER
‘ (é .
SAMPLE ION ION FRAGMENTATION PRODUCT ION ION
IONIZATION SELECTION Von REACTION SELECTION DETECTION

Figure 1.1 Schematic of Triple Quadrupole Mass
Spectrometer

Ions generated in the source are mass-selected by the
first quadrupole. The selected ion then enters the
pressurized RF-only quadrupole at low energies of
typically 0-50 eV where characteristic products are
produced through the interaction of the selected ion with
a neutral target gas. These products may be formed by a
collisionally activated dissociation (CAD) or a
collisionally activated reaction (CAR). The ionic products
of the collision are focused into and mass-analyzed by the
third quadrupole. An RF-only quadrupole was chosen as the
collision chamber since it exhibits no mass discrimination

over a wide range of mass to charge (m/z) values and

strongly focuses all scattered ions resulting in an
extremely high transmission efficiency of nearly 100%. The
spectra which are produced via CAD or CAR in a quadrupole
MS/MS are easy to interpret since the unit mass resolution
of the quadrupole is independent of ion energy over a

fairly wide value.

The collision processes in the high energy sector—
type MS/MS instrument and the low energy quadrupole MS/MS
instrument are markedly different. In high energy
instruments the collision process is well characterized
and involves a conversion of kinetic to internal energy
through a direct vertical Franck-Condon type of electronic
excitation (1). The selected ion thus excited may fragment
after relaxation of the electronic energy into vibrational
levels. Product ions resulting from dissociation, charge
stripping, charge inversion, and charge exchange may be
formed. No product ions resulting from association or
substitution reactions with the target gas may be formed
since the electronic excitation occurs over a time much
shorter than a vibrational period. In contrast, the
collision process in low energy quadrupole instruments is
not well characterized. The selected ion energy is only a
small fraction of that in the high energy instruments,
much less than that needed for a similar electronic
excitation process. Excitation is likely the result of a
transfer of momentum to vibrational excitation, followed

by subsequent fragmentation (2). The interaction time

between the ion and neutral molecule at these low energies
can be several vibrational periods long. This long
interaction time permits association and substitution
reactions to occur in addition to the fragmentation and
charge exchange reactions seen at both high and low
energies. Charge inversion reactions begin to appear at

collision energies of ca. 100 - 200 eV.

Important variables which affect observed products in
the low energy collision process of TQMS are the collision
energy and the collision gas type. While the fragmentation
spectra observed in the high energy system are usually
invariant with respect to the collision energy over a
large range of energies, the type and abundance of
products observed in the low energy system are quite
sensitive to the collision energy. The nature of the
collision gas can markedly affect production formation.
The use of a noble gas results primarily in CAD products.
The use of a "reactive" collision gas may result in CAR
products formed from the ion-neutral molecule complex. The
pressure of the selected collision gas also affects the
type of products observed. At low collision gas pressures
only products resulting from single collisions are
observed. As the collision gas pressure is increased,
multiple collisions become significant and products
arising from second and higher order reactions can be

observed.

The analytical utility of TQMS can be easily
realized. The ability to fragment each selected source ion
in the second quadrupole has obvious advantages in
structure determination by TQMS. With a source which
provides molecular ions, the ability to mass select the
different molecular ions and subject them to CAD or CAR
provides spectra characteristic of each selected
component. This permits mixture analysis of unprepared
samples as long as sufficient quantities are present for

detection.

The research described in this dissertation focuses
on exploring the characteristics of the low energy
collision process in TQMS and utilizing these
characteristics to aid in maximizing the analytical

applications of TQMS.

The thesis is divided into five chapters. This first
chapter serves to provide background information on the
collision processes in the techniques of MS/MS, especially
those occurring in TQMS. An overview of the Michigan State
University TQMS instrument is presented including a
discussion of the ion optical path, system capabilities,

modes of operation and information obtainable.

The second chapter explores in detail the
characteristics of the collision process in TQMS as
observed with the collision of the methane molecular ion
with various target gas molecules. In-depth studies on the
pressure and energy dependence of fragmentation and
reaction products are reported and certain kinetic data
are determined. The characteristics observed with these
studies provide the background for the analytical

applications of TQMS which follow.

A comparison of different TQMS instrument
configurations is the topic of the third chapter. This
chapter is divided into two parts. In the first part, the
pressure and energy dependence of CAD products from the
collision of bromobenzene molecular ion with nitrogen
collision gas is determined. In the second part, the
observation of charge inversion reactions is briefly
investigated. The results obtained from these studies are
compared with those from a commercial TQMS instrument and
conclusions are drawn regarding the relative instrumental

configurations.

In the fourth chapter an application of CAD to the
identification of compounds in an eleven-component
phenolic mixture is presented. The relative merits of
different ionization methods for this analysis are

discussed as is the usefulness of multiple reaction

monitoring for simultaneous determination of each

component.

The application of CAR for differentiating between
positional isomers is described in the fifth chapter.
Positional isomers of various hydroxy and halo aromatic
compounds of environmental significance were studied. The
reactions of the selected ion with an appropriate
collision gas were shown to provide spectra which differed

significantly among the isomers studied.

Historical Perspectiyes 9n Exosource Fragment Formation
inMaLSSpeoizmetrx

Perhaps the first observance of fragment ion
formation external to the ion source occurred with the
peaks arising from metastable decompositions. These
decompositions are slow (rate constants of 10*5 - 10'3

5‘1) so the ions responsible for metastable peaks may be
sufficiently stable to leave the ion source, but undergo
unimolecular decomposition prior to analysis. The first
use of metastable peaks to elucidate fragmentation
pathways was proposed in 1945 (3). The location of peaks
from metastable decompositions observed in double focusing
mass spectrometers allowed both the precursor ion and the
fragment ion to be determined, resulting in selected ion
fragmentation spectra. The scope of metastable peak

observations has been recently reviewed (4).

Selected ion fragmentation spectra which are richer
in information content can be obtained by the use of a
collision gas to increase the internal energy of ions in a
selected ion beam (5,6). The resulting fragmentation
spectrum often resembles the spectrum resulting from
electron impact. The collision gas is introduced into a
collision cell which may be in a field free region of the
mass spectrometer. The collision between the selected ion
and the neutral target gas molecule converts some of the
kinetic energy of the ion into internal energy which may
then be randomized throughout the internal degrees of
freedom of the ion. The thus excited ion may then
dissociate. Early applications of this collisional
activation process were demonstrated for dissociation
products of simple molecules CO+ (7) and 02+ (8). This
technique which is called CID or CAD came into popular use
following a report on the CID of aromatic molecular ions
by Jennings in 1968 (5). At first, the collision gas was
present throughout the entire field-free regions of a mass
spectrometer (5,6). Later developments confined the
collision gas to a small region by use of specially
designed collision cells (9,10). Additional versatility
was introduced with collision cells which could be set to
a chosen potential relative to the ion source (11,12).
Until 1978 all of the CID spectra were obtained with
sector type mass spectrometers. Most prominent was the use

of a reversed geometry double focusing mass spectrometer

10

with the technique called MIKES (13). The collision
between the selected ion and the neutral target gas occurs

at ion kinetic energies in the kilovolt range.

In 1978 the technique of performing selected ion
fragmentation in a triple quadrupole system was
demonstrated (2,14). A quadrupole mass filter operated in
an RF-only mode is used as the collision cell. In this
system the collision between the selected ion and the
neutral target gas occurs at low ion kinetic energies (ca.
2 - 20 eV). These low energy collisions provided a highly
efficient method for fragmenting ions. The excitation
process is different than that in the MIKES type
instrument and involves a vibrational excitation of the
selected ion by momentum transfer. Scattering losses were
greatly reduced by the strong positive focusing fields of

an RF-only quadrupole.

Collision induced dissociations have since been
observed in a number of mass spectrometers with differing
configurations. A double quadrupole instrument with a
short (2.5 mm) grounded collision cell in the
interquadrupole region has been shown to be an efficient
device for performing selected ion fragmentation (15). A
"hybrid" instrument consisting of a magnetic sector
followed by an RF-only quadrupole collision cell and a
quadrupole mass filter was also shown to be effective for

collision induced dissociation studies (16). A Fourier-

11

transform (FT) mass spectrometer has been shown to be an
interesting alternative to MIKES type and quadrupole type
instruments for obtaining collision induced dissociation
spectra (17). Stepwise dissociation of CID products may be
monitored by using the appropriate sequence of double-
resonance pulses. Such studies are not possible with the
MIKES type and quadrupole type instruments. Finally,
collisional activation studies performed in a high
pressure drift tube source designed by Munson have

recently been reported (18).

Cnilisienfimessesinflassfieesfrenatrx

At the heart of the collision processes responsible
for collisionally activated dissociations and reactions
are ion-molecule reactions. Studies of ion-molecule
reactions began with the advent of mass spectrometry and
the formation of H3+ from the collision of H2+ with H2,
reported by Dempster in 1916 (19). The significance of
studying ion-molecule reactions is realized when one
considers that they are a significant part of the
chemistry of the upper atmosphere, flames, electrical
discharges, radiation phenomena and interstellar
processes. Chemical reactions studied in the gas phase are
free from the effects of solvation, so a better
understanding of the interaction in terms of mechanism and

rates is possible. Ionic reactions in solution are often

12

influenced by the chemical properties of the solvent
molecules. Large compilations of thermochemical data on
ionic and neutral species have been obtained from measured
appearance potentials (20). Major treatises concerning the
nature of ion-molecule reactions have been published
(21,22). The formation of CH5+ from the reaction of CH4+
with CH4 reported by several groups in the 1950’s (23-25)
lead to the first analytical application of ion—molecule
reactions in the technique of chemical ionization (CI)

developed by Munson and Field in 1966 (27).

The rapid growth in ion-molecule studies since the
1960’s can be attributed to several factors: external
control of ions is possible in terms of the internal
energy they possess which is dependent on their method of
formation, their direction and their kinetic energy; the
cross section for ion-molecule interactions is an order of
magnitude greater than for neutral-neutral molecule
interactions; the pressure and temperature of neutrals is
under control of the experimenter allowing kinetic data
such as rate constants and cross sections to be obtained;
and the effects of various molecular parameters and
reaction conditions on the rate constant can be probed.
The greater reactivity of ions and neutrals versus
neutrals and neutrals arises from ion-permanent dipole and
ion-induced dipole forces (26); the ion and neutral

species approach along a reactive surface which gives rise

13

to sufficient energy to overcome many traditional reaction

barriers.

Many techniques have been developed for studying ion-
molecule reactions (21). Among them are the pulsed ion
source, high pressure and controlled temperature sources,
flowing afterglow, flow drift tubes and selected ion flow
tubes (SIFT). Along with these are the techniques of ion
cyclotron resonance (ICR), crossed beam, merged beam, and
tandem mass spectrometry with a collision chamber between

the separate mass analysis stages.

With the tandem mass spectrometry technique the
collisional activation which gives rise to dissociation or
reaction results from a conversion of the part of the
selected ion’s translational energy into internal energy.
The processes responsible for this conversion can arise
from either electronic or vibrational excitation.
Electronic excitation involves a vertical transition
predominant at ion energies above 1000 eV with small
target molecules while vibrational excitation is adiabatic
and favored at low ion energies with large target
molecules (1,28). The collision process in sector type
MS/MS instruments is a result of electronic excitation and
this process has been studied in detail (29,30). No long-
lived collision complexes are formed in the sector type
instruments as the excitation occurs much faster than a

vibrational period. Ions resulting from changes in

14

electronic state such as charge exchange and charge
inversion products are possible. The collision process in
quadrupole MS/MS instruments is typically a low energy
process and electronic excitation is not likely.
Vibrational excitation is favored at these low energies
and products arising from ion-molecule reactions are
possible, along with fragmentation products of the excited
selected ion. The low energy collision process in
quadrupole MS/MS has not been probed as deeply as the high
energy collision process and is an area of active

research.

With regard to the variables of the collision
process, the internal energy of the selected ion has been
reported to have little effect (6,30) or substantial
effect (31,32) on the fragmentation spectrum resulting
from collisional activation depending on whether or not
the peaks arising from low energy pathways such as
metastables are included. The translational energy of the
selected ion can control the which pathways are accessible
for dissociation. The mass of the target gas can control
the energy of the collision in terms of center of mass
coordinates. The nature of the target gas can control
whether reactive collisions occur as well as the
efficiency of collisional activation with polarizable
targets expected to give a higher cross section for

effective collisions (33). The pressure of the target gas

15

determines whether one or more collisions are responsible

for the resultant spectrum.

A more thorough understanding of the low energy ion-
molecule collision process is necessary so that analytical

applications of these processes may be developed.

Minds}: 9_f_ 191$ mtrumentation

The research for this dissertation was performed on
the TQMS instrument developed in the author’s laboratory
at Michigan State University. The TQMS first became
operational in Fall 1978 and since that time research has
led to a number of papers concerned with the basic
collision process and its analytical application (2,34-
39). During that period the TQMS instrument control and
acquisition hardware and software, ion optical path,
detection system, vacuum system, pressure and temperature
control, and inlet systems all underwent development and
improvement. The resultant instrument is an extremely
flexible research tool well suited for basic research into

the low-energy collision process.

Each device along the ion path is under keyboard
control and its potential is variable over the range of -
200 to +200 volts. The chemical ionization (CI) ion source
is from a Finnigan 3000 and the electron impact (EI)

source is of in-house design (40). The source lens

16

elements have been improved through extensive modeling
with ion trajectory simulations (41). The final source
lens element for both sources is an ELFSTH ferroceramic
cylinder (42) which is inserted a short distance into the
first quadrupole. The ELFST“ stone, which is a leaky
dielectric, aids in passing ions through the fringing
fields present at the ends of a quadrupole as it allows
the RF fields to pass and effectively blocks the mass

discriminating DC fields.

The first and third quadrupole mass filters are
Extranuclear Laboratories model 162-8 with 0.95 cm rod
diameter and overall length of 20.0 cm. Their mass range
covers 0 - 1000 atomic mass units and their resolving
power is adjustable to one part in 1500. The quadrupole
collision chamber was constructed in-house and is of
similar dimensions (0.954 cm X 21.6 cm). Currently, the
three quadrupoles are operated at different RF frequencies
of 1.7, 2.35 and 1.55 MHz for the first through third
quadrupole, respectively. Efforts are underway to drive
all three quadrupoles at the same RF frequency. The dc rod
offset voltages of each of the three quadrupoles may be
varied over a wide range and the difference between their
value and the ion source potential determines the ion’s
axial kinetic energy; and in the case of the quadrupole

collision chamber, the lab collision energy, Elab.

17

A major theoretical (41) and experimental effort
resulted in the present design of the interquadrupole
lenses for maximizing ion transmission between
quadrupoles. The two sets of interquadrupole lenses each
consist of four elements with the first and last element,
and ELFSTM ferroceramic cylinder, connected together. The
second lens element is a thick (0.350 inch) plate with a
large aperture (0.250 inch) and the third lens element is
a thinner plate (0.125 inch) with a reduced aperture (0.20
inch). The spacing between all elements is constant at
0.050 inch. In operation the interquadrupole lenses
provide efficient focusing of ions over a wide mass range

into the following quadrupole.

A large aperture lens is present at the end of the
third quadrupole followed by a conversion dynode (43) and
an off-axis Channeltron?H electron multiplier. The output
of the detector is fed into an electrometer amplifier and
the output of the amplifier is connected to a 12 bit
analog to digital converter. A detector dynamic range
covering five orders of magnitude is possible with this

configuration.

All of the ion path components with the exception of
the source are mounted off of the rear vacuum chamber

flange in the third chamber as shown in Figure 1.2.

18

Chamber 1 Chamber 2

01:le

 

Figure 1.2 Vacuum Chamber and Quadrupole Arrangement.

All necessary electrical feedthroughs and collision
gas inlet and pressure measurement lines are also
positioned on the rear flange. This design allows easy
removal and disassembly of the ion path and facilitates
future instrument development and modifications. The ion
source is easily interchanged by opening the vacuum

chamber at the Chamber 1/Chamber 2 interface.

The forty liter volume vacuum chamber consists of
three differentially pumped regions. In this fashion the
area near the ion source, the external environment of the
pressurized second quadrupole and the detector are each
maintained at a safe and good vacuum despite heavy gas
loads under CI and CAD operations. The turbomolecular
pumps, 500 l/s, 270 l/s and 270 1/5 in chambers 1, 2 and 3
respectively, provide a clean, oil free vacuum and permit

rapid switching of ion sources.

19

Absolute measurements of the source, second
quadrupole and system pressure are made with a model 310-
BHS-l MKS capacitance manometer. Bayard-Alpert type ion
gauges monitor the pressure of each vacuum chamber and of
the quadrupole collision chamber. An automatic pressure
controller and servo driven valve from Granville-Phillips

maintains a set collision gas pressure.

Current sample introduction methods include a heated
batch inlet system for gases and volatile liquids, a
direct insertion probe for introduction of solids and low

volatility liquids and a Varian 3700 capillary column GC.

For efficient and productive instrument operation a
modular microprocessor system based around Intel’s 8085
microprocessor was developed (44). The hardware required
for the TQMS system is described in detail elsewhere (45)
and can be broken down into six categories: control system
computer; mass selection; data acquisition; ion lens
control; mass storage; and display. The software system
developed for the microprocessor system is based on the
extensible programming system FORTHTH and is ideally
suited for instrument control. The capabilities and
performance of this system are described elsewhere (45)
but in brief, the flexibility of this system allows any
experimental mode to be accessed and permits

straightforward design and programming of new experiments

20

without requiring the operator to have any programming

expertise.

Medesandfltilitxefgneraiien

The TQMS instrument may be operated in a wide variety
of modes. The first and third quadrupoles can function as
mass filters when both the RF and DC fields are applied.
They may function in a fashion analogous to the second
quadrupole if only the RF field is applied. The
combination of these various modes of quadrupole operation

give rise to the following modes of TQMS operation:

A normal mass spectrum of the ions present in the
source may be obtained by scanning either the first
quadrupole or the third quadrupole in its mass filtering
RF/DC mode and having the other quadrupole passing all
ions in an RF-only mode. No collision gas is present in
the quadrupole collision chamber. In either of these cases
the TQMS instrument functions as would a conventional

single quadrupole MS.

A daughter spectrum of a selected parent ion is
obtained when both the first and third quadrupoles are in
their mass filter modes with the first quadrupole set to
pass only the desired parent ion and the third quadrupole

scanned to observe all CAD products resulting from

21

collision of the selected parent ion with the collision

gas in the quadrupole collision chamber.

A parent spectrum of a selected fragment ion is
obtained when both the first and third quadrupoles are in
their mass filter mode and the first quadrupole is scanned
while the third quadrupole is set to pass a particular CAD
fragment ion generated in the quadrupole collision

chamber.

A neutral loss spectrum is obtained when the first
and third quadrupoles are in the mass filter mode and both
are scanned with a constant mass difference maintained
between them. This type of spectrum reveals all of the
parents which lose a particular neutral fragment selected
by the magnitude of the mass offset via CAD in the
quadrupole collision chamber. In reactive collision
systems it may be possible to scan for the addition of a
selected neutral group via CAR in the quadrupole collision

chamber.

The utility of these various scanning modes is
realized in the following selected applications. Structure
elucidation can be effected by the conventional EI
ionization of a pure sample in the source. The source-
generated ions are then mass-selected in turn by the first
quadrupole and the daughter spectrum of each is obtained.
By this process the various fragment ions may be

identified allowing the structure of the original parent

22

ion to be deduced. If a softer ionization procedure such
as CI or low energy E1 is utilized, a sample consisting of
many components would yield molecular ions of each
compound present. The selection of each molecular ion by
the first quadrupole and the recording of its
characteristic daughter spectrum allows for direct
analysis of mixtures. Rapid targeted analysis of complex
mixtures for specific compounds is easily performed by
employing the TQMS instrument to monitor specific
characteristic reactions. Both the first and third
quadrupoles are preset to pass only a specific parent ion
and daughter ion respectively. Only those compounds which
pass the first quadrupole as the selected parent and
produce the selected daughter ion in the quadrupole
collision chamber give a response at the detector. An
unlimited number of parent/daughter pairs may be monitored
in the technique of multiple reaction monitoring.
Functional group type analysis of a complex mixture is
facilitated by the use of the neutral loss scan, where the
neutral loss selected is characteristic of the desired

functional group (45-47).

In addition to the use of the above-mentioned modes
of operation, the TQMS instrument through variation of
select collision variables may yield other useful
experimental results. A further understanding of the gas
phase ion-molecule interactions in the collision chamber

may be obtained by monitoring the abundance of the CAD

23

fragment ions as a function of the collision gas pressure.
The reaction order and the reaction cross section for
formation of the different fragments can be determined.
Metastable ions, if present, are easily spotted by their
zero order pressure dependence (40). Variation of the
selected ion’s lab collision energy, the ion’s kinetic
energy in the collision chamber, and its effect on the
type and abundance of fragment ions seen can provide
information concerning the bond strengths and internal
energy of the selected parent ion (40). Variation of the
dc offset in the final quadrupole and its effect on
transmission of a selected CAD fragment ion can indicate
the amount of translational energy lost in the low energy
collision process under favorable conditions (48). A
clever mode of operation is one in which the chemical
resolution of the CAD process in TQMS is applied to
perform high resolution analysis. Resolution of one part
in 23,000 has been demonstrated by observing small peak
shifts for daughter ion peaks from a pair of isobaric
compounds (49). For this technique to work the two parents
must have identical peak shapes and different

characteristic daughters.

The versatility of the TQMS instrument is evident by
the variety of experiments which may be performed. The
vast amount of information obtainable through this
technique is leading to its rapidly growing use in

chemical analysis.

CHAPTER THO

CHARACTERISTICS OF THE COLLISION PROCESS IN TQMS

The low energy ion-molecule collision process in the
quadrupole collision chamber is responsible for formation
of CAD and CAR products. A thorough understanding of this
collision process is necessary to maximize the potential
of TQMS for chemical analysis. The goal of the research
reported in this chapter to explore the characteristics of
the collision process in TQMS. The knowledge gained
regarding ion-molecule reactions in the collision chamber
may afford analytical benefits in terms of selectivity
similar to those which were experienced in the ion source
through the use of the ion-molecule reactions involved in

chemical ionization (CI) (26).

The use of a quadrupole collision chamber in TQMS is
the major reason for the high sensitivity demonstrated
with a TQMS instrument (2). The ion-molecule processes
that lead to CAD or CAR products proceed in a highly
efficient fashion, much more so than in a sector
instrument high energy collision cell. In the high energy
collision process, losses due to scattering and
neutralization via ion-electron recombination in the
plasma formed by the collisionally ionized target gas can

be severe.

24

25

Early studies on the low energy collision process
were performed in our laboratory to observe the effects of
the following variables: ion internal energy; ion axial
energy; ion transverse energy; collision gas species; and
collision gas pressure (37). Observation of the CAD
fragments from the molecular ion of cyclohexane as a
function of the ionizing energy in the source revealed
that excess internal energy is not an essential factor in
the CAD process. The axial energy dependence of the CAD
fragments of the methane molecular ion demonstrated that
greater collision energies were required for production of
the lighter fragments. The transverse kinetic energy, a
result of the quadrupole collision chamber RF field,
contributes only an average of a few electron volts and
was determined to have little effect on the collision
process (2). These results indicate that the fragmentation
process proceeds by an efficient conversion of a portion
of the selected ion’s kinetic energy into internal energy.
The efficiency of the CAD process for methane was shown to
increase with increasing mass of the noble collision gases
used from helium to xenon. The efficiency of nitrogen as a
collision gas was between that of argon and krypton. The
center-of—mass collision energy increases with increasing
target mass according to Ecm = Elub [m2/(m1 + m2)] where
Fhab is the lab collision energy, m1 is the mass of the
reactant ion and m2 is the mass of the collision gas. The

pressure of the collision gas determines whether single or

26

multiple collisions occur. Fragmentation efficiency for
methane increases with increasing pressure but so do

scattering losses in the collision chamber.

The first evidence for formation of CAR products was
observed with the collision of selected ions from ethane
with methane collision gas resulting in the formation of
products containing three carbons. Use of deuterated
ethane ions revealed extensive hydrogen-deuterium
scrambling. Other associative reactions have been
reported. The products of the reactions of CaHe+ and C3H7+
with n-hexane have been reported (50). Long-lived
collision complexes of the reaction of SiHa+ with C2H4 have
been observed (51). Also, very endothermic associative
reactions such as that of Ar+ incident on N2 to form ArN+

have been reported (52).

Collision of the CCl+ ion from carbon tetrachloride
with the collision gases Ar, N2, 02, H20 and He produced
three types of products: simple fragmentation products C+
and Clt; stable complex ions resulting from a CAR
involving the selected ion with all collision gases but
He; and charge exchange ions from all collision gases
except He. The magnitude of the complex ions and charge
exchange ions were three to four orders of magnitude less
intense than the parent ion. The charge exchange ions are

of low abundance since they are formed with little, if

27

any, forward momentum and are therefore mostly lost within

the quadrupole collision chamber.

In an effort to increase our understanding of the low
energy collision process, the mechanisms of the model ion-

molecule reactions 2.1 and 2.2 were studied (38).
Csz+ + CH4 ---> CzHa+ + H2 + CH4 (2.1)

CH3+ + CH4 ---> C2H5+ + Hz (2.2)
I

'—--> 0233+ + 2H2

With only the selected reactant ion entering the collision
chamber and no collision gas present, a scan of the second
quadrupole rod offset voltage provides an ion transmission
versus retarding potential curve. The value of the rod
offset which excludes one half of the parent ions
represents the median zero energy point and was used as a
reference point for determining relative ion axial energy.
The derivative of this curve indicates an approximately
symmetrical energy spread of two to three volts for ions
from the source. The reaction products were studied at

various ion energies relative to the reference point.

Reaction 2.1 was studied as both Csz+ + CD4 and C2D5+
+ CH4. At low collision energies the loss was H2 in the
first case and D2 in the second which confirms that the H2

loss is from the Csz+ ion exclusively. At higher energies

28

(10-20 V) a slight mixing of H and D is observed in the

products.

Reaction 2.2 produces CzHa+ even at low energies.
Since this is not a product under thermal conditions, it
results from the energy spread of ions entering the
collision chamber or the off—axis energy in the quadrupole
collision chamber. As the average ion energy is increased
to five volts the C2H3+ product becomes dominant. Studies
with deuterated ion and deuterated collision gas provided
results which were in good general agreement with those
obtained by Huntress (53) who used ion cyclotron
resonance, Abramson and Futrell (54) who used tandem MS
techniques and others. The study of Reaction 2.2 served to
illustrate that the energy spread of reactant ions
prevents a detailed study of branching ratios at near-

thermal energies.

Recent work into the mechanism of the low energy
collision process by Douglas (48) has supported the theory
that the CAD of large organic ions proceeds through
separate activation and unimolecular fragmentation steps.
In the activation step, a major portion of the
translational energy of the reactant ion is converted into
internal energy. The excited parent ion with many degrees
of freedom may survive many vibrations before
dissociating. This two-step mechanism is supportive of the

observation that CAD spectra are usually similar to E1

29

spectra since the internal energy may randomize throughout

the excited ion prior to dissociation.

The use of TQMS for the study of ion-molecule
collisions will certainly lead to a better understanding
of the processes involved. The ease with which the
mechanism can be studied, and cross sections determined
make the TQMS instrument well suited for these studies.
The results obtained may generate an arsenal of selective
techniques useful for certain difficult analytical

problems.

Collisions of Methane Molecular Ion

Introdugtion

In order to gain a better understanding of the
collision process in TQMS a comprehensive study of the
interaction of the methane molecular ion with a target gas
was undertaken. Previous work by this group (34,37) and by
others (55) have shown the major collision parameters to
be the following: energy in the system in the form of the
ion’s internal energy, ion axial kinetic energy and the
transverse or off-axis energy supplied by the RF field in
the quadrupole collision chamber; and the nature and the

pressure of the selected collision gas.

The internal energy possessed by a selected ion is

dependent on the ionization technique utilized in the ion

__I - d

 

30

source. The effect which the precursor ion’s energy has on
the CAD process has been explored (37). The ion axial
energy determines the amount of collision energy
available. The RF-only field gives rise to a transverse
ion energy which has been shown by calculations to be an
average of several electron volts, independent of the
magnitude of the applied RF field (2). However, while the
average may be small, recent simulations have revealed
that the instantaneous off-axis energies can be over ten
volts and thus be a significant source of energy for an

ion (41).

The type of collision gas chosen may have a dramatic
effect on the collision product ions seen. The mass of the
collision gas determines the fraction of available lab
collision energy which may be converted into center-of-
mass collision energy. The pressure of the collision gas
will determine whether single collision or multiple

collision conditions exist.

The study reported here is concerned with observing
the effect of these variables on the collision of the
selected ion, CH4+, and a discussion of the thus available
information regarding the characteristics of the collision

process .

31

Experimental

All experiments were performed on a TQMS instrument
which was developed in our laboratory (34,39). The
instrument utilizes a closed quadrupole collision chamber
and interquadrupole lenses for improved ion transmission.
The methane molecular ion was generated in a source from a
Finnigan 3000 under 70 eV electron impact conditions. The
CH4+ ion selected by the first quadrupole was allowed to
interact with a number of target gases in the collision
chamber. The collision gases used in this study, nitrogen,
argon, helium, oxygen, and nitrous oxide, were used
directly without additional drying or purification. The
target pressure was measured with a Bayard-Alpert ion
gauge connected to the collision chamber. The ionic
products resulting from the collision were focused into
the third quadrupole and mass analyzed. Analog detection
was used. All data were collected and stored on disc using

a microcomputer system developed in our laboratory (44).

The collision of CH4+ with a nitrogen target was
studied at a collision energy of 17.5 eV (lab). The
pressure dependence of the CAD products formed upon
collision with nitrogen was observed. The efficiency of
nitrogen, oxygen and helium target gases for CAD ion
formation were determined as was the effect of the RF
voltage in the collision chamber. The variation of the

second quadrupole rod offset revealed the energy

32

dependencies for dissociation to the several fragment ions
upon collision with nitrogen. The pressure dependence of
the collision process was also observed at a lower
collision energy of 4.5 eV with a nitrogen target gas.
Comparative studies with CD4+ as the reactant ion were

also made for the low collision energy system.

Besultsand ' uS'o

The collision products of CH4+ and nitrogen for 17.5
eV lab collision energy are shown in Figure 2.1. Note that
the intensity is displayed on a logarithmic scale. The
target pressure is high enough (6.6 x 10“ torr) so that
ions produced by multiple collisions (e.g. C+) are seen.
The multiple collisions permit stepwise dissociation of
the CAD fragments to occur. In addition to the fragments
of CH4+, ions of m/z 28-30 are evident. Charge exchange
formation of N2+ (m/z 28) is a low probability endothermic
process and accounts for only 0.1% of the base peak
signal. Proton transfer to generate HN2+ (m/z 29) is less
endothermic and its relative abundance is 1.6%. The ion at
m/z 30 (0.2%) is probably NO+. A detailed discussion of
these and other types of collisionally activated reaction

(CAR) products will be presented in a later section.

33

 

 

 

 

0-
C’ _
3
-Q _
<E
_' —2-
Q)
0:
0')
O
_I _
-4 I . l

 

 

 

 

Moss (omu)

Figure 2.1 CAD (N2) spectrum of CH4+ at 17.5 eV lab
collision energy. 6.6 x 10’4 torr N2.

34

Dependence of CAD Products on Collision Chamber Pressure,
RF Voltage and Target Gas Species

Measurement of the fragment ion yield as a function
of the collision gas pressure can provide much information
regarding a particular collision system. The efficiencies
of the collision system may be evaluated as shown in
Figure 2.2. The upward arrow in Figure 2.2 represents the
in loss due to scattering. An equation for the combined
collection efficiency of the collision chamber and
transmission through the third quadrupole,

(IP + EIE1)/IPo, where 190 is the parent ion intensity
incident on the second quadrupole and IP and 1?: are the
transmitted parent ion and CAD fragment ion intensities,
respectively. A reduction in the collection efficiency
occurs at higher target pressures due to increased
scattering and charge exchange losses. The total
fragmentation efficiency is given as thi/(IP + ZIPI). The
overall CAD efficiency is then Elrl/Ipo. These efficiencies
for the collision of CH4+ with N2 at 17.5 eV lab collision
energy are shown in Figure 2.3. The reduction in collected
ions and the increase of fragmented ions with increasing
target gas pressure combine to give a maximum CAD
efficiency of 12% at 5 x 10" torr. At a pressure of

3.7 x 10" torr only 50% of the ions are lost. At the
highest pressure used in this study (1.04 x 10-3 torr) the

total fragmentation efficiency is 53%.

35

IPo IP
in
CAD IS “F1
Quad In

Fragmentation Efficien cy

E =£h_
F IP+£IFI
Collection Efficiency
IP+£IFI
EC = ——
IPo

Overall CAD Efficiency
81,-,

IPo

 

ECAD = Er“ Ec=

Figure 2.2 Efficiency terms for the CAD process in the
quadrupole collision chamber.

36

Total
. F ragmentation
Efficiency

Collection
Efficiency

EFFICIENCY
.0
I

CAD Efficiency

 

 

0°01 I 1 n I l l l I
10'5 10‘4 10'3
COLUSION CHAMBER PRESSURE (Torr)

Figure 2.3 Efficiencies for the collision of CH4+ with
N2 at 17.5 eV lab collision energy.

37

The relative abundances of all CAD product ions
observed are displayed as a function of target pressure in
Figure 2.4. The lower axis in Figure 2.4 can also be
represented by the amount of average time between
collisions. The inverse of the Langevin collision rate,
which is dependent on the polarizability of the target and
the reduced mass of the ion-molecule pair (62), times the
number density of target gas molecules provides the time
interval expected between collisions (63). Calculated
values range from 3.2 x 10'3 s at 10‘5 torr to 3.2 x 10‘5
s at 10'3 torr. The residence time of the CH4+ ion in the
collision chamber , calculated from t = 0.707 1
(m/ELnb)°-5, where t is in 10‘6 seconds, and l is the path
length in cm (40), is approximately 1.5 x 10‘5 s at 17.5
eV lab collision energy. Loss of hydrogen to give CH3+ is
the predominant ion. Formation of NzH+ rivals that of CH+.
Charge exchange to produce N2+ is the least abundant. The
change in slope for the CAD fragment ions is likely the
result of the onset of multiple collisions at pressures
above 2 x 10'4 torr. The slopes of the curves in the
single collision region can give the reaction order. A
metastable ion would be first order in CH4+, and zeroth
order in N2. A simple fragmentation collision is first
order in N2. Products of multiple collisions are second,
third and higher order in N2. CAD fragments which have
non-integer slopes may be due to combined processes of

different orders. The amount that the slope of the curve

 

100

Ir/Ip(%)

0.1

0.01

Figure 2.4

38

TIME BETWEEN COLLISIONS (sec)

 

 

3 3 k i h b 6
6 6 ’ 6 6 ’ 6
«,rv“ (sf 6‘90 «,m“ (at 649° 3.16
l I I I l I
16+e15+
+ +

    

 

/ 16 +4 1.3 +

/ 16 +4 29+
- 4% Metastable/
correction
/

’ 16++30+
16 +—> 12 +

16++28+

fi’ fi 1 l I U 1 I I
10'5 1o" 10'5

COLLISION CHAMBER PRESSURE (Torr)

Pressure dependence of CAD (N2) product ions
at 17.5 eV lab collision energy.

 

39

for formation of CH3+ in the single collision region
differs from one, the expected value for a reaction first
order in N2, is a measure of the metastable ion
contribution. In the absence of collision gas only a very
small abundance ion for CH3+ is observed. From the amount
of CH3+ signal needed to subtract from the curve to give a
line with slope of one (shown by the dashed line in Figure
2.4), a metastable ion contribution of 4% can be
estimated. The signal for C+ was near the detection limits
at the low target gas pressures. Formation of N2H+ is
first order in nitrogen pressure at low collision gas
pressures. Although the pressures are too high for
accurate reaction order determination, formation of N2+
appears first order and formation of NO+ is a multiple

order process.

These results can be summarized in Table 2.1 where
the slopes of the curves under both single collision and
multiple collision pressures are presented. With the
exception of CHaf, which has a metastable contribution,
the slopes for formation of CH2+ through C+ exhibit
decreasing slopes. This behavior is unexplained. At
1.04 x 10-3 torr N2, the single fragmentation efficiency

to form CH3+ is 40% and that to form NzH+ is 1.5%.

The cross sections for the various processes involved
at a fixed collision energy may be determined from the

pressure dependence of the products. The cross section for

4O

 

 

 

 

 

 

Table 2.1 Slopes of the pressure dependence curves for
CAD product formation at 17.5 eV lab collision
energy.

Slopes
Reaction to Form Single Collision Multiple Collision
Pressure Pressure
CH3+ 0.49 1.25
CH2+ 0.83 1.5
CH+ 0.62 2.0
C+ 0.3 2.0
MM” --—- 1.3
N2H+ 1.05 1.85
NO+ ---- 3.1

41

loss of the parent ion due to fragmentation, scattering
and nondetected charge transfer may be derived from the
slope of In (IP/IPo) versus target pressure at low target
gas pressures. At 17.5 eV lab collision energy, a value of
6.0 x 10-14 cm3 for the product cl results. An estimate of
1, the path length, is provided by use of an equation
derived by Dawson which estimates the effect the axial
energy of a given mass ion has on increasing the actual
path length in an RF-only quadrupole (56). The correction
is only 2.3% in this case and results in a value of 22.1
cm for 1. Assuming this value for l, a value of

27.1 x 10‘13 cm2 is obtained representing the cross section
for loss of incident parent ion. The cross sections for
the specific fragmentation reactions may be determined
from the slope of a plot of

In [1 - IP/(IP + 2121)] versus target pressure. The values
obtained are presented in Table 2.2. The 20.5 x 10‘19 cm2
difference between the cross sections for total parent ion
loss and for parent ion loss resulting in ionic products,
GT, is then a direct measure of the cross section for
scattering loss and charge transfer loss within the
collision chamber. The majority of the loss is probably
due to scattering since the charge exchange reaction to

form N2+ is endothermic by 66.1 kcal/mole (66).

The effect of the RF voltage level on the various
collision process efficiencies is displayed in Table 2.3.

The RF voltage for optimum transmission of all ions

42

Table 2.2 Cross sections for CAD product formation at
17.5 eV lab collision energy.

arotala : 6.61 x 10'16 cm2
016—->15 = 5.51 x 10'13 cm2
ale-—>14 = 9.2 x 10‘17 cm2
manna : 6.3 x 10-18 cm2
016-->12 = 7.3 x 10‘19 cm?
015-->29 = 1.0 x 10‘17 cm2

a. dTotal is the cross section for CH4+ to give all ionic
products detected, C+, CH+, CH2+, CH3+, N2H+, NO+ and
N2+.

 

 

43

Table 2.3 Collision process efficiencies as a function
of the RF voltage on the quadrupole collision
chamber.a

 

RF CAD Collection Fragmentation IP/IPo
Efficiency Efficiency Efficiency

(V) % (%) (%)

40 4 7 38.8 12 3 O 34

67 6 1 39.3 15 6 0 33

93 6.8 38.9 17.5 0.32

107 6.5 38.6 16.9 0.32

a. Nitrogen target at 4 X 10'4 torr and collision energy
of 10.5 eV lab.

44

through the second quadrupole is indicated by the slight
maximum for the collection efficiency at 67 V. The
differences in the CAD and fragmentation efficiencies and
IP/IPo with increasing RF voltage are a result of the
different transmission characteristics of CH4+ and CH3+
through the second quadrupole. At an RF setting of 65 V,
CH3+ transmission is maximized. Transmission of the parent
ion, CH4+, is maximum around 30 V RF and displays the
effect of the presence of an aperture prior to the second
quadrupole. The effect of restrictive apertures on parent
ion transmission was treated by Dawson (56). Briefly, the
presence of a restrictive aperture imposes the condition
that the ion entering the quadrupole collision chamber has
a node in its trajectory at the aperture. It must also
have a corresponding node at the aperture present at the
other end of the quadrupole to exit. Ions generated
through CAD or CAR within the collision chamber do not
reveal this selective behavior since they may be formed at
any distance along the quadrupole axis. This serves to
average out the effects of the preceeding restrictive
aperture. This effect will not change the conclusions of
this study. The parent ion transmission decreased smoothly
to 90 V RF where it then parallels the CH3+ transmission
until near the low mass limit for the applied RF voltage
(130 V for CH3+). These transmission differences are
present when the collision chamber has a fixed RF voltage

applied. If this RF voltage is instead a fixed fraction of

45

the RF voltage applied to the third quadrupole, both
parent and fragments will have similar transmission curves

(56-58).

Changing the mass of the target gas leads to a change
in the center-of—mass collision energy. For a nitrogen
target and a 10.5 eV Elab a value of 6.7 eV Ecm results.
With helium and argon as targets, values of 2.1 and 7.5 eV
Ecm result, respectively. The effect that the RF voltage
level has on the collision efficiencies with He and Ar
targets at the same lab collision energy (10.5 eV) and
pressure (4 x 10“ torr) is similar to that experienced
with an N2 target. The efficiencies for all three targets
at 93 V RF are compared in Table 2.4. The apparent
increased CAD efficiency for helium results from a
doubling of the collection efficiency. Scattering losses
are decreased with a light target, and charge exchange
channels are less favored with targets with a higher
ionization potential (59). The higher fragmentation
efficiency with molecular nitrogen leads to a higher CAD
efficiency for N2 when compared to Ar and may be a result
of the increased polarizability of N2 over Ar
[1.76 x 10'24 cm3 (60) vs. 1.63 x 10‘24 cm3 (61)] (33).
With the target gases studied, the CAD efficiency
increases with increasing ionization potential of the
target, in agreement with the results of others (5,59)

obtained with sector type MS/MS instruments.

 

 

.Awmv .mmn Soum mmsdm> Hmflpcouom sowmeasoH .n

.Honemno GOMmflHHoo

0:9 GM mm > mm pom >0 m.oH mo seam .Huop «IOH N v mo mononuoum pcmnma .m

 

 

 

«s.o m m m as s s H.N m am we
mm.o m.sH m.mm m m s.m m.mH «z
6
4
mm.o m NH m.mm o m m.s a ma u<
$3 $3 $3 :3
soaoaoamee soaoaoamme so=6H6secm A>6c an was was
oaHxaH aoapmucoammue cospoossoo a<o sum pmmums pastas

Eoflao£ tam comonpfln

«.mpmwumv
Spas mo«o:0«ofimmo mmOUOHm Godmfiaaoo v.N OHQMB

47

Energy Dependence of CAD Products

The energy dependence for formation of C+ through CH3+
and NzH+ was observed. In general, the shapes of the
energy dependence curves for the CAD fragment ions were
similar. All exhibited thresholds for formation a few eV
(ca. 3) more negative than the median parent ion energy
zero reference point, as determined by use of the
retarding potential technique (65). As the collision
energy was increased, production of all CAD fragments
increased to a maximum value and then dropped off slightly
(5-10%) at higher collision energies. The collision energy
corresponding to maximum formation of each CAD product ion
at a nitrogen target pressure of 3.5 x 10'4 torr and 67
volts RF in the second quadrupole are listed in Table 2.5.
The collision energies reported are relative to the median
parent ion energy zero reference point. The enthalpies of
formation required to produce the fragments in a step-wise
manner, as might be expected under multiple collision
conditions are also in Table 2.5. It should be noted that
the lowest energy pathways to form the CAD fragments
directly from CH4+ are much less energetic than the values
reported in Table 2.5. Formation of CH2+ and H2 is
endothermic by only 2.56 eV; that of CH+, H2 and H, 7.12
eV; and that of C+ and 2H2, 6.7 eV (66). The greater the
energy that is required to form a particular fragment, the

greater is the collision energy maximum for production of

 

48

Table 2.5 Collision energy dependence of CAD fragment

ion yield.a

Position of maxima

 

 

 

Reaction Elab (eV) Ecm (eV) Alhb (eV)
16+ ---> 15+ 4.1 2 6 1 7
16+ ---> 14* 8.9 5.7 7.1
16+ ---> 13+ 19.3 12.3 11.6
16+ —--> 12+ 19.7 12.5 15.7
a. Nitrogen target at 3.5 X 10‘4 torr.

AHR values calculated from ref. (66) and assume
stepwise formation of H (i.e. CH4+ ---> C+ + 4H).

49

that fragment. Combinations of direct and consecutive
formation mechanisms would explain the observed energy
dependences. In sharp contrast to the energy dependence of
a CAD fragment ion is that of a CAR fragment ion, N2H+.
Maximum NzH+ formation occurs at the lowest collision
energies. The interaction times at these lower energies
are slightly longer and the formation of the more

energetic fragment ions is greatly reduced.

Pressure Dependence of CAR Products

 

A daughter scan of CH4+ at low collision energy
(Elab = 4.5 eV) and 6.6 x 10'4 torr N2 is shown in Figure
2.5. A wide variety of products are evident. The nitrogen
collision gas was analyzed in the source and revealed the
minor contaminants of Ar, 02, and H20 as shown in Table
2.6. Given the wide variety of ion~molecule reactions
possible, the collision of CD4+ with the target gas was
used to reveal the number of reactant ion protons present
in a product ion. A comparison of the CH4+ and CD4+
reaction product spectra was further aided by the
determination of the pressure dependence of the product

ion.

First, the efficiencies for the collision of CH4+
with N2 at 4.5 eV and 17.5 eV lab collision energies are
compared. The results for 4.5 eV lab collision energy are

shown in Figure 2.6. A number of differences between this

50

 

 

 

 

 

 

 

 

 

 

0—
C. -
3
-O .
<1:
_'—2-
Q)
n:
C)
o
_J _
_4 I ' l l

 

Moss (amu)

Figure 2.5 CAR (N2) spectrum of CH4+ at 4.5 eV lab
collision energy. 6.6 x 10‘4 torr N2.

51

Table 2.6 Percent total ion current of components
present in nitrogen collision gas. 70 eV EI.

 

 

Component % TIC
N2 99.748
Ar 0.15b
02 0.05
H2O 0.06

a. Includes N2+ (95.24 %), N+ (4.5 %).
b. Includes Ar+ (0.13 %), Ar++ (0.02%).

52

 

 

1 - Total
~ Fragmentation
Efficiency
Collection
. Efficiency
>. CAD Efficiency
0
z
E
9 0.1 —
u.
L .
In
.1
‘I
0'01 l I 1 I I t T l I
10‘5 10" 1o-3

COLLISION CHAMBER PRESSURE (Torr)

Figure 2.6 Efficiencies for the collision of CH4+ with
N2 at 4.5 eV lab collsion energy.

 

53

figure and Figure 2.3 exist. The CAD efficiency (which
includes CAR products) is now maximized 9% higher at 21%
around 8 x 10'4 torr. The maximum total fragmentation

efficiency is up 16% to 69%.

The 50% reduction in collected ion does not occur
until 1 x 10" torr higher target pressure. The increased
collection efficiency, 31% at 10'3 torr versus 18% at 10—3
torr, indicates fewer scattering and charge exchange

losses at reduced collision energy.

The predominant ion in the CAR spectrum (Figure 2.5)
is still the CAD fragment CH3+. Formation of N2H+ (m/z 29)
is four times larger than the next most abundant CAD
fragment CH2+. The pressure dependence of the CAR products
yield to the following observations. Formation of m/z 18
parallels that of m/z 19. Formation of m/z 42 and 43 have
pressure dependencies similar to that of N2H+, a first
order product. Finally, the formation of m/z 30, a
multiple order product, is paralleled by that of m/z 31
and 45. Plots of the intensity of 30+ and 45+ against the
square of the pressure reveal a second order dependence on

nitrogen pressure.

This information, coupled with that presented in
Figure 2.7, where the CAR product spectra of CH4+ and CD4+
are shown, allow the following assignments to be made.
Product ions at m/z 18 and 19 are H20+ and H30+,

respectively. They are formed mostly through the

g__

 

54

 

 

 

 

 

 

 

 

 

 

 

 

CH4+
100 -
I x 4 x 100
. 80-
C _
3
43 50 -
< -
. 40—
a) _
01 20- l
I . II I
0 I ' I ' I I
10 20 30 40 50
Moss (emu)
CD4+
100 - x 4 x 100
. 80 -
C .
3
43 50 -
<( _
40-
a) _
D: 20 _
I . -l -I
0 T ' ' I ‘ I
10 20 30 40 50

Moss (omu)

Figure 2.7 CAR (N2) spectra of CH4+ (upper) and CD4*
(lower) at 4.5 eV lab collision energy.
6.6 x 10‘4 torr N2.

 

55

exothermic reaction of CH4+ and H20 to give H20+

( AHR = -1.1 kcal/mole) and H30+ (.AHR = -42.8 keel/mole).
All heats of reaction were calculated using values
obtained from references (66) and (67). They may also be
formed by the reaction of the major product ion, N2H+ with
H20. The reaction to give H30+ by this alternate route is
exothermic by 54.4 kcal/mole. These two sources for H20+
and H30+ formation would explain their observed pressure
dependence between that of first and second order. The
product ions at m/z 29, 42 and 43 are assigned to first
order products. One reactant ion proton is incorporated
into 29+ and two and three reaction ion protons are
incorporated into 42+ and 43+, respectively. The reaction
between CH4+ and N2 to produce NzH+ (m/z 29) is endothermic
by only 11.4 kcal/mole (0.5 eV). The 31+ in the CD4
spectrum is of the proper magnitude to be 15NND+. Charge
exchange to give N2+ is endothermic by 66.1 kcal/mole
(2.87 eV) and is not observed. The 4.5 eV lab collision
energy provides only 2.86 eV of center-of—mass collision
energy. Even if N2+ is formed it may not be observed.
Charge exchange products would possess little forward
momentum and may be lost within the quadrupole collision
chamber. The ions at m/z 42 and 43 may have two and three
reactant ion protons, respectively. Associative reaction
products CH2N2+ and CH3N2+ may account for these ions.
Their formation from CH4+ and N2 is exothermic by slightly

less than 1 eV. The ions at m/z 30, 31 and 45 have their

 

 

56

major contribution from ions with zero, one and one
reactant hydrogens respectively. Contributions of ions
with two reactant ions protons is possible for m/z 30 as
is three reactant ion protons for m/z 31. Following these
assignments the ion at m/z 30 may be composed of NO+ and
CH20+. A multiple order reaction would account for NO+
formation. The formation of CH20+ from the reaction of CH4+
with 02 or H20 may add a contribution to m/z 30. The
specific reaction(s) responsible would require knowledge
of the neutral products formed. In a similar fashion

m/z 31 may result from a multiple order reaction to form
HNO+ and a reaction of CH4+ with 02 or H20 to form CHaod.
These tentative assignments are in agreement with the CD4+
spectrum. A small amount of m/z 32 is formed by the
exothermic (0.65 eV) charge exchange reaction of 02 with
CH4+ to form 02. The ion at 45+ is a multiple order
reaction likely the result of further reaction of N2H+
with 02 or H20 to form N2HO+. Both of these reactions are
endothermic by approximately 2.3 eV. A small contribution
to m/z 45 from the reaction of CH4+ with 02 to produce

CHOz+ is also possible.

The cross sections for the processes involved at this
low collision energy may be determined and compared to the
values obtained at higher collision energies. The cross
section representing loss of the parent ion due to
fragmentation, scattering and charge exchange was measured

at low collision gas pressures. The correction factor for

 

57

the path length at 4.5 eV collision energy is 8.6% (56).
The path length is then 23.5 cm in the second quadrupole
and the value for the cross section representing loss of
parent ion is 25.7 x 10’16 cm2. The cross sections for the
production of all CAD and CAR ions formed and the cross
sections for formation of CH3+, N2H+ and NO+ are presented
in Table 2.7. The cross section for NzH+ formation is 15
times larger at this lower energy. The cross section to
give all ionic products is similar at both energies as is
the cross section for scattering loss and charge transfer

loss.

Studies of the CAR products obtained with other
collision gases (Ar, 02, N20) demonstrated that the major
CAR product ion formed in most cases is the protonated
target gas, in agreement with the CAR (N2) results
reported here, with the exception that charge exchange to
form 02+ is energetically more favored and no protonated 02

is observed.

Conclusiens

The study of the collisions of methane with a target
gas allow several conclusions to be reached. The
excitation process in the collision cell involves an
efficient conversion of ion kinetic energy into internal
energy. CAD fragment ions such as 0* from CH4+, that

require more energy to produce than is available from a

 

58

Table 2.7 Cross sections for CAD/CAR product formation
at 4.5 eV lab collision energy.

dTotal = 6.68 x 10-13 cm2

016 --> 15 4.6 x 10'16 cm2

016 --> 29 1.48 x 10’16 cm2

016 --> so 3.5 x 10-18 cm2

59

single collision at low collision energies may be produced
at higher collision energies. The CAR products,
particularly the protonated target gas, are favored at
very low collision energies where the transit time through
the second quadrupole collision chamber is longer and the
pathways for extensive CAD fragment formation are not
accessible. Higher CAD (including CAR) efficiencies are
possible at lower collision energies as a result of
increased CAR product formation and decreased scattering
and charge transfer losses. Ion losses due to scattering
and neutralization become significant at higher target gas

pressures.

The CAD efficiency of a target gas (He, N2, Ar) was
found to increase with increasing ionization potential of
the target gas. The increased efficiency of N2 over Ar as
a collision gas may result from the fact that N2 is a
slightly more polarizable target and polarizable targets
are more efficient for effecting vibrational excitation
(33). This would support the argument that the excitation
process in the collision chamber involves a vibrational

excitation through momentum transfer.

Kinetic data such as cross sections may be determined
from the pressure dependence of the CAD and CAR ion yield.
The total fragment ion/product ion cross section remains
essentially unchanged at the different collision energies

employed in this study. There is, however, a shift from

60

CAD fragment formation to CAR product formation as one

lowers the collision energy.

Under fixed RF operation of the collision chamber
there is an optimum setting for maximized transmission of
each ions of interest. The transmission characteristics of
selected ions are altered by the presence of the
restrictive apertures in the interquadrupole lenses;
removal of the apertures would eliminate the artifacts
observed and the resulting transmission curves would be
similar to those seen in TQMS instruments with a close-

coupled quadrupole design.

Many CAR products, both exothermic and endothermic,
are formed in the collision process. A comparison of the
CAR (N2) spectra of CH4+ and CD4+ along with the reaction
orders which may be determined from a pressure study of
CAR product yield provides an effective method for
identifying many of the products of ion-molecule reactions

in the collision chamber of a TQMS instrument.

 

CHAPTER THREE

COMPARISON OF DIFFERENT TQMS INSTRUMENT CONFIGURATIONS

A comparison between two TQMS instruments with
differing instrumental configurations is presented in this
chapter. The main differences in the instrumental
configurations are in the method of ion transmission
between quadrupoles and in the design of the collision
chamber. Two separate studies were performed to aid in the
comparison and these are presented in two parts. In part
A, the pressure and energy dependence of CAD products from
the collision of bromobenzene molecular ion with nitrogen
collision gas are determined. In part B, the factors
governing the use of the quadrupole collision chamber for

performing charge inversion reactions are investigated.

A- Collisions of Bromobenzene Molecular Ion

Introdugtion

The triple quadrupole mass spectrometer was utilized
to observe the characteristics of the collisionally
activated dissociation of the bromobenzene molecular ion
with a nitrogen target gas. The TQMS instrument is a

flexible tool for investigation of this process as the

61

62

collision energy, collision gas pressure and all
potentials along the ion path can be varied. Bromobenzene
was chosen for the study as a large data base exists for
this compound (68,69) and also to provide a means for
comparison of the current ion optical path with that of a
TQMS instrument of close coupled design in which no
interquadrupole lenses are used (57). Studies performed in
another lab on the close coupled instrument with
bromobenzene have provided evidence for a two step CAD
process (48), an activation step followed by a separate

unimolecular from the dissociation.

The relative ion abundances as a function of the
collision gas pressure for the fragments of C6H5Br+ are
observed in this present study and have previously been
explained in terms of consecutive and competing reactions
(48), each possessing a characteristic cross section. The
collision energy dependence of the cross section for
formation of the three major fragment ions is explored.
The relative contribution of the cross section for
formation of each fragment ion to the total cross section
is presented. The comparison of results between the two
types of triple quadrupole MS instruments allow some
conclusions to be reached regarding the ion optical path

arrangement in our TQMS instrument.

63

Experimental

All experiments were performed on the TQMS instrument
developed in our lab and described elsewhere (34,39). The
ion optical design includes two sets of four lenses to aid
in ion transmission in the interquadrupole region. The
first and last element in this lens set contain
ferroceramic cylinders connected together so their
potentials are equal. These cylinders partially penetrate
the preceding and following quadrupole and serve to
effectively block the direct current component of a
quadrupole operated in a RF/DC (mass filter) mode and
allow the positive focusing RF fields to pass. The effect

of the fringing fields is greatly reduced thereby.

The source, a CI source from a Finnigan 3000, served
to produce the CsHsBr+ molecular ion via charge exchange
with a benzene chemical ionization gas. This method
generates ions with low internal energies since the
ionization potential for benzene is only 0.26 eV above
that of bromobenzene. The source pressure was maintained
at an optimal value of 0.25 torr of benzene as measured by
a MKS model 310-BHS-1 capacitance manometer. The
potentials on the source lens elements relative to the
source were kept low to prevent possible excitation of the
parent and were still sufficient in magnitude to exhibit

good focusing.

64

The nitrogen collision gas was grade 4.5T“, with
claimed purity of 99.995%, and used without further
purification. The collision gas pressure was maintained
with an automatic pressure controller and measured with a

Bayard-Alpert ion gauge.

The collision energy in the second quadrupole is
variable by changing the DC rod offset imposed on that
quadrupole. The magnitude of the collision energy (lab) is
given by the difference in potential between the ion

volume in the source and the second quadrupole rod offset.

An effect of the second quadrupole RF voltage on the
transmission of parent ion and fragment ions was observed
and an "optimum" setting was chosen. The choice of
potentials for the interquadrupole lens elements between
the second and third quadrupoles and the DC rod offset
applied to the third quadrupole determine the amount of
drawout potential applied. The drawout potential is an
attractive potential applied to aid in the collection and
transmission of CAD/CAR products. Drawout potentials are
especially helpful for detection of collision products
which possess little forward momentum. Large drawout
potentials on the third quadrupole serve to decrease the
resolution of this mass filter. For the results reported
here, drawout potentials of 5 V and 35 V were applied to

the first element of the interquadrupole lens and the

65

third quadrupole, respectively, without serious

degradation of resolution.

To probe the characteristics of the ion path of our
instrument and for comparison with those of a close
coupled TQMS design, several studies were conducted. The
pressure dependence of the fragment ions resulting from
collision of the parent ion with nitrogen was determined.
Collision gas pressures were varied from low pressures
(ca. 10'5 torr) where single collision conditions exist to
pressures where multiple collisions predominate (ca. 10“

to 10'3 torr).

The dependence of fragment ion yield on the collision
energy was observed for a collision gas pressure in the
single collision regime, 2.7 x 10-5 torr. From this data,
the cross section for the fragmentation of CsHsBr+ was
determined as a function of collision energy. Also the
relative contribution of each fragmentation product to the

cross section was determined.

A conversion dynode prior to the detector was used to
reduce any mass discrimination effects of the continuous

dynode detector which was operated in an analog fashion.

Resultg and Discussion

A daughter scan of bromobenzene obtained at 25 eV lab

collision energy and 3.8 x 10“ torr nitrogen collision

 

66

gas is shown in Figure 3.1. In addition to the reported
peaks of C4H3+ at 51 and C6H5+ at 77, peaks are present
notably at m/z 94, also at m/z 116, 128, and 129. At
slightly higher pressures of N2, 5 x 10“ torr, fragments
at m/z 65, 105, 107 and 130 are seen. The ion at m/z 94 is
unique in that it arises from a reaction with a

contaminant molecule:

CeHSBr+ + H20 ---> CsHSOH+ + HBr (3.1)

The enthalpy change for this reaction is exothermic by 9.9
kcal/mole. The same product may also arise from the

reaction

CsH5+ + H20 -—-> C6H50H+ + H (3.2)

which is endothermic by 10.9 kcal/mole. The water may be
present from an undried tank of nitrogen and a "wet" gas
manifold transfer line. All of the ions observed are

tabulated in Table 3.1.

The effect of the quadrupole collision chamber RF
voltage on the loss of bromine from the parent ion is
shown in Figure 3.2. The transmission curve of the parent
ion does not display the step and is nearly flat on a log
scale from 5 to 90 with a slight maximum near 25 and a
sharp cutoff just below 100. The step occurs at the point
where the RF voltage is high enough to cut off
transmission of the 77+ ion through the second quadrupole,

the portion of the curve at RF values greater than this

67

.Nz HHOP vIoH x m.m

and >0 mm 90 +ummmeo Mo Eaupooen

Azccov

o:

.smuoso sofinaaaoo

H.m magmas

on

 

 

 

 

 

was 601

'uan

 

68

Table 3.1 Ions resulting from collision of bromobenzene
with nitrogen collision gas at 25 eV lab
collision energy.

m/z Ion Assignmenta
156 M+ CsHsBr+
130 (M—26)+ (CAHaBr)+
129 (M-27)+ (CIHzBr)+
128 (M-28)+ (CIHBr)+
116 (M-40)+ (CaHBr)+
107 (M-49)+ (C2H4Br)+
105 (M-51)+ (C2H2Br+)
94 (M-62)+ C6H50H+
77 (M-79)+ CsHst
65 (M-91)+ (CsHsf)
51 (M-105)+ C4H3+

a. Assignments in parentheses are tentative.

 

 

 

69

 

 

-7 _

C, _ .

:1

.o

< —9 — MM,“ .....

OW . . -—""K‘..I~Vf'\fiv_\,"_ .

O ' .Nflcg

__I ”A.

_11 ‘I'l'l'l'l'l'l'l'l'l'l‘IIl:I
0 200 400 600 800 1000 1200
RF Voltage

Figure 3.2 Dependence of bromobenzene CAD (N2) fragment
ion 77+ (CeHst) on second quadrupole RF
setting.

 

 

70

cutoff is indicative of ions which left the collision
chamber as collisionally activated parent and fragmented
prior to mass analysis by the third quadrupole. It may
also be explained by collision of the transmitted parent
ion with collision gas in the interquadrupole lens prior
to mass analysis. If the unimolecular fragmentation of the
excited parent ion is the major source of this signal,
this would provide additional support for the theory of
the two step nature of the CAD process. At an RF value of

60 (800 V) the signal level is 2.5% of the maximum.

The cutoff voltages observed are in close agreement
with that predicted for the lower mass limit of a RF-only
quadrupole (2). For all other spectra recorded the RF
voltage was set at 330 V with predicted upper and lower
mass limits of 1600 and 33 amu, respectively, assuming
5 eV off-axis energy (2). The fixed RF setting maintains a
consistent transmission for each m/z value, although it

may not provide optimized transmission for each m/z.

The pressure dependence for the formation of the
various observed fragments at 25 eV of lab collision
energy is shown in Figure 3.3. The shape of the curve for
CeHs+ (m/z 77) and C4H3+ (m/z 51) are qualitatively similar
to that reported with the close coupled instrument (48).
However, operation with a fixed RF voltage on the second
quadrupole prevents a direct comparison with the pressure

dependence observed when the RF voltage is

71

 

 

 

 

 

10-71 77 +
. 156+
10"-
1 94+
. 51+
107*
. 65+
’2
E 10-“- 128+
a d
2 .
E I +
- 116
10“°- /
«I
.7
10." I I n I I l
O 2 4 6 8 1O 12

COLLISION CHAMBER PRESSURE (Torr x 10‘ )

Figure 3.3 Pressure dependence of bromobenzene parent
and CAD (N2) fragment ions at 25 eV lab
collision energy.

72

scanned as a function of the RF voltage on the third
quadrupole. A notable difference between the two are that
in the present study the m/z 51 formation is of lower
abundance relative to m/z 77 formation due to the partial
discrimination of 51+ by the fixed RF voltage on the
second quadrupole. The pressure dependence for ions 94*,
107+, 65+, 128+, and 116+ is also reported in this study. A
plot of the log (relative abundance) versus the log
(target pressure) for the fragments at m/z 77, 94, 51 and
65 is presented in Figure 3.4 and that of fragments at m/z
107, 128 and 116 in Figure 3.5. The slope of the curve for
formation of 77+ (C6H5+) at low collision gas pressures is
1.2 and supports a first order dependence on nitrogen
pressure. All other fragments, except m/z 94 (C6H50H+), are
observed at too high a pressure to accurately determine
the reaction order. Multiple collisions are significant at
these pressures. However, a number of observations can be
made. Formation of ions at m/z 128 and 116 exhibit similar
slopes of 1.0 in the pressure region where they are first
detected; this supports a first order reaction for loss of
neutral C2H4 and C3H4, respectively. Formation of ions at
m/z 51, 65, and 107 show similar multiple order N2
pressure dependence, all with slopes of approximately 3.6.
These ions probably represent the combined loss of several
neutral species among which are Csz and Br. The pressure
dependence of 94+ formation illustrates a marked change in

slope in the low 10'4 torr

73

 

 

4 77+
100 -
1
1
94+
1° " 51*
+
+ 65
R 1 —
a d
H d
‘\
ll.
H
0.1 .1
q
0.01
j T 1 l 1 I U T T 1
IO" 10"4 10'3

COLLISION CHAMBER PRESSURE (Torr)

Figure 3.4 Pressure dependence of bromobenzene CAD (N2)
fragment ions 77*, 94+, 51+ and 65* at 25 eV
lab collision energy.

74

 

 

10 -
‘1 107+
128+
N 1—
O.
H a ‘l'
\ 116
IL
H
0.1—
0.01
U 1 V l 1 1 I I T T U
10'5 10“ 10'3

COLUSION CHAMBER PRESSURE (Torr)

Figure 3.5 Pressure dependence of bromobenzene CAD (N2)

fragment ions 107+, 128+ and 116* at 25 eV
lab collision energy.

75

range (from 0.2 to 3.2). This may be accounted for by two
sources of water, the water present in the unbaked
collision chamber area leading to the approximate zero
order pressure dependence observed at low target pressures
and the water present in the nitrogen collision gas
leading to multiple order pressure dependence at higher N2
pressures. The major factors for the observed behavior of
94+ at higher N2 pressures are likely the onset of
multiple collisions and the opening of a reaction channel

for formation of 94+ from the CeHs+ ion as in Reaction 3.2.

The variation of collision energy has a marked effect
on the formation of the 77+ fragment as shown in Figure
3.6. The source block potential is +30 V. The
discrimination of 77+ at more negative second quadrupole
rod offsets can be alleviated if all of the following ion
optics, the interquadrupole lens set and the third
quadrupole rod offset, are varied concurrently. The
resulting linked sweep is shown in Figure 3.7. The linked
sweep prevents the fragment ions from being "trapped" in a
deep potential well. In the same manner, the energy
dependencies for transmission of the parent ion, 156 f,
and for formation of fragment ions 51+, 77+ and 94+ were
measured as shown in Figure 3.8. These reactions were
observed at a N2 pressure of 2.7 x 10-5 torr, well within
single collision conditions. A number of features of this
figure deserve comment. Fragments arising from

dissociation of M+ have their maxima for formation at

76

 

 

100 -
80 -
c'
3 60 -
D .
< _
—' 40 a
0)
0C .
20 - -‘
0 ..,~..~....,.~,-.o,-.eM-x~»~,-«-~w-‘-'-‘-w';' Ink—k
TII'IIIIT'IIITITI'I'IfT
-140 -100 -60 -20 2° 60

Second Quad. DC Offset (V)

Figure 3.6 Dependence of 77+ formation on second
quadrupole dc offset.

 

 

77

 

100-
80—-
c'
3 60- 7‘
.Q
< "‘.' :"
— 40- ,.,
(D
m -I
20-l
0 I'ITT'H‘I'I'ITI'Vr'I'I'I
—140 -100 -60 ~20 20 60
Second Quad. DC Offset (V)
Figure 3.7 Dependence of 77+ formation on second

quadrupole dc offset with concurrent sweep
of interquadrupole lens and third quadrupole
dc offsets.

78

 

 

h
- 156+ ,»
u. ‘
2' I.
I w? wfi‘.‘ Tl. ..
f " .0, '-
d \ v- e‘
~M 0‘ HM”: '«~' ”I
U . : 0’
k; a, x
I
a:
e
C
3 - 8 -
.0 77+ 1.........~.-.w...--.,.-
‘. ’ - ~ . -'.§.
< .4 F.- » ’w‘V’. '-'~'I' fl~ "" .‘I.
51+ '
e e .’e..~.. '8‘." .v’a,’--,Jc~ ‘. . e
0‘ “I ' . ' ' "-9-“. ~
A s s '-
O x I
A o’ ’
-I '~ ~.
' V
94+
’ I
_ 1 0 _ .‘o . . 9.. . , ‘
' ' .e ' ’. 2“. .e~ :0 {‘50. 0 .‘. '3. ..vfln- U e...‘e I. . I .Q
. ~ . . ~ g. 1‘5“:"t’N-‘Q'J.“I'V'I'fe'? ~.o.o , *2, ~.- . . . . . I . . _ 9.... ~ A. . V...“ ,‘
r‘..0 '~’_fi. an . e o .v‘
I s ,
$ .
- - \ o .’ .e‘
' O . e u
.u . ‘. °.’. =-
. I... . .. . l 1 ' . . ..’.'."
.'-:A.:x‘ .- ° 1 A 0 0 .3: t . i
f I I T 1 r I l f 1 U I I 1 U I I I

-100 -60 -20 20 60
Second Quad. DC Offset (V)

Figure 3.8 Dependence of bromobenzene parent (156*) and
CAD (N2) fragments (77*, 51+ and 94*) on
collision energy.

 

79

large values of lab collision energy whereas the ion
resulting from reaction of M+ with water has its maximum
at low collision energy, typical of most collisionally
activated reaction (CAR) products. At second quadrupole
rod offset potentials high enough to retard any parent ion
from passing, the exothermic reaction to form 94+ still
occurs. The reaction to form 77+ is endothermic by 2.89 eV
and this energy is readily obtained by a very efficient
conversion of the parent ion translational energy into
internal energy. For this system 19 eV of lab collision
energy corresponds to 2.89 eV center-of—mass collision
energy and the onset for 77* formation trails behind the
parent ion transmission by roughly this amount. At higher
collision energies the reaction channel to form CAH3+ (m/z
51) directly from M+ opens. The onset for 51+ formation
follows the parent ion by approximately 8.4 eV center-of-
mass collision energy. The energy dependence curves for
the fragment are all relatively smooth, in sharp contrast
to the transmission curve obtained for the parent. The
presence of restrictive apertures at both ends of the
quadrupole collision chamber can give rise to this type of
curve for simple ion transmission. The same amount of M+
is incident upon the second quadrupole for all collision
energies. This type of transmission behavior is not
evident with a TQMS instrument of close coupled design.

The fragment ions are less inclined to exhibit this

80

behavior since they are formed at a continium of positions

in the second quadrupole.

From the data in Figure 3.8, the cross section for
fragmentation of C6H5Br+ versus collision energy may be
derived. The results for 0 to 45 eV of lab collision
energy are similar to previously reported values obtained
with the TQMS of close coupled design (48). The shape of
the curves are identical with the curve obtained in this
study shifted to 5-6 volts lower energy. Data above 45 eV
collision energy provided values for cross sections which
fluctuated with the parent ion transmission. The relative
cross sections were calculated from -ln (1 - ZIri/(IP +
2121)) where 2121 is the total fragment ion current and IP
is the transmitted parent ion current. This is a

simplified form of the equation

a = (-2.83 x 10 '17/1 P) * 1n [1 - ZIFi/(IP + 2I31)] (3.3)

where d, the cross section, is in cmz; l, the path length,
in cm; and P, the collision gas pressure, in torr. A large
relative cross section was observed at 0 eV collision
energy and this results from the contribution of 94+
formation at this low energy. A value for comparison is
the lab energy required for producing half the maximum
cross section and from this work a value of 31.5 eV is
obtained; a value of approximately 37 eV obtained with the
close coupled instrument (48). The fact that this work was

performed with a fixed RF voltage on the second quadrupole

 

 

81

and the resultant different transmission characteristics
for the different fragment ions affects the comparison.
The close coupled instrument sweeps the second quadrupole
RF voltage with the mass of the third quadrupole and does

not suffer differing transmission characteristics.

The relative contribution of each fragment to the
total cross section at each collision energy may also be
derived from this data. These relative cross sections from
0 to 120 eV of lab collision energy are shown in Figure
3.9. The importance of the 94+ reaction product for the
total cross section decreases rapidly with increasing
collision energy. At sufficiently high collision energies
direct formation of 51+ is responsible for 30% of the

total cross section.

A measure of the fragment ion median kinetic energy
and energy distribution can be obtained by applying a
retarding potential to the third quadrupole rod offset at
a fixed collision energy. However, these measurements do
not provide meaningful results in the TQMS instrument used
in this study due to the energy selective nature of the
electrostatic interquadrupole lenses. Overall the fragment
ion energy distributions as determined with a close
coupled TQMS instrument are shown to be broadened by the
scattering process upon collision in the second quadrupole
(48). The energy of the parent ion following an inelastic

collision has a term which is directly dependent on the

82

 

 

 

 

1.0-
0.8“ 156 + 4 77 +
0.6-
9.1
U} 0’4” 156+ -) §I+
0.2
156+ -> 94*
o.oj ,
o 20 40 so so 100 120

LAB COLLISION ENERGY (0v)

Figure 3.9 Relative cross sections for bromobenzene CAD
(N2) as a function of collision energy.

83

cosine of the scattering angle (70), and the fragment ion
energy is related to the parent ion energy by a factor of
fragment mass/parent mass (48). With use of the retarding
potential technique it may be possible to observe kinetic
energy releases on fragmentation in properly configured
TQMS instruments, provided the effect is not obscured by
the width in distribution arising from the scattering

angle.

conclusions

The results reported in this study are beneficial not
only for gaining knowledge regarding the collision process
in TQMS, but they also serve to illustrate differences
between a TQMS instrument with interquadrupole lenses and
one with close coupled quadrupoles. The operation of the
second quadrupole with a fixed RF voltage does not permit
an exact quantitative comparison between the two different
instrument designs due to differing transmission
characteristics, although qualitative and semiquantitative

comparisons can be made.

Another major difference between the two types of
TQMS instruments is the design of the quadrupole collision
chamber. In the TQMS used for the studies reported here,
the enclosed collision chamber is uniformly pressurized
with target gas over its entire length. Consequently, ions

are exposed to target gas throughout their residence time

84

in the second quadrupole. In the close coupled TQMS
instrument the target gas is introduced in the center of
an open, transparent (wire mesh) quadrupole assembly by
means of a molecular jet, so the collision region is much
shorter and the ions have less time to interact with the
target gas. This might explain why a larger number of
fragment peaks are observed with the MSU TQMS instrument.
The only fragment peaks reported with the close coupled

TQMS instrument were CeHs+ (m/z 77) and C4H3+ (m/z 51).

The observed effect of the RF voltage in the
collision chamber on the fragmentation of C6H5Br+ to CsHs+
may provide evidence for separate activation and
fragmentation steps. The interquadrupole lenses between
the second and third quadrupoles may provide an effective
drawout potential for observation of fragment ions
generated with low translational energy. A variation of
the collision energy reveals that fragment ions formed
from a CAD process occur at higher collision energies than

those from a CAR process.

The activation step which involves the conversion of
translational energy to internal energy is shown to be a
highly efficient process. The energy dependent-
transmission of the parent ion through the second
quadrupole may be the result of the presence of the
interquadrupole lenses. This has an adverse effect on

determining kinetic data from measured ion intensities

85

such as cross sections. Over the energy range studied,
results similar (5-6 eV difference) to those reported with
a close coupled instrument were obtained. The TQMS
instrument is a versatile tool for studying collision
processes between ions and neutral molecules at energies
above thermal. The large range of scattering angles in the
collision may serve to obscure some kinetic effects.
Future improvements to the ion path and collision chamber
of our instrument will improve its suitability for studies

of the collision process.

B. Investigation of Charge Inversion in the Collision
Chamber in Triple Quadrupole Mass Spectrometry

Introduction

The triple quadrupole mass spectrometer was utilized
to probe the characteristics of our collision chamber for

observing the energetic charge inversion reaction

M‘ ---> M+ + 26‘ (3.4)

This reaction has been observed only in systems in which
the center-of—mass collision energy has been quite high.
Most investigations of the charge inversion reaction have
utilized sector MS/MS instruments (71). The energy for the
inversion reaction arises from collision of the energetic

(ca. 3-10 kV) parent ion beam with a light target such as

86

helium. For an ion of mass 150 accelerated by 3000 volts,
a center-of-mass collision energy of 78 eV is obtained.
Similar energy may be obtained by collision of the same
mass ion, with axial kinetic energy in the 100-200 eV
range, on a heavy target gas. In fact, charge inversion
reactions have recently been reported in a commercial
triple quadrupole MS/MS instrument which utilized a close
coupled quadrupole design (72). The current design of our
research instrument differs significantly from the close
coupled instrument design. The RF-only quadrupole
collision chamber in our system is of closed design and
has two sets of four element interquadrupole ion lenses

developed with the aid of ion trajectory simulations (41).

The effects of ion path potentials and collision gas
pressure on the charge inversion reaction for the compound
hippuric acid (CoHoNOa) were observed. The collision gas in
the second quadrupole was sulfur hexafluoride. This target
gas has a mass of 148 and a high electron affinity.
Previous work in our laboratory and that of others has
failed to show charge inversion products with conventional
target gases (i.e. Ar, N2). Hippuric acid was chosen in
part to aid in the comparison of the capabilities of
different TQMS instrument designs. The positive charged
fragment ions formed were of low abundance relative to the
negative charged parent ion, indicating a much lower
reaction cross section than that for noninverted

dissociation. Fragment ion intensities were large enough

87

for analog detection electronics to successfully record

the spectra.

Experimental.

All experiments were performed on the TQMS instrument
in our lab which is described elsewhere (34,39). A CI
source from a Finnigan 3000 utilizing a reagent gas
mixture of methane and nitrous oxide produced an abundant
hydrogen abstracted molecular anion, (M - H)", at m/z 178
for hippuric acid. The gas mixture was in the ratio of
1:0.4 as measured by a source pirani tube and maintained
at 1.4 torr. The source lens elements were operated in a
standard acceleration/deceleration mode. The selected ions
after passage through the first mass filter enter a four
element interquadrupole lens. This lens set provides
effective focusing of ions over a wide mass range into the

following quadrupole.

The amount of collision energy available in a
quadrupole collision chamber depends on the magnitude of
the difference between the source potential where the ions
were formed and the DC rod offset applied to the
quadrupole collision chamber. In our system all ion path
potentials are variable over a range of i 200 volts. This
limits the maximum lab collision energy accessible with
our system to less than 400 eV. The results reported in

this study are those of 170 eV collisions. Industrial

 

88

grade SFs was maintained at constant pressure in the
collision chamber with the aid of an automatic pressure
controller. The collision gas pressure was measured with a
Bayard-Alpert ion gauge tube calibrated for SFs. The RF
voltage was maintained at a constant value of 160 V in the
second quadrupole. This setting provided good transmission
of ions in the mass range of interest (20 to 180 amu) and
had a low mass limit just above the abundant OH“ reagent

ion.

The setting for the ion path potentials following the
second quadrupole is critical, both for the type of
product ions observed as well as for the analyzing
resolution of the third quadrupole. The effect of
collision energy on the charge inversion fragment ions was
observed. The optimization of the interquadrupole lens
between the second and third quadrupoles for the desired
types of product ions was performed. Also, the choice of
the third quadrupole rod offset and its effect on peak

shape and fragment ion type was studied.

The pressure of the SFe target gas and its effect on

the charge inversion ion abundance was investigated.

All ions, both positive and negative, were detected
with the aid of conversion dynodes coupled with the

Galileo ChanneltronTM model 4770 detector.

89

Results and Discussion

The lab collision of 170 eV was chosen since it is
above the threshold for formation of all observed
fragments and also to facilitate comparison between
instruments. A charge inversion spectrum obtained at
2.6 x 10" torr SF6 and the structure of hippuric acid are
is shown in Figure 3.10. The threshold for formation of
fragments C6H5+ (m/z 77), and CsHsCO+ (m/z 105), was
observed to be approximately 50 eV, in agreement with a
previously reported value (72). Other fragments can be
seen at m/z 51, 28, 89, and 127. C4H3+ is partially
responsible for 51+. SFe undergoes dissociative ionization
in the collision process to produce SF+ (m/z 51), SFa+ (m/z

89), and SFS+ (m/z 127).

The RF voltage on the collision chamber was
maintained at 160 volts. Although this fixed setting does
not provide optimized transmission for all of the ions, it
provides a consistent and adequate transmission for each

m/z value of interest.

The choice of potentials on the interquadrupole lens
elements between the collision chamber and the final
analyzing quadrupole has a marked effect on the type of
ions observed. Optimum daughter spectra are obtained when
the quadrupole collision chamber is treated as the source

of the fragment ions, which possess forward momentum as a

90

.mmm
GOfinflHHoo 9mm anon Tod x m.N .pHoo oeusmmfls
Ho Esupoomn coflnuosbw. omnono one ousposupm 0H6 ousmflm

AsEov mmofi

 

 

 

 

 

8— 9: cm? 2: am so 3 cm
— _.P.b.—.-p— hp—._»F.—L ._.—. NI
1. Fl
0.0.1 OEDQEI
. ._. o
2 1
O

 

 

 

was 601

°unqv

 

 

91

result of the CAD process, and the following ion path
elements are adjusted accordingly Such a spectrum is shown
in Figure 3.11. The spectrum in this figure and that in
Figure 3.10 were acquired with the second quadrupole rod
offset at 150 V, a collision gas pressure of 2.6 x 10-4
torr, the interquadrupole lens elements set at 150, 70 and
200 V and the third quadrupole rod offset at 140 V. The
source potential was maintained at -20 V. An effective
drawout potential may be applied to the interquadrupole
lens set by dropping the potential of the first lens
element, the ferroceramic cylinder, which is inserted into
the back of the second quadrupole. The spectrum shown in
Figure 3.12 was obtained with this lens set at 5, -5 and
50 V. The charge inversion product ion (M - H)+ is nearly
the base peak. Fragment ions resulting from this parent
are reduced in intensity. The intensity of ions resulting
from SFe are all increased relative to the charge
inversion fragment ions. The ability to apply a drawout
potential to this lens set allows one to probe for ion-
molecule collision product ions which may, by nature of
their formation, possess little forward momentum. In many
cases, these low forward momentum product ions are formed
in the second quadrupole but are either detected at
reduced intensities or not detected at all simply because
some or all of these ions fail to exit the second
quadrupole. The fragment ions arising directly from a

selected parent ion are easily seen since they possess

92

.naoH
OHOQSHoasvuopcw co HoapGOpoa QSOZGuo 950:9“:
pace causamaz Ho asupooan Godunoifi 0922.8 :6. madman

Aanv mmofi

of 2: at cm. 2: on em 3 cu
_._..._._._L_.__.__..r._._._...

 

_ . _ _

 

 

 

cm

3

8

cm

2:

'Ieéi

'UIWCIV’

 

93

.mcoa
oHomsupoSUHopcfi so HQGOHQCOpom psozonp Spa:
neon cannons: ac asupoonn communes/g.“ owns—.8 Nam ousmfim

AsEoV mmofi

of of o: of E: on 8 3 cm
— . — p P p — p — b — p — h — — p h b — P P b — P — . P p — . D

 

 

 

 

 

 

l 2:

'Ietl

'Ufiqv

 

94

energy related to the parent’s lab energy after collision
by a factor of the daughter mass divided by the parent

mass and they travel at the same speed as the parent (48).

The rod offset of the final analyzing quadrupole
serves mostly to effect the resolution obtainable in the
final MS stage and to a smaller part, impose a drawout
potential. For maximum resolution the third quadrupole rod
offset should be held near that of the second quadrupole.
Table 3.2 demonstrates the effect of decreasing the third
quadrupole rod offset on the peak width at half height for
selected ions: 51*, 77+, 105+ and 127+. The greater the
axial energy possessed by an ion in the third quadrupole,
the fewer RF cycles it experiences and the poorer the mass
discrimination. Most notable with the decrease in mass
resolution is a lift-off of the leading edge of the peak.
The relative abundance of these selected ions as a
function of the third quadrupole rod offset is shown in
Table 3.3. Decreasing the rod offset leads to an
increasing drawout potential of the third quadrupole
relative to the second quadrupole. This is shown by the
increase of the relative abundance for ions which arise
from fragments of SFe, 51 (SF+) and 127 (SF5+). The
abundance of 51+ at high third quadrupole rod offsets may
be due in a large part to the C4H3+ fragment ion. As the

rod offset decreases more SF+ (m/z 51) is collected.

95

Table 3.2 Effect of decreasing third quadrupole rod
offsets on peak shape.

Third Quadrupole Peak Widths at half height (amu)
Rod Offset 51+ 77+ 105+ 127+
160 0.75 1.20 1.40 -~--

100 2.00 1.50 1.50 1.55

50 2.65 1.85 2.20 2.13

0 3.50 2.25 3.25 1.75

-50 3.40 2.90 3.60 1.75

 

96

Table 3.3 Relative abundances of ions formed from charge
inversion of hippuric acid at various third
quadrupole rod offsets.

Third Quadrupole Relative Abundance Intensity
Factora
Rod Offset 51+ 71+ 105* 127+

160 15.4 100 46.5 ---- x 1.0

100 46.0 100 76.3 7.8 x 1.4

50 42.2 100 55.4 10.8 x 2.2

O 48.2 100 47.2 24.1 x 2.0

-50 73.9 100 49.5 29.4 x 2.2

a. Times more intense

97

The abundances of the inversion product ions are
sensitive to the collision gas pressure. The pressure of
SFs was varied from a low of 1.5 x 10'5 torr up to
1.05 x 10'3 torr. The results of this variation are shown
in Table 3.4. The intensity of 89+ and 127+ was not
recorded for a pressure of 6 x 10’5 torr. No charge
inversion ions were seen for the highest pressure
attempted. At very low pressures single collision
conditions exist and 105 + is more abundant than 77*. As
the pressure is increased and multiple collisions become
significant, the abundances of the two ions undergo a
reversal, in agreement with results by Douglas (72). As
the pressure is raised higher, the fragment ions from SFe

predominate.

The investigation of the charge inversion process of
hippuric acid reported here serves to illustrate that the
optimum daughter spectrum is obtained when the second
quadrupole is treated as an ion source. The inclusion of
interquadrupole lenses allow a drawout potential to be
applied without serious degradation of resolution as would
be the case in a tandem quadrupole instrument of close
coupled design where increasing the drawout potential (the
difference between quadrupoles two and three rod offsets)

would severely decrease the resolution. The overall result

98

Table 3.4 Relative abundances of ions formed from charge
inversion of hippuric acid at various second
quadrupole pressures.a

Second Quadrupole

 

 

 

Pressure Relative Abundance (%)
SF6(torr) 28+ 39+ 51+ 77+ 89+ 105+ 127+
1.5 x 10"5 --- 3.4 14.5 95.3 1.7 100 1.9
6.0 x 10-5 --- --- 11.7 77.4 --- 100 ---
2.6 x 10'4 9.4 --- 16.6 100 6.6 70.4 7.9
6.6 x 10-4 --- --- 23.7 27.4 --- --- 100
1.05 x 10-3 --- --- --- --- --- --- ---

a. ELAB = 170 eV.

99

is the ability to impose a drawout potential to detect low

forward momentum product ions.

The major improvements likely to increase the
performance of our TQMS instrument for charge inversion
studies would be the implementation of pulse counting
detection for increased sensitivity. Also beneficial will
be the results of research underway to improve the design
of the collision chamber to facilitate collection of all

ionic products formed.

 

CHAPTER FOUR

APPLICATION OF COLLISIONALLY ACTIVATED DISSOCIATIONS
FOR COMPONENT IDENTIFICATION IN A PHENOLIC MIXTURE

Introduction

The eleven component acid fraction of the priority
pollutants was chosen to demonstrate an application of
collisionally activated dissociation (CAD) in triple
quadrupole mass spectrometry (TQMS) to aid in mixture
analysis. The potential of this technique for the rapid
screening of targeted compounds in complex mixtures will

be discussed.

Mass spectrometry has been applied for mixture
analysis of complex petrochemical mixtures since the
1940’s. By itself, mass spectrometry was of limited use
for these analyses. Improved analysis resulted when a
preseparation of the mixture via gas chromatography was
performed prior to mass analysis. The application of GC/MS
for the analysis of organic pollutants has been reviewed
(73-75). The restrictions of chromatography regarding the
types of compounds which are stable in and can pass
through a column and the time required for chromatographic
analysis led to the deve10pment of a new technique in

which a mass spectrometer is utilized for the separation

100

101

stage, that of MS/MS. With this technique came a
substantial reduction in the chemical noise problems

experienced in GC/MS due to column bleed.

Early applications of MS/MS used only the metastable
ions from a selected parent for mixture analysis. These
experiments were performed on a double-focusing mass
spectrometer of conventional (E/B) and reversed (B/E)
geometry (76-79). The inclusion of a collision cell into a
reversed geometry instrument resulted in a versatile mass-
analyzed ion kinetic energy spectrometer (MIKES) where the
ion selected by the first sector (B) could be induced to
fragment by collision with a neutral target gas, usually
helium, prior to analysis by the final sector (E). The
information contained in these collision induced spectra
is much greater than that contained in the metastable
spectra. The application of MIKES for mixture analysis has
been the subject of many papers (13,80-82). Different
methods of ionization were found beneficial for certain
types of analysis. Soft ionization techniques such as
chemical ionization (CI) (13) and field ionization (83,84)
have found successful application with MIKES for mixture
analysis. MIKES analysis with negative chemical ionization
(NCI) often provides information complimentary to that
presented by studies of positive ions (85,86). Specific
examples of the MIKES technique for mixture analysis have
demonstrated its inherent suitability for detecting small

amounts of a specific compound in complex matrices. This

 

102

is dramatically shown by the analysis of cocaine and its
metabolite in human urine (87). Other impressive
applications include MIKES analysis of deoxyribonucleic
acid pyrolysis products (88), and underivatized peptide
mixtures (89,90) where isomer identification and sequence
information were obtained. Further examples of MIKES for
the analysis of mixtures are presented elsewhere (13,80-
82,85,86,91). The MIKES technique allows for direct
mixture analysis with little or no sample pretreatment.
The efficiency of the collision process and the mass
resolution can be improved significantly if quadrupoles

are used as the mass analyzers (2,14,34).

Double quadrupole instruments were developed for
their application to the analysis of isomer mixtures and
targeted compounds in complex mixtures (92). Recent
reviews have treated the application of triple quadrupole
mass spectrometry for analysis of mixtures and select
compounds (93,94). The high transmission of the RF-only
quadrupole collision cell in TQMS enables the high
sensitivity of the technique (34). For the analysis of
parathion on ivy and lettuce leaves, a linear calibration
curve down to 10'11 gram was shown (94). Analysis of TCDD
in the presence of 3000 fold larger amounts of isobaric
species, DDE and PCB demonstrates the selectivity of TQMS
(95). The TCDD CAD spectrum is unique and accurate
quantitation is achieved. The application of NCI and CAD

for ionization and subsequent dissociation of components

 

103

in complex mixtures such as industrial sludge in a TQMS
instrument is shown to be a sensitive technique (55). For
trace analysis the use of multiple reaction monitoring
provides very selective and fast determination since only
a minimal amount of spectral information need be recorded.
Functional group analysis is facilitated either by
scanning both mass filters to observe a characteristic
neutral fragment loss or by selecting a characteristic
fragment ion with the second mass filter (46,47,96,97).
These methods are easier to implement on a quadrupole
MS/MS instrument than are the corresponding scans on a

sector MS/MS instrument.

In the present study, the application of the MSU TQMS
instrument for analysis of the compounds shown in Figure
4.1 is discussed. Present are two phenols, five
chlorophenols and four nitrophenols. The utility of CAD
for the direct analysis of very small samples is

discussed.

Experimental

The TQMS instrument used in this study was developed
at MSU and is described elsewhere (34,39). Negative ions
were detected using a conversion dynode (+3000 volts)
prior to the detector. All samples were reagent grade.
Four different ionization techniques were compared for

production of molecular ions. These included electron

 

104

.npcspsHHom apfiuoflum canopooupxo cfio<

H.v assess

 

 

 

 

 

 

  

 

 

 

 

 

 

NN- .2.)

n
so2m=m4>mem2_nuvxw I

em .2.)

JOZMIQ

mEZflHDAAom whmmofimm
mamflhoaflmfixm Q~O<

so so
woo 3 5 o. 40 mm. .k.) mm. .3.)
AOZmzm A no 40mmmolo AOmem n no
Iomcgzo<92mi Iomtzaloé nomoamoghlcém
o o
no go
vw. .Buz we. .;.5 Nv— .3.) n1
No. Q 4 40mm”;
somemomtzEIEm 402m=mom04:o_ole.m IEIomoazolm
o 0
on. it: an. .>>.z mm. .2.)
AOZmEmomtznm mozm—mmomtzuo No: ‘ moznmzmomommolo n
o o
_o

 

 

 

 

 

 

 

105

impact (EI) positive ion CI with methane and isobutane and
negative ion CI with a mixture of methane and nitrous
oxide. All were performed in a Finnigan 3000 CI ion
source. The results of El and NCI ionization methods for

the analysis are presented in detail.

The source parameters for the El study were 20 eV
ionizing electron energy and 0.5 mA emission current. The
second quadrupole collision chamber was pressurized to
2 x 10'4 torr argon target gas and the collision energy

was 15 eV (lab).

The source parameters for the NCI study were 150 eV
electron energy, 1.0 mA emission current and a CH4/N20
reagent gas mixture in a 1.6:1 ratio maintained at 1 torr
as determined by a source pirani tube. The quadrupole
collision chamber was pressurized to 1 x 10‘3 torr argon

with a lab collision energy of 10 eV.

For each component, a CAD spectrum obtained from the
molecular ion of a pure sample was compared to a CAD
spectrum obtained from the mass-selected characteristic
ions present from a laboratory mixture sample. Examples of
mixture analysis for both positive and negative molecular
ions are shown in Figure 4.2. Positive molecular ions
formed through E1 of a mixture of three of the components
present are shown in the upper portion of Figure 4.2. The
different molecular ions are separated with the first

quadrupole mass filter. The selected ion, m/z 184

106

.mcoe 0>Hpcmoc one o>fipflnom £902 mo mflmhamco

ousteE .uom

 

 

A/

 

 

INN. INN.
two—ATIIUWWWV.Imo.
Nm INm

 

 

 

 

 

 

 

 

 

 

 

 

 

29.850 zo_.<m<$m
zo_ mm<2
+vm. +vm.
+50. +no.
+.m +.m

 

 

 

 

 

 

 

 

 

 

 

 

onwasnoop m2\mz “Odomshowsa m.¢ ousmwh
O INN. ® ® ® :3. In... Infill ..m.
umo.wwwll.umm. new..m|||mwwwv.nmm. nem..¢|||. on.
® :2. ® 0 ® :3. .- :3. LEAII. .3.

29.22%

20:55:04.5 mm; zo:<~._zo_
® +1... ® ® @ +3. +3. +3.11 mm.
+so. +«m. +vm.ATIIuwwwv.+vm. +vm.ATII. vm.
® +8 @ ® @ +3. +3. +~m.A.I| um.

 

 

 

107

(2,4-dinitrophenol) enters into and is fragmented within
the second quadrupole, an RF-only quadrupole collision
chamber. The characteristic fragment ions formed through
CAD are then analyzed by the third quadrupole and
detected. An analogous case is depicted for the proton
abstracted negative ions shown in the lower portion of
Figure 4.2 with the selected molecular ion from

orthonitrophenol.

A direct insertion probe was used to transfer samples

into the source.

gggults and _issussion

A major factor necessary for successful mixture
analysis with MS/MS lies in the ability to form molecular
or characteristic ions of each of the components to be
analyzed. The 20 eV EI source often generates suitable
abundances of parent ions for their analysis while
minimizing the amount of fragment ions generated from each
compound. However, some fragment ions are generated and
these can give rise to interferences in the analysis of
closely related compounds. For example, trichlorophenol
upon EI ionization, even at low ionizing energies, will
generate a (M - Cl)+ ion which will interfere with the
analysis of dichlorophenol. The NCI technique provides OH-
reagent ions which abstract a proton to form abundant

(M - H)‘ ions (98). The (M - H)‘ ions carry a majority of

 

108

the ion current. Few interfering fragments are generated
with NCI and their effect on the analysis can be minimized
as discussed later. Positive ion CI with both methane and
isobutane was found unsuitable as ionization methods for
this application due to the degree of fragment ions and
addition product ions generated which interfere with the
analysis of the many closely related components in the

mixture.

The reference CAD spectra for the 20 eV EI generated
positive ions are presented in Table 4.1. The table gives
the spectra in terms of losses from the selected parent to
emphasize that the CAD fragments are characteristic of the
functional groups present on these compounds. For example,
2,4-dichlorophenol exhibits losses of: HCl; CO and C1; C0
and HCl; CO, Cl and C2H2; and CO, HCl and Cl. Similarities
are evident among components possessing the same
functional groups. The majority of interfering fragment
ions such as a dichlorophenol fragment ion from
trichlorophenol dissociation in the source can be avoided
by choosing the proper isotope as the ion characteristic
of that component. With the dichlorophenol example,
selection of m/z 164 (CSH4O35C137Cl+) as the characteristic
ion greatly reduces the interference since very little
164+ is formed from trichlorophenol. While this is
sufficient for the chlorine containing components, the
slight interferences with the nitro containing components

are unavoidable. Even though these

109

 

 

.OHomswvnsu
pnhwm 050 an common :0« omm mo uczoad Hanan Eoum Axe.flv emu .coon onH< .U
.Axm.m. mm .Axv.mv an .Axm.mv mm
.Axo.nv mm .Axb.v. we .Aao.mv am .Axm.mv be .Axo.v. ma .Auo.nv vm .Goou ouH< .m
.Axm.H. me .coon onH< .o
.Axo.nv mm .Axb.mv mm .Axa.mv no .Axo.nv ow .A&>.Hv ma .coon onH< .o
.Axm.uv mm .ARN.HV mm .ANN. wm .coon ouH< .0
ANN.NV mm .Axo.Hv mm nA<me N\a .Goon onH< .n
.moocouomuovca oao>u ow sonogo nu: xuom N + < 0:» awnhauso 0:9 Rom .6
«mo
.00 amno HOE Ho oz 00m 00 on: no «mo “Ame anon oHnunowm
II II o.> II II II II II II II vow uHOQO£mouoH£ooucom RAH
II II II II o.m II II m.n II II 0mm umououoIoIouaficaoIo.v AoH
II II >.m II II II II II II II mmfioHononmouoHSOAupIm_v.N Am
II II II II m.mfi II II II II II voH «HocoamouvfinwnIv.N .0
II II m.w II II II II II II II ammu HocosmouoHsomva.N Ab
II II II >.mm II II II II II II Nvu HanouoIEIouoHAOIm Am
II II II II n.m II II II o.H II mnH uaosogmouoficIo Am
II II II II m.>a II II II II II mma nHosognoupchA Av
II II H.v II II II o.w II II II amNH HosozmouoHnoIo Am
m.H II II II II II m.o II II m.mH «we Hocosaamnnoewva.m AN
II N.H II II II o H N.> II II II vm Hocosm AH
vaz mmIZ wMIZ mmIE onus mmIz mNIZ mHIz BHIE mHIZ +2 ncaomfioo

oooooaon sown nnoa unusuauh .ncod oopnhogou HM >oom «o anaconn Q<o

.A<m a. nod

H.v oHnuh

Oo
.mwo
.«02
.~02

I- I- m.mm I- I- I-
0
1 II II II II II III
1

mVHIEONHIE mmIE vaE mmIZ NmIZ

Ozm

.«OZ

.nmo

INOI

Oo 00

.«02 .Ho

.«02 .Hom

II 0.0m
¢.N II

N02
.N02
.mN

NOz
.N02

N02
.«02

«mac N02
.00 .02 02
.Ho .m .«Oz

mmIz beE mbIz

00
.«Oz

«FIE

Oom
.Hom

mmIZ

Noz
.Omm
IHOI

Oo
.Hom

m.mH
fi.N

m.om

m.>N
m.H

wmIZ

00 OD «mao

.Ho .Oz .00 Noz Adm

II II O.H ll

mmIE mmIZ vaE mVIE

..u.p:oo. ..v magma

A:
SH

111

interferences exist, they are small enough that they

little affect the analysis.

The reference CAD spectra of the characteristic ions
generated by NCI with an OH' reagent ion are presented in
Table 4.2. In most cases NCI with OH' provided (M - H)’
anions which carried almost all of the ion current. Only
in the cases of o-nitrophenol, 2,4—dinitrophenol, 4,6-
dinitro-o-cresol and pentachlorophenol were any fragment
ions formed in the source. This was made advantageous for

this analysis since the [(M - H) - O]‘ ion (m/z 181) was

 

chosen as the characteristic ion for 4,6—dinitro-o-cresol,
avoiding interference with trichlorophenol isotopes at its
parent (M - H)‘. Also the [(M - H) - Cl]- ion (m/z 230)
was chosen as the characteristic ion for pentachlorophenol
as the parent (M - H)- was stable to collision and did not
fragment under the conditions employed in this study. The
terminology [(M - H) - O]‘ and [(M - H) - Cl]‘ is
introduced since these represent losses from the
respective parent (M — H)" ions generated in the source
and the terminology (M - OH)‘ and (M - HCl)‘ is not
indicative of their origin. The CAD spectra displayed in
Table 4.2, as with the previous table, show that the
losses are characteristic of the functional groups
present. All nitro containing components lose NO and all
chlorine containing (M - H)“ ions lose HCl. The extent of
fragmentation is generally less with the negative ions.

However, the CAD fragments are unique to the CAD process

 

112

.cou0£0 mm: condom 03v Gd uoahom cow IHHUIAmIzvu

 

 

one .Hdoo sewnwaaoo 0:0 a“ :vaupcoemnum omhovcd #0: can IAmIzV 0:9 .n
.uopmsHo HocoznouoHLo«np 03p Baum mucouOHHOHGM Uwo>d op mdmhangu 05¢ how
cmnoso mm: GOA mass .00hsom 03¢ Ca Had no xwom <m Rm.m n no: pssomfioo mags n
02
IuoI
Ozm «Annoy Co: 00 Cam :0 0 mac Aqmv whoa oHanOHm
IHHI IHHI IHHI mwww IHHI IHHI IHHI IHHI 0mm IHHUIomIsz AHOEUSQOAoflxomvsom AHA
>.H N.mm II II II II II II «ma IHOIAmIZVHaaomouoIoIouvficfich.v AoH
II II II II II II II II mm. -.mIz. HocosmouogosupIoJ§ 8
II m.m II II II II II II mmH IAmIZV HoaosnoupflsHoIv.N Aw
II II II II II II II II Hma IAEIZV HOGOSQOHoflsoflUIv.N as
II II II II II II II II HVH IAEIZV HonouOIEIOHoHsoIm Am
II m.mN m.H II N.H N.v m.mH II and IAmIzv Hocosmouv«:Io Am
II H.mm m.H II II II II II mmfi IAmIEv Hocosnouvchn Aw
II II II II II II II II hNH IAmIz. accosmouOHsoIo an
II «.0 II II II II II m.o HNH IAmIZV Hoconudhnpoeflva.N AN
II II II N.H II II II II mm IAmIZV Hocosm AH
HmIE ome mNIz mNIZ mHIE baIz mHIZ mHIZ N\E :oH pouooaom pgsonaoo
.A<m x.

sea conceaon sown nnoH pcosmcuh .n:o« ooosuocou Huz no oupoonn Q<o N.v canoe

113

Oz 00 0.02
.uoz .Oz 00 00 00 00 Oz IMOI
.NAOOV .«Oz .NAOZV .Noz «AHUEV «A02. .02m .Oz .30 Noz Hum HO Adm

I- - - - - II I- - I- - - - .22

II II II II II s.H m.s m.m II II I- II .oH
II II II II II II II I- II II «.mm N.H .m
o.m e.m s.mH v.m II e.oH II II II m.ofi II - .e

II II II II II II II II II -- o.H II .m
II II II II II I- - II m.m H.m II I- .m
II I- II II II II - II II N.N II II .v

- I- -I - - II - II I- I- m.s - .m

- I- I- - -I I- - - - I- - - .m
- - - II I- - I- - - - II - is

NmHIE voHIE mmIZ vblz NFIZ omlz mmIE mmIz vaE ©VIE mmIZ mmIZ

.Ap.pcoov N.v ®HQMB

114

since these fragments are not usually generated in the ion
source under NCI, unlike the fragments generated from
positive ion EI parents. The collision gas pressure of
10'3 torr argon was five times higher than that employed
with positive ions showing the increased stability the
selected negative ions have towards collisionally

activated dissociation.

A mixture prepared of equal amounts of each component
was ionized in the source by both 20 eV EI and OH' NCI.
The spectra obtained are displayed in Figure 4.3 and
Figure 4.4. The EI spectrum in Figure 4.3 shows detectable
amounts of each component coupled with significant
fragmentation as evidenced by the abundance of lower m/z
fragments. A much simpler NCI spectrum is displayed in
Figure 4.4. Far less fragmentation is evident and this
results in a simpler separation process to be performed by
the first quadrupole. Ideally, there is only one peak per
component. Fragment peaks cannot be distinguished from

component peaks with those lighter masses.

The correlation between the CAD spectrum of a mass-
selected characteristic ion from the mixture and the CAD
spectrum of a pure sample of that specific component leads
to positive identification of that component. Examples of
this technique are shown in Figures 4.5 and 4.6 for El
generated p-chloro—m-cresol and 2,4,6-trichlorophenol

ions, respectively. With NCI the mixture and reference CAD

 

.OHSQxflE mo Esmuooan Hm >5 ON m.¢ ousmah

AsEov mmofi

cow of so. 3.. cm. 8. cm cm 3
— n I— L p — — — n n u — h n _

 

 

115

I. . Id..:fi .

 

 

 

 

 

'IQH

'unqv

 

116

.owspxae mo Enhpooam Hoz ImO e.¢ oucmah

AJEOV mmofi

8N SN CNN com cm. 2: 3. cm. co. on 8 S.
_ L P. _ .—

 

 

 

 

' I

 

om

S.

8

on

8.

'ISH

‘unqv

 

117

 

 

 

 

 

(a)100 -
. 80-
C _
3
13 60 -
<1 -
40-
m .
Q: 20 _
0 l.
l'j'l'l'l'l'l'l‘l'l
SO 70 90 110 130 150
Moss (amu)
(b)100-
. 80-
C .
3
43 50 -
<1 _
40-
m _
D: 20-
0 L In
I l' l ' l T l 'I l l' I 'l
50 70 90 110 130 150
Moss (omu)
Figure 4.5 20 eV EI CAD (Ar) spectra of mixture (a) and

reference (b) m/z 142 M+ ion for p-chloro-m-
cresol.

(a)100

. 80
C.

3
43 50
<(

. 40
(D
Q: 20

0
(b)100

. 80
C
3
J] 60
<(

. 40
Q)
CK20

0

118

 

 

 

 

 

 

 

 

111111'lrl'I.'l'l':'l'l'l'
70 90 110 130 150 170 190
Moss (omu)
a I
'l'l'rTl'l'l'j'l'l'l I I
70 90 110 130 150 170 190
Moss (amu)

Figure 4.6 20 eV EI CAD (Ar) spectra of mixture (a) and

reference (b)
trichlorophenol.

m/z

142 M+

ion for

2,4,6-

 

119

spectra for the isomeric ortho and para nitrophenol and
2,4-dinitrophenol are compared as shown in Figures 4.7,
4.8 and 4.9, respectively. The slight differences in peak
heights arise from the sample introduction on a direct
probe and the acquiring of the reference and mixture CAD
spectra on separate days. The isomeric nitrophenols may be
distinguished by their different CAD spectra and the fact
that the more volatile isomer (o—nitrophenol) comes off of
the direct probe well in advance of the less volatile para

isomer.

For a targeted analysis of these compounds in a small
and less concentrated sample the technique of multiple
reaction monitoring (MRM) is advantageous. The selectivity
of the process and the rapid acquisition of data for each
component would permit simultaneous determination and
quantitation of all monitored species. In most cases,
monitoring a primary reaction (i.e. most abundant) would
provide the needed information. However, in the case of
the isomeric nitrophenols a second confirming reaction to
observe loss of O is required. Table 4.3 presents the
primary and confirmatory reactions to be monitored for
analysis of the phenolic mixture. With the lack of
additional CAD fragments for some of the chlorophenols the
appropriate loss from an isotopic peak can provide
confirmation. The application of MRM for this analysis

will be undertaken in the near future.

 

120

 

 

 

 

 

 

(a)100 -
. 80-
C .
3
43 60 -
<1 _
. 40-
m -
D: 20 _
0 n * r ' l ' l I ‘ l ' I
90 100 110 120 130 140
Moss (omu)
(b)100-
. 80-
C _
:5
13 60 -
< _
40-
Q) -
D: 20 _
0 II I ' 'I ' il ' I ' I
90 100 110 120 130 140
Moss (omu)
Figure 4.7 NCI CAD (Ar) spectra of mixture (a) and

reference (b) m/z 138 (M-H)‘ ion for
o-nitrophenol.

[1|-

121

 

 

 

 

 

 

(a)100 -
. 30 -
C _
3
J] 50 -
<1 _
. 40 -
o
05 20-
J
0 ' l ' l ' I ' I ' I
90 100 110 120 130 140
Moss (omu)
(b)100 -
. 80 -
C _
3
43 60 -
<I: _
4o -
m -
0: 20 _
O . T I fi I ' I ' I‘r I
90 100 110 120 130 140

Moss (omu)

Figure 4.8 NCI CAD (Ar) spectra of mixture (a) and
reference (b) m/z 138 (M—H)' ion for
p-nitrophenol.

 

 

(a)100
. 80
C
3
13 60
<(
40
Q)
frzo
0
(b)100
. 80
C
3
43 60
<(
. 40
Q)
0: 20
0

122

 

 

 

 

 

 

I .41 I A] I I l
' I ' I ' I ' I 1 I j I ' I ' I ' T 1 I ' I T 1 ' l ' I
50 70 90 110 I30 150 I70 190
Moss (amu)
I I I l I I I
' I I I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I
50 7O 90 110 130 150 I70 190

Mass (omu)

Figure 4.9 NCI CAD (Ar) spectra of mixture (a) and

reference (b) m/z 183 (M—H)' ion for
2,4-dinitrophenol.

123

Table 4.3 Multiple reaction monitoring reactions for

 

 

Compound Primary Reaction Confirming Reaction
Phenol 93' ---> 65‘ ------
2,4-dimethylphenol 121' ---> 106' 121' ---> 91‘
o-chlorophenol 127‘ --—> 91‘ 129‘ -—-> 91'
p-nitrophenol 138‘ ---> 108' 138* ---> 92-
o—nitrophenol 138' ---> 108' 138' ---> 122-
p-chloro-m-cresol 141' —--> 105‘ 143‘ ---> 105'
2,4-dichlorophenol 161' ---> 125‘ 161‘ —--> 89-
2,4-dinitrophenol 183' ---> 95' 183" ---> 123-
2,4,6-trinitrophenol 195‘ ---> 159' 195- -——> 160‘
4,6-dinitro-o-cresol 181' ---> 151‘ 181' ---> 123'

Pentachlorophenol 230' ---> 202‘ 228‘ ---> 200'

124

Conclusions

Both low energy EI and OH“ NCI were successfully
employed as ionization techniques for component
identification in the chosen mixture. Some interfering
species are generated in the source but proper choice of
the isotopic composition of the selected ion can reduce
their effect. In this study NCI provided the most abundant
parent ion species coupled with the least fragmentation in
the source. As such, it is an ideal ionization method for
this analysis. The CAD losses observed for both positive
and negative ion species are quite characteristic of the
functional groups present in the compound. The negative
ion parents are more stable to collision than are the

positive ion parents.

The use of CAD in TQMS for complex mixture analysis
has several advantages over other techniques. Compared to
GC/MS, the time delays associated with chromatographic
separation are eliminated. The selectivity is enhanced and
chemical noise is effectively reduced. Extensive sample
preparation is not required. A targeted analysis, in which
only the components of interest are analyzed, may be
rapidly performed. Continuous analysis of flowing process
streams (e.g. atmospheric analysis) is possible without
having to collect batch samples for subsequent
chromatographic analysis. A wide variety of samples

including nonvolatile or thermally labile compounds may be

125

introduced into the source of a mass spectrometer for
ionization by chemical desorption, field ionization or
fast atom bombardment. Also the samples have minimum

exposure to extraneous material in the TQMS instrument.

GC/MS still remains the better method for the
identification of all components present in a complex
mixture. The TQMS instrument is best suited for a rapid
screening survey for select compounds or compound types in
complex mixtures. The additional sensitivity, the unit
mass resolution of both mass filters and the simplicity of
operation make TQMS a preferred MS/MS technique for the

analysis of mixtures.

CHAPTER FIVE

APPLICATION OF COLLISIONALLY ACTIVATED REACTIONS FOR
DIFFERENTIATING BETWEEN POSITIONAL ISOMERS

Introduction

Chemically specific ion-molecule reactions may be
used with triple quadrupole mass spectrometry (TQMS) for

the analysis of positional isomers.

Opportunities for chemically specific reactions exist
in both the source and the collision cell. The benefits of
employing ion-molecule reactions in a chemical ionization
(CI) source for the analysis of isomers have been reported
(99,101,105,106,108). Isomers may be analyzed since the
rate constants for thermoneutral ion-molecule reactions
are likely to be significantly affected by small
structural differences in a molecule (100). The thermal
energy techniques of CIMS, flowing afterglow, ion
cyclotron resonance and high pressure MS are all well
suited to study gas—phase ion chemistry. Isobutane CI was
shown to be effective for the determination of the
relative configuration of substituents in various cyclic
epimeric molecules (99). The same study had also compared
the helium charge exchange spectra with the isobutane CI

spectra to demonstrate that spectral changes from

126

127

differences in stereochemistry are most likely at low ion
internal energies. Water CI of isomeric hydroxybiphenyls
to form hydrogen-bonded clusters of the protonated parent
ion with water, MH+ * (H20)n (n = 1-3), show gas-phase
hydrogen bond strengths to be a sensitive probe of the
structure of protonated molecules (101). The ratios of
hydrated parent ions in water CI are indicative of
individual isomers. Ammonia and methyl amine as CI
reagents were shown to undergo electrophilic aromatic
substitution on monohalobenzenes (102). Subsequent work
using ion cyclotron resonance techniques (103) revealed
the ion-molecule reaction which leads to substitution to
be between either NHa+ and Csst or CSHEX+ and NH3. These
reactions may serve as structural probes for
polysubstituted samples, although this was not attempted

in reported efforts.

Hydrogen-deuterium exchange with deuterated reagent
gases in a CI source was shown to be an analytical
technique to determine the number of active hydrogens in
an organic compound (104). Exchange with D20, C2H50D or ND3
CI reagent gases were observed for both positive and
negative molecular ion species (105). With NDa or CH30D as
the reagent gas, differentiation of the primary, secondary
and tertiary amines was possible. Further applications of
isotope exchange to locate and count protons in different
chemical environments and to provide information similar

to that available from NMR chemical shift and integration

 

128

measurements was presented (106). Observation of exchange
reactions with D20 and protonated benzenes demonstrated
the first reported exchange of aromatic hydrogens (107).
Sequential exchanges were observed with a rate of exchange
dependent on the sample structure. The isomer
m-difluorobenzene incorporated only one deuterium while
the o- and p-difluorobenzene exchanged all ring hydrogens.
Similar differences exist with the xylenes. The site of
protonation on a substituted benzene can also be
determined by the amount of exchange (108). Extensive
exchange indicates protonation on the ring. Lack of
exchange coupled with the presence of hydrated parent ions

indicate protonation on the substituent.

Studies have been conducted with flowing afterglow
and selected ion flow tube (SIFT) techniques where
hydrogen/deuterium exchange reactions of carbanions are
observed and shown to be a useful tool for probing gas
phase ion structure (109-111). The extent of exchange was
shown to depend on the pressure of the deuterating agent
and the reaction time. In summary, regarding
hydrogen/deuterium exchange, the exchange for positive
ions depends on the difference in basicities for the
deuterating agent and the conjugate base of the protonated
ion undergoing exchange. No exchange occurs if the
difference in proton affinities of the bases is greater
than 20—25 kcal/mole (105). With negative ions the

exchange depends on the difference in acidities for the

129

deuterating agent and the conjugate acid undergoing
exchange; no exchange is observed if the difference in

acidities is greater than 20 keel/mole (111).

Benefits for isomer differentiation similar to those
from chemical ionization in the source can be realized by
employing ion-molecule reactions in the quadrupole
collision cell of a TQMS instrument. In the collision
cell, the collision gas species, its pressure and the
collision energy may be readily varied. The collision
gases used in the second quadrupole and the collisionally

activated reaction for each are presented in Table 5.1.

The collision of the selected molecular ion with a
target gas (N2) at low collision energy results in a
vibrationally excited molecular ion (2,48) which then
undergoes subsequent dissociation. The spectrum obtained
at these low collision energies is qualitatively similar
to that obtained with high energy collisional activation
(CA) in a sector MS/MS instrument. Both techniques are of
limited use for differentiating positional isomers. With
the isomeric monohydroxybiphenyls both the high energy CA
spectrum (112) and the low energy CAD spectrum (113)
demonstrate spectral differences, but no masses were

unique for a particular isomer.

The use of NH3 as a collision gas can lead to

dissociations of the selected ion (NHa functioning simply

130

Table 5.1 Collision activated reactions for different
collision gases.

 

 

Collision Gas Collision Activated Reaction
N2 M+ + N2 ---> Fragments
NHa MX+ + NH3 -—-> (MNH2)H+ + X
D20 (M — H)‘ + D20 --—>

(M - H - HY+ DY)‘

SF6 (M - H)- + SF6 --->

(M - H)+* + SF6 + 2e’

131

as a target) and also to electrophilic substitution

reactions at low collision energies.

Deuterium exchange of the active hydrogens of a
selected ion upon collision with D20 in the quadrupole
collision chamber provides insight into the selected ion’s
structure. The exchange products from the reaction of
selected (M - H)‘ ions of carboxylic acids with deuterium
exchange reagent gas have recently been reported (114).
Differences in the extent of exchange are observed between

some selected isomeric carboxylic acids.

The collision of selected (M — H)‘ ions with a heavy
target, SF6, at high collision energy (180 eV lab) may
result in a charge inversion reaction in which two
electrons are stripped from the negative parent ion to
produce a positive parent ion. These inverted ions are
generally formed with a greater proportion of higher-
energy states than positive molecular ions formed by
conventional ionization and usually undergo extensive
fragmentation characteristic of the positive ion (71). The
charge inversion process studied in sector MS/MS
instruments has been shown to be quite sensitive to fine
details of molecular structure (115). Different charge
inversion spectra were observed for isomeric
dihydroxybenzoic acids (115) and isomeric polycyclic
aromatic hydrocarbons (116). The first observation of

charge inversion in TQMS was reported recently (72).

 

132

Charge inversion may routinely be performed in a
quadrupole collision cell provided sufficient lab
collision energy is available to give similar center-of-
mass collision energies between sector MS/MS where the
selected ion is incident on a light target (helium) and
quadrupole MS/MS where the selected ion is incident on a

heavy target (SF6).

Other CAR reactions in a quadrupole collision cell
which have been reported include: the reaction of the Cal-15+
fragment ion, generated by collision of protonated
substituted benzenes with a non-reactive target gas, with
a small amount of reactive gas (H20) present in the target
gas to provide characteristic adduct ions (117), and the
reaction of selected ions with an acetylene collision gas

to give several interesting addition products (118).

The isomers selected for this study are shown in
Figure 5.1. The compounds included the three
dichlorobenzenes (Ia - c), the three monohydroxybiphenyls
(IIa - c), the three monobromobiphenyls (IIIa - c),
2,2’-dibromobiphenyl (IVa) and 4,4’-dibromobiphenyl (IVb).
These isomers were chosen to assist in developing
methodology for distinguishing between isomers of the more
highly substituted polybromobiphenyls and their

metabolites.

133

CL CL

II HWN
III" BM“
NH

Figure 5.1 Isomers selected for this study.

134

Experimental

All experiments were performed on the TQMS instrument
developed in our laboratory and described elsewhere
(34,39). The sample compounds were of the highest purity

available and were used without further purification.

For both N2 and NH3 collision gas studies the
molecular ions were generated by 70 eV E1. The collision
gas pressure was high enough to permit some multiple
collisions to occur (3 x 10-4 torr). The collision energy,
the difference in potential between the ion source and the
second quadrupole DC rod offset, was maintained at zero.
The distribution of energies (less than 3 eV) of source-
generated ions allow those ions with greater than the zero
median energy to pass through the collision chamber. The
level of RF voltage supplied to the RF-only quadrupole
collision chamber was held at a constant value for each
daughter scan. Moderate drawout potentials (ca. -20 volts)
were applied to the first element of the interquadrupole
lens set following the second quadrupole and maintained
through the third quadrupole without seriously degrading

resolution.

The negative molecular ion species for the deuterium
exchange and charge inversion studies was generated with a
reagent gas mixture of 1:0.2 CH4/N20, maintained at 1.2

torr as determined by a source pirani tube. The D20

135

collision gas pressure, as measured by an ion gauge tube
calibrated for H20, was high enough for some multiple
collisions to occur (2.5 x 10“ - 5 x 10" torr). The
collision energy for the exchange reaction was typically

3 eV and drawout potentials of 10 volts and 15 volts were
applied to the first interquadrupole lens element and the
third quadrupole rod offset, respectively. The SF6
collision gas pressure was 2.5 x 10" torr as measured by
an ion gauge tube calibrated for SF6. The second
quadrupole rod offset was 178 volts more positive than the
source, producing a lab collision energy of 178 eV. Only a
10 volt drawout potential on the third quadrupole rod
offset was applied. Again for all daughter scans the RF

voltage on the collision chamber was held constant.

Results and Discussion

The utility of collisionally activated reactions for
the differentiation of positional isomers is demonstrated
with the following examples: CAD (N2) of the
dibromobiphenyls (IVa and IVb); CAR (NHa) of the
dibromobiphenyls (IVa and IVb), the monobromobiphenyls
(IIIa - IIIc), and the dichlorobenzenes (Ia - Ic); CAR
(D20) of the dichlorobenzenes (Ia- Ic), the
monohydroxybiphenyls (IIa -IIc), and the dibromobiphenyls
(IVa and IVb); and CAR (SF6) of the monohydroxybiphenyls
(IIa ~IIc).

 

136

Reaction with N2

The products resulting from the collision of
compounds IVa and IVb with N2 are shown if Figure 5.2. The
selected parent ion at m/z 312 contains both 79Br and 81Br,
although an ion containing the same isotopes could have
also been chosen. The intensities of the CAD fragment ions
at m/z 231 and 233, loss of bromine, and m/z 152, loss of
both bromines, are characteristic for each isomer. The (M
- Br)+ ion is 60 times larger for the 2,2’-dibromobiphenyl
(IVa) isomer relative to the 4,4’-dibromobiphenyl (IVb)
isomer. The (M - 2Br)+ ion is 14 times larger for IVb
relative to IVa. These spectral differences would permit

the two isomers to be differentiated.

Reaction with NHa

The electrophilic substitution products resulting
from the reaction of NH3 with the dibromobiphenyls (IVa
and IVb) are evident in Figure 5.3. The CAR product ions
correspond to (M — Br + NH2)H+, which is slightly larger
for IVa, and (M - 2Br + NH2)H+, which is five times larger
for IVb. The differences in the CAR spectra would allow
these isomers, occurring individually, to be readily

distinguished.

The CAR (NHa) spectra of the monobromobiphenyls (IIIa

- IIIc) are presented in Figure 5.4. The collision gas

137

No 2,2'-DIBROMOBIPHENYL

Log Rel. Abun.
I

 

 

 

 

150' '1'9'0' '230' '270' '310
4.4'-DIBROMOBIPHENYL

Log Rel. Abun.
I

 

 

d h_
v v 1 1

150' mléd ...2.30.....2.}O.....3id
Moss (amu)

 

Figure 5.2 CAD (N2) spectra of IVa and IVb m/z 312 M+
ion (C12H879Br81Br). 93 V RF second quadrupole.

 

138

2,2'-DIBROMOBIPHENYL

Log Rel. Abun.
. El
[\3 c: Q

L A

 

 

 

 

 

'— TTTII'IIITII'I' t'v Ili'v'l'lll'll

150 190 230 270 310
4.4'-DIBROMOBIPHENYL

Log Rel. Abun.
I

 

 

7| "l

- I‘llvlevjv'v'vrviw

150 190 230"'2'?b'7"'31'd
Moss (amu)

 

Figure 5.3 CAR (NHs) spectra of IVa and IVb m/z 312 M+
ion. 93 V RF second quadrupole.

Figure 5.4

139

mu 2-BROMOBIPHENYL

o

5

2

53—2

.8.“
-4,,,.|

 

15 "160' ' ' ' éid’fiflo

Illb 3-BROMOBIPHENYL

O

-2

Log Rel. Abun.

 

150 ' ' 17150. ' ' ' élo" 24o
4—BROMOBIPHENYL

 

file
0
§
3 —2
§
_ , . , .J, ,,,,,,,,,,,, ,
150 180 210 240
Mass (amu)

CAR (NHa) spectra of IIIa - IIIc m/z 234 M+
ion (012H981Br). 93 V RF second quadrupole.

 

140

pressure was measured at 3.8 x 10-4 torr. All three
isomers lose equivalent amounts of bromine to give m/z
153, (Cle9)+. The substitution reaction product at m/z
170, (Clzl-th)+ for IIIb is three times as large as that
for 111a and six times as large as that for IIIc. Since
the CAR spectra are characteristic for each isomer, a
mixture of these isomers could be analyzed provided the
proper application of the superposition principle is

followed (119).

The products of the collision of M+ (CsH435Clz) from
the dichlorobenzenes (Ia— Ic) with NHa are shown in Figure
5.5. The CAD fragments for loss of Cl and C12 are at m/z
111 and 75, respectively. The CAR products, (CsH’rClN)+ and
(CGH7N)+, are evident at m/z 128 and 93. The (M ~ 201)+ CAD
fragment was detected at 8 x 10'5 relative abundance for
Ib. There are many marked differences among the isomers as
exemplified by the loss of both chlorines in the ratio of
12:1:150 for the ortho, meta and para isomers
respectively. More substitution product (M - Cl + NH3)* is
formed from the meta isomer as is the case for the 3-
bromobiphenyl isomer in Figure 5.4. The intensity of the
same substitution product from the ortho and para isomers
show a similar correlation with the intensity of the
substitution product for the other bromobiphenyl isomers
(IIIa and IIIb). From the observed differences in the CAR
spectra for the dichlorobenzenes it would be possible to

analyze a mixture of the three isomers.

141

I0 o-DICHLOROBENZENE

0
g 6.5%
< I.5°/.
e —2 .
0‘ o.l°/. 02"
8‘
._l
_4 IIIIIIIII

 

 

 

70 90 llOl'lIO'filgO
1b m—DICHLOROBENZENE

I0°/°
_2 0.8%
o.I°/.
_4....,J, ,1.
70 90 110 130 I50
p-DICHLOROBENZENE

Log Rel. Abun.

 

Ic
d ISO/o
B
4%

<r |.2°/. 2
B -2 0.4%
D:

0"

O

_I

-4

 

 

7'0" 9b T lIO ' 130 ' 150
Mass (amu)

Figure 5.5 CAR (NHa) spectra of Ia, Ib and Ic m/z 146 M+
ion (CeH435C12). 67 V RF second quadrupole.

142

Reaction with D20

The amount of hydrogen/deuterium exchange for the
dichlorobenzenes is evident by observing the mass region
centered around the selected (M - H)‘ ion (m/z 145) as
shown in Figure 5.6. The D20 collision gas pressure was
5 x 10“4 torr, high enough to permit multiple collisions.
The collision energy for all CAR (D20) spectra displayed
was 3 eV. The ortho (Ia) and para (10) readily undergo
hydrogen/deuterium exchange, under these experimental
conditions, to give as the base peaks three deuterium and
one deuterium incorporation, respectively. The parent ions
are less than 20% relative abundance. For the meta isomer
(Ib) exchange is not as favored since the (M - H)‘
selected ion is the base peak. However, a preferential
exchange resulting in the incorporation of two deuteriums
at 70% relative abundance is observed. The unique product
distribution for these isomers would permit their analysis

in a mixture.

The CAR (D20) spectra of the hydroxybiphenyls (IIa -
IIc) are shown in Figure 5.7. The collision gas pressure
was 2.5 x 10" torr. IIb and 110 cannot be distinguished
by this method since they behave similarly: the single
hydrogen/deuterium exchange product at 12% relative
abundance. For IIa the single exchange product is 30 times
less and a small amount of a dual exchange product is

observed.

143

10 o-DICHLOROBENZENE
l

Rel. Abun.

0, ll . . .
140 150
lb m-DICHLOROBENZENE

 

 

E

$3

0. .ll . .
140 150
10 p—DICHLOROBENZENE
l

E.

312'
0' II n A 1
140 150

Mass (amu)
Figure 5.6 CAR (D20) hydrogen/deuterium exchange

spectra of Ia, Ib and 1c m/z 145 (M-H)‘
67 V RF second quadrupole.

 

144

no 2—HYDROXYBIPHENYL
o

—2

Log Rel Abun.

 

-4, .D .
16 175

Kb 3-HYDROXYHPHENYL

O

-2

Log Rel Abun.

 

—4l 1 I
165 175

4-HYDROXYEPHENYL

Log Rel Abun.
.1,

 

165 ' 1'75
Moss (onnu)

Figure 5.7 CAR (D20) hydrogen/deuterium exchange

spectra of IIa - IIc m/z 169 (M-H)’ ion. 67
V RF second quadrupole.

 

145

The amount of hydrogen/deuterium exchange for the
dibromobiphenyls (IVa and IVb) can be extensive and
characteristic. As seen in Figure 5.8, a maximum of seven
protons of the selected parent ion (m/z 311) are exchanged
for both isomers. For the IVa isomer the favored product
contains only one deuterium at a comparable intensity to
the selected parent. The IVb isomer reveals a surprising
incorporation of seven deuteriums as the base peak, twice
as large in intensity as the selected parent ion. These
isomers could be readily differentiated. An investigation
of the pressure dependence of the exchange products for
IVb shown in Figure 5.9 demonstrated the effect of
multiple collisions at higher D20 pressures. The parent
ion (m/z 311) is rapidly consumed with increasing D20
pressure. Exchange of one deuterium for hydrogen is
dominant under single collision conditions at low
collision gas pressures. Additional exchange occurs as the
pressure is raised (multiple collision conditions) until
the product ion containing the maximum number of
exchangeable protons is exclusively formed. The maxima for
formation of product ions with greater than one and less
than seven deuteriums were at values intermediate to those
for m/z 312 and 318. The sensitive dependence of the
hydrogen/deuterium exchange on the lab collision energy is
shown in Figure 5.10. Exchange of all hydrogens occurs
only at low collision energies and occurs only over a

narrow energy range. Exchange of a single hydrogen is

146

No 2,2 -D|BROMOBIPHENYL

Rel. Abun.

 

 

 

[I 11111111

0 .
305 31 5

 

Nb 4.4'-DIBROMOBIPHENYL

M 111

0 r
305 315
Mass (amu)

Figure 5.8 CAR (D20) hydrogen/deuterium exchange
spectra of IVa and IVb m/z 311 (M-H)‘ ion.
160 V RF second quadrupole.

Rel. Abun.

 

 

 

 

 

147

 

IOO f l l l l I l l l l I
- 0311' —>3|8‘
0311‘ -» 311'
’ (lair->312

 

 

 

 

:50
(I)
C
(D
E
O

Ix 10‘4 5x 10‘4 Ix 10'3

DZO, Pressure ,torr

Figure 5.9 Pressure dependence of CAR (D20) reaction
products for IVb. 3 eV lab collision energy.

148

T

100

3ll’—->3|8‘ 3||"—>3|2'

50

 

Intensity

 

l

0
-lO 0 IO 20
Lab Collision Energy

Figure 5.10 Collision energy dependence of CAR (D20)
reaction products for IVb. 2.5 x 10'4 torr
D20.

149

preferred at higher collision energies and has a broader
energy dependence. A longer lifetime for the selected ion-
D20 collision complex coupled with additional collisions
of the selected ion with D20 molecules lead to increased
hydrogen/deuterium exchanges at lower collision energies.
The differences in CAR (D20) products for IVa and IVb may
reflect a difference in acidities of the (M - H)‘ anions,

with IVb more acidic.

Reaction with SF6

Charge inversion products formed from the energetic
(178 eV lab) collision of the (M - H)" ions of the
hydroxybiphenyls (IIa - IIc) with a heavy target, SF6, are
displayed in Figure 5.11. The collision energy in center-
of-mass coordinates is 82 eV, comparable to collision
energies in sector MS/MS instruments. The inverted ion
(M - H)+ is the base peak in all three spectra. Fragments
arising from a dissociative ionization of SF6 are evident
at m/z 51 (SF+), 89 (SF3+) and 127 (SF5+). The process for
formation of positively charged SF6 fragments is not
clear. Upon the high energy collision, the SF6 may form F’
and SF5+, F', F2 and SFa+ and F', 2F2 and SF+. However,
the formation of negatively charged ions from SF6 was not
investigated. Fragments of the excited inverted ion are
present at m/z 39 (CaH3+), 5O (C4H2+), and 63 (C5H3+) and

there are several Ce-containing fragments. Although

150

110 2-HYDROXYBIPHENYL

 

 

 

 

 

 

1-
:3
E
30 60 90 120 150 180
IIIb 3-HYDROXYBIPHENYL
1l
5
E
0-
30 60 90 120 150 180
EC 4-HYDROXYBIPHENYL
1l
.4?
Ill
30 60 90 120 150 180

Mass (amu)

Figure 5.11 CAR (SF6) charge inversion spectra of IIa -
IIc m/z 169 (M-H)’ ion. 2.5 x 10'4 torr SF6.

67 V RF second quadrupole.

 

151

differences exist for the inversion spectra of the
different isomers, further optimization of the process may
be made resulting in the development of charge inversion
in a quadrupole collision chamber as a useful tool for

differentiating isomers.

cl ' n

Isomer analysis can be effected by use of
collisionally activated reactions in the collision chamber
of a TQMS instrument. The utility that CAR provide for
dramatically enhancing spectral differences between
positional isomers has been demonstrated. The large
dynamic range, the selectivity and specificity, the
efficiency of the collision process, and the freedom from
chemical noise are characteristics of TQMS which make it

well suited for these studies.

In select cases collision with nitrogen may provide
sufficient information for distinguishing isomers. With an
ammonia collision gas, greater substitution of ammonia for
a halogen occurs in compounds with meta substituents. Also
the substitution reaction is that of M+ with NHa. The
reaction of NH3+ with M would be unlikely in a CI source.
Hydrogen/deuterium exchange in a quadrupole collision
chamber may serve as a sensitive probe of ion structure
and in addition provide information on the basicity of gas

phase ions. The sensitivity of charge inversion spectra to

152

molecular structure, demonstrated with high energy sector
MS/MS instruments, is also demonstrated with charge

inversion performed in a quadrupole collision chamber.

LIST OF REFERENCES

10.

11.

12.

13.
14.

LIST OF REFERENCES

J. Durup in "Recent Developments in Mass Spectrometry,"
K. Ogata, T. Hayakawa, Ed., University Park Press,
Baltimore, MD, 1969, p. 921.

R.A. Yost, C.G. Enke, D.C. McGilvery, D. Smith,

J.D Morrison, Int. J. Mass Spectrom. Ion Phys., 30,
127 (1979).

J.A. Hipple, E.U. Condon, Phys. Rev., 68, 54 (1945).

J.L. Holmes, J.K. Terlouw, Org. Mass Spectrom., 15,
383 (1980).

K.R. Jennings, Int. J. Mass Spectrom- Ion Physics, 1,
227 (1968).

F.W. McLafferty, R. Koornfeld, W.F. Hadden, K. Levsen,
I. Suksi, P.F. Bente, S-C. Tsai, I. Howe, J. Am. Chem.
Soc., 95, 2120 (1975).

C.E. Melton, G.F. Wells, J. Chem. Phys., 21, 1132
(1957).

F.M. Rourke, J.C. Sheffield, W.D. Davis, F.A. White, J.
Chem. Phys., 31, 193 (1959).

J.H. Beynon, R.G. Cooks, T. Keough, Int. J. Mass
Spectrom. Ion Phys., 13, 437 (1974).

T. Wachs, C.C. Van de Sande, P.F. Bente, P.P. Dymerski,
F.W. McLafferty, Int. J. Mass Spectrom. Ion Physics,
Zfi. 21 (1977).

J.H. Beynon, R.G. Cooks, Int. J. Mass Spectrom. Ion
Phys., 19, 107 (1976).

T. Wachs, C.C. Van de Sande, F.W. McLafferty, Org. Mass
Spectrom., 11, 1308 (1976).

R.W. Kondrat, R.G. Cooks, Anal. Chem., 50, 81A (1978).

R.A. Yost, C.G. Enke, J. Am. Chem. Soc., 100, 2274
(1978).

153

 

15.

16.

17.

18.

19.
20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

154

D. Zakett, R.G. Cooks, W.J. Fies, Anal. Chem. Acta,
113, 129 (1980).

G.L. Glish, S.A. McLuckey, T.Y. Ridley, R.G. Cooks,
Int. J. Mass Spectrom. Ion Phys., 11, 157 (1982).

R.B. Cody, B.S. Freiser, Int. J. Mass Spectrom. Ion
Phys., 11, 199 (1982).

K. Blom, B. Munson; Presented at the 30th Annual
Conference on Mass Spectrometry and Allied Topics;
Honolulu, Hawaii, June 6-11, 1982.

A.J. Dempster, Philos. Mag., 31, 438 (1916).

J.L. Franklin, Ed., "Ion-Molecule Reactions, Part I,”
Dowden, Hutchinson and Ross, Stoudsburg, PA, 1979,
p. 2.

J.L. Franklin, Ed., ”Ion-Molecule Reactions, Parts I
and II," Dowden, Hutchinson and Ross, Stroudsburg, PA,
1979.

M.T. Bowers, Ed., "Gas Phase Ion Chemistry, Volumes 1
and 2,” Academic Press, New York, 1979.

V.L. Tal’roze, A.K. Lyubimova, Kokl. Akad. Nauk SSSR,
33, 909 (1952).

D.P. Stevenson, D.O. Schissler, J. Chem. Phys., 33,
1353 (1955).

F.H. Field, J.L. Franklin, F.W. Lampe, J. Am. Chem.
Soc., 13, 5697 (1956).

M.S.B. Munson, F.H. Field, J. Am. Chem. Soc., 33, 2621
(1966).

M.T. Bowers, Ed., ”Gas Phase Ion Chemistry, Volume 1,”
Academic Press, New York, 1979, Chapter 8.

H. Yamaoka, P. Dong, J. Durup, J. Chem. Phys., 31, 3465
(1969).

T. Wachs, F.W. McLafferty, Int. J. Mass Spectrom. Ion
Phys., 23, 243 (1977).

M.S. Kim, F.W. McLafferty, J. Am. Chem. Soc., 133, 3279
(1978).

C.J. Proctor, B. Kralj, A.G. Brenton, J.H. Beynon, Org.
Mass Spectrom., 13, 619 (1980).

32.
33.
34.
35.

36.
37.

38.

39.

40.

41.

42.
43.

44.

45.

46.
47.

48.

155

P.A.M. van Koppen, A.J. Illies, S. Liu, M.T. Bowers,
Org. Mass Spectrom., 11, 399 (1982).

A. Russek, Physica, 15, 165 (1970).
R.A. Yost, C.G. Enke, Anal. Chem., 31, 1251A (1979).

R.A. Yost, C.G. Enke, Org. Mass Spectrom., 13, 171
(1981).

R.A. Yost, C.G. Enke, Am. Lab., June, 1981, p. 88.

C.G. Enke, J.A. Chakel, E.J. Darland; C.G. Enke, R.A.
Yost, R.K. Latven; C.G. Enke, R.A. Yost; Papers
Presented at te 27th Annual Conference on Mass

Spectrometry and Allied Topics; Seatle, Washington,
June 3-8, 1979.

C.G. Enke, J.A. Chakel, J. Allison, H.R. Gregg; C.G.
Enke, R.K. Latven, B.H. Newcome; Papers Presented at

the 28th Annual Conference on Mass Spectrometry and
Allied Topics; New York, New York, May 25-30, 1980.

J.A. Chakel, C.A. Myerholtz, C.G. Enke; H.R. Gregg,
J.W. Chai, J.A. Chakel, P.A. Hoffman, R.K. Latven, B.S.
Matthews, C.A. Myerholtz, B.H. Newcome, C.G. Enke; R.K.
Latven, C.G. Enke; Papers Presented at the 29th Annual
Conference on Mass Spectrometry and Allied Topics;
Minneapolis, Minnesota, May 24-29, 1981.

R.K. Latven, Ph.D. Diss., Michigan State Univ., East
Lansing, MI (1981).

J.W. Chai, Ph.D. Diss., Michigan State Univ., East
Lansing, MI (1982).

Extranuclear Laboratories, Inc., Pittsburgh, PA.
D.F. Hunt, F.W. Crow, Anal. Chem., 33, 1781 (1978).

B.H. Newcome, C.G. Enke, Rev. Sci. Instrum., 33(12),
2017 (1984).

C.A. Myerholtz, M.S. Thesis, Michigan State Univ., East
Lansing, MI (1982).

D.F. Hunt, J. Shabanowitz, Anal. Chem., 35, 574 (1982).

D.Zakett, P.H. Hemberger, R.G. Cooks, Anal. Chem. Acta,
113, 149 (1980).

D.J. Douglas, J. Phys. Chem., 33, 185 (1982).

 

49.

50.

51.

52.

53.
54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

156

P.H. Dawson; Presented at the 30th Annual Conference on
Mass Spectrometry and Allied Topics; Honolulu, Hawaii,
June 6-11, 1982.

G.L. Glish, P.H. Hemberger, R.G. Cooks, Anal. Chem.
Acta, 119, 137 (1980).

R.W. Abernathy, F.W. Lampe, J. Am. Chem. Soc., 03,
2573 (1981).

G.L. Glish, Ph.D. Diss., Purdue Univ., West Lafayette,
IN (1980).

W.T. Huntress, J. Chem. Phys., 33, 5111 (1972).

F.P. Abramson, J.H. Futrell, J. Chem. Phys., 13, 1925
(1966).

'D.F. Hunt, J. Shabanowitz, A.B. Giordani, Anal. Chem.,

£2, 386 (1980).

P.H. Dawson, J.E. Fulford, Int. J. Mass Spectrom Ion
Phys., 12, 195 (1982).

P.H. Dawson, J.B. French, J.A. Buckley, D.J. Douglas,
D. Simmons, Org. Mass Spectrom., 11, 205 (1982).

P.H. Dawson, J.B. French, J.A. Buckley, D.J. Douglas,
D. Simmons., Org. Mass Spectrom., 11, 212 (1982).

J.A. Laramee, D. Cameron, R.G. Cooks, J. Am. Chem.
Soc., 133, 12 (1981).

J.O. Hirschfelder, C.F. Curtiss, R.B. Bird, "Molecular
Theory of Gases and Liquids," John Wiley and Sons, New
York, 1954, p. 950.

G.A. Cook, Ed., "Argon, Helium and the Rare Gases,"
Interscience, New York, 1961, p. 151.

W.C. Gardiner, Jr., "Rates and Mechanisms of Chemical
Reactions, " W.A. Benjamin, Inc., Menlo Park, CA, 1972,
p. 75.

J. Allison, Ph.D. Diss., Univ. of Delaware, Newark, DE
(1977).

F.H. Field, J.L. Franklin, "Electron Impact Phenomena,”
Academic Press, New York, 1957, Chapter 4.

E. Telay, D. Gerlich, Chem. Phys., 4, 417 (1974).

 

hm

 

66.

67.

68.

69.

70.

71.

72.

73.
74.

75.

76.

77.

78.

79.
80.

81.

82.
83.

157

H.M. Rosenstock, K. Draxl, B.W. Steiner, J.T. Herron,
”Energetics of Gaseous Ions," J. Phys. Chem. Ref. Data,
3, Supplement 1 (1977).

K.N. Hartman, S. Lias, P. Ausloos, H.M. Rosenstock,
S.S. Schroyer, C. Schmidt, D. Martinsen, G.W.A. Milne,
"A Compendium of Gas Phase Basicity and Proton Affinity
Measurements," National Bureau of Standards, Washinton,
DC, 1979, NBSIR 79-1777.

T. Baer, B.P. Tsai, D. Smith, P.T. Murray, J. Chem.
Phys., 35, 2460 (1976).

H.M. Rosenstock, R. Stockbauer, A.C. Parr, J. Chem.
Phys., 13, 773 (1980).

J. Appel in ”Collision Spectoscopy," R.G. Cooks, Ed.,
Plenum Press, New York, 1978, p. 252.

J.E. Szulejko, J.H. Bowie, I. Howe, J.H. Beynon, Int.
J. Mass Spectrom. Ion Phys., 33, 99 (1980).

D.J. Douglas, B. Shushan, Org. Mass Spectrom., 11, 198
(1982).

R.A. Hites, Adv. Chromatogr., 13, 69 (1977).

Q.V. Thomas, J.R. Stork, S.L. Lammert, J. Chromatogr.
Sci., 13, 583 (1980)-

W.C. Schnute, D.E. Smith, Am. Lab., Dec., 1980, p. 87.
R.G. Cooks, J.H. Beynon, R.M. Caprioli, G.R. Lester,
"Metastable Ions," Elsevier, Amsterdam, 1973,

Chapter 6.

K. Levsen, H.K. Wipf, F.W. McLafferty, Org. Mass
Spectrom., 3, 117 (1974).

D.H. Smith, C. Djerassi, K.H. Mauer, U. Rapp, J. Am.
Chem. Soc., 33, 3482 (1974).

E.J. Gallegos, Anal. Chem., 13, 1348 (1976).

T.L. Kruger, J.E. Litton, R.W. Kondrat, R.G. Cooks,
Anal. Chem., 13, 2113 (1976).

F.W. McLafferty, F.M. Bockhoff, Anal. Chem., 33, 69
(1978).

R.G. Cooks, Amer. Lab., Oct., 1978, p. 111.

J.H. McReynolds, M. Anbar, Int. J. Mass Spectrom. Ion
Phys., ZA, 37 (1977)-

84.

85.
86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

158

K. Levsen, H.D. Becky, Org. Mass Spectrom., 3, 570
(1974).

G.A. McClusky, G.A. Kondrat, R.G. Cooks, J. Am. Chem.
Soc., 133, 6045 (1978).

R.W. Kondrat, G.A. McClusky, R.G. Cooks, Anal. Chem.,
33, 1222 (1978).

R.W. Kondrat, G.A. McClusky, R.G. Cooks, Anal. Chem.,
33, 2017 (1978).

K. Levsen, H.R. Schulten, Biomed. Mass Spectrom., 3,
137 (1976).

C.V. Bradley, I. Howe, J.H. Beynon, J. Chem. Soc.,
Chem. Commun., 12, 562 (1980).

C.V. Bradley, I. Howe, J.H. Beynon, Biomed. Mass
Spectrom., 3, 85 (1981).

T.L. Kruger, R.G. Cooks, J.L. McLaughlin, R.L. Ranieri,
J. Org. Chem., 12, 4161 (1977).

G.L. Glish, R.G. Cooks, Anal. Chem. Acta, 113, 145
(1980).

F.W. McLafferty, P.J. Todd, D.C. McGilvery, M.A.
Baldwin, F.M. Bockhoff, G.J. Wendel, M.R. Wixom, T.E.
Niemi, "Adv. Mass Spectrom.," 8B, 1980, p. 1589.

J.R.B. Slayback, M.S. Story, Ind. Res. Dev., 23, 128
(1981).

T. Sakuma, W.R. Davidson, B.A. Thomson, L.M.
Danylewych, B. Shushan, J.E. Fulford, N.M. Reid, G.A.V.
Rees, H. Tosine, 32nd Pittsburgh Conf., Atlantic City,
NJ (1981) p. 134.

T.R. Henderson, R.E. Reyer, C.R. Clark; presented at
the 29th Annual Conference on Mass Spectrometry and
Allied Topics; Minneapolis, Minnesota, May 24-29, 1981.

C.A. Myerhholtz, C.G. Enke; Presented at the 30th
Annual Conference on Mass Spectrometry and Allied
Topics; Honolulu, Hawaii, June 6-11, 1982.

A.L.C. Smit, F.H. Field, J. Am. Chem. Soc., 33, 6471
(1977).

C.C. Van deSande, F. Van Gaeuer, P. Sandra, J.
Monstrey, Z. Naturforsch, 323, 573 (1977).

100. B. Munson, Anal. Chem., 33, 772A (1977).

I.

 

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.
112.

113.
114.

115.

116.

117.

118.

159

D.P. Martinsen, S.E. Buttrill, Jr., J. Am. Chem. Soc.,
193. 6559 (1978).

W.C.M.M. Luijten, W. Okenhout, J. Van Thuijl, Org. Mass
Spectrom., 13, 329 (1980).

W.J. Van der Hart, W.C.M.M. Luijten, J. Van Thuijl,
Org. Mass Spectrom., 13, 463 (1980).

D.F. Hunt, C.N. McEwen, R.A. Upham, Anal. Chem.,
1292 (1972).

4.

D.F. Hunt, S.K. Sethi, J. Am. Chem. Soc., 132, 6953
(1980).

D.F. Hunt, S.K. Sethi; Presented at the 27th Annual
Conference on Mass Spectrometry and Allied Topics;
Seattle, Washington, June 3-8, 1979.

B.S. Freiser, R.L. Woodin, J.L. Beauchamp, J. Am. Chem.
Soc., 31, 6893 (1975).

D.P. Martinsen, S.E. Buttrill, Jr., Org. Mass
Spectrom., 11, 762 (1976).

J.H. Stewart, R.H. Shapiro, C.H. DePuy, V.M. Bierbaum,
J. Am. Chem. Soc., 33, 7650 (1977).

C.H. DuPuy, V.M. Bierbaum, G.K.King, R.H.Shapiro, J.
Amer. Chem. Soc., 133, 2921 (1978).

C.H. DuPuy, V.M. Bierbaum, Chem. Res., 15, 146 (1981).

R. Rabbiani, T. Kuster, J. Seibi, Agnew. Chem., 33, 115
(1977).

Unpublished results.

A.B. Giordani, Ph.D. Diss., Univ. of Virginia,
Charlottesville, VA (1982).

G.A. McClusky, R.W. Kondrat, R.G. Cooks, J. Am. Chem.
Soc., 199. 6045 (1978).

D. Zakett, J.D. Ciupek, R.G. Cooks, Anal. Chem., 33,
723 (1981).

D.I. Carroll, J.G. Nowlin, R.N. Stillwell, E.C.
Horning; Presented at the 30th Annual Conference on
Mass Spectrometry and Allied Topics; Honolulu, Hawaii,
June 6—11, 1982.

R.A. Yost, personal communication.

160

119. L.M. Bass, M.T. Bowers, Org. Mass Spectrom., 11, 229
(1982).

 

 

 

  

llllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll