11:: .1:
.s)...: V
r.)

29...:
4....

v... I
.v) A
.‘I.\, ,

 

uiiiiiigiliiWn

1" on I} ‘I
i. I 3‘ r

This is to certify that the

dissertation entitled

UNIT RESOLUTION AND PHOTO-FRAGMENTATION
IN A TANDEM TIME-OF-FLIGHT
MASS SPECTROMETER

presented by

Paul Raymond Vlasak

has been accepted towards fulﬁllment
of the requirements for

Ph . D . degree in 9! 12m

ﬂaw 54.4

Major professor

 

ﬂ}. 2&3 Hi:

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE ll RETURN BOX to roman this checkout 1mm your ncord.
TO AVOID FINES mum on or bdon dd. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l
ﬂ

usu loAn Afﬁrmative Action/Equal Opportunlty Institution
W

 

 

 

m1

 

 

 

 

 

 

UNIT RESOLUTION AND PHOTO-FRAGMENTATION IN A
TANDEM TIME-OF-FLIGHT MASS SPECTROMETER

By

Paul Raymond Vlasak

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1995

UMI NUmber: 9619918

 

UMI Microform 9619918
Copyright 1996, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized
copying under Title 17, United States Code.

 

UMI

300 North Zeeb Road
Ann Arbor, MI 48103

ABSTRACT

UNIT RESOLUTION AND PHOTO-FRAGMENTATION IN A
TANDEM TIME-OF-FLIGHT MASS SPECTROMETER

By

Paul Raymond Vlasak

The goal of this work is the development of a dual reﬂectron tandem
time-of-ﬂight (TOF) mass spectrometer (MS) capable of collecting complete
tandem mass spectrometry data sets on the chromatographic time scale. A
system which can meet this goal must use sample eﬁciently and be capable
of high mass spectral generation rates. In addition, the instrument must
maintain unit resolution over the mass range of interest in order to be
analytically useful.

The instrument described here uses an electron ionization source
which accumulates ions between source extractions. Precursor ion separation
occurs in a reﬂectron TOF analyzer, with a mass resolution of 1500 at the
reﬂectron focal point.

Precursor selection and fragmentation are accomplished by
intersecting a focused, high-energy laser pulse with a single, iso-mass ion
packet at the focal point of the ﬁrst mass analyzer. Unit mass resolution for

selection has been demonstrated to m/z 158, while calculations indicate that

the real limit of unit mass resolution is about m/z 1200. The high degree of
ion/photon overlap results in fragmentation efﬁciencies as high as 79% (for
the molecular ion of bromobenzene) and emciencies for a variety of ions
similar to those observed for collisionally-induced dissociation in commercial
tandem MS instruments. Product spectra are qualitatively similar to those
obtained using commercial instrumentation.

Prior to fragmentation, the precursor ion packet is isolated by an
interleaved comb ion deﬂection gate. Unit mass resolution for isolation has
been demonstrated to m/z 167. Calculations indicate that the real limit of
unit mass resolution is about m/z 300.

Upon dissociation, product ions retain a fraction of their kinetic energy
equal to the fraction of mass retained. Thus, a relative kinetic energy range
of nearly 100% results for the product ions. Using an acceleration stage to
reduce this relative kinetic energy range and a novel, broad energy range
focusing reﬂectron, a mass resolution of 300 for product ion spectra has been
demonstrated.

This tandem mass spectrometer is capable of generating product
spectra at the repetition rate of the laser while maintaining unit mass

resolution over a mass range appropriate for gas chromatographic analyses.

for Deena, Thomas, and Dad

iv

ACKNOWLEDGEMENTS

Many people contributed to this project, providing everything from
labor to inspiration. First and foremost, I thank and acknowledge my
advisor, Dr. Chris Enke. His guidance has been excellent, it has been
appropriate, and it has been appreciated. Dr. Enke has taught me that
conﬁdence not only means knowing that one can do what others have done,
but knowing that one can do what others have not done. I will strive to
emulate his enthusiasm and creativity in my own pursuits.

Working with other students on this project has been both enjoyable
and productive. I thank Dr. David McLane, Dr. Mary Seeterlin,
Doug Beussman, Tina Erickson, Qinchung Ji, Ben Gardner, and
Eric Hemenway, all of whom are esteemed friends and have my greatest
respect as scientists.

I have learned from many excellent teachers over the years and I am
grateful to all of them. I have been especially fortunate to have had
Ms. Miesner, Ms. Davis, Mr. Costigan, Mr. McDougall, Mr. Detviler,
Mr. Tressel, Mr. Young, Dr. Walters, Dr. Miessler, Dr. Hardgrove,

Dr. Hanson, Dr. Abdella, Dr. Failletaz, Dr. Allison, Dr. Crouch, Dr. Enke,

Dr. Leroi, and Dr. McGuﬁn. I am especially grateful to Dr. Crouch, who
served as my second reader, and the other members of my guidance
committee, Dr. Leroi, and Dr. Reusch.

In addition to the students and faculty who have contributed to this
project, I would like to acknowledge a number of other support persons.
Mike Davenport, Russ Geyer, Harold Hilbert, and Marty Rabb all
contributed to the design as well as the construction of the tandem TOF
instrument. In addition, I would like to thank Tom Clarke, Ron Haas,
Sam Jackson, Dick Menke, and Scott Sanderson.

Finally, I thank my family and friends for the support and
encouragement that have allowed me to pursue this goal. My parents and
grandparents taught me to be proud of myself and my work. My father
instilled in me scientiﬁc curiosity. My brothers grew, shared, and explored
with me. Deena, my patient and supportive wife, and Thomas, my child, are
everything.

I did not get here on my own, and I am grateful to all of you.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... xi
LIST OF FIGURES ........................................................................ xii
Chapter 1: Introduction .................. 1
1.1 Goal ......................................................................................................... 1
1.2 Tandem mass spectrometry ................................................................... 1
1.2.1 Ion structure data ............................................................................. 2
1.2.2 Mixture data ................. 3

1.3 MS for chromatographic detection ......................................................... 4
1.3.1 Overlapping chromatographic peaks ............................................... 4
1.3.2 Instrumentation ............................................................................... 5
1.3.2.1 Mass ﬁlters ......................................................................................... 5

1.3.2.2 Batch array detectors ......................................................................... 7

1.3.2.3 True array detectors .......................................................................... 9

1.3.3 Greater challenge of chromatography/MS/MS ............................ 10
1.3.4 Evolving chromatographic methods .............................................. 11

1.4 Overview of the tandem TOF instrument ........................................... 1 1
1.4.1 TOF provides high spectral generation rate and throughput ...... 15
1.4.2 PID provides high efficiency fragmentation .................................. 16'
1.4.3 Unit resolution for TOF/PID/TOF analysis ................................ 16

1.5 Historical background .......................................................................... 17
1.5.1 Instrumentation for high speed MS/MS ....................................... 1 7

1.5.2 Tandem TOF and hybrid instruments .......................................... 18

1.5.3 Photodissociation as a fragmentation technique for MS .............. 22

1.6 References ............................................................................................. 24
Chapter 2: Instrument design ................................................... 30
2.1 Chamber & vacuum system ................................................................. 31
2.2 Electronics ............................................................................................. 32
2.3 Sample introduction ............................................................................. 33
2.4 Source .................................................................................................... 34
2.5 First mass analyzer .............................................................................. 36
2.6 Instrument control and data collection system ................................... 39
2.7 Detectors ............................................................................................... 40
2.8 References ............................................................................................. 41
Chapter 3: Precursor selection and fragmentation ......... 42
3.1 Photo-induced dissociation as a fragmentation method ..................... 42
3.2 The laser, optics, and light tube .......................................................... 43
3.3 Ion/photon overlap ................................................................................ 45
3.3.1 Calculated overlap ......................................................................... 45
3.3.2 Observed overlap ............................................................................ 49

3.4 Photo-fragmentation efﬁciency ............................................................ 5 1
3.5 References ............................................................................................. 55

‘ Chapter 4: Precursor isolation .. ............................................ 56
4.1 Elimination of interference .................................................................. 56
4.2 Other applications of ion gates ............................................................ 59
4.2.1 Precursor selection .......................................................................... 59
4.2.2 Preventing detector saturation ....................................................... 59

4.3 Gating methods .................................................................................... 60

4.3.1 Deﬂection plates ............................................................................. 60

4.3.2 Interleaved comb ion deﬂection gate ............................................. 61
4.3.3 Comparison .................................................................................... 62
4.4 Instrumentation ................................................................................... 64
4.4.1 Interleaved comb ion deﬂection gate device .................................. 65
4.4.2 Electronics ...................................................................................... 69
4.4.2.1 Description ....................................................................................... 69
4.4.2.2 Pulse shapes ..................................................................................... 71

4.5 Performance .......................................................................................... 72
4.6 References ............................................................................................. 77
Chapter 5: The second mass analyzer ..................................... 79
5.1 Requirements ........................................................................................ 79
5.2 Reﬂectron TOF ..................................................................................... 80
5.2. 1 Background .................................................................................... 80
5.2.2 One-stage and two-stage reﬂectrons .............................................. 80
5.2.3 Grid-free reﬂectrons ....................................................................... 81
5.2.4 Perfectron ........................................................................................ 83
5.3 Reﬂectron TOF for product ion analysis ............................................. 83
5.3.1 Separation within the reﬂectron .................................................... 84
5.3.2 Separation by post-dissociation acceleration ................................ 85
5.4 Product ion analysis in the tandem TOF instrument ......................... 86
5.4.1 Post-dissociation acceleration ........................................................ 86
5.4.2 Elimination of interference due to MPI ......................................... 87
5.4.3 Grid-free reﬂectron for product ion analysis ................................. 89
5.4.4 Linear TOF mass analyzer ............................................................ 90
5.5 A broad kinetic energy range focusing reﬂectron ............................... 91
5.5.1 Calculation of a profile .................................................................. 94
5.5.2 SMON modeling to achieve desired profile .............................. 103

ix

5.5.3 Construction ................................................................................. 1 1 1

5.5.4 Performance .................................................................................. 112
5.5.5 Possible improvements ................................................................. 116
5.5.6 Signiﬁcance of this design method .............................................. 119
5.6 Calibration of product ion spectra ..................................................... 120
5.7 References ........................................................................................... 12 1
Chapter 6: TOF/TOF as an analytical tool ........................... 124
6.1 Unit resolution and eﬁcient photo-fragmentation ........................... 124
6.2 Analytical utility of PID product spectra .......................................... 126
6.3 Feasibility of GC/MS/MS .................................................................... 127
6.4 TOF/TOF in other analytical applications ........................................ 128
6.5 References ........................................................................................... 130

LIST OF TABLES

Table 3.1 PID efﬁciencies for a variety of precursor ions ........................... 54

Table 5.1 Reﬂectron conﬁguration by segment .......................................... 99

Figure 1.1

Figure 2.1

Figure 3.1

Figure 3.2

Figure 3.3

Figure 4.1

Figure 4.2

Figure 4.3

LIST OF FIGURES

Schematic diagram of the tandem TOF mass
spectrometer with instrument control and data

collection system ...................................................................... 12
The electron ionization storage source ................................... 34

Selection of precursor iso-mass ion packet by laser light
at a) m/z 157, and b) m/z 1200. (note diﬂerent scales) ........ 47

Precursor selection by pulsed-laser photodissociation ........... 50

Product spectrum for m/z 156 (molecular ion) of

bromobenzene .......................................................................... 53

Product spectra of m/ z 91 of toluene a) with convoluted
normal spectrum, and b) with normal spectrum

subtracted. (* indicates product ion peak) .............................. 57

SIMION models showing equi-potential contours and
ion trajectories for a) the interleaved comb ion
deﬂection gate, and b) a deﬂection plate gate (note
different scales) ........................................................................ 63

The interleaved comb ion deﬂection gate device .................... 66

Figure 4.4 SIMION models showing equi-potential contours for a)
the planar gate device, and b) the non-planar gate

device ........................................................................................ 68

Figure 4.5 Overview of electronics used to produce complementary

gate pulses ............................................................................... 70
Figure 4.6 Typical pulse shapes produced by the DEI pulsers ............... 72

Figure 4.7 TOF mass spectra of the molecular ion region of
diphenylmethane for gate pulse width of a) 1.28 us,
b) 140 ns, and c) 140 ns with gate trigger signals

reversed .................................................................................... 74

Figure 5.1 A plot of 2-AT/T as a function of ion energy for

homogeneous ﬁeld reﬂectrons ................................................. 82

Figure 5.2 Molecular ion region of bromobenzene a) without MPI
elimination ﬁeld and b) with NIPI elimination ﬁeld .............. 88

Figure 5.3 Product spectrum from m/z 156 of bromobenzene

collected using a short, linear TOF analyzer ......................... 92

Figure 5.4 Calculated proﬁle for the non-linear portion of the
reﬂectron ................................................................................ 100

Figure 5.5 A plot of 2-At/t as a function of ion energy for the fully
gridded reﬂectron .................................................................. 102

Figure 5.6 SIMION a) fully gridded reﬂectron, b) two-grid
reﬂectron, and c) two-grid reﬂectron after adjustment

of ring voltages ....................................................................... 104

Figure 5.7 SIMION ion trajectories for a, a') the fully gridded
reﬂectron, b, b') the two-grid reﬂectron, and c, c') the
two-grid reﬂectron after adjustment of the ring

electrode voltages ................................................................... 106
Figure 5.8 Comparison of MathCad® calculation and SIMION
simulation .............................................................................. 107

Figure 5.9 A plot of 2-At/t as a function of ion energy for the two-
grid, empirically adjusted reﬂectron over the range of
1950 to 2700 eV at intervals of 25 eV ................................... 1 10

Figure 5.10 Product spectrum of m/z 91 from toluene ............................ 113

Figure 5.1 1 Comparison of resulting reﬂectron proﬁles for diﬂ'erent
lengths of the ﬁrst stage ........................................................ 117

Chapter 1: Introduction

1.1 Goal

The goal of this research is to continue the development of a tandem
mass spectrometer (MS/MS) capable of greatly increasing the amount of
MS/MS data obtainable from sample components as they elute ﬁ-om a
chromatographic column. The characterizing power of MS/MS for pure
sample introduced to the ion source for a sufﬁcient period of time is
tremendous. However, the characterizing power of mm is limited for
sample introduced chromatographically since current commercial
instrumentation allows the collection of only a few mass scans during the
elution time of a single component. This dissertation describes the
development of a high speed tandem time-of-ﬂight (TOF) mass spectrometer,
with emphasis on the achievement of unit mass resolution for all aspects of

instrument performance.
1.2 Tandem mass spectrometry

Tandem mass spectrometry (MS/MS) is the linking of two mass
analyzers. Ions are initially created from solid, liquid, or gaseous sample in
an ion source. These ions may consist of intact molecular ions or ionic

fragments of the molecule. The ﬁrst mass spectrometer is used to select ions
1

2
of a particular mass to charge (m/z) ratio from those produced in the source.
Following this selection, a m/z-changing reaction occurs (in most cases a
dissociation), either spontaneously or in response to induced energy in a
reaction region. Energy can be induced by collision of the ions with a
background gas (collision-induced dissociation or CID), collision of the ions
with a surface (surface-induced dissociation or SID), or by irradiating the
ions (photo-induced dissociation or PID). Ions which decay spontaneously
outside of the ion source are referred to as metastable ions. The charged
products of the reaction are then analyzed by the second mass spectrometer.
The ions of a particular m/z value selected in the ﬁrst mass analyzer are
referred to as precursor ions. Ions resulting from the m/z-changing reaction
are referred to as product ions, and ions resulting from fragmentation in the
ion source are referred to as fragment ions. In the case of quadrupole ion
traps or Fourier transform mass spectrometers (FTMS), the entire sequence
occurs within one mass analyzer which performs both the initial selection
and the analysis of product ions. When magnetic sector instruments, electric
sectors, quadrupoles, or time-of-ﬂight mass analyzers (TOF) are used,
precursor selection and product analysis are performed in sequential mass

analyzers.
1.2.1 Ion structure data

While the basic experiment remains unchanged, the role of the ﬁrst

mass analyzer varies depending on the type of ionization used and whether

3
or not the sample is a pure compound. In one case, a pure compound is
introduced to the mass spectrometer and is ionized by a ‘hard’ ionization
method (one which results in a high degree of fragmentation in the source)
such as electron-impact ionization. Each m/z value observed in a normal
mass spectrum for the pure compound can then be selected, the ion
fragmented, and a mass spectrum for that m/z value ion obtained. In this
case, the beneﬁt of MS/MS is the increased structural information obtained
and the ability to determine ﬁ'agmentation pathways [1, 2]. Both mass
analyzers provide ion structure data. In the majority of analyses, the
complete MS/MS data set provides valuable additional information regarding

the molecular structure of the sample compound.
1.2.2 Mixture data

If a ‘soft’ ionization method (one which results in little or no
fragmentation of the sample molecule) such as chemical ionization or
electrospray ionization is used, then the ﬁrst mass analyzer can serve to
separate different molecular ions from one another. This allows the
introduction of a mixture directly to the mass analyzer, thereby eliminating
the need to use chromatography to separate the compounds and greatly
reducing analysis time. The fragmentation and subsequent mass analysis
provide the structural information required to identify the compound [2, 3].
Operating in this way, tandem mass spectrometry is especially suited to trace

analysis of targeted compounds in mixtures [2, 4, 5]. The ﬁrst mass analyzer

4
is used to provide separation of a mixture rather than to provide ion
structure data as was described in the previous section. Of course, isomers or
other compounds having the same molecular weight cannot be separated

with the ﬁrst mass analyzer.
1.3 MS for chromatographic detection

Often gas chromatographs (GC) or high performance liquid
chromatographs (HPLC) are interfaced to mass spectrometers and to tandem
mass spectrometers. The combination of chromatographic resolving power
and the sensitivity and selectivity of MS make these systems extremely
powerful, explaining the popularity of these instruments in the scientiﬁc
community. In this work, a GC was interfaced to the instrument as described

below, and this thesis will refer primarily to GC applications.
1.3.] Overlapping chromatographic peaks

If a mixture is not fully separated by chromatography, the normal
mass spectra observed are a convolution of the spectra of the coeluting
compounds. These overlapping spectra are exceedingly difﬁcult to interpret
unless mathematical methods can be employed to separate the spectra.
Spectral subtraction can be used to remove background which may consist of
column bleed, overlapping elutions, or both [6, 7]. Reverse library search
algorithms (in which known spectra are compared to the sample spectrum to

determine whether or not the compound is present) are effective when the

5
identities of compounds expected to be present are known [8-10]. In some
cases, mathematical deconvolution of the overlapping spectra is possible
based on a pure elution proﬁle for a particular m/z value unique to each
compound [1 1-14]. Alternatively, factor analysis has been used to determine
components of coeluting mixtures [15, 16].

In chromatography/MS/MS, if compounds coelute, the ﬁrst mass
analyzer can be used to provide further separation, thereby increasing
resolving power. If the chromatographic separation is complete, both mass
analyzers can provide structural data and increase selectivity. For many
applications, GC/MS or LC/MS provides adequate selectivity and sensitivity,
whereas GC/MS/MS or LC/MS/MS is most often applied to complex mixtures
in complex matrices such as analyses of environmental [17], food [18], or

pharmaceutical [19, 20] samples.

1.3.2 Instrumentation

1.3.2.1 Mass filters

Most commercially available MS and MS/MS . instruments are
scanning mass spectrometers. These scanning instruments are also referred
to as mass ﬁlters because they allow the passage of a single m/z value at a
time, and the conditions of the mass spectrometer are scanned as a function
of time to collect a mass spectrum. The scanning mass ﬁlters include

quadrupoles, magnetic sectors, and combined electric and magnetic sectors

6

for normal mass spectrometry. The most common type of tandem mass
spectrometer is the triple quadrupole mass spectrometer in which the ﬁrst
and third quadrupole are the mass ﬁlters and the center quadrupole serves
as the reaction region. Tandem mass spectrometers based on electric and
magnetic sectors are of varied geometry and involve 3 to 4 sectors in general,
with the reaction region simply consisting of a ﬁeld-free region between
sectors.

Mass ﬁlters exhibit a trade-off between scan rate and sensitivity since
increasing scan rate reduces the amount of time spent collecting each m/z
value. Even when a mass ﬁlter is scanned slowly, the majority of ions
produced in the source are always lost since all but one m/z value are being
ﬁltered away continuously. Most GC/MS analyses are performed at a scan
rate that allows the collection of a few mass spectra per second in order to
provide adequate sensitivity for monitoring capillary GC eﬂluent.

An important consideration is the compatibility of sampling rate and
chromatographic peak width. To maintain less than 0.1% error in peak
height determination, at least 2.6 samples are required per peak (full width
at half maximum height) [21]. This assumes that the peak shapes are
Gaussian, that noise is negligible, and that the same sampling rate is
maintained on either side of the peak for a considerable time. Description of
the entire shape of the elution proﬁle across a real chromatographic peak

(not necessarily Gaussian) with the same level of accuracy in the presence of

7

noise may require 100 or more data points [22]. Since capillary GC typically
results in peaks having widths on the order of a few seconds, the sampling
rate normally used (a few Hz) is barely sufﬁcient to allow quantitation based
on peak height at a useful level of accuracy. This rate is not sufﬁcient to
allow effective mathematical deconvolution, which requires accurate
description of the peak proﬁle [12]. The fact that commercial MS detectors for
GO barely provide adequate sampling rates illustrates an important obstacle
in the development of a tandem mass spectrometer for GO detection.

Another diﬂiculty associated with the relatively slow scan of mass
ﬁlter instruments is that the sample pressure in the ion source can change
signiﬁcantly during the collection of a single spectrum. This results in a
skewed spectrum. For example, if a spectrum is collected on the rising edge of
a chromatographic peak, then the high m/z portion of the spectrum is
enhanced, whereas the opposite is true if the spectrum is collected on the

falling edge of a peak.
1.3.2.2 Batch array detectors

Batch array detection by ion trap mass spectrometry or time-of-ﬂight
(TOF) mass spectrometry can eliminate the skewing problem since ionization
occurs for a period of time followed by detection of all of the ions formed
during the ionization period. Unfortunately, the spectral generation rate is
not signiﬁcantly different for quadrupole ion trap mass spectrometers than

for quadrupole. mass spectrometers. Potentials are applied to the quadrupole

8

ion trap mass spectrometer in such a way as to sequentially eject ions of
increasing m/z values through holes in one of the trap electrodes where a
detector records ion signal. While sensitivity for the quadrupole mass
spectrometer is inversely proportional to scan rate, the ion trap mass
spectrometer is limited by the fact that the trap can hold a limited number of
ions, therefore allowing a limited time for sample ionization, followed by an
analysis period during which chromatographic eﬂluent is lost. Also, scan rate
is inversely proportional to mass resolution.

Ion trap mass spectrometers are capable of mm experiments
through the use of an excitation sequence which eliminates m/z values from
the trap volume above and below the desired precursor m/z. Fragmentation
by any of the common fragmentation techniques can then be implemented,
followed by the collection of product spectrum in the usual manner.

TOF mass spectrometry is a pulsed technique where ions are
periodically extracted from an ion source and subsequently separated
according to m/z value in a ﬁeld-free ﬂight tube. A spectrum is collected as a
function of time as the ions strike a detector at the end of the ﬂight tube.
TOF has the highest spectral generation rate of any mass analyzer, usually
on the order of several kilohertz, limited only by the ﬂight time of the largest
m/z value. Collection of the large number of spectra produced by TOF for a
sustained period of time, as is required for chromatographic analyses,

requires the use of sophisticated data collection electronics. The integrating

9
transient recorder developed at Michigan State University [23-2 5] averages a
speciﬁed number of individual transients (single spectra) and stores the
summed spectra to disk for a spectral generation rate compatible with
accurate description of elution proﬁles for GC/MS. A GC/TOF-MS instrument
with a data collection system that operates on the TOF time scale has

recently become commercially available [26].
1.3.2.3 True array detectors

True array detection can be accomplished either in a sector instrument
with the use of a spatial array detector or by Fourier transform mass
spectrometry (FTMS), again with the elimination of the spectral skewing
problem. Spatial array detectors are very effective in improving sensitivity
and spectral generation rate [2 7, 28]. These detectors are expensive, however,
as are the sector instruments with which they are used. Either mass
resolution or mass range is limited by the number of elements in the array.
FTMS requires extremely low pressures within the instrument, so that the
gas load of chromatography is an obstacle to this combination. Spectral
generation rate for FTMS is low due to a trade-off between acquisition time
and mass resolution. FTMS can be used to accomplish MS/MS in a manner
similar to that described for ion trap mass spectrometers, with a period of

time required for isolation of the precursor.

10

1.3.3 Greater challenge of chmmatography/MS/MS

Using MS/MS as a chromatographic detector as opposed to MS
ampliﬁes the spectral generation rate requirements since the analyst would
ideally collect a complete MS/MS data set during an elution. For tandem
mass ﬁlter instruments, this would require a scan of the product analyzer for
each m/z value in the normal mass spectrum. For commercially available
chromatography/MS/MS instruments, only a small portion of the MS/MS
data set can be collected during the elution of a single component.

Chromatography/MS/MS is usually performed in selected reaction
monitoring (SRM) mode where a single or a few speciﬁc precursor/product
combinations are monitored with the exclusion of the rest of the data set in a
targeted analysis [2]. This method requires previous knowledge of the
sample. If the pure compound is available, it can be characterized by MS/MS
to ﬁnd a unique fragmentation pathway for that compound which will allow
its identiﬁcation when analyzed in its matrix. Using SRM, entire
chromatographic runs are dedicated to the detection of a single or a few
compounds, while ability to monitor any other compound in the matrix is
sacriﬁced. Despite these limitations and laborious method development
requirements, the high degree of selectivity and sensitivity of MS/MS has
proven itself extremely useful in many areas of research for both GC and LC

applications.

11

1.3.4 Evolving chromatographic methods

Related to the spectral generation rate requirements associated with
current MS and MS/MS instrumentation are the characteristics of the
sample from the GC column. Speciﬁcally, the improvements in GC as a result
of the development of capillary columns in the early 1980’s dramatically
aﬁ'ected the requirements of the mass spectrometer. Peak widths were
reduced from around twenty seconds to a few seconds, and the amount of
material introduced into the mass spectrometer was reduced to picograms.

The performance of scanning mass spectrometers barely meets the challenge.
1.4 Overview of the tandem TOF instrument

Figure 1.1 is a diagram of the tandem TOF mass spectrometer
developed in Professor Enke’s laboratory. The electron-impact source ionizes
gaseous sample introduced from a GC oven (not pictured) either continuously
by direct-leak from a sample vial or by injection and chromatography.
Outside of the oven, a heated transfer line (not pictured) serves to heat the
column through which the sample stream enters the instrument. The
spectral acquisition sequence begins with a function generator which
generates a square wave. This signal simultaneously triggers the source
pulsing circuit and a delay generator. In response to the trigger signal from

the function generator, a 200 V pulse is applied to the back-plate of the

MPI Retardation Interaction
. . . Region Re 'on
Source Post-Dissocmtion ‘9 81 Gate Reflectron 1
Acceleration Region \ / /

 

 

 

 

I. 0,000....

~|IEI \§\

 

 

// \llllllll\
Milli

 

 

 

 

 

 

 

 

 

 

 

 

 
    
 
   

 

     

 

 

l M? a
3. \WW _: _____s
E _=F —: —-——-——.
Reﬂectron 2 Detector
Excimer Gate
Laser Pulser
In
Function 3 Trigger
Generator 5 Trigger Transient
Recorder
Acquisition

 

 

 

 

 

Delay 1 Delay Times
Generator

Figure 1.1 Schematic diagram of the

      
   

 

tandem TOF mass spectrometer

with instrument control and data collection system.

13

source, causing extraction of the ions into a secondary, static acceleration
region of 550 V. The resulting ion beam is collimated with an einzel lens and
steered with two sets of orthogonally oriented steering plates. The ion beam
is then reﬂected by reﬂectron one and directed towards the interaction
region. Isa-mass ion packets separate from one another through the ﬁrst
mass analyzer and are energy-focused by the reﬂectron at the position of the
interaction region. The instrument chamber and vacuum system, sample
introduction system, ion source, ﬁrst mass analyzer, electronics, detectors,
and instrument control and data collection systems are discussed in more
detail in chapter 2.

Just prior to arrival at the interaction region, a gate device is used to
isolate a single, chosen iso-mass ion packet by deﬂection of all other m/z
values onto a path which will not result in collision with the detector. This
step is not necessary for precursor selection since selection is determined
later by the interaction of a laser pulse with the precursor ion packet. Rather,
the gate serves to eliminate background. The ﬁring of the gate is controlled
by a delayed pulse from the delay generator, which is programmed from the
computer. The gate device and its control electronics are discussed in detail
in chapter 4.

The pulsed laser is timed (in a similar fashion) to irradiate the ion
packet as it passes through the interaction region. Focusing of the laser pulse

and the precursor iso-mass ion packet to a small volume results in high

l4
eﬁciency fragmentation of the precursor ions. Pulsed-laser selection and the
resultant photodissociation eﬁciencies are the subject of chapter 3.

After interaction with the laser pulse, the resulting product ions and
unfragmented precursor ions enter the second mass analyzer. Besides the ion
packet itself, a considerable population of neutral background gas exists in
the interaction region. Multi-photon ionization of this background gas results
in a large background signal. To prevent these ions ﬁ'om reaching the
detector, a 20 to 30 V retardation ﬁeld is created at the start of the second
mass analyzer. This ﬁeld does not discriminate against the sample ions.
Immediately following this retardation ﬁeld is an acceleration ﬁeld of 1950 V
which causes a separation of the product ions through the subsequent ﬁeld-
free ﬂight. Ion optical elements, such as the ion gate and the grids used to
create the retardation and acceleration ﬁelds, are placed on an optical rail
which has been mounted parallel to the ion ﬂight path in the interaction
region. After dissociation and subsequent acceleration, a second reﬂectron is
used to energy-focus and reﬂect the product ions onto the detector. A detailed
description of the reﬂectron developed for the second mass analyzer is given
in chapter 5. The signal as a result of each source pulse is referred to as a
transient and is recorded by the transient recorder, where it can be averaged
with other transients to improve signal-to-noise ratio. The averaged or
single-shot spectra are then transferred to the computer hard drive for

storage.

15
It is not evident in Figure 1.1 that the ion ﬂight path is actually tipped
with respect to the plane of the page (or the bottom of the instrument
chamber). In other words, the source is higher than the interaction region
which is higher than the detector. This arrangement avoids the possibility of

the laser light striking ions in locations other than at the interaction region.

1.4.1 TOFprovides high spectral generation rate and

throughput

The achievement of more MS/MS data without a reduction in
sensitivity requires the abandonment of tandem mass ﬁlters in favor of a
mass array analyzer, at least for the product ion analysis stage. Also, a very
high spectral generation rate is desirable. Due to their considerable isolation
and analysis periods, FTMS and ion trap mass spectrometry are not suitable.
As already discussed, the mass ﬁlters (sectors and quadrupoles) are also not
suitable, with the possible exception of a sector instrument employing a
spatial-array detector. TOF offers the advantage of unrivaled spectral
generation rate and unrivaled efﬁciency when used in conjunction with an
ion source that stores a considerable portion of the ions formed between
pulsed extractions. Not only is it possible to perform array detection of
product ion spectra with a tandem TOF arrangement, but it is possible to
collect normal spectra at a similar rate simply by opening the gate and not
ﬁring the laser. Product spectrum generation rate is limited only by the

repetition rate of the laser. The laser currently used has a maximum

16
repetition rate of 250 Hz, although other commercially available excimer
lasers offer higher rates. A multiple sector MS/MS instrument utilizing a
spatial array detector for product ion analysis can collect normal spectra only

by scanning the ﬁrst mass spectrometer.
1.4.2 PID provides high efﬁciency fragmentation

TOF is a pulsed technique which can produce dense iso-mass ion
packets. Pulsed excimer lasers provide high energy pulses of light which can
be focused to approximately the volume of the ion packet when the ion beam
and light path are arranged orthogonally. As is discussed in chapter 3, this
dense ion-photon overlap results in high efﬁciency fragmentation and high

resolution selection of iso-mass precursor ion packets.
1.4.3 Unit resolution for TOF/PID/TOF analysis

Achieving the primary goal of this project is a useful demonstration
only if unit mass resolution can be maintained for both stages of analysis
over a mass range appropriate to GC analyses. The major portion of my work
and of this thesis is devoted to the development of an ion optical arrangement
that can meet this criterion despite the traditional reputation of TOF as a
low resolution technique. Unit mass resolution has been demonstrated for
precursor ion separation (chapter 2), isolation (chapter 4), and selection

(chapter 3) as well as for product ion analysis (chapter 5).

17

1.5 Historical background

Despite the limitations of commercially available tandem mass
spectrometers, the development of faster chromatography/MS/MS systems
has been the focus of relatively few publications in the scientiﬁc literature.
These are reviewed in section 1.5.1, below. Numerous tandem TOF
instruments have been constructed for a variety of goals, as well as hybrid
instruments which incorporate one TOF stage. These instruments are
reviewed in section 1.5.2. Finally, since photo-induced dissociation is by no
means a routine fragmentation method for MS/MS, relevant history

regarding PID will be provided in section 1.5.3.
1.5.1 Instrumentation for high speed MS/MS

In 1983, Stults, et al. [29] introduced the concept of time-resolved ion
momentum spectrometry (TRIMS). This method couples a magnetic sector
with a TOF mass spectrometer. This MS/MS instrument is non-traditional in
that the observed dissociation occurs between the source and the magnetic
sector. The source is operated in pulsed mode while the magnet is scanned.
The m/z value of the product ion and that of the precursor from which it
originated is obtained from the known ﬁeld strength of the magnet and the
ion packet’s ﬂight time. When later combined with the MSU integrating
transient recorder, the true power of this system was demonstrated by

collecting a complete MS/MS data set for caffeine in 5 seconds, a spectral

18

generation rate compatible with packed-column GC [30]. This instrument is
fundamentally limited by the scan rate of the magnet. A similar MS/MS
method which uses an electric sector in place of the magnetic sector is called
time-resolved ion kinetic energy spectroscopy (TRIKES) [29, 31]. This system
offers advantages and capabilities similar to those of the TRINIS system.

Another approach to fast MS/MS was the coupling of a quadrupole
analyzer with a TOF analyzer by Glish, et al. [32]. This instrument
demonstrated a throughput advantage over tandem mass ﬁlter instruments,
but gave poor product spectrum resolution and was limited in speed by its

data acquisition system.
1.5.2 Tandem TOF and hybrid instruments

A variety of tandem TOF instruments and hybrid MS/MS instruments
containing one TOF stage have been constructed for purposes other than for
use as a chromatographic detector. The ﬁrst use of TOF for analysis of
product ions was the reported observation of metastable ions using linear
TOF instruments (containing no reﬂectron) [33-39]. In these instruments,
several grids were used to create a short retardation stage near the end of the
ﬂight tube which causes a slight time shift for ions which have undergone
metastable decay. The product ions retain a fraction of the kinetic energy of
the precursor ion equal to the fraction of mass retained by the product ion.
Therefore, at a given retardation amplitude, the lower m/z portion of the

product ion spectrum will be reﬂected by the ﬁeld, while the higher m/z

19
portion will be transmitted. Slowly increasing the amplitude of the
retardation ﬁeld allowed the determination of the m/z of the product ion.
While these demonstrations did not involve discrete, multiple stages of mass
analysis, they did demonstrate the collection of product spectra using TOF.
The product spectra collected in this manner are, of course, convoluted with
the normal spectrum and are poorly resolved.

The next category of TOF instruments which allow the analysis of
metastable ions is a modiﬁcation of a single-stage reﬂectron TOF mass
spectrometer. This category is differentiated from that already discussed by
the use of a reﬂectron rather than a retardation stage to mass disperse
product ions. A reﬂectron consists of an electric ﬁeld which ions penetrate to
a depth (or height if one considers a potential hill) corresponding to their
kinetic energies. Ions with smaller kinetic energies penetrate the ﬁeld less
deeply than ions with greater kinetic energy. Reﬂectrons have primarily been
used to compensate for ﬂight time differences of iso-mass ions having a
distribution of kinetic energy, as discussed in the instrument overview above.
However, reﬂectrons can also (simultaneously) time disperse product ions on
the basis of their fractionated kinetic energies. This method has been used
with a variety of reﬂectron types to analyze metastable decay after multi-
photon ionization (MPI) [40], 2$2Cf ﬁssion fragment ionization or plasma
desorption mass spectrometry .(PDMS) [41-43], ~ fast ion bombardment

ionization or secondary ion mass spectrometry (SIMS) [44-45], electron

20

ionization [46], and, most recently, matrix-assisted laser desorption
ionization (MALDI) [47-49]. Boesl, et al. [40] dropped the amplitude of the
reﬂectron ﬁeld slightly, causing undissociated ions to not be reﬂected,
thereby differentiating between metastable and normal ions. In the SIMS
and PDMS instruments [41-45], decompositions were identiﬁed by the
detection of a neutral at a detector placed behind the reﬂectron, with the
arrival time of the neutral ‘correlated’ to the precursor m/z which
decomposed and the product ion which is then detected at the normal
detector position. This technique requires a very low ion ﬂux. The more
recent instruments [46-49] employ an ion gate to isolate a precursor m/z
(and products which have dissociated from that precursor m/z) so that any
signal other than the precursor in the resulting spectrum can be attributed to
metastable decay. Details regarding reﬂectrons and their use for product ion
analysis are reviewed and discussed in more detail in chapter 5.

Single reﬂectron TOF analyzers have also been used with a variety of
fragmentation techniques, where the dissociation is accomplished within the
reﬂectron. Kaufmann, et al. [50] observed product ions caused by collision of
precursor ions with a stainless steel mesh electrode near the front of the
reﬂectron in a SIMS instrument. Later, Williams, et al. [51] introduced a
solid plate near the back of the reﬂectron to cause SID of ions created by

MPI, while Duncan’s group [52, 53] employed laser PID at the turn-around

2 1
point of the ions in the back of the reﬂectron in their studies of metal cluster
ions.

The true tandem TOF instruments consisting of discrete TOF
analyzers can be classiﬁed according to whether they employ sequential
linear TOF analyzers, one linear TOF and one reﬂectron TOF analyzer, or
two reﬂectron TOF analyzers. Several linear TOF / linear TOF instruments
have been constructed for a variety of purposes which use gate selection of
ions prior to fragmentation and an acceleration stage in the second mass
analyzer to accomplish mass dispersion of the product ions. Smalley’s group
[54, 55] and Geusic, et al. [56] accomplished photodissociation of cluster ions,
Schey, et al. [57] performed SID of ions generated by electron ionization, and
Jardine, et al. [58] demonstrated CID of ions produced by fast atom
bombardment (FAB).

Two groups have constructed instruments which use a linear TOF
analyzer followed by a reﬂectron TOF analyzer. Schey, et al. [59] describe an
improved version of the instrument already mentioned above for SID of ions
produced by electron ionization. Schlag’s group [60, 61] has described two
versions of an instrument which allows PID of ions produced by MPI. The
two versions involve two different methods of reﬂectron focusing which are
described in chapter 5.

Finally, the last classiﬁcation of tandem TOF instruments is those

which use two reﬂectron TOF analyzers. Only two instruments have this

22
geometry, one of which is the subject of this thesis. The other was constructed
by Cornish and Cotter [62-65]. Cornish and Cotter use laser desorption
ionization followed by CID via gas introduced to the chamber through a
pulsed valve.

In general, the tandem TOF arrangements described above exhibit
poor resolution for either precursor separation, product ion resolution, or
both. The one exception is the tandem reﬂectron TOF instrument described
by Cornish and Cotter [62-65]. This instrument is not compatible with GC,
however, due to its laser desorption ion source.

Several hybrid instruments appear in the literature which use one
TOF stage. In addition to those described in section 1.5.1 above, a magnetic
sector / TOF instrument has been described by Gandy, et al. [66] which uses
an electron ionization source and SID as the fragmentation technique.
Russell’s group [67] has demonstrated PID in a hybrid electric/magnetic
sector / linear TOF instrument. More recently, this group substituted a
reﬂectron TOF instrument for the linear TOF [68]. This instrument has been
used in the neutral-ion correlation mode for collection of product spectra
resulting from metastable decomposition and CID and has demonstrated

excellent resolution for both precursor ion selection and product ion analysis.
1.5.3 Photodissociation as a fragmentation technique for MS

Photo-induced dissociation of organic ions is not routinely used as a

fragmentation technique primarily due to the relatively low cross-sections

23

(about 10'17 to 10:19 cm?) for the PID process as compared to those for CID
(about 10'15 to 2 X 10'“ cm2) [2]. Dunbar [69, 70] is credited with much of the
pioneering work in the area of photodissociation of organic ions, performed in
ion cyclotron resonance (ICR) mass spectrometers where the trapped ions can
be irradiated for up to 3 seconds. A variety of light sources have been used to
accomplish PID in ICR cells in the infrared [70-73], visible [74], and
ultraviolet [7 5-7 7] regions. Unfortunately, due to the background of neutrals
in the ICR cells, it is difﬁcult to differentiate between ions that result from
PID versus ions that result from ionization of the neutrals or CID of ions
with the neutrals. ICR cells were once the only realistic option for PID
experiments due to the nature of the light sources available, so that
researchers were forced to deal with the problems associated with these
experiments.

As light sources improved, photodissociation experiments in beam
instruments became possible, reducing the problems associated with neutral
1 background which hampered ICR experiments. Beynon’s group [78, 79] ﬁrst
demonstrated ion photodissociation in a sector instrument arranging the ion
beam coaxial to the light path, and Wagner-Redeker and Levsen [80] later
reported a similar apparatus. McGilvery and Morrison [81] and Krailler and
Russell [82, 83] have arranged the laser perpendicularly to the ion beam in
quadrupole and sector instruments, respectively. Both groups later

implemented a coaxial geometry to improve sensitivity.

24

Modern pulsed lasers have become an attractive option for PID due to
their compatibility with pulsed ion sources [84] in sector instruments and
with TOF mass spectrometers. As already mentioned in section 1.5.2,
Smalley’s group [54, 55] and Geusic, et al. [56] accomplished
photodissociation of cluster ions in simple linear TOF / linear TOF
instruments, while Duncan’s group [52, 53] employed laser PID at the tum-
around point of the ions in the back of a reﬂectron TOF instrument in their
studies of metal cluster ions. Russell’s group [67] has demonstrated PID in a
hybrid electric/magnetic sector / linear TOF instrument, and Schlag’s
group [60, 61] has employed laser PID in a linear TOF / reﬂectron TOF
instrument.

More complete histories of the utility of PID in mass spectrometry are
available [12, 84, 85]. To date, the majority of work done with PID can be
described as fundamental. Studies such as determining PID cross-section,
investigating kinetic energy release upon dissociation, or investigating the
effect of laser wavelength and power on photodissociation are among the
most common [84]. Chapter 3 describes the implementation of laser PID in
this dual reﬂectron tandem TOF instrument for the primary purpose of

providing efﬁcient fragmentation of precursor ions.
1.6 References

1 Yost, R. A.; Enke, C. G. Amer. Lab., 1981, 88-95.

10
11
12

13

14

15
16

17

18
19

25

Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry / Mass
Spectrometry: Techniques and Applications of Tandem Mass
Spectrometry, VCH Publishers, Inc.: New York, 1988.

Yost, R. A.; Enke, C. G. J. Am. Chem. Soc., 1978, 100:7, 2274.

Hunt, D. F.; Shabanowitz, J .; Harvey, T. M.; Coates, M. L. Anal. Chem.
1985, 57, 525.

Glish, G. L.; Shaddock, V. M.; Harmon, K.; Cooks, R. G. Anal. Chem.,
1980, 52, 165.

Watson, J. T. Introduction to Mass Spectrometry, Raven Press: New
York (1985).

Blaisdell, B. E.; Sweeley, C. C. Anal. Chim. Acta 1980, 117, l.
Abramson, F. P. Anal. Chem. 1975, 47(1), 45.

Dromey, R. G.; Steﬁk, M. J .; Rindﬂeisch, T. C.; Dufﬁeld, A. M. Anal.
Chem. 1976, 48(9), 1368.

Blaisdell, B. E.; Sweeley, C. C. Anal. Chim. Acta 1980, 117, 17.
Biller, J. E.; Biemann, K. Anal. Letters, 1974, 7(7), 515.
McLane, R. D. Ph. D. Thesis, Michigan State University, 1993.

McLane, R. D.; Vlasak, P. R.; Enke, C. G.; Yefchak, G. E.;
Rodriguez, P. A; Eddy, C. L.; Mazzone, M. A.; Pinkston, J. D.
Proceedings of the 40th ASMS Conference on Mass Spectrometry and
Allied Topics, Washington, DC, May 31 - June 5, 1992, 1171.

McLane, R. D.; Vlasak, P. R.; Enke, C. G.; Yefchak, G. E.;
Rodriguez, P. A.; Eddy, C. L.; Mazzone, M. A; Pinkston, J. D. (in
preparation for submission to Anal. Chem.)

Malinowski, E. R.; McCue, M. Anal. Chem. 1977, 49, 284.
Knorr, F. J.; Futrell, J. H. Anal. Chem. 1979, 51, 1236.

Gale, B. C.; Fulford, J. E.; Thomson, B. A; Ngo, .A; Tanner, S. D.;
Davidson, W. R.; Shushan, B. 1. Adv. Mass Spectrom. 1986, 10, 1467.

Trehy, M. L.; Yost, R. A; Dorsey, J. G. Anal. Chem. 1986, 58, 14.
Nelson, C. C.; Foltz, R. L. Anal. Chem. 1992, 64, 1578.

20
21
22
23

24

25

26

27

28

29
30

31

32

33

34

35

26

Zirrolli, J. A; Murphy, R. C. J. Am. Soc. Mass Spectrom. 1993, 4, 223.
Kelly, P. C.; Horlick , G. Anal. Chem. 1973, 45(3), 518.
Chesler, S. N.; Cram, S. P. Anal. Chem. 1971, 43, 1922.

Erickson, E. D.; Enke, C. G.; Holland, J. F.; Watson, J. T. Anal. Chem.
1990, 62(10), 1079.

Tecklenburg, R. E., Jr.; McLane, R. D.; Grix, R.; Sweeley, C. C.;
Allison, J.; Watson, J. T.; Holland, J. F.; Enke, C. G.; Gruner, U.; Gotz,
H.; Wollnik, H. Proceedings of the 38th ASMS Conference on Mass
Spectrometry and Allied Topics, Tucson, AZ, June 3-8, 1990.

Holland, J. F.; Newcombe, B.; Tecklenburg, R. E., Jr.; Davenport, M.;
Allison, J.; Watson, J. T.; Enke, C. G. Rev. Sci. Instrum. 1991, 62(1),
69.

Meridian Analytical Systems, Okemos, MI (recently purchased by
Laboratory Equipment Company (LECO), St. Joseph, MI).

Tuithof, H. H.; Boerboom, A J. H.; Meuzelaar, H. L. C. Int. J. Mass
Spectrom. Ion Processes 1975, 17, 299.

Louter, G. J .; Boerboom, A J. H.; Stalmeier, P. F. M.; Tuithof, H.H.;
Kistemaker, J. Int. J. Mass Spectrom. Ion Processes 1980, 33, 335.

Stults, J. T.; Enke, C. G.; Holland, J. F. Anal. Chem. 1983, 55, 1323.

Eckenrode, B. A; Watson, J. T.; Enke, C. G.; Holland, J. F. Anal.
Chem. 1990, 62, 1362.

Pinkston, J. D.; Rabb, M.; Watson, J. T.; Allison, J. Rev. Sci. Instrum.
1986, 57(4), 583.

Glish, G. L.; McLuckey, S. A; McKown, H. S. Anal. Instr. 1987, 16,
191.

Hunt, W. W., Jr.; Huffman, R. E.; McGee, K. E. Rev. Sci. Intrum. 1964,
35 (1), 82. ‘

Hunt, W. W., Jr.; Huffman, R. E.; Saari, J .; Wassel, G.; Betts, J. F.;
Paufve, E. H.; Wyess, W.; Fluegge, R. A Rev. Sci. Intrum. 1964, 35 (1),
88.

Hunt, W. W., Jr.; Huffman, R. E.; McGee, K. E. J. Chem. Phys. 1964,
41 (9), 2709.

36

37
38
39

40

41

42

43

44

45

46

47

48

49

50

51

27

Ferguson, R. E.; McCulloh, K. E.; Rosenstock, H. M. J. Chem. Phys.
1965, 42(1), 100.

Dugger, D. L.; Kiser, R. W. J. Chem. Phys. 1967, 47(12), 5054.
Haddon, W. F.; McLafferty, F. W. Anal. Chem. 1969, 41(1), 31.

Chait, B. T.; Field, F. H. Int. J. Mass Spectrom. Ion Processes 1981, 41,
17.

Boesl, U.; Neusser, H. J .; Weinkauf, R.; Schlag, E. W. J. Phys. Chem.
1982, 86', 4857.

Della Negra, 8.; Le Beyec, Y. Int. J. Mass Spectrom. Ion Processes
1984, 61, 21.

Della Negra, 3.; Le Beyec, Y. Anal. Chem. 1985, 57, 2035.

Brinklman, G.; Hakansson, P., Kjellberg, J .; Demirev, P.;
Sundqvist, B. U. R. ; Ens, W. Int. J. Mass Spectrom. Ion Processes 1992,
114, 183.

Standing, K. G; Beavis, R.; Bolbach, G.; Ens, W.; Lafortune, M.;
Main, D.; Schueler, B.; Tang, X.; Westmore, J. Anal. Instrum. 1987,
16, 173.

Tang, X.; Beavis, R. ; Ens, W.; LaFortune, F.; Schueler, B.;
Standing, K. G. Int. J. Mass Spectrom. Ion Processes 1988, 85, 43.

Haberland. H.; Kommeier, H.; Ludewigt, C.; Risch, A Rev. Sci.
Intrum. 1991, 62( 10), 2368.

Spengler, B.; Kirsch, D.; Kaufmann, R.; Jaeger, E. Rapid Commun.
Mass Spectrom. 1992, 6, 105.

Kauﬁnann, R.; Spengler, B.; Lutzenkirchen, F. Rapid Commun. Mass
Spectrom. 1993, 7, 902.

Kaufmann, R.; Kirsch, D.; Spengler, B. Int. J. Mass Spectrom. Ion
Processes 1994, 131, 355.

Kaufmann, R.; Kirsch, D.; Rood, H.-A; Spengler, B. Rapid Commun.
Mass Spectrom. 1992, 6, 98.

Williams, E. R.; Fang, L; Zare, R. N. Int. J. Mass Spectrom. Ion
Processes 1993, 123, 233.

52

53

54

55

56

57

58

59

60

61

62

63
64
65

66

67

28

LaiHing, K.; Cheng, P. Y. ; Taylor, T. G.; Willey, K. F.; Peschke, M.;
Duncan, M. A. Anal. Chem. 1989, 61, 1460.

Cornett, D. S.; Peschke, M.; LaiHing, K.; Cheng, P. Y.; Willey, K. F.;
Ducnan, M. A Rev. Sci. Instrum. 1992, 63(4), 2177.

Brucat, P. J .; Zheng, L.-S.; Pettiette, C. L.; Yang, S.; Smalley, R. E.
J. Chem. Phys. 1986, 84(6), 3078.

Liu, Y.; Zhang, Q.-L.; Tittel, F. K.; Curl, R. F.; Smalley, R. E. J. Chem.
Phys. 1986, 85(12), 7434.

Geusic, M. E.; Jarrold, M. F.; McIlrath, T. J .; Freeman, R. R.;
Brown, W. L. J. Chem. Phys. 1987, 86(7), 3862.

Schey, K.; Cooks, R. G.; Grix, R.; Wollnik, H. Int. J. Mass Spectrom.
Ion Processes 1987, 77, 49.

Jardine, D. R.; Morgan, J.; Alderdice, D. S.; Derrick, P. J. Org. Mass
Spectrom. 1992, 27, 1077.

Schey, K. L.; Cooks, R. G.; Kraft, A; Grix, R.; Wollnik, H. Int. J. Mass
Spectrom. Ion Processes 1989, 94, l.

Weinkauf, R.; Walter, K.; Weickhardt, C.; Boesl, U.; Schlag, E. W.
Z. Naturforsch. 1989, 440, 1219.

Boesl, U.; Weinkauf, R.; Schlag, E. W. Int. J. Mass Spectrom. Ion
Processes 1992, 112, 121.

Cornish, T.; Cotter, R. J. Rapid Commun. Mass Spectrom. 1992, 6,
242.

Cornish, T. J.; Cotter, R. J. Org. Mass Spectrom. 1993, 28, 1129.
Cornish, T. J.; Cotter, R. J. Anal. Chem. 1993, 65, 1043.

Cornish, T. J .; Cotter, R. J. Rapid Commun. Mass Spectrom. 1993, 7,
1037.

Gandy, R. M.; Ampulski, R.; Prusaczyk, J .; Johnsen, R. H. Int. J. Mass
Spectrom. Ion Physics 1977, 24, 363.

Tecklenburg, R. E., Jr.; Russell, D. H. presented at the Pittsburgh
Conference, Atlanta, Georgia, March 1989.

68

69

70

71

72

73
74

75

76

77

78

79

80
81

82
83

84
85

29

Strobe], F. H.; Preston, L. M.; Washburn, K. 8.; Russell, D. H. Anal.
Chem. 1992, 64, 754.

Dunbar, R. C.; In Gas Phase Ion Chemistry, Bowers, M. T., Ed.
Academic: New York, 1979, Volume 2.

Dunbar, R. C.; In Gas Phase Ion Chemistry, Bowers, M. T., Ed.
Academic: New York, 1984, Volume 3.

Baykut, G.; Watson, C. H.; Weller, R. R.; Eyler, J. R. J. Am. Chem. Soc.
1985, 107, 8036.

Watson, C. H.; Baykut, G.;Battiste, M. A; Eyler, J. R. Anal. Chim.
Acta 1985, 178, 125.

Watson, C. H.; Baykut, G.; Eyler, J. R. ACS Symp. Ser. 1987, 359.
Dunbar, R. C. Anal. Instrum. 1988, 17, 113.

Bowers, W. D.; Delbert, S.-S.; McIver, R. L. J. Am. Chem. Soc. 1984,
106, 7288.

Hunt, D. F.; Shabanowitz, J.; Yates, J. R. J. Chem. Soc., Chem.
Commun. 1987, 548.

Dunbar, R. C.; J. Chem. Phys. 1989, 91, 6080.

Kingston, E. E.; Morgan, T. G.; Harris, F. M.; Beynon, J. H. Int. J.
Mass Spectrom. Ion Processes 1972, 43, 261.

Grifﬁths, I. W.; Mukhtar, E. S.; Harris, F. M.; Beynon, J. H. Int. J.
Mass Spectrom. Ion Processes 1981, 38, 333.

Wagner-Redeker, W.; Levsen, K. Org. Mass Spectrom. 1981, 16, 538.

McGilvery, D. C. ; Morrison, J. D. Int. J. Mass Spectrom. Ion Physics
1978, 28, 81.

Krailler, R. E.; Russell, D. H. Anal. Chem. 1985, 5 7, 1211.

Krailler, R. E.; Russell, D. H. Int. J. Mass Spectrom. Ion Processes
1985, 66, 339.

Tecklenburg, R. E., Jr.; Russell, D. H. Mass Spec. Rev. 1990, 9, 405.
Beussman, D. J. Ph. D. Thesis, Michigan State University, 1995.

Chapter 2: Instrument design

The tandem reﬂectron TOF instrument has been developed and
continues to develop as a team effort. I am the third student to contribute to
this project, following Mary A. Seeterlin [1] and R. David McLane [2]. Mary’s
contributions involved primarily the design and construction of the vacuum
chamber and pumping system, the instrument’s electronics, the source, the
ﬁrst mass analyzer, and other ion optical elements. David’s contributions
involved the selection and installation of the laser, the development of the
instrument control system and data collection system, and the development
of the gas chromatograph / instrument interface. Although I have
maintained, modiﬁed, and, in some cases, aided in the design and
construction of these aspects of the instrument, included below is just a brief
description of these developments. The aspects of instrument development for
which I was primarily responsible are described in the following chapters.

Since my involvement, two more students have joined the tandem TOF
project. Douglas J. Beussman joined the project shortly before David and
Mary ﬁnished their degrees, taking over responsibility for laser maintenance
and contributing to several aspects of instrument development that are
described in the chapters below. Douglas’ primary area of interest is the

study of the photodissociation process itself, observing the effect of laser
30

3 1
power and wavelength on the efﬁciency of photodissociation and the speciﬁc
fragmentation pathways for a variety of precursor ions. Tina Erickson has
recently joined the tandem TOF effort on an ofﬁcial basis; however, her
interest in the project has led her to assist with experiments for some time.
Speciﬁcally, she has facilitated the evolution of the instrument control and
data collection systems as various components are added and replaced. She
plans to continue Douglas’ efforts in the study of the photodissociation

process.
2.1 Chamber & vacuum system

The instrument chamber is constructed of 0.75 inch stainless steel and
has inner dimensions of 60 inches long by 11 inches wide by 9 inches tall.
The top of the instrument chamber has three removable access panels to
allow easy access to all areas of the instrument. A variety of ﬂanges contain
appropriate electrical and mechanical feed-throughs.

The source housing is a 6-way cross mounted on the end of the
instrument chamber. A viewport is attached to the top of the source housing.
Two turbomolecular pumps (Pfeiﬂ‘er, Hudson, NH) are used to evacuate the
chamber. One, a TPH 270, is mounted directly underneath the source while
the second, a TPH 510, is mounted on the underside of the chamber itself,
accomplishing differential pumping of the source chamber and main

chamber. These pumps are backed by two D16A rotary-vane rough pumps

32
(Leybold Vacuum Products, Export, PA). The pressures achieved are
measured with two ionization gauges mounted on the instrument, one on the
source housing and the other on the chamber itself. The normal working
pressure of the instrument is between 3 and 6 x 10-7 torr for the main
chamber. The pressure in the source chamber varies widely depending on the
rate of sample introduction. Without any sample ﬂow, the pressure is
typically between 5 x 108 and 1 x 10'7 torr. The design and construction of

the vacuum chamber is described in more detail elsewhere [1].
2.2 Electronics

The electronics which supply power for various instrument
components were designed and constructed in-house by Mary Seeterlin and
Martin Rabb. High voltage supplies provide DC voltages for the instrument’s
ion optical elements. Most of these voltages are achieved through the use of a
course and ﬁne adjust on the front panel. The course consists of a switch
which allows a choice of 5 ranges. The ﬁne consists of a high performance
potentiometer which allows adjustment of the voltage within 2 volts of the
desired value. Most of the electrode voltages are achieved by the voltage
dividers between three power supplies, one of which is a positive supply,
usually adjusted to about 350 V. The others correspond to the ‘ﬁeld-free’
voltages of the ﬁrst and second mass analyzers (generally -550 and -2500 V.

Other instrument components which require voltages more negative than the

33
second ﬁeld-free, such as the einzel lens, or cause a signiﬁcant load, such as
the detectors, incorporate additional supplies.

A 200 V pulse is applied at a ﬁxed frequency to the source back plate
and represents the only dynamic voltage with the exception of the gate
circuitry which are described in more detail in chapter 4. The source pulse is
of adjustable width and amplitude, and the source pulser itself is mounted in
a box directly behind the source housing. In addition to the high voltage
supplies and the source pulsing circuitry, a low voltage, high current supply

provides power for the source ﬁlaments.
2.3 Sample introduction

Gaseous samples are introduced into the mass spectrometer by use of a
gas chromatograph (model 5890, Hewlett-Packard, Palo Alto, CA), which is
interfaced to the source of the tandem TOF by a heated transfer line
(Finnigan, San Jose, CA) mounted orthogonal to the source housing. A
column is placed into the transfer line with the end of the column just inside
the source region. The gaseous molecules are allowed to spray into the source
region, where electron impact ionization occurs. Inside the GC oven, the
column can either be attached to the heated GC injection port or to a glass
sample vial. The ﬁrst mode allows controlled injections of sample followed by
gas chromatographic separation. When the column is attached to the vial, a

continuous ﬂow of sample is introduced into the mass spectrometer,

34
generally without carrier gas, on the basis of pressure differential between
the vacuum chamber and the (heated) vial. This is useful when a sample is
required to be present for long periods of time, such as for instrument tuning

and studies on the laser PID spectroscopy of ions.

2.4 Source

 

Electron
Filament - focusing
/ element
F. I
Grid
Storage // two

volume \ g

/ . F
/ Ll \

Focusing
Grid one aperture

 

 

 

 

Backplate

Figure 2.1 The electron ionization storage source.

The ion source currently used in the TOF/TOF instrument is based on
a design introduced by Grix, et al. [3] and is illustrated in Figure 2.1. Two
opposing, electron-emitting, 0.018 cm diameter, rhenium ﬁlaments are

connected in series and positioned within electron focusing elements. The

35

electron focusing elements are held at a negative voltage with respect to the
body of the source and the ﬁlament, while the ﬁlament is held at about 70 V
more negative than the body of the source in order to direct the electrons
between the back-plate and grid one of the source with about 70 eV energy.
Sample is continuously introduced into the source where it is ionized in the
electron beam. Periodically, the back plate of the source is pulsed to
approximately 200 V, causing ions to be accelerated out of the source into an
acceleration ﬁeld formed by grid one and grid two.

Essential to the achievement of high sensitivity is the ability of the
source to accumulate a signiﬁcant fraction of the ions that are produced
between the extraction pulses. It is believed, as Studier postulated in
1963 [4], that ion loss is limited by the creation of a potential well between
the back-plate and grid one as a result of the negative space charge of the
dense, well-focused electron beam. The efﬁciency of ion storage has been
demonstrated and measured experimentally [1, 5]. At an extraction rate of
500 Hz, the maximum rate for the Questek excimer laser, storage is about
25% efﬁcient. In addition to the sensitivity improvement by ion
accumulation, the ion bunching caused by simultaneous extraction and
focusing results in a great improvement in signal-to-noise ratio. All ions of a
particular m/z value that are accumulated in the source between extractions
arrive at the detector within a few tens of nanoseconds. The peak detector

current is therefore several orders of magnitude greater than the average ion

36
current. The development, design, and construction of the source is described
in more detail elsewhere [1].

Source tuning is accomplished by iterative, empirical adjustment of
the voltages applied to each electron repeller, the source back plate (between
extractions), the ﬁrst grid, the focusing aperture between the ﬁrst and second
grid, and the bias voltage of the ﬁlament while monitoring peak intensity
and peak width. The voltage of the second grid is set at the ﬁrst ﬁeld-free
voltage and is not adjusted. The ﬁlament current is generally held at about

3.0 A
2.5 First mass analyzer

About 21 cm beyond the source, an einzel lens is employed to collimate
the ion beam produced in the source. The voltage on the einzel lens is
adjusted empirically based on ion signal intensity. In addition, a set of
orthogonally oriented pairs of steering plates just beyond the einzel lens are
used for minor steering of the ion beam into the ﬁrst reﬂectron, also
empirically tuned to achieve mammum ion signal at a detector placed at the
focal point of the reﬂectron. The potential of the ﬁrst mass analyzer is held at
-550 V. To accomplish this, shields were constructed from stainless steel
mesh. These are isolated physically (and electrically) from the inner surfaces

of the source chamber using ceramic spacers.

37

A nearly linear acceleration ﬁeld is established during the source
extraction pulse. The potential dr0p from the center of the source to the ﬁeld-
ﬁ'ee region is 650 V. Due to the spatial distribution of ions in the source, a
kinetic energy distribution is created since ions at the back of the source
receive more energy (and therefore higher velocity) than ions of the same
mass near the front of the source. At some point in the ﬂight tube, a minimal
ﬂight time distribution is created at a position called the space-focus plane
where ions originating near the back of the source overtake ions created near
the front. In addition to the spatial distribution in the source, there is an
energy distribution which degrades resolution in linear TOF. Wiley and
McLaren demonstrated resolution improvements in linear TOF by using an
adjustable two-stage acceleration to space-focus ions at the detector [6].

Mamyrin, et al. [7] demonstrated a method of improving resolution
using a reﬂectron. In this approach, ions are accelerated using a steep
extraction ﬁeld, resulting in a space-focus plane near the source. As iso-mass
ion packets travel toward the reﬂectron, the ions with more energy and
higher velocity lead those of lower energy. The reﬂectron employed by
Mamyrin consisted of a retarding ﬁeld and a reﬂecting ﬁeld formed by two
grids and a back-plate. Ions of higher energy penetrate the reﬂecting ﬁeld to
a greater depth than iso-mass ions of lower energy and spend more time in
the reﬂectron. Upon reﬂection, higher energy ions are again 'behind' lower

energy ions. Eventually an image of the distribution at the space-focus plane

38
is created at the detector. The reﬂectron allows an extended effective ﬂight
length, providing more time for iso-mass ion packets to become separate from
one another without a corresponding spread in the iso-mass ion packet width
as observed in linear TOF.

Wollnik, et al. [8] further improved upon the reﬂectron TOF
instrument by using grid-free reﬂectrons. These devices do not suﬁ’er from
ion transmission losses due to collisions with the grid material or loss of
resolution associated with the ﬁeld perturbations in the vicinity of the wire
mesh [9]; and they also serve to focus the ion packets radially. The ﬁrst stage
of mass analysis in the tandem TOF instrument was designed and
constructed using a combination of the above technologies. Upon placing a
detector at the interaction region of the TOF/TOF, a mass resolution of 1500
(FWHM) is routinely observed [1]. This effective bunching of ions at the
interaction region is essential to achieving maximum efﬁciency in the
tandem TOF mass spectrometer.

The ﬁrst reﬂectron consists of 9 stainless steel electrodes and a
stainless steel back-plate. Each electrode is a 0.051 cm thick stainless steel
ring with an outer diameter of 10.2 cm and an inner diameter of 6.35 cm. The
electrodes are separated from each other by 1.05 cm ceramic spacers for a
total reﬂectron length of 10.5 cm. Potentials are supplied to four of the nine
electrodes, with 5 MO resistors linking the others in series. The reﬂectron is

attached to the instrument chamber by three posts, two of which are

39
connected to external motion feedthroughs (MDC Vacuum Products,
Hayward, CA) which allow the direction of the ads of the reﬂectron to be

adjusted. Tuning is accomplished by an empirical, iterative process.
2.6 Instrument control and data collection system

The timing sequence of the experiment is started by a square wave
pulse from a function generator (Datapulse, Culver City, CA). This wave
simultaneously triggers the pulser for the source back-plate, a delay
generator (model 4222, LeCroy, Chestnut Ridge, NY), and a digital storage
oscilloscope (model 9450, LeCroy, Chestnut Ridge, NY). The four-channel
delay generator is controlled through a General Purpose Interface Bus
(GPIB) connection (National Instruments, Austin TX) by a 486/33 computer
(Zenith, St. Joseph, Ml) running LabWindows® (National Instruments,
Austin, TX). A LabWindows® instrument control program was written [2],
which allows the delay on each channel to be independently set in 1 ns
increments. The delay generator provides a 5 V pulse with a rise time of less
than 1 ns and pulse width of 100 ns to both the gate pulsers and the external
trigger of the excimer laser. The gate pulser delay is set such that the ion
packet or range of ion packets desired is allowed to pass, while others are
deﬂected. The laser discharge pulse is timed such that the photons arrive at
the interaction point at the same time as the ion packet of interest. The

detector anode is connected to the input of the digital storage oscilloscope

40
through a terminated 50 Q BNC cable, allowing signal transients to be
captured, averaged, and stored. A GPIB connection between the oscilloscope
and the computer allows the resulting averaged spectrum to be captured and
stored to disk, using the LabWindows® control program. More detail on the
gate pulser control system is provided in chapter 4 below, while the
instrument control and data collection systems in general are described in

more detail elsewhere [2].
2.7 Detectors

Detection of the ions at the end of the second mass-analyzer is
achieved through the use of a dual microchannel, chevron-type detector
(modiﬁed model TOF-2003, Galileo, Sturbridge, MA). Two 40 mm diameter
microchannel plates (MCPs) are used to provide a gain of approximately 107.
A conical anode is contained in a shield, tapering to a BN C vacuum
feedthrough such that the entire anode and following cable are matched
at 50 Q.

A second, removable detector with 25 mm microchannel plates has
been constructed and mounted on a sliding carrier such that it can be placed
anywhere along the optical rail beyond the interaction region. This allows
detection and veriﬁcation of ions after the interaction region components, ion
focusing at the interaction region, and crude spectra to be collected just

beyond the interaction region in order to test the photodissociation process.

41
This simple detector does not have an impedance-matched anode, and

signiﬁcant ‘ringing’ is observed after the individual mass peaks.
2.8 References

1 Seeterlin, M. A. Ph. D. Thesis, Michigan State University, 1993.
2 McLane, R. D. Ph. D. Thesis, Michigan State University, 1993.

3 Grix R.; Gruner, U.; Li, G.; Stroh, H.; Wollnik, H. Int. J Mass
Spectrom. Ion Processes 1989, 93, 323.

4 Studier, M. H. Rev. Sci. Instrum., 1963, 34(12), 1367.

5 Yefchak, G. E.; Puzycki, M. A; Allison, J.; Enke, C. G.; Grix, R.;
Holland, J. F.; Li, G.; Wang, Y.; Wollnik, H. Proceedings of the 38th
ASMS Conference on Mass Spectrometry and Allied Topics, Tuscon,
Arizona, June 3-8, 1990, 540.

6 Wiley, W. C.; McLaren, I. H. Rev. Sci. Instrum. 1955, 26(12), 1150.

7 Mamyrin, B. A; Karataev, V. I.; Shmikk, D. V.; Zagulin, V. A. Sov.
Phys-JET?J 1973, 37(1), 45.

8 Wollnik, H.; Przewloka, M. Int. J Mass Spectrom. Ion Processes 1990,
96, 267.

9 Bergmann, T.; Martin, T. P.; Schaber, H. Rev. Sci. Instrum. 1989,
60(3), 347.

Chapter 3: Precursor selection and fragmentation

3.1 Photo-induced dissociation as a fragmentation method

As described in chapter 1, photo-induced dissociation is not routinely
used as a fragmentation method for tandem mass spectrometry. However,
PID is uniquely suited to the fragmentation needs of our tandem time-of-
ﬂight mass spectrometer for several reasons. With this instrument, iso-mass
ion packets are focused to temporal and spatial dimensions that are very
similar to those of our focused laser pulse. The resulting high degree of
spatial and temporal overlap provides optimal PID efﬁciency despite the
relatively low cross-sections for the PID process. Simultaneously, PID yields
high resolution selection of the precursor ion packet of interest. As opposed to
collision-induced dissociation (CID) and surface-induced dissociation (SID),
the product ions formed by PID do not have the associated additional kinetic
energy distribution acquired ﬁ‘om a collisional process. The amount of energy
introduced to precursor ions is precisely known when PID is used, while CID
and SID are characterized by broad distributions of deposited energy [1]. In
addition, there is no gas load introduced to the mass spectrometer by the PID

process as there is when CID is used.

42

43

Photo-induced dissociation has been widely observed, predominantly
in conjunction with photon ionization, but also as a process separate from the
ionization event. A brief history of the use of PID in mass spectrometry is
given in chapter 1. The wide-spread adoption of PID as a fragmentation
technique in the future will depend greatly on the type of instrumentation
which is in routine use. Since the scanning mass spectrometers prevalent
today do not generally operate in pulsed mode, they cannot exhibit the
performance advantage with the pulsed laser PID that is described below for
the tandem TOF instrument. While the trapping instruments may be
conﬁgured to facilitate high efﬁciency PID by pulsed laser, they are not
capable of mass spectral acquisition rates comparable to the repetition rates
of modern excimer lasers (500 Hz or more). To date, the only system which
accomplishes this level of efﬁciency for photo-fragmentation while collecting
product spectra at a rate compatible with the GC time scale is the tandem

TOF instrument presented here.
3.2 The laser, optics, and light tube

The laser used for the studies described in this thesis is a Questek
(Lambda-Physik, Acton, MA) 2580vB excimer. This laser is capable of
producing radiation of several wavelengths depending on the gas mixture
used. All photodissociation data presented in this thesis were collected using

the ArF line at 193 nm (6.4 eV per photon). Operating at 20 Hz, this laser

44
produces approximately 150 mJ per pulse. The pulse width of the laser is
approximately 17 ns.

A cylindrical plano-convex fused silica (Dynasil 1100) lens (Newport,
Irvine, CA) of 300 mm focal length is used to focus the laser beam to an area
approximately 2 cm high by 1 mm wide at the interaction region inside the
mass spectrometer. Two fused silica interferometer ﬂats (Newport,
Irvine, CA) are positioned on ﬂanges mounted to the outside of the mass
spectrometer in order to allow photons to pass into the instrument and exit
the other side where they are collected in a beam dump device.

Since the laser beam is of sufﬁcient power to ablate metal, holes have
been cut in the shield material to create an unimpeded path for the laser
beam. To prevent ﬁeld leakage through these holes, a tube has been
constructed and placed through the holes such that the focused laser beam
can pass through the tube without interacting with the shields or any other
metal component inside the instrument. This tube is held at the potential of
the ﬁrst ﬁeld-free region (—550 V). Since the ‘light tube’ spans both ﬁeld-free
regions, a second, shorter tube is slid over the ﬁrst tube which spans the
width of the second ﬁeld-free region. This tube is held at the same voltage as
the second ﬁeld-free region shield (—2500 V). The two tubes are electrically
isolated from each other by a thin sheet of Kapton® (E.I. du Pont de
Nemours, Wilmington, DE). Two 22 mm high by 34 mm wide holes have been

cut out of opposite sides of the light tube where it intersects the ion ﬂight

45
path. Ions traveling from the ion source to reﬂectron 1 pass over the light
tube, ions traveling from reﬂectron 1 to reﬂectron 2 pass through the light
tube, and ions traveling from reﬂectron 2 to the detector pass under the light

tube.
3.3 Ion/photon overlap

For precursor ion selection we rely on the laser pulse interacting with
exactly one precursor m/z value. Two factors affect our ability to accomplish
unit resolution for selection over the m/z range of interest. The ﬁrst is the
precision and accuracy with which we are able to time the laser pulse with
respect to the ion packet. The instrument control system which accomplishes
the precision timing of the laser pulse is brieﬂy described in chapter 2 and is
described in detail elsewhere [2]. Laser delay time can be controlled in 1 ns
increments, and the jitter in the laser thyratron discharge is less than 2 ns.
The second factor is the relative dimensions of the ion packet and the laser
pulse and their overlap over the period of the laser pulse in the interaction

region.
3.3.1 Calculated overlap

In order to predict the resolution for precursor selection that is possible
with our system, a calculation of the interaction of the laser pulse and the ion
packet is required. The laser pulse is focused to a width of apprommately

1 mm at the interaction region. It is about 17 ns in duration. The mass

46

resolution of the ions at the interaction region is about 1500 (t/2At deﬁnition,
full width at half maximum). The kinetic energy of the ion beam is
approxdmately 650 eV. The effective ﬂight distance of the ﬁrst mass analyzer
is 2.4 m. Given these values, one can easily determine that any iso-mass ion
packet (regardless of its m /2 value) is about 0.8 mm in length (FWHM) at the
interaction region. Next, one needs to consider two things: ﬁrst, adjacent
m/z values are closer together in space (and time) as the m/z value
increases; and, second, the ion packet moves a signiﬁcant distance during the
17 ns laser pulse. Lower m/z value ions travel farther during the 17 ns laser
pulse than higher m/z value ions. The limit of unit resolution for selection
occurs when the separation between adjacent ions (measured from peak
center to peak center) is equal to one half the sum of the laser pulse width at
the interaction region, the iso-mass ion packet width, and the distance the
iso-mass ion packet moves during the laser pulse. Using this formula, one
can determine that the upper m/z limit for unit resolution of selection
(FWHM) occurs at around m/z 1200. For GC applications, this mass range is
more than adequate. Focusing the laser beam to less than 1 mm in width or
improving precursor separation would allow greater resolution for precursor
selection.

Figure 3.1a illustrates the interaction of a laser pulse with a m/z 157

ion packet. If the laser pulse is timed such that the photons at the midpoint

47

 

 

m/z 158

m/z 157

m/z 156

 

 

 

 

,.aut...re

 

 

 

 

 

10 ll 12 l3 l4 15 16 17 18 19

 

 

«i-
.3-

zummnouau ciao—cm

Ab-

20

23456789

Position (mm)

b)

 

 

m/z 1200 m/z 1201

111/: 1199

 

 

 

 

p

 

 

 

 

p p p
_ a 1 «

3%.:an 9,533—

.0-

15

14

13

 

Position (mm)

Selection of precursor iso-mass ion packet by laser light at

a) m/z 157, and b) m/z 1200. (note different scales).

Figure 3.1

48
of the laser pulse reach the interaction region at the same time as the
midpoint of the ion packet, approximately 97% of the selected ion packet is
overlapped by some portion of the laser pulse, with no overlap or interference
of neighboring m/z ion packets. The laser pulse is indicated by the shaded
rectangle. The solid line indicates the position of the ion packet at the
beginning of the laser pulse, while the dotted line indicates the position of
the ion packet at the end of the laser pulse (17 ns later). At m/z 1200, as
depicted in Figure 3.1b, 92% of the precursor ion packet is overlapped by
some portion of the laser light. However, 11% of each of the adjacent m/z
values is overlapped by some portion of the laser light as well. (Note that
peak width was measured at half maximum rather than at baseline in the
above calculation.) This model assumes Gaussian ion peak shapes and makes
no compensation for the varying intensity of the laser light across its width
or ﬁom the front of the laser pulse to the back. Since the excimer laser pulse
is expected to have ‘hot spots’ in the beam cross-section and to have a
maximum intensity in the temporal center of the pulse, these assumptions
are not strictly valid. The assumptions made about the laser pulse
characteristics will tend to result in a conservative estimate for the limit of

unit resolution for precursor ion selection.

49
3.3.2 Observed overlap

Figure 3.2A is a spectrum of the molecular ion region of bromobenzene.
Ions in this spectrum have traveled through the entire instrument to the
normal detector position. With the laser on and the delay time between
source extraction and laser ﬁring set such that the laser pulse arrives at the
interaction region at the same time as the precursor ions, interaction
between the ions and photons occurs. Figures 3.2B-3.2D are the spectra
obtained when the laser delay times are set at 86.40 as, 86.70 as and 86.97
us to coincide with the ion packet arrival times for m/z values 156, 157 and
158, successively. The depletions associated with each of these times, being
limited to the selected ion, demonstrate the eﬁ‘ectiveness of precursor
selection by laser timing adjustment. Calibration of delay times is easily
accomplished due to the linear dependence of ﬂight time on the square root of
m/z. The ion arrival time jitter and laser thyratron discharge jitter were both
measured to be less than 2 ns. Once the instrument is calibrated, selection of
a new precursor ion simply involves changing the laser delay time to the
calculated value for the desired precursor m/z value. In addition, the day to
day arrival time drift is less than 100 ns. Recalibration requires only 3-4
attempted laser delays. The bromobenzene example demonstrates that we

are able to select only one precursor m/z value with the laser, as is expected

5O

m/z 156 m/z 158

0 .0 4 -- a.
l l

l A

O

O

CO
%

O

O

N
1

\ m/z 157

 

wwutﬁw" ”TL*“‘“22

Intensity (V)

\
f/ \\ MW

.1 .3 ., M f
O “ Aggy-"'T'X’ I W’v - - 4L. -A._J/ A “I“ A .4 .L “f" l ‘ M‘” a M.” M
j T v v r Y Y Y

 

 

 

0 .0 2 --
/
0 .0 1 i» 7 y /
.1 I \
‘ Max: \M

. / “ \m ,l’m"\

MR_ 4...!" ‘9.

140.9 141.3 141.7 142.1 142.5

Flight Time (us)

Figure 3.2 Precursor selection by pulsed-laser photodissociation.

51
based on the calculations outlined above. Precursor ion areas have been

depleted by 55.2%, 56.3%, and 51.8% for m/z 156, 157, and 158, respectively.
3.4 Photo-fragmentation efficiency

Besides providing high resolution of selection, focusing the laser pulse
and iso-mass ion packet to a small volume results in high photodissociation
eﬁciencies. The photodissociation efﬁciency can be expressed by to, where d)
is the photon density and o is the photodissociation cross-section. When the
excimer laser, operating with ArF (193 nm), provides 150 mJ of energy per
pulse focused to a 2 cm by 1 mm area, the photon density is 7.3 x 1018
photons/cm? Given that PID cross-sections for small organic ions are on the
order of 10'17 to 10'19 cm2 [1, 3], photodissociation efﬁciencies near 100% are
expected for some ions in this instrument.

The efﬁciency of the PID process is determined by the eﬂ'iciency of
fragmentation and the efﬁciency of collection of the ions. We can let P and Po
represent the precursor ion current with and without laser excitation,
respectively, and BF. the total current of all fragment ions. The
ﬁagmentation efﬁciency is Er = 2F: / (P + XFi), the collection eﬁciency is
Ec = (P + SE) / Po, and the overall PID efﬁciency is the product
Epm = Er x Ec [4]. To perform these efﬁciency studies, a detector was placed

approximately 20 cm beyond the interaction region, with the acceleration

52
region in place as described in chapter 1. This arrangement allows us to
study the PID process independent of the effects of the second reﬂectron.

Using this conﬁguration, the product spectrum for the molecular ion
(m/z = 156) of bromobenzene was collected (Figure 3.3). The precursor ions
have been separated by their m/z-dependent velocities and focused at the
detector. This product spectrum was obtained by subtracting the laser 03
spectrum from the laser on spectrum. Ion peaks at t =93.60 us, 93.39 us,
92.95 us, 92.50 us and 91.50 as correspond to products ions with 6, 4, 3, 2,
and 1 carbon atoms, respectively. The calculated fragmentation efﬁciency,
collection efﬁciency and overall PID efﬁciency are 56%, l 16%, and 65%,
respectively. The fragments observed in this product spectrum are similar to
those observed in a normal 70 eV electron impact mass spectrum. Compared
to the E1 spectrum, the relative intensities of the lower m/z fragment ions
are enhanced in this PID product spectrum.

Product spectra for precursor ions from bromobenzene, toluene,
nitrobenzene, acetophenone, triethylamine, N ,N-diethylformamide,
N -methylacetamide, glycine anhydride, and cyclohexene were collected. The
compounds and precursor ions chosen demonstrate photo-fragmentation of
odd-electron, even-electron, aromatic and non-aromatic species. Efﬁciencies
are given in Table 1. In most of these studies, reﬂectron one was tuned to

energy-focus at the detector rather than at the interaction region to improve

53

 

 

 

 

 

 

 

 

 

 

 

 

0.015 .-
001 + C4
E 0.005 -
B:
’8 .
,3 0 l}
s i
:1 y
i-t
-0005 «~
.001 .. precursor
depletion
-0.015 4. + t ¥n
91.25 92.25 93.25 94.25 95.25

Flight time (microseconds)

Figure 3.3 Product spectrum for m/z 156 (molecular ion) of

bromobenzene.

54

ace—Same ma? $2: mo 5.885.... 538:8 E.

god—ovum“. mam..— aw MO Vmwumﬂﬂ GOROH GOBUGHQH—hm On: am wwmﬂoom Ohm? mac“ Homuﬁoohnmk

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

: a m a 03 no ocoxoaoﬁomol
385.358
ms. .52 mm 2: a: 5:53“
253882».
hm em mm ca mp -352? Z
2.ngqu
3.. 3 3 2: 2 22552.2
mm m: ea c: ww mam—mseors
mm 2: en 03 m3 massage]
2w 5: mm O: «NH oncncmnoaﬁ:
3. me mm 6: Ha 2833
on 2 H 3. :2 1E.
2: am: an cm: 99.:
mm mm cm 3: S.
mm m: cm on“ $4 econconoEoS
:5 Q: G.; :25
monomoam Accomoam monomoem— zuacam «\E
GE deﬁes—30 non-5:039; 0.5.15 home..— no— homesccum 6:25:30

 

2:: 53:83 me 3.3.2; a new moﬁqomommo OE ﬁn 935,—.

 

55

product ion resolution. This results in reduced PID efﬁciencies. ‘When
reﬂectron 2 is used in the ﬁnal instrument conﬁguration to focus the product
ions, the optimal focus will be achieved at both the interaction region and at
the detector. To determine the effect of improved focusing on PID efﬁciency,
two precursor ions (indicated by asterisks in Table 1) were studied with the
focus at the interaction region. As hypothesized, PID efﬁciencies were
increased. Collection efﬁciencies of greater than 100% were sometimes
observed. This is believed to be the result of greater detector gain for ions
with higher velocity (lower m/z) [5].

The overall PID efﬁciencies of between 11% and 100% are comparable
to the high eﬁciencies achievable in the CID process in a triple quadrupole
instrument (65% for n-hexane) [4]. These results give evidence that PID can
be an analytically useful fragmentation technique when implemented in a

system with high ion/photon overlap.
3.5 References

1 Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry / Mass
Spectrometry: Techniques and Applications of Tandem Mass
Spectrometry, VCH Publishers, Inc.: New York, 1988.

2 McLane, R. D. Ph. D. Thesis, Michigan State University, 1993.

3 Dunbar, R. C.; In Gas Phase Ion Chemistry, Bowers, M. T., Ed.
Academic: New York, 1979, Volume 2.

4 Yost, R. A; Enke, C. G.; McGilvery, D. C.; Smith, D.; Morrison, J. D.
Int. J. Mass Spectrom. Ion Physics 1979, 30, 127-136.

5 Meier, R. ; Eberhardt, P. Int. J. Mass Spectrom. Ion Processes 1993,
123, 19-27.

Chapter 4: Precursor isolation

4.1 Elimination of interference

As demonstrated in chapter 3, when pulsed-laser photodissociation
(PID) is used as the fragmentation method, it is possible to interact the laser
pulse with a single m/z value ion packet, thereby accomplishing selection of
the precursor ion packet from an ion beam in TOF mass spectrometry [1].
However, the resulting product spectrum may overlap the portion of the
normal ion spectrum having lower m/z values than the laser-selected
precursor as well as (potentially) product ion spectra due to metastable
decomposition of other m/z values [2, 3]. Figure 4.1a is a convoluted
spectrum of toluene containing the normal spectrum of toluene and the
product spectrum of m/z 91 of toluene. This interference can be effectively
eliminated by subtracting a ‘laser oﬁ’ spectrum from a ‘laser on’ spectrum.
Such a difference spectrum is shown in Figure 4.1b, where the peaks due to
the normal spectrum have been subtracted away and the product spectrum
can be clearly seen. These product spectra were collected using a reﬂectron
TOF analyzer which is described in chapter 5, accounting for the much
improved mass resolution as compared to Figure 3.3. The precursor ion

signal appears as a negative peak due to the subtraction.

56

a)

0.012 “r
0.01 «-
0.008 T
0.006 T

0.004 +

Intensity (V)

0.002 a

 

-0.002
78

b)

0.012 ~~

0.01 *r
0.008 ~-
0.006 ..

0.004 +

Intensity (V)

0.002 1*

01

 

-0.002
78

Figure 4.1

 

57

65*

91+

65

63
89

39*

63*

 

 

 

 

 

 

 

 

 

1 L 1 1 1 1
1 i T Y i 1 r T 1 T

80 82 84 86 88 90 92 94 96 98 100 102

Time (microseconds)

65*

39* precursor
depletion

63‘“

41*

 

 

1 J l L I I i
l 7 I I l I T

80 82 84 86 88 90 92 94 96 98 100 102

T

J
1

J»

Time (microseconds)

Product spectra of m/ z 91 of toluene a) with convoluted
normal spectrum, and b) with normal spectrum subtracted.

(* indicates product ion peak).

58

The subtraction of spectra requires that the ‘laser oﬂ’ spectrum be
collected very near in time to the ‘laser on’ spectrum in order to avoid peak
intensity ﬂuctuations associated with the changing sample pressure in the
source region. This could be accomplished by operating at twice the repetition
rate of the laser and alternately collecting product and normal spectra.
However, this scheme introduces several diﬁiculties that would increase the
complexity of the data collection system because alternate transients would
have to be collected in alternate channels for summing in real-time. Also,
either the subtraction of spectra would have to occur in real-time or else
twice the usual number of Spectra would have to be collected.

A more eﬁ'ective approach is to isolate the selected precursor ion
packet prior to fragmentation by elimination of all others. This can be
accomplished through the use of an ion gate, which deﬂects unwanted ions
on a trajectory which will not result in collision with the detector. In this
mode, the product spectrum can be collected without normal spectrum
interference or interference from metastable decomposition of other m/z
values as long as unit resolution for isolation is maintained. Ultimately,
precursor selection is accomplished through the overlap of the laser pulse
and the ion packet of interest as described in chapter 3. This method of
precursor isolation has the added beneﬁt of reducing chemical noise which is

often limiting in mass spectrometry [4].

59

4.2 Other applications of ion gates

4.2. 1 Precursor selection

When collision-induced dissociation (CID) [5, 6] or surface-induced
dissociation (SID) [7] is used as the activation method for tandem TOF mass
spectrometry, an ion gate must be used to select a particular m/z packet for
dissociation with rejection of all others. Without m/z selection, unlike the
case for pulsed-laser PID, many m/z value ions may be fragmented, and the
resulting spectra are convolutions of multiple, overlapped product and
normal spectra. A pulsed gas valve has been used to introduce collision gas
for CID in a tandem TOF instrument [6], but the duration of the burst in the

interaction region (about 200 as) is too long to accomplish selection.
4.2.2 Preventing detector saturation

Elimination of low-mass ions or ions of little interest may be important
to avoid detector saturation which can limit detector response to ions of
interest. Examples of this include elimination of ions formed from the gas
chromatography (GC) carrier gas for GC/MS applications and elimination of
matrix ions in matrix-assisted laser desorption / ionization (MALDI) mass
spectrometry. Another use of gate devices for TOF instruments is the

temporal encoding of continuous ion beams [8].

60

4.3 Gating methods

Two deﬂection devices have been described in the literature for ion
beam modulation and/0r gating purposes in TOF mass spectrometry. These
devices are deﬂection plates and interleaved comb ion deﬂection systems,
both of which will be described and compared below. Three distinct methods
for ion beam modulation have been identiﬁed [8] for use with deﬂection plate
gating devices. The work described below is concerned only with the normal
‘gate mode’ of operation, and the other modes of operation are not considered.
However, the advantages of the interleaved comb ion deﬂection gate device
over deﬂection plate devices as outlined below should apply to all three
modes of operation.

In order to achieve m/z selection in TOF mass spectrometry, the gate
device is held in its closed or deﬂecting state until the iso-mass ion packet of
interest is about to reach the deﬂecting ﬁeld created by the gate device. At
this time the gate is electronically switched to its open or non-deﬂecting
state, the ion packet passes through the volume that the deﬂecting ﬁeld

previously occupied, and then the gate is switched to its closed state again.
4.3.1 Deﬂection plates

Early devices for ion beam modulation consist of parallel plates
between which the ion beam passes. The potential of one or both of the plates

are modulated to create, alternately, a deﬂecting ﬁeld and a non-deﬂecting

61
ﬁeld in order to achieve selection of the desired portion of the ion beam.
Descriptions and examples of the use of this type of ion gate in TOF mass

spectrometry are prevalent in the literature [5-15].
4.3.2 Interleaved comb ion deﬂection gate

The interleaved comb ion deﬂection gate consists of a plane of closely
spaced parallel wire segments. Alternate wire segments are connected,
creating two distinct wire sets which are electrically isolated from one
another. An ion beam traveling orthogonal to the plane of wire segments is
unaffected by the device (excepting collisional interactions) when the wire
sets are held at the potential of the ﬁeld-free ﬂight path. However, when the
wire sets are held at opposite potentials of equal magnitude with respect to
the ﬁeld-free potential, approximately half of the ions are deﬂected to one
side and the rest of the ions to the other side, depending on the polarity of
the ﬁeld which the particular ion passes through.

This interleaved comb device was ﬁrst proposed by L. B. Loeb to
H. F. Lusk for use as an electron ﬁlter [16]. Lusk’s preliminary studies
showed that if a high frequency RF potential was applied between the
alternating wire sets, that electrons should be eliminated from an ion beam
with little effect on the slower ions. Use of this device as an electron ﬁlter
was demonstrated by A. M. Cravath [17] and later by N. E. Bradbury and
R. A. Nielsen [18]. The device was ﬁrst used for m/z selection in the ion gate

mode for TOF mass spectrometry by R. Weinkauf, et al. [2]. Several groups,

62
including this one, are using the interleaved comb ion deﬂection gate for m / 2

selection in TOF mass spectrometry [2, 3, 19-2 1].
4.3.3 Comparison

The inherent advantage of the interleaved comb ion deﬂection gate
over the deﬂection plate gate is related to the physical dimensions of the ﬁeld
which is produced by this device in the closed (or deﬂecting) state. In TOF
mass spectrometry, consecutive iso-mass ion packets of increasingly higher
m/z values are closer together in time and space. Decreasing the length of
the deﬂection ﬁeld without reducing the angle of deﬂection is therefore very
important in improving resolution of selection. Using SIMION [22], a
computer program which predicts electric ﬁeld shapes and ion motion, the
interleaved comb ion deﬂection gate was compared with deﬂection plates.
Both devices were conﬁgured to accomplish an angle of deﬂection of
about 10°. With wire spacing of 1 mm and i 250 V applied to alternate wire
segments (relative to the ﬁeld-free potential) in the interleaved comb ion
deﬂection gate, the ﬁeld produced extends about 1.1 mm in both directions
from the device center. The limit of ﬁeld effect was deﬁned as the i 5 V equi-
potential contour lines. Equi-potential contours (at i 5 V, i 50 V, and
i 100 V) and ion trajectories through the device as modeled in SIMION are

shown in Figure 4.2a.

63

a)

 

. C Q ’7
+ 250 V ere\_\ >3“:
3:2?“ + 5 V contour

,1

- 250 V wir L. f) \ - 5 V contour Ion trajectories

 
 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1!)
//,//// <\«.\ /F‘/§\\\ + 5 V contour /
/ L» \d/ lv/ \
K v/ \_\ / Field-free
\\ /~’\ + 170 V plate shield
:ﬂ /

170 V plate Ion trajectories

\ the.

 
 

 

Figure 4.2 SIMION models showing equi-potential contours and ion

trajectories for a) the interleaved comb ion deﬂection gate,

and b) a deﬂection plate gate (note different scales).

64

For a set of deﬂection plates 15 mm wide and 25 mm apart with
i 170 V applied, the ﬁeld extends at least 18 mm in both directions from the
device center, again using the i 5 V equi-potential contours as the limit of
ﬁeld eﬂ‘ect. Equi-potential contours (at i 5 V, i 50 V, and i 100 V) and ion
trajectories through the device as modeled in SIMION are shown in
Figure 4%. The model presented here depicts a limiting case of placing the
shield electrodes (at ﬁeld-ﬂee potential) very near to the deﬂection plates. As
the distance between the shield electrodes and the deﬂection plates
increases, the ﬁeld produced by the deﬂection plates extends even farther in
both directions.

The large difference in the length of the deﬂection ﬁeld for the two
devices can be explained by two factors. First, the large ﬁeld strength
produced by the closely spaced wire segments results in efﬁcient deﬂection
for the interleaved comb device. Also, the ﬁeld produced by this device is
canceled as the distance from the wires increases due to the equal and
opposite potentials on the adjacent wire segments. It is crucial that the wires

are perfectly co-planar for effective cancellation of the ﬁeld.
4.4 Instrumentation

Little description has previously been provided in the literature
concerning interleaved comb ion deﬂection gate devices. Here a detailed

account of the device itself and suitable electronics which achieve the desired

65
pulse shapes for the gate mode of m/z selection are presented, similar to the

information presented in reference [23].
4.4.1 Interleaved comb ion deﬂection gate device

A drawing of the interleaved comb ion deﬂection gate constructed in
our laboratory is shown in Figure 4.3. Two continuous, stainless-steel,
0.076 mm diameter wires (California Fine Wire, Grover City, CA) form a
plane of 26 wire segments as shown in Figure 4.3b (front view). Two Vespel®
(El. du Pont de Nemours, Wilmington, DE) blocks (A & B) are held apart by
#6-32 threaded rods (C & D) inserted in holes in the Vespel blocks with nuts
at both ends. The tension on the wires serves to hold the blocks together on
the threaded rods. As can be seen in Figure 4.3a (top view), a row of 0.38 mm
holes are drilled in the Vespel block to hold the wire segments 1 mm apart.
Since the two wires must be electrically isolated, each wire is wound around
a different #3-48 threaded rod (E & F) suspended above the holes on either
side of the row of holes. A congruent assembly is employed at the bottom of
the gate. The ends of the wires are anchored by pinching them between two
nuts and two washers on the suspended rods (E & F). Electrical contact is
made to the wires through threaded rods E & F. The effective area of the gate
is approximately 25 mm by 50 mm.

It was observed that the wire segments did not form a perfect plane

due to the effect of one wire being wound around threaded rod E, to the back,

66

a) Top View

fail/1i

 

b) Front View

 

n u g '
WWW : ! WMWQ‘W/[Mﬂ’ll/f/{WI’VfV/Wll/W/lll

/.\/1

   
 
  

 

 

 

 

 

   

\

\WW

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.3 The interleaved comb ion deﬂection gate device.

67

and the other wire being wound around threaded rod F, to the front. The
result was that alternate wire segments arced out of the plane in opposite
directions very slightly. The ﬁeld produced by this imperfect device was not
canceled effectively as the distance from the wires increased. In fact, a net
positive ﬁeld was generated on one side of the gate and a net negative ﬁeld
was generated on the other. The ﬁelds produced by a planar gate device and
a non-planar gate device are compared in Figure 4.4 as modeled in SIMION.
Experimentally, the imperfect device caused an arrival-time shift in the
selected iso-mass ion packet. Depending on which wire set received the
positive deﬂecting voltage and which the negative, the ion packet was either
accelerated by the negative and positive sloping ﬁelds before and after the
gate, respectively, or decelerated by the positive and negative sloping ﬁelds
before and after the gate, respectively. The direction of the arrival-time shift
was reversed by reversing the deﬂecting voltages supplied to the two wires in
the gate.

To eliminate this problem, 2.6 mm diameter ceramic rods (G & H) were
inserted in front of threaded rods C & D, but behind the plane of wires, so
that all of the wire segments are in the same direction and form a plane. The
ceramics are effectively held in place by the tension of the wires and the
threads on rods C & D (see Figure 4.3). After the ceramic rods were inserted,

no arrival-time shift was observed.

a)

 

    
  

ﬂ /’7\

/ I .f\

/ l. "‘v”/’

’ A
\.

l
+ 100 V \

Field-free shield + 5 V /m
+ 250 V Wire\ &@

+50Vm

 

,/
,

‘ ,
‘v

V/ /

\ia/ - 250 VWire

Kﬁ-wv

/.=-=<\

// 1”;\\\ \
,. , \
1 (5.,» .> ) - 5 V
\. ~- / ,//
/:“f <\/

( 1Z'('\/l\ _ 100 V

\ x

1))

 

+250 V wire

/

Field-free shield

 

 

Figure 4.4 SIMION models showing equi-potential contours for a) the

planar gate device, and b) the non-planar gate device.

69
Tension on the wires is adjusted by rotating the nuts on rods C & D,
sliding the ceramic rods G & H up or down, or adjusting the position at which
the ends of the wires are anchored. To achieve uniform tension across the

plane, the individual wire segments are manually massaged.

4. 4.2 Electronics

4. 4. 2. 1 Description

Figure 4.5 is an overview of the electronics built for gate control. A
TTL level trigger from the ion source extraction rate time base is routed
through a LeCroy 4222 programmable delay generator (PDG) in a CAMAC’
frame which is interfaced to the host computer using a general purpose
interface bus (GPIB). Delayed trigger pulses from the PDG range from 170 us
to 16.7 ms with lns resolution. This delayed pulse is used to trigger a
programmable pulse width generator (PPWG) circuit, which is described in
detail elsewhere [23]. The PPWG circuit produces complementary pulses of
variable width from 11 us to 2561 ns in 10 ns steps. Due to unequal
propagation delays throughout the system, an edge alignment circuit was
built into the PPWG circuit. The pulse width is speciﬁed in the
Lab Windows® (National Instruments, Austin, TX) instrument control
program as an 8-bit word which is applied to the PPWG through a National

Instruments model AT-MIO- 16F-5 board.

Programmable Delay
Generator

 

#1:. Tower P3 _...

 

 

 

 

km 4222 PDG

 

8-blt mrd from
AT-MlO-16F-5

Figure 4.5

Programmable Pulse
Width Generator

 

~—H—Tngger Q I.

70

 

 

 

 

 

 

 

 

 

5

5

 

 

 

 

 

 

pulses.

 

 

 

 

 

 

 

 

------

 

PoworSupply
1.5m2W

 

 

_ a

 

 

 

 

 

 

 

 

 

 

 

High Volhge Pulsar

Gate Pulse

'Hiléﬁféméiii "it EWW

    

1
3

Overview of electronics used to produce complementary gate

7 l

The complementary pulses from the PPWG trigger a pair of
GRX-l.5K-E pulse generators (Directed Energy, Inc. (DEI), Fort Collins, CO)
which supply the high voltage pulses applied to the gate device. One pulser
switches from a positive deﬂection voltage to the ﬁeld-free region potential.
The other unit switches from the negative deﬂection voltage to the ﬁeld-free
region potential. Positive and negative deﬂection voltages are supplied by
separate positive and negative high voltage power supplies ﬂoating at the
ﬁeld-free region potential. The DEI units do not have full rail-to-rail swings
and therefore the low (or high) input must be offset below (or above) the ﬁeld-
free region voltage. Low voltage power supplies with adjustable linear
regulators provide the i 5 to 10 volts needed. The outputs from the DEI
pulsers are routed through impedance-matched 50 Q cables about 60 cm long
to an impedance-matched vacuum feed-through. Inside the vacuum chamber,
insulated wires (not impedance matched) about 20 cm long connect the

pulser output to rods E & F of the gate device (see Figure 4.3).
4. 4. 2.2 Pulse shapes

Typical pulse shapes are shown in Figure 4.6, where output signals
from the DEI pulsers are plotted versus elapsed time since ion source
extraction. The line labeled Vs indicates the ﬁeld-free potential. Rise and fall

times are on the order of 30 us for the i 250 V pulses.

72

 

-100 “

-200 “ Deﬂection Deﬂection
-300 “

-4... \g ,r r
f

 

 

-500 ~— fo w
-600 T
-700 *—
-800 —»_ _, Transmission WW-WWW-W

-900ururl..w.:...rlﬁ..rrrr..
85.1 85.2 85.3 85.4 85.5 85.6

 

J.;-A

Voltage (V)

 

 

 

Time (microseconds)

Figure 4.6 Typical pulse shapes produced by the DEI pulsers.

4.5 Performance

Performance of the interleaved comb deﬂection gate was characterized
in the tandem TOF mass spectrometer [3, 20, 23]. The ﬁrst reﬂectron TOF
analyzer was focused at the position of the gate device by placing a detector
4 cm beyond the gate (as near as possible). A mass resolution of about 2000
was observed at this position for ions having energies of approximately
650 eV. The detector was then placed 15 cm beyond the gate device. Just
prior to the gate device, an aperture limited the beam diameter to 12.7 mm,
while the diameter of the active area of the detector is 25 mm. Figure 4.7a is

a spectrum showing the molecular ion region of diphenylmethane

73
(MW = 168). During the gate closed period, the wires are set to potentials of
i 250 V with respect to the ﬁeld-free voltage. The gate was operating with an
appropriate time delay and a wide gate pulse (1.28 us) in order to select the
entire region.

Figure 4.7b is the same spectral region with the time delay adjusted
appropriately and the gate pulse narrowed to 140 us in order to select the
m/z 167 ion packet. Optimization of the gate delay and gate pulse width was
accomplished sequentially. First, the gate delay was adjusted to as high a
value as possible without reducing the intensity of the peak corresponding to
the chosen m/z value. Next, the gate pulse width was set to as small a value
as possible without reducing the intensity of the peak. In this manner, both
edges are optimized for maximal resolution of isolation on both sides of the
object peak. One might consider use of the gate to actually make the
precursor peak narrower and hence improve mass resolution. However, in
this experimental arrangement, it was observed that setting the gate delay
too high or the pulse width too narrow resulted in both a narrowing of the
peak and a reduction in its intensity.

‘ The spectrum in Figure 4.70 was collected under the same conditions
as for Figure 4.7b except that the trigger signals to the two pulsers were

reversed in order to eliminate just the m/z 167 ion packet. By integration of

a)

Intensity (mV)

b)

Intensity (mV)

0)

Intensity (mV)

74

 

 

 

162 163 164 165 166 167 168 169 170

 

 

 

162 163 164 165 166 167 168 169 170

16-T

12 ~

 

m/z

i

l J l
l j l

«r—
«>—

162 163 164 165 166 167 168 169 170

m/z

Figure 4.7 TOF mass spectra of the molecular ion region of

diphenylmethane for gate pulse width of a) 1.28 us,
b) 140 ns, and c) 140 ns with gate trigger signals reversed.

75
the m/z 167 for the spectra in Figures 4.7a and 4.7b, the peak area for the
isolated peak in Figure 4.7b was 88% of the area for the corresponding peak
in Figure 4.7a. Hence, little reduction in transmission occurs.

The potential difference required to achieve complete elimination of
ions for this experimental arrangement was investigated by slowly increasing
the potentials of the deﬂecting voltage supplies. It was observed that below
i 175 V, signal due to coherent ion arrival could be observed. The 350V
potential difference between the wire segments of the interleaved comb ion
deﬂection gate, according to SIMION modeling, should result in a deﬂection
angle of about 85°. The angle expected to be required by simple geometrical
considerations is 7 .2°. This investigation represents fairly good agreement
between the theoretical model and the experimental results. Several factors
could account for the minor discrepancy, such as uncertainties in the
positions of the gate, aperture, and detector. With greater signiﬁcance, it is
likely that the ion beam is not collimated and/or that the beam is not
perfectly parallel with the rail on which the ion optical elements are
mounted. Since the detector is placed at a much greater distance from the
gate during normal operation of the tandem TOF instrument, the deﬂection
angle required is smaller than that required for the previous study.

The data presented in Figure 4.7 demonstrates unit resolution for
selection over a mass range of at least 167 in our tandem TOF mass

spectrometer. Based on ion packet width, distance between the ion packets,

76

the ﬂight distance inﬂuenced by the gate deﬂection ﬁeld (as determined by
SIMION), and the pulse rise and fall times, unit resolution over a mass range
of at least 300 is possible. The limit of unit resolution for isolation occurs
when the separation between adjacent ions (measured ﬁ'om peak center to
peak center) is equal to the sum of the width of the gate deﬂection ﬁeld, the
iso-mass ion packet width (FWHM), and the distance the iso-mass ion packet
moves during the pulse rise (or fall) time. The work presented here
represents a signiﬁcant improvement over previously demonstrated
capabilities in ion packet isolation. The performance of the interleaved comb
ion deﬂection device is dependent on a number of factors, including mass
resolution at the gate position, energy of the ions, angle of deﬂection
required, distance between wire segments in the device, jitter in timing
delays, pulse height, pulse width, pulse rise-times, pulse stability (extent of
ringing and droop), and extent to which the pulse shapes are complementary.
We expect that the interleaved comb ion deﬂection device will be widely used
for precursor ion selection (or isolation) in tandem TOF instruments and for
other applications where modulation of an ion beam is necessary.

For GC analyses, a mass resolution for isolation of 300 should be
adequate for the majority of applications. For this tandem TOF instrument,
pulsed-laser PID allows unit resolution for selection to m/z 1200 (see
chapter 3). Isolation of ion packets above m/z 300 will still be an eﬂective

method of eliminating the vast majority of interferences along the time

77
domain. In the worst case, two neighboring m/z values on either side of the
ion packet of interest would be allowed to pass, and these would be reduced
in intensity. The neighboring m/z value(s) of lower mass than the precursor
ion packet would obscure only the upper few percent of the product spectrum
mass range. The neighboring m/z value(s) of greater mass than the
precursor ion packet will not interfere unless signiﬁcant metastable decay of
these ions occurs. A product spectrum for m/z 91 of toluene for which the
gate was used to eliminate interference from the normal ion spectrum is
presented in chapter 5 (see Figure 5.10).
4.6 References
1 Seeterlin, M. A.; Vlasak, P. R.; Beussman, D. J .; McLane, R. D.;

Enke, C. G. J. Am. Soc. Mass Spectrom. 1993, 4, 751.

2 Weinkauf, R.; Walter, K.; Weickhardt, C.; Boesl, U.; Schlag, E. W.
Z. Naturforsch. 1989, 44a, 1219.

3 Beussman, D. J.; Vlasak, P. R.; McLane, R. D.; Seeterlin, M. A.;
Enke, C. G. (submitted to Anal. Chem.)

4 Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry / Mass
Spectrometry: Techniques and Applications of Tandem Mass
Spectrometry, VCH Publishers, Inc.: New York, 1988.

5 Jardine, D. R.; Morgan, J .; Alderdice, D. S.; Derrick, P. J. Org. Mass
Spectrom. 1992,27, 1077.

6 Cornish, T. J.; Cotter, R. J. Anal. Chem. 1993, 65, 1043.

7 Schey, K.; Cooks, R. G.; Grix, R.; Wollnik, H. Int. J. Mass Spectrom.
Ion Processes 1987, 7 7, 49.

8 Yefchak, G. E.; Schultz, G. A.; Allison, J.; Enke, C. G.; Holland, J. F.
J. Am. Soc. Mass Spectrom. 1990, 1, 440.

10

ll

12

13

14

15

16

17
18
19

20

21

22

23

78

Baker, J. M. B. J. Phys. E: Sci. Instrum. 1974, 7, 364.

Bloomﬁeld, L. A.; Geusic, M. E.; Freeman, R. R.; Brown, W. L. Chem.
Phys. Lett. 1985, 121, 33.

Pinkston, J. D.; Rabb, M.; Watson, J. T.; Allison, J. Rev. Sci. Instrum.
1986, 5 7(4), 583.

Brucat, P. J.; Zheng, L.-S.; Pettiette, C. L.; Yang, S.; Smalley, R. E.
J. Chem. Phys. 1986, 84(6), 3078.

Haberland, H.; Kornmeier, H.; Ludewigt, C. ; Risch, A Rev. Sci.
Instrum. 1991, 62( 10), 2368.

Cornett, D. S.; Peschke, K.; LaiHing, K.; Cheng, P. Y.; Willey, K. F.;
Duncan, M. A. Rev. Sci. Instrum. 1992, 63(4), 217 7 .

Kaufmann, R.; Spengler, B.; Lutzenkirchen, F. Rapid Commun. Mass
Spectrom. 1993, 7,902.

Loeb, L. B. Basic Processes of Gaseous Electronics, University of
California Press: Berkely, 1961.

Cravath, A. M. Phys. Rev. 1929, 33, 605.
Bradbury, N. E.; Nielsen, R. A. Phys. Rev. 1936, 4.9, 388.

Boesl, U.; Schlag, E. W.; Walter, K.; Weinkauf, R.
U. S. Patent # 5,032,722, 1991.

Beussman, D. J .; Vlasak, P. R.; Erickson, T. A.; Davenport, M. R.;
Enke, C. G. presented at the 43rd ASMS Conference on Mass
Spectrometry and Allied Topics, Atlanta, Georgia, May 22, 1995.

Hoyes, J. B.; Curbishley, S. G.; Doorbar, P.; Tatterton, P.;
Bateman, R. H.; Beer, B.; Lockett, J. Proceedings of the 42nd ASMS

Conference on Mass Spectrometry and Allied Topics, Chicago, Illinois,
May 29 - June 3, 1994, 682.

Dahl, D. A.; Delmore, J. E. SMION PC/PS2 Version 5.0 (real made),
1988, Idaho Falls National Engineering Laboratory, EG&G Idaho Inc.,
PO. Box 1625, Idaho Falls, ID 83415.

Vlasak, P. R.; Beussman, D. J.; Davenport, M. R.; Enke, C. G.
(submitted to Rev. Sci. Instrum.)

Chapter 5: The second mass analyzer

5.1 Requirements

The requirements of the second mass analyzer are different from that
of the ﬁrst due to the nature of the sample ions. When acceleration occurs
from an ion source, each iso-mass ion packet has approximately the same
kinetic energy distribution as other iso-mass ion packets. In contrast, the
ions that are accelerated after PID in the tandem TOF instrument have
m/z-dependent energies. Upon dissociation, the ions and neutrals retain
approximately the velocity distribution of the precursor ion packet assuming
that the translational energy release upon dissociation is not signiﬁcant.
These ions have a relative kinetic energy range of nearly 100%, however,
ranging from a small fraction of the precursor energy for low m/z product
ions to the entire precursor energy for undissociated precursor ions,

according to the equation:
= u . ——"'°1 (1)

where Uprod and Uprec are the energies of the product and precursor ion, and

mprod and 111me are the masses of the product and precursor ion, respectively.

79

80

5.2 Reflectron TOF

5.2.1 Background

In reﬂectron time-of-ﬂight (TOF) mass spectrometers employing
electron ionization sources, the application of an extraction ﬁeld to the source
volume creates a minimal spatial/temporal distribution for iso-mass ions
outside of the ion source at a position called the space-focus plane [1-4].
There is a signiﬁcant distribution of the kinetic energy of these ions as a
result of their initial spatial and energy distributions. The reﬂectron, or ion
mirror, is used to reproduce the spatial/temporal distribution of the space-
focus plane at the surface of an ion detector, thereby improving mass
resolution by increasing ﬂight time (and the separation between adjacent iso-
mass ion packets) without increasing signiﬁcantly the temporal width of the
iso-mass ion packets [2-6]. Generally, the relative kinetic energy range of
ions accelerated from the source is small (< 10%), and most reﬂectron time-of-
ﬂight instruments described in the literature have been designed to

effectively focus these small energy ranges.
5.2.2 One-stage and two-stage reﬂectrons

The most common reﬂectrons in use are one-stage and two-stage
reﬂectron designs which provide ﬁrst-order and second-order energy
compensation, respectively. Homogeneous ﬁelds (stages) are created with the

use of wire mesh ‘grids’. The prescribed geometry of the reﬂectron ﬁeld, in

81
both cases, is an analytically derived function of instrument variables [2, 6].
Figure 5.1 is a plot of Z-AT/T versus ion energy for three reﬂectron solutions
where T is the arrival time of an ion having average velocity and AT is the
difference between T and actual arrival time. Taking the inverse of the
difference between the extreme vertical values over a speciﬁc energy range
on this chart gives an estimate of mass resolution for that energy range.
Curve A is for an optimized single-stage reﬂectron, curve B is calculated
according to reference 2 where simplifying assumptions were made in
deriving the solution for a two-stage reﬂectron, and curve C is a solution
calculated using MathCad 5.0 [7], a numerical solving program, giving a
result shilar to that described in reference 6 for a two-stage reﬂectron. This
chart is calculated using values which match the reﬂectron focusing problem
for the second mass analyzer in the tandem TOF instrument. It can be seen
that these simple reﬂectrons provide excellent energy compensation for small
deviations in ion energy. Boesl, et al. [8] note that for larger energy ranges
(>10%), the two-stage reﬂectron can be re-conﬁgured to give somewhat

improved resolution over that speciﬁc energy range.
5.2.3 Grid-free reﬂectrons

Several grid-free reﬂectron designs have also been described [5, 9-13].
The ability to energy focus without the use of grids is an advantage since the

grids limit transmission via ion collisions with the wires and scattering of

82

 

0.008
0.007
0.006
0.005
E 0.004
E4 0.003
<.1
N 0.002
0.001
0
-0.001
-0.002
8 O O O O O O O O O O 8 O O O O
0 LO 0 LO 0 LO O ID 0 L0 10 O to O
03 O 0 Fl H N N CO CO ‘d‘ V to LO (0 Q0 L‘
H N N N N N N N N N N N N N N N
Ion Energy (eV)

Figure 5.1 A plot of 2-AT/T as a function of ion energy for homogeneous

ﬁeld reﬂectrons.

83
ions which pass near the wires [14]. In addition, throughput can be improved
by the radial focusing of the ion beam in the cylindrically-symmetric,
inhomogeneous reﬂectron ﬁeld. Grix, et al. [5] report that an energy range of
greater than 10% can be effectively focused using a grid-ﬁes design. This is
the type of reﬂectron which is used in the ﬁrst mass analyzer of the tandem

TOF instrument.
5.2.4 Perfectron

Rockwood [15] has developed a method giving inﬁnite order correction
which uses a parabolic ﬁeld in a device termed a ‘perfectron’. The perfectron
contains no ﬁeld-free region, and the focal points of the perfectron are at the
reﬂectron entrance. A preliminary version of this instrument resulted in a
resolution of 125, because of poor pulse response of the electronics and
deviations from the ideal parabolic ﬁeld in a grid-free design. The difﬁculty
of achieving a parabolic ﬁeld and the unconventional design (lacking ﬁeld-

free distance) prohibit its use in the current tandem TOF instrument.
5.3 Reflectron TOF for product ion analysis

Several reﬂectron focusing methods have been used to improve mass
resolution for product ion analysis in TOF mass spectrometers. These
methods fall into two main categories segregated by whether or not the
product ions (resulting from either an induced dissociation or a metastable

decomposition) are accelerated after dissociation.

84

5.3.1 Separation within the reﬂectron

Without post-dissociation acceleration, separation of product ions from
one another and from the precursor ion packet from which they originated is
accomplished entirely within the reﬂectron on the basis of the reduced
energy for product ions (see Equation 1). This results in a correspondingly
smaller residence time in the reﬂectron. N 0 separation occurs in the ﬁeld-free
regions since here the product ion packets have approximately the same
velocity distribution as the precursor ion packet from which they originated.

Early demonstrations involved metastable decompositions occurring
after ionization by 3 2”Cf plasma desorption source [16, 17] and a secondary
ion (SIMS) source [18] using a single reﬂectron for normal ion and product
ion analysis. Decompositions were identiﬁed by the detection of a neutral at a
detector placed behind the reﬂectron, with the arrival time of the neutral
‘correlated’ to the precursor m/z which decomposed and the product ion
which is then detected at the normal detector position. In general, one-stage
reﬂectrons are advantageous in this conﬁguration since mass calibration is
accomplished as a linear function of time. Schlag’s group [8, 19] used a linear
TOF / reﬂectron TOF arrangement, in which dissociation was accomplished
by intersecting a laser pulse with the precursor ion packet at the space-focus
plane of the linear TOF analyzer. This instrument incorporated an ion gate
to isolate a particular precursor m/z thus eliminating the need for the

correlation technique. The potentials of a two-stage reﬂectron were scanned

85
to consecutively focus product ions of diﬁerent m/z values at a detector with
a ﬁxed arrival time. Cornish and Cotter [20] also incorporated an ion gate in
their dual reﬂectron TOF instrument in which precursor ions are fragmented
by collisionally-induced dissociation (CID) at the focal point of a ﬁrst
reﬂectron TOF analyzer, and the products are analyzed in a second reﬂectron
TOF analyzer. Initially, a single-stage reﬂectron was used. A later
modiﬁcation involving an empirically-generated, curved reﬂectron ﬁeld
improved resolution and resulted in a non-linear time vs. mass relationship,
with calibration accomplished by ﬁtting the data to a third-order polynomial

by means of non-linear regression analysis [2 1, 22].
5.3.2 Separation by past-dissociation acceleration

The other major product ion analysis method involves a post-
dissociation acceleration step in which m/z-dependent velocities are
imparted to the product ions. This acceleration results in separation of iso-
mass ion packets in the ﬁeld-free region with the primary role of the
reﬂectron being energy-focusing.

Schey, et al. [23] focused product ions produced by surface-induced
dissociation in a linear TOF / reﬂectron TOF tandem mass spectrometer. The
reﬂectron was grid-ﬂee and allowed energy focusing over a range of greater
than 10%. Because the target surface was at a higher potential than the
ﬁeld-free region, the ions were essentially decelerated prior to dissociation

and accelerated after dissociation. Since the reﬂectron could not compensate

86
for the energy spread due to partitioning, overall resolution was improved by
optimizing for one product ion at a time. Schlag’s group [8] also incorporated
a post-dissociation acceleration stage in the linear TOF / reﬂectron TOF
instrument described above. The acceleration reduced the relative kinetic
energy range to the point where a two-stage reﬂectron could adequately
compensate. This focusing method regained the ability to collect a complete
mass spectrum from each ionization pulse, which was lost in the scanning
method described above, but resulted in lower mass resolution than for the

scanning method.

5.4 Product ion analysis in the tandem TOF instrument

5.4.1 Post-dissociation acceleration

In the tandem TOF instrument, our goal is to collect a complete
product spectrum from every source extraction. For this reason, the
incorporation of a scanning process such as that used by Schlag’s
group [8, 19] is not desirable. We chose to use an acceleration stage beyond
the dissociation step, such that separation occurs within the entire ﬂight
path beyond the acceleration. Here, a reﬂectron serves primarily to focus the
energy distribution rather than to act as both a separation and an energy-
focusing device.

The acceleration is accomplished by means of a pair of grids placed

about 1 cm apart approximately 2 cm beyond the interaction region. These

87
grids are aligned on the previously mentioned optical rail which is mounted
parallel to the ion ﬂight path below the interaction region. The ﬁrst grid is
held at the ﬁeld-free potential of the ﬁrst mass analyzer (-550 V) while the
second grid is held at the ﬁeld-free potential of the second mass
analyzer (-2500 V). Besides imparting m/z-dependent velocities to the
product ions, this acceleration reduces the relative kinetic energy range from

nearly 100% (see Equation 1) to about 32%.
5.4.2 Elimination of interference due to W]

In addition to providing efficient dissociation of precursor ions, the
laser pulse causes multi-photon ionization (MP1) of neutral background gas
in the interaction region. We observed that some portion of these ions drift
toward the acceleration stage and are accelerated into the second mass
analyzer. Figure 5.2a illustrates the background that is generated as a
result. The large, noisy signal at the left is the electrical ringing associated
with the discharge of the laser thyratron. The large, broad signal following is
due to MPI interference, while the sharp peaks are the molecular ion region
of bromobenzene (m/z 156-159). In the example shown, the laser was timed
so that it did not cause photodissociation of the bromobenzene. The detector
was placed on the optical rail about 20 cm beyond the interaction region.

In order to eliminate this background, a third grid was introduced

beyond the interaction region between the acceleration grids such that the

88

 

 

 

 

 

 

 

 

 

 

a)

.2.

O
3

>.
9‘:

U}

r:

Q)

H

s:
H

Time (microseconds)
b)

*3.

O

3,

>.

3:3

U)

c:

G)

4.3

c: I
H

84 89 94 99 104

Time (microseconds)

Figure 5.2 Molecular ion region of bromobenzene a) without MPI
elimination ﬁeld and b) with MPI elimination ﬁeld.

89
grids are spaced about 1 cm apart, with the ﬁrst grid placed about 2 cm
beyond the interaction region. A potential about 20 V more positive than the
ﬁrst ﬁeld-free potential is applied to the center grid. The ions produced by
MPI have only thermal velocities, while product ions and undissociated
precursor ions retain some percentage of the acceleration energy introduced
by the ﬁrst mass analyzer. This retardation ﬁeld stops the ions produced by
MPI, but allows the product ions to pass with the exception of approximately
the lowest 3% of the mass range. Figure 5.2b illustrates the effective

elimination of the NIPI interference.
5.4.3 Grid-free reﬂectron for product ion analysis

Initially, a grid-free reﬂectron was incorporated in the second mass
analyzer [24]. This analyzer was expected to focus a larger energy range than
homogeneous ﬁeld _ designs. The ﬁrst photodissociation experiments
performed in this instrument involved the molecular ion of bromobenzene
(m/z 156 or 158). After depletion of either of these precursor ion signals by
proper adjustment of laser timing, no product ions were observed. It was
expected that the major ﬁagment ion resulting from the photodissociation
process would be m/z 77. Because m/z 77 had a signiﬁcantly different
energy than the precursor ion, for which reﬂectron 2 had been optimized, it
was postulated that the resolution was simply too poor for any product peak

to be observed [24].

90
In response to this hypothesis, the next experiments involved the
fragmentation of the molecular ion of toluene (m/z 92), which was expected
to yield a product ion at m/z 91. Since the loss of a very small mass
percentage gives a product ion having nearly the same kinetic energy as the
precursor, it was expected that this product ion packet should be nearly as
well focused as the precursor. This product ion was indeed observed. It was
also found that adjusting the potentials of the reﬂectron to bring lower
energy ions into focus allowed the observation of m/z 77 from the
bromobenzene molecular ion [24]. Recognizing that the relatively slow
process of scanning reﬂectron voltages was inconsistent with the goals of this

project, we set out to ﬁnd or develop a diﬂ‘erent focusing solution.
5.4.4 Linear TOF mass analyzer

Although a reﬂectron capable of meeting the focusing needs of the
second mass analyzer was eventually developed as described in section 5.5
below, the investigation of photodissociation as a fragmentation method
continued using a short, linear TOF mass analyzer. This analyzer was
implemented simply by placing a detector about 20 cm beyond the interaction
region. Depending on the nature of the investigation, reﬂectron 1 was either
tuned at the interaction region by temporarily placing a detector there, or
tuned at the position of the detector 20 cm beyond. The former conﬁguration

provides maximal ion/photon overlap and therefore maximal

91
photodissociation efﬁciency, while the latter conﬁguration results in
improved resolution for both precursor and product spectra.

Operating the instrument in this manner allowed the observation of
entire product ion spectra and the determination of PID efﬁciencies for a
number of ions as discussed in chapter 3. As expected, however, the
resolution of the product spectra was quite poor. Figure 5.3 is a product
spectrum for bromobenzene obtained while the focus of reﬂectron l was at
the detector position. This spectrum is a subtraction of the ‘light off spectrum
from the ‘light on’ spectrum having been collected prior to the introduction of
the ion gate. In the subtracted spectrum, the precursor ion depletion is
observed as a negative peak. The positive product ion peaks are labeled with
the number of carbon atoms present. The resolution of this spectrum is not

adequate to differentiate between product ions varying by one mass unit.
5.5 A broad kinetic energy range focusing reﬂectron

As outlined above, the second mass analyzer in the tandem TOF
instrument must be capable of focusing a relative kinetic energy range of
about 32%. Coincidentally, another instrumentation project in this research
group also requires focusing of ions having a relatively broad energy
range (~30%). This project is the construction of a mass spectrometer in

which ions are extracted from an ion trap storage device and analyzed by

92

0.015 5—
0.01 + C4

0.005 «~ Cl C2 Cs C6

0 . vi

-0005 4+

-. precursor
0' 01 depletion

Intensity (V)

 

 

 

 

 

-0.015 . + . 4
91.25 92.25 93.25 94.25 95.25

Flight Time (us)

Figure 5.3 Product spectrum from m/z 156 of bromobenzene collected
using a short, linear TOF analyzer.

93

time-of-ﬂight (IT/TOF) [25, 26]. The storage device serves to accumulate
continuously produced ions between extractions, thus increasing sensitivity.
Ions in the trap occupy a relatively large volume compared to traditional
electron ionization sources and are subjected to a large extraction potential.
The spatial distribution of ions within the ion trap storage device becomes a
broad relative energy distribution upon extraction and subsequent
acceleration. In addition, energy induced by the trapping ﬁeld will contribute
to the energy distribution. In the IT/TOF, each iso-mass ion packet has a
broad energy distribution (~30%), unlike the situation in the tandem TOF
instrument. In the latter case, each iso-mass ion packet has a narrow
distribution, but the energy range is broad over the mass range of the
product ions. Despite this difference, the goal for reﬂectron design in the
IT/TOF is very similar to that in the tandem TOF instrument.

Since our goals were similar, the development of the broad kinetic
energy reﬂectron design method described below was a collaborative effort
between Qinchung Ji,, whose Ph. D. research is the development of the
IT/TOF, and the tandem TOF group. This method involves a numerical
calculation step to establish an electrostatic voltage proﬁle followed by a
computer simulation step which allows us to achieve the desired ﬁeld proﬁle

using a practical device.

94

5.5.1 Calculation of a profile

The reﬂectron ﬁeld proﬁle is established in a front to back manner,
beginning with the lowest energy ions that need to be focused and persisting
until the entire energy range of interest is covered. The proﬁle used for the
tandem TOF instrument contains a linear and a non-linear ﬁeld region. The
proﬁle for the non-linear ﬁeld region is determined by speciﬁc, practical
instrument parameters, namely: ﬁeld strength and length of the linear ﬁeld
region, kinetic energy range of the ions to be focused, angle between the
ﬂight path of the ions in the ﬁeld-free region and the axis of the reﬂectron,
and length of the ﬁeld-free region. This proﬁle is generated as a series of
linear segments, such that the nodes of the segmented curve correspond to
the positions of virtual grids in a theoretical reﬂectron. The potential applied
to each grid is calculated so that iso-mass ions having exactly enough energy
to reach a grid, stop there, and change direction will have the same total
ﬂight time as iso—mass ions reaching each of the other grids. Ions which turn
around at positions other than at grids will have slightly different ﬂight
times.

Suppose the lowest energy ion to be focused has energy U1. Our
strategy is to construct a ﬁrst reﬂectron segment from two parallel grids
placed a distance L1 apart so that an ion having energy U1 would stop exactly
at the second grid. In order to achieve this, the ﬁeld-free voltage is applied to

the ﬁrst grid, and a voltage V: is applied to the second:

V1 = E cosza (2)

where a is the angle between the ﬂight path in the ﬁeld free region and the
axis of the reﬂectron, and z is a single, positive charge.

The length of the ﬂight path outside of the reﬂectron (ﬁeld-free region)
is Ls, which is the sum of the distance from the initial space-focus plane to
the face of the reﬂectron and the face of the reﬂectron to the reﬂectron focal
point (often the detector position). The ﬂight time (t) of a hypothetical, singly-
charged positive ion having energy U1 (which would come to a complete stop
with respect to the reﬂectron axis at the position of grid two) can be

calculated from:

1=—+'_ (3)

 

where vff is the velocity of the ion in the ﬁeld-free region given by:

v, = LEE: (4)
m

where m is the mass of the ion. The average velocity (V1) of the ion in the

space between grid one and two (parallel to the reﬂectron axis) is given by:

V1 = -Vﬂ‘ ~COS a (5)

l
2
The factor of two in the second term of Equation 3 comes from the fact

that the ion traverses the ﬁeld between grids one and two twice. The average

velocity of the ion in the space between grid one and two (V1) is simply the

96

average of its velocity parallel to the reﬂectron axis at the position of grid one
(vrr -cos a), and its velocity at the position of grid two (zero). All terms in
Equation 3 (fully expanded) are known or measurable physical quantities
with the exception of m, which may be arbitrarily chosen as long as its value
is kept constant throughout. Therefore a ﬂight time (t) can be calculated for
any given set of conditions.

Having established a ﬂight time (t) for an ion of energy U1, we wish to
establish a second ﬁeld between grid two and a third grid placed a distance

In behind grid two. The ﬂight time for an ion stopping at this third grid is

 

 

 

 

 

 

 

given by the equation:
t = £1: + 2;L1 + 2;L2 (6)
vs V1 V2
where the velocity terms are now:
. +
V11“ = J2 (U1 U2) (7)
m
2-U2
vs-cosa + m -cosa
V = 8
1 2 ( )
V2 = —l— 2-U2 -cosa (9)
2 m

where U2 is the additional energy that the ion must have (above U1) to
exactly reach grid three. Also, V2 is the potential difference between grid two

and three where:

V2 = U” cosza (10)

 

Using the ﬂight time (t) calculated from Equation 3, all terms in Equation 6
(fully expanded) are known values with the exception of U2 which we solve
for numerically using MathCad® [7].

The remainder of the reﬂectron ﬁeld segments are calculated using
MathCad® according to the general equation for the nth stage (n > 2):

__ Lﬂ' “.1 2Li 2Ln

 

 

 

 

 

 

t _ — + _ + _ (ll)
VE 1:1 vi Va
(12)
vs -cosa +
V. = (13)
‘11 = ~1— VZ ' U" -cosa (l4)
2 in

Note that the Un term (the unknown variable) appears 2-n times in
Equation 11 when this equation is fully expanded, illustrating the need to
use a numerical solving program such as MathCad® to determine the

unknown. V1. is found according to:

V1. = —U—" cosza (15)
z

98

For the tandem TOF instrument, a potential proﬁle was calculated for
the second reﬂectron containing 12 segments using the above method. This
reﬂectron was conﬁgured for a drift length (La) of 188.6 cm, the distances
between the grids (L1 to L12) as given in Table 5.1, the angle (a) between the
ﬂight path and the reﬂectron axis of 2 degrees, 1650 eV as the lowest energy
to be focused (U 1), and 2700 eV as the highest energy that can be focused.
The actual calculated potentials to be applied to the grids (relative to the
ﬁeld-free potential) and the energies of hypothetical ions reaching each grid
are given in Table 5.1, while Figure 5.4 graphically shows the proﬁle beyond
grid two. Generally, the proﬁle is curved through the ﬁrst several grids and
becomes more linear towards the back of the reﬂectron. To obtain a more
exact ﬁeld curvature, electrodes were placed more densely in the initial
curved portion of the reﬂectron, as can be seen in Table 5.1 and Figure 5.4.
As will be discussed later, the shape of the calculated proﬁle depends on the
values chosen for La, U1, and L1 through L12. The proﬁle shown in Figure 5.4
is the result of several iterations of input parameters, restricted by the size
and physical shape of our vacuum chamber (originally designed around a
relatively short, grid-free reﬂectron) and the actual energy spread of product
ions given the possible range of accelerating voltages for this instrument. The
proﬁle shown in Figure 5.4 is one which can focus ions in the energy range of
interest and meets the physical size restrictions in our tandem TOF

instrument.

99

Table 5.1 Reﬂectron conﬁguration by segment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Segment Length Energy of Ion Calculated Adjusted
(cm) 8 (eV) 5 Potential (V) ° Potential (V) d
1 1.054 1650.00 1647.99 1647 .99
2 0.527 1773.79 1771.63 2590
3 0.527 1865.44 1863.17 1910
4 0.527 1945.10 1942.73 1920
5 0.527 2018.35 2015.89 2015
6 0.527 2087.51 2084.97 2095
7 0.527 2153.81 2151.19 2175
8 1.054 2279.87 2277.09 2278
9 1.054 2401.28 2398.36 2403
10 1.054 2519.53 2516.46 2526
1 1 1.054 2635.59 2632.38 2635.5
12 1.054 2750.14 2746.79 2745
a distances of linear ﬁeld regions between grids
b energies of ions that would exactly reach each grid
2 calculated potentials for the fully gridded reﬂectron

empirically adjusted potentials used in the ﬁnal, two—grid reﬂectron

 

100

 

2800

Potential (V)

 

 

 

1600 v + + + 1 1 1 1 1

 

O 10 20 30 4O 5O 60 7 0 80 90

Distance Beyond Grid Two (mm)

Figure 5.4 Calculated proﬁle for the non-linear portion of the reﬂectron.

101

As mentioned previously, only iso-mass ions which stop exactly at grid
positions are expected to have exactly the same ﬂight times. The ﬂight times
for ions having energies which cause them to turn around between electrode
positions can be calculated. Figure 5.5 is a plot of 2-At/t (where t is the time
for an ion which turns around at a grid position, and At is the difference
between the ﬂight time for a particular ion and the time for an ion which
turns around at a grid position) for ions in the range of 1650 to 2700 eV for
the calculated reﬂectron ﬁeld proﬁle. Calculations were made at 1 eV
intervals for the purpose of preparing this plot. Positions where the curve
dips toward zero indicate grid locations. It can be seen that in the front of the
reﬂectron, where the curvature of the voltage vs. distance proﬁle is greatest
(see Figure 5.4), 2-At/t exceeds 0.016 which corresponds to a mass resolution
of about 50. In the back of the reﬂectron, where the voltage vs. distance
proﬁle is more linear, the ﬂight time deviations for ions turning around
between grids are smaller. However, there is a jump to larger deviations
between 2100 eV and 2200 eV. This corresponds to the change from closely
spaced grids in the front of the reﬂectron to more widely spaced grids in the
back (see Table 5.1).

For the purposes of developing an eﬂ'ective focusing proﬁle, U1 was set
to a value 300 eV lower than the lowest possible energy for the product ions.

This was done to take advantage of the smaller deviations in the back of the

102

 

0.0175 1
0.015

0.0125 \

0.01 \

l

lp

l \l\
0 \I WWW v

T

1650 1800 1950 2100 2250 2400 2550 2700

 

 

 

 

2-At/t

 

 

 

 

 

 

 

 

 

 

 

 

 

Ion Energy (eV)

Figure 5.5 A plot of 2-At/t as a function of ion energy for the fully gridded

reﬂectron.

103
reﬂectron and to avoid the large deviations expected for the front part of the
reﬂectron. In Figure 5.5, the portion of the reﬂectron in which ions turn
around is from 1950 eV to 2700 eV. It can be seen that 2-At/t is less than
0.0025 (Resolution > 400) for this energy range. This predicted resolution is
greater than that predicted for optimized two-stage reﬂectrons. Also, this

reﬂectron is about half as long as the two-stage design.
5.5.2 SIMON modeling to achieve desired profile

Having calculated a potential proﬁle expected to focus iso-mass ions in
a kinetic energy range of 1950 eV to 2700 eV, we next constructed a model of
the 12 stage reﬂectron in SIMION [27]. SIMION is a program designed to
predict electric ﬁelds based on position, shape, and voltage applied to
graphically entered electrodes and to predict the motion of charged particles
in the ﬁelds produced. Figure 5.6a is the SIMION reﬂectron, with the heavy
lines indicating electrodes and the light lines indicating equi-potential
contours. The reﬂectron is cylindrically symmetric. Grid one (at ﬁeld-free
potential) is on the left, with the solid back-plate at the right. A shroud
extends from the ﬁrst grid around the other, higher potential electrodes to
preserve the ﬁeld-free drift space. The potentials corresponding to the equi-
potential contour lines are the applied potentials as given in Table 5.1. The
contours, of course, follow the conductive electrodes, with minimal leakage of

the ﬁeld-free potential into the spaces between the electrodes near the edges.

104

b)

a)

 

1III-IIII-IIIIIIII-III-IIIIIIIIIIIIUIIIII

 

 

 

 

 

 

 

 

 

 

 

 

/_/f/i//
\\K\\

 

 

 

 

 

A

 

 

 

 

e)

 

 

 

 

 

 

 

 

 

SIMION a) fully gridded reﬂectron, b) two-grid reﬂectron,

Figure 5.6

and c) two-grid reﬂectron after adjustment of ring voltages.

105

Figure 5.7a shows the trajectories of 6 iso-mass ions having kinetic
energies of 1950, 2100, 2250, 2400, 2550, and 2700 eV. These ions begin at
the same point (a perfect space-focus plane at upper left) with a trajectory 2
degrees from the reﬂectron axis. Prior to entering the reﬂectron, the ions
follow the same path, as would be expected for a collimated ion beam in a
ﬁeld-free region. The low energy ions spend less time in the reﬂectron, have a
smaller velocity component perpendicular to the reﬂectron ﬁeld, and do not
penetrate the reﬂectron ﬁeld as far as the high energy ions. Upon exiting the
reﬂectron, the iso-mass ions follow parallel paths, eventually becoming
focused temporally at a second space-focus plane. The tick marks in the ion
trajectories are equal time markers. During the ﬁrst leg of ﬁeld-free ﬂight,
the ions having higher energy (and higher velocity) move ahead of the lower
energy ions, but spend more time in the reﬂectron, exiting the reﬂectron later
than the lower energy ions. At the second space-focus plane, the higher
energy ions have caught up to the low energy ions, resulting in a narrow
temporal distribution. Figure 5.7a' is an expanded view of the portion of
Figure 5.7 a contained in the rectangle.

For purposes of comparison, the SIMION simulation results (indicated
by the squares) are plotted along with the MathCad® calculated results
(indicated by the line) in Figure 5.8. Simulations have been made for ions in
the range of 1950 to 2700 eV at 25 eV intervals. It can be seen that these

predictions are virtually identical.

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

111111111

111M 1.

 

 

 

 

 

 

 

 

L1lll|11

 

 

 

 

 

 

 

 

 

 

b')

0)

THE-M 1111111111
Wt“;

WM W.,— W ---- ﬁﬁrﬁz

;::;::::£11,_____H_;LLL HI I

”M M 11___1-__1111.

 

 

 

 

 

 

 

Figure 5.7 SIMION ion trajectories for a, a') the fully gridded reﬂectron,
b, b') the two-grid reﬂectron, and c, c') the two-grid

reﬂectron after adjustment of the ring electrode voltages.

 

 

1:1 SIMION
-— MathCad

 

 

 

107

 

0.0025 ,
0.002 ‘

 

 

l l

A

l

 

 

Ion Energy (eV)

calculation and SIMION

of MathCad®

Comparison

Figure 5.8

simulation.

108

In the SIMION simulation, the grids are assumed to have 100% ion
transmission. However, because the grids actually have limited open area
(88%), ion transmission through a real reﬂectron with many grids would be
unacceptably low. Also, ﬁeld distortions in the vicinity of grids can result in
deﬂection of ions and loss of resolution [14]. To eliminate this limitation we
replaced most of the virtual grids in the SIMION reﬂectron with open rings
with the exception of the ﬁrst two grids, which deﬁne the end of the ﬁeld-free
region and the steep initial rise, and the solid plate at the back of the
reﬂectron. Removal of these grids results in distortion of the ﬁeld in the
reﬂectron. Figure 5.6b is a SIMION plot of the equi-potential contours that
result from this new conﬁguration using the calculated potentials given in
Table 5.1. The equi-potential contours follow the ring electrodes but deviate
from the original grid position as the distance from the ring electrodes
increases. Also, the potential vs. distance proﬁle along the reﬂectron axis is
no longer that which was calculated.

These ﬁeld distortions result in three major consequences. First, the
focusing properties observed for the totally gridded reﬂectron simulation are
degraded since the potential proﬁle has changed. Second, the curved ﬁeld
affects radial ion motion, so that ions entering the reﬂectron at different
positions or at slightly different angles are steered differently (spherical
aberration). Third, ions entering the reﬂectron at the same position and

angle, but having different energies are also steered differently (chromatic

109
aberration). The loss of temporal focus resulting from the ﬁeld distortions can
be observed in Figure 5.7b and Figure 5.7b', which show trajectories for ions
with kinetic energies in the range of 1950 to 2700 eV. Also apparent is the
chromatic aberration, which in this case results in a narrower beam at the
reﬂectron focal point. This reﬂectron yields a predicted resolution of less
than 50.

In order to compensate for the ﬁeld distortions seen in Figure 5.6b, the
ring electrode voltages can be manually adjusted in SIMION by an iterative
process. With each adjustment of electrode voltages, a contour plot is
observed and trajectories simulated to determine effect. This procedure
allows the calculated potential vs. distance proﬁle to be achieved quite closely
in the two-grid reﬂectron for the path that the ions are expected to travel in
the reﬂectron. This results in improved temporal focusing, but increases the
spherical and chromatic aberration by further increasing the curvature of the
reﬂectron ﬁeld. This effect can be seen in Figure 5.6c, which shows the equi-
potential contours for the manually adjusted reﬂectron. The voltages applied
to the rings in this simulation are the adjusted potentials given in Table 5.1.
Figure 5.7c and Figure 5.7c' show trajectories for ions in the range of 1950 to
2700 eV. The eﬂects of chromatic aberration in this case result in a broad ion
beam. The temporal distribution has decreased considerably as a result of the

manual adjustment. Figure 5.9 is a plot of 2-At/t vs. ion energy Because the

110

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.002
+
0.0015 A
_..o_B
ll
0.001 +C
0.0005
V A
Z 0—
-0.0005
-0.001
-0.0015
0.00211111111111411
OOOOOOOOOOOOOOOO
LOOLOOLOOtQOlOOlOOlOOl-OO
CDOOHHNNO‘DCGV‘VIQLQQDCOL‘
HNNNNNNNNNNNNNNN
IonEnergy(eV)

Figure 5.9 A plot of 2-At/t as a function of ion energy for the two-grid,
empirically adjusted reﬂectron over the range of 1950 to
2700 eV at intervals of 25 eV.

1 11
ion beam is not inﬁnitely narrow as modeled above, but rather has a ﬁnite
diameter of about 1 cm, three traces are drawn in Figure 5.9. Curve B
corresponds to the center of the beam (as modeled above), giving a theoretical
resolution of approximately 800. Curve A corresponds to the same
trajectories begun from a position 5 mm nearer the center of the reﬂectron
while curve C indicates trajectories begun 5 mm further from the center of
the reﬂectron, for an overall beam diameter of 1 cm. The differences between
these traces are a result of spherical aberration. Including the effect of the
beam width, an overall resolution of approximately 400 is predicted. This is
the reﬂectron conﬁguration we chose to emulate for the second reﬂectron in

the tandem TOF instrument.
5.5.3 Construction

Reﬂectron 2 is constructed from stainless steel rings (outer
diameter = 10.2 cm, inner diameter = 6.35 cm, thickness = 0.051 cm).
Stainless steel mesh (90 LPI, 88% open area, Buckbee-Mears, St. Paul, MN)
was spot-welded to the face of the ﬁrst two rings to form the linear portion of
the reﬂectron. The last electrode in the reﬂectron stack is a solid stainless
steel plate. Ceramic spacers are used to hold the electrodes together in a
stack and isolate them from one another electrically. The distances between
the rings are given in Table 5.1 as are the adjusted voltages applied to the
electrodes. The total length of the reﬂectron is 9.5 cm. The reﬂectron is

mounted in the instrument by three posts attached to the back. Two of the

112
posts are connected to linear motion feedthroughs which allow optimization

of mirror position while under vacuum.
5.5.4 Performance

Figure 5.10 is a product spectrum obtained using the two-grid,
empirically-adjusted reﬂectron in the tandem TOF mass spectrometer. The
precursor ion was m/z 91 from toluene which was isolated by the gate prior
to dissociation. This spectrum represents a relative kinetic energy range of
about 32%.

The mass resolution (m/Am, FWHM) is about 300 for product spectra
and about 1500 (t/(21At), FWHM) for normal spectra. This resolution for
normal spectra through reﬂectron 2 (collected without use of the gate or laser
PID) is approximately the same resolution that is observed when a detector is
placed near the position of the interaction region. One would normally expect
resolution to improve with increased ﬂight distance since the separation
between adjacent iso—mass ion packets increases. Our result shows that the
arrival time distribution for each iso-mass ion packet is being broadened at
approximately the same rate as the distance between adjacent iso-mass ion
packets increases. There are two factors which are believed to be primarily
responsible for this effect. First, some ﬂight time deviation is expected as a
function of energy and beam width as depicted in Figure 5.9. Second, the

introduction of seven grids in the second mass analyzer (the reﬂectron grids

113

 

 

 

 

 

 

 

 

0.013 1
0.011 «— product peaks used
for calibration
0.009 ~~ / \\
, \
S 14/ s1
>, 0.007 -~
43
-1-1
”:3
a.) 0.005 1
43
c:
l—Il
0.003 1
0.001 ~~ [ MW
. [i ll” Wl I [[1ka 1‘
' 1' WWW hm...
‘0.001TTIIIierIifrwiiIIIIiIIIIiIWIIiI—r—IiiilItiﬁrrri

 

10 20 30 40 5O 60 70 80 90 100

Figure 5.10 Product spectrum of m/z 91 from toluene.

1 l4
counted twice) can also result in broadening of peaks since ﬁeld distortions
occur in the vicinity of these grids [14]. This effect is especially pronounced
when the ﬁeld strength changes drastically across the grid, as is the case for
both the post-dissociation acceleration grids and the reﬂectron grids.

The fact that the product spectra are not as well resolved as normal
spectra is not primarily a result of temporally broadened product ion peaks,
but the compression of this spectrum in time as compared to the normal
spectrum. Normal (undissociated) ion packets separate from one another
continuously through the entire ﬂight path beginning at the source, while
product ion packets do not begin to separate from one another until they are
accelerated after dissociation in the interaction region. The mass resolution
of 300 observed for the product spectra is similar to the resolution of 400
predicted by the SIMION model. Because each iso-mass product ion packet
has a small kinetic energy range rather than the entire energy range (which
was used in making the resolution prediction), one would expect that the
actual peak widths would be narrower than those predicted in this manner.
However, the loss of resolution as a result of grid effects, spherical and
chromatic aberration, uncertainties in the dimensions of instrument
components, uncertainties in applied voltages, and uncertainties in the
modeling itself apparently are offsetting. Uncertainties in the dimensions of
the instrument are important since the reﬂectron voltages are calculated

based on these dimensions, which include ﬁeld-free drift length and distance

115

between reﬂectron electrodes. Distance between the two grid electrodes is of
particular importance because of the large ﬁeld applied between them.
Unfortunately, this distance is particularly difﬁcult to control since the wire
mesh attached to the ring electrodes does not tend to stay ﬂat. Attempts to
apply tension to the wire mesh worsens the situation as the material then
tends to ripple. The reﬂectron voltages are not tuned in any way to improve
performance for two reasons. First, adjustment of individual voltages
degrades resolution. Second, any change in the proﬁle which does not make
the proﬁle more like that calculated is likely to aﬁect applicability of the
derived mass calibration equations described below. Indiscriminate tuning is
effective for the ﬁrst reﬂectron, however, since all m/z values have
approximately the same energy distribution and therefore travel similar
paths through the ﬁrst reﬂectron.

It was previously mentioned that due to chromatic and spherical
aberration, beam de-focusing effects may result in ion loss and m/z
discrimination (see Figure 5.7). Evidence of this problem presents itself in
the tandem TOF instrument since relative intensities of product m/z’s are
aﬂected by the adjustment of the reﬂectron’s angular position. When mirror
position is optimized for maximum throughput of low m/z product ions
rather than normal, undissociated ions, their intensity can be improved by
approximately a factor of two, with a corresponding reduction in the normal

ion peak intensities.

116

5.5.5 Possible improvements

As mentioned above, the speciﬁc reﬂectron used in the tandem TOF
instrument was optimized within the constraints of pre-existing size
limitations of the vacuum chamber. When a greater range was considered for
the independent variables in the design calculation, it was found that
extending the linear ﬁeld region at the front of the reﬂectron resulted in a
longer non-linear region with less curvature. Figure 5.11 is a comparison of
two calculated proﬁles for the same choice of U1, (1, and La, but curve A is
calculated using L1 of 1.054 cm as used previously in the tandem TOF
reﬂectron 2 development, and curve B is calculated using L1 of 10.54 cm.

There are several reasons why the longer proﬁle is a more effective
solution. Because the proﬁle for the longer ﬁrst stage is less curved, the ﬂight
time deviations for ions that turn around between grid positions is smaller.
There is far less ﬁeld distortion as a result of removal of the grids in the
longer reﬂectron as compared to the shorter one, reducing the extent to
which ring electrodes need to be adjusted to restore the calculated proﬁle.
The more homogeneous ﬁeld of the longer reﬂectron causes less spherical and
chromatic aberration. Because the change in ﬁeld strength across the
reﬂectron grids is smaller for the long linear stage than for the shorter linear
stage, the detrimental ﬁeld distortions in the vicinity of the reﬂectron grids

will be smaller [14]. Besides the improvements in using a longer reﬂectron

117

2800

 

2600 4

[0

sh

O

O
1

2200 +

 

Potential (V)

 

 

2000 1» + B

 

I

1800 —

 

 

 

1600 1 1 1 1 1 1
O 20 4O 60 80 100 120

Depth of Reﬂectron Beyond First Stage (mm)

Figure 5.11 Comparison of resulting reﬂectron proﬁles for different
lengths of the ﬁrst stage.

118
design, it is expected that using a smaller angle, a, will result in improved
performance. Since the ﬁeld proﬁle in the non-linear portion of these
reﬂectrons varies from the outer edges to the center, with large curvature in
the equi-potential lines near the edges (see Figure 5.6c), using only the
center portion of the reﬂectron (by reducing a) should reduce aberration
eﬁ‘ects.

Unfortunately the physical size limitations of the tandem TOF vacuum
chamber do not allow the testing of a signiﬁcantly longer reﬂectron design.
However, the IT/TOF instrument design incorporates the suggested
improvements. The tandem TOF instrument and the IT/TOF instrument pose
slightly different challenges, making a direct comparison of results difﬁcult.
The tandem TOF reﬂectron 2 must focus individual iso-mass ion packets
having small energy ranges, but with a broad relative energy range over the
mass range. In the IT/TOF, the relative kinetic energy range over each
speciﬁc m/z ion packet is large (~30%). In the tandem TOF instrument,
macroscopic inaccuracy in the ﬁeld proﬁle used will result primarily in
calibration inaccuracy. In the IT/TOF instrument, inaccuracy in the ﬁeld
proﬁle will result in reduced mass resolution. A direct comparison of the
performance of a short and long reﬂectron can be performed in the IT/TOF

instrument by constructing one reﬂectron of each type.

119

5.5.6 Signiﬁcance of this design method

The reﬂectron design method presented here results in non-linear ﬁeld
reﬂectrons capable of focusing large energy ranges with the use of only two
grids. The systematic progression of design results in a practical device in
terms of performance and ease of construction despite geometric constraints.
This design method provides an alternative to analytical solutions for energy-
focusing and a new approach to ion optical design.

The method described above has been successfully used to design a
reﬂectron which provides unit resolution for product ions having a relative
kinetic energy range of about 32% in the tandem reﬂectron TOF mass
spectrometer. While the performance of this speciﬁc reﬂectron is satisfactory
in its application, we expect that the longer design under construction for the
IT/TOF instrument will result in improved performance for broad energy
range focusing for reasons mentioned above.

This design method is quite versatile. While one and two-stage
reﬂectrons must be constructed in accordance with analytical solutions,
reﬂectrons designed by this method have much more ﬂexible geometry. As
discussed above, however, longer reﬂectrons with smaller angles between ion
ﬂight path and reﬂectron axis are expected to perform better. In both the
tandem TOF and the IT/TOF instrument, our needs require focusing of ions
with relative energy spreads around 30%. This is certainly not the limit of

this reﬂectron design method. We expect that any energy range can be

120
compensated for by adding stages to the back of the reﬂectron during the
calculation step.

It is interesting to consider the case of reducing the distance between
adjacent grids in the calculation to the point where the ﬂight time deviations
for ions stopping between grids becomes negligible. This is not useful as a
practical device due to the large number of grids; however, the resulting
proﬁle would be an excellent solution to the broad energy range focusing
problem for instruments with ﬁeld-free regions, unlike the 'perfectron'
solution described by Rockwood [15] which has no ﬁeld-free region. Although
our solution is more consistent with current reﬂectron instrument designs
than the ‘perfectron' design, the challenge again becomes the design and
construction of a practical device which accurately produces this ﬁeld proﬁle

without serious aberration.
5.6 Calibration of product ion spectra

Calibration of product ion spectra is easily accomplished using the

 

 

 

derived relationship:
m m
t _ t : _E°_ _ PM,
W A + B Jim”, ] (16)
——-A + B
mime

where tprec is the precursor arrival time, tpmd is the product arrival time, with:

 

B = —' (18)

where Up... is the average kinetic energy of the precursor ions, and Uacc is the
energy gained by post-dissociation acceleration.

Since A and B are diﬁcult to determine by direct measurement, two
product ion arrival times for known m/z values along with the arrival time
and m/z value of the precursor ion are used in Equation 16 to solve for the
two unknowns (A and B) using MathCad®. These constants can be used to
calibrate product spectra collected from any m/z value precursor ion for the
same instrument tune. For the spectrum in Figure 5.10, the peaks at m/z
values 39 and 65 were used for calibration. Mass assignment is accurate to
within 0.1 a.m.u. for most of the spectrum, with slightly larger error for

masses in approximately the lowest 25% of the spectrum.
5.7 References

1 Wiley, W. C.; McLaren, I. H. Rev. Sci. Instrum. 1955, 26(12), 1150.

2 Karataev, V. I.; Mamyrin, B. A; Shmikk, D. V. Sov. Phys. Tech. Phys.
1972, 16(7), 1177.

3 Mamyrin, B. A; Karataev, V. I.; Shmikk, D. V.; Zagulin, V. A. Sov.
Phys.-JETP 1973, 37(1), 45.

4 Grix, R.; Gruner, U.; Li, G.; Stroh, H.; Wollnik, H. Int. J. Mass
Spectrom. Ion Processes 1989, .93, 323.

5 Grix, R.; Kutscher, R.; Li, G.; Gruner, U.; Wollnik, H. Rapid Commun.
Mass Spectrom. 1988, 2(5), 83.

10

11

12

l3

14

15

16.

17

18

19

20

21

122

Gohl, W.; Kutscher, R.; Laue, H. J .; Wollnik, H. Int. J. Mass Spectrom.
Ion Processes 1983, 48, 411.

MathCad 5.0 ® for Windows; Copyright © 1991-1994 by MathSoft,
Inc., 101 Main Street, Cambridge, MA 02142.

Boesl, U.; Weinkauf, R.; Schlag, E. W. Int. J. Mass Spectrom. Ion
Processes 1992, 112, 121.

Berger, C. Int. J. Mass Spectrom. Ion Processes 1983, 46, 63.

Frey, R.; Weiss, G.;Kaminski, H.; Schlag, E. W. Z. Naturforsch. Teil A.
1985, 40, 1349.

Walter, K.; Boesl, U.; Schlag, E. W. Int. J. Mass Spectrom. Ion
Processes 1986, 71, 309.

Kutscher, G.; Grix, R.; Li, G.; Wollnik, H. Int. J. Mass Spectrom. Ion
Processes 1991, 103, 117.

Bergmann, T.; Martin, T. P.; Schaber, H. Rev. Sci. Instrum. 1990,
61(10), 2592.

Bergmann, T.; Martin, T. P.; Schaber, H. Rev. Sci. Instrum. 1989,
60(3), 347.

Rockwood, A. L. Proceedings of the 34th ASMS Conference on Mass
Spectrometry and Allied Topics, Cincinnati, Ohio, June 8-13, 1986,
173.

Della Negra, 8.; Le Beyec, Y. Int. J. Mass Spectrom. Ion Processes
1984, 61, 21.

Della Negra, 8.; Le Beyec, Y. Anal. Chem. 1985, 5 7, 2035.

Tang, X.; Beavis, R.; Ens, W.; LaFortune, F.; Scheuler, B.; Standing, K.
G. Int. J. Mass Spectrom. Ion Processes 1988, 85, 43.

Weinkauf, R.; Walter, K.; Weickhardt, C.; Boesl, U.; Schlag, E. W. Z.
Naturforsch. 1989, 44a, 1219.

Cornish, T. J.; Cotter, R. J. Org. Mass Spectrom. 1993, 28, 1129.

Cornish, T. J.; Cotter, R. J. Rapid Commun. Mass Spectrom. 1993, 7,
1037.

22

23

24

25

26

27

123

Cornish, T.J.; Cotter, R. J. Proceedings of the 42nd ASMS Conference
on Mass Spectrometry and Allied Topics, Chicago, Illinois, May 29 -
June 3, 1994, 1046.

Schey, K. L.; Cooks, R. G.; Kraft, A; Grix, R.; Wollnik, H. Int. J. Mass
Spectrom. Ion Processes 1989, .94, 1.

Seeterlin, M. A. Ph. D. Thesis, Michigan State University, 1993.

Ji, Q.; Vlasak, P. R.; Holland, J. F.; Enke, C. G. Proceedings of the
42nd ASMS Conference on Mass Spectrometry and Allied Topics,
Chicago, Illinois, May 29 - June 3, 1994, 1042.

Ji, Q.; Davenport, M. R.; Holland, J. F.; Enke, C. G. presented at The
43rd ASMS Conference on Mass Spectrometry and Allied Topics,
Atlanta, Georgia, May 22, 1995.

Dahl, D. A; Delmore, J. E. SMON PC/PS2 Version 5.0 (real mode),
1988, Idaho Falls National Engineering Laboratory, EG&G Idaho Inc.,
PO. Box 1625, Idaho Falls, ID 83415.

Chapter 6: TOF/TOF as an analytical tool

6.1 Unit resolution and efficient photo—fragmentation

The goal of this research project, as outlined in chapter 1, is the
development of a tandem mass spectrometer capable of greatly increasing the
amount of MS/MS data obtainable from sample components as they elute
from a chromatographic column. Several steps along the road to this goal
have been described in this thesis.

Time-of-ﬂight mass spectrometry is especially suited to any application
which requires a high spectral generation rate. The characteristically poor
mass resolution of TOF has been an impediment to its development, however.
This is especially true for applications which require a gas-phase ion source.
In the TOF/TOF instrument, we have demonstrated a mass resolution of
1500 for separation of precursor ions (see chapter 2). We have demonstrated
that focusing a laser pulse to the approximate dimensions of a single, iso-
mass precursor ion packet allows unit resolution of selection at least to m/z
158. Calculations show that the limit of unit resolution for selection is
around m/z 1200 (see chapter 3). Isolation of the precursor ion packet (prior
to PID) is important in order to eliminate background. We have

demonstrated that the use of an interleaved comb deﬂection gate allows unit

124

125

resolution for isolation to at least m /z 167 with calculations indicating that
the actual limit is around m/z 300 (see chapter 4). Using a novel reﬂectron
which can focus a broad energy range, we have demonstrated a mass
resolution of 300 for the collection of product ion spectra. The development
method used to create this reﬂectron represents a new approach in ion
optical design (see chapter 5). Unit resolution over a mass range of at least
300 a.m.u. is consistent with the requirements of a detector for gas
chromatography, showing that tandem TOF is an alternative to the more
traditional tandem mass analyzers based on mass ﬁlters.

Given the known photodissociation cross-sections for organic ions and
the photon density that can be generated by focusing a high-energy excimer
laser pulse to a small volume, it follows that photo-fragmentation eﬁciencies
should be high for ions which occupy this volume. This was demonstrated
experimentally in the tandem TOF instrument (see chapter 3) when a
fragmentation efﬁciency of 79% was observed for the molecular ion of
bromobenzene. The efﬁciencies for a variety of ions were shown to be similar
to the efﬁciencies observed for CID in a triple quadrupole mass spectrometer.
A large number of precursor ions are focused to the interaction region

volume, thereby making efﬁcient use of sample as well as laser power.

126

6.2 Analytical utility of PID product spectra

The implementation of the novel reﬂectron in the second mass
analyzer results in unit mass resolution for the product spectra. This
development allows a more complete charactrization of photo-induced
dissociation (at 193 nm) as a fragmentation technique in the tandem TOF
mass spectrometer [1-4]. Photodissociation cross-sections for 45 ions
originating from 27 different compounds have been determined. These cross-
sections range from 2.9 x 10'19 to 1.4 x 10'18 cm2 [2, 4]. Aliphatic ions
originating from both saturated and unsaturated molecules are among those
successfully photo-fragmented. This provides evidence that virtually all
organic ions should photodissociate when 193 nm radiation is used and '
further indicates that PID will provide efﬁciencies that are useful in
analytical applications when a system with high ion/photon overlap is
employed.

The product spectra from photodissociation in the tandem TOF mass
spectrometer were compared to product spectra collected using CID in a
triple quadrupole mass spectrometer [1, 2, 4]. All ions observed in the CID
spectra were also observed in the corresponding PID spectra. In general, the
low m/z portions of the PID (spectra were enhanced with respect to the
corresponding CID spectra. In addition, peaks due to loss of hydrogen from
various product ions were more frequently observed in the PID spectra than

in the CID spectra. These factors both indicate that more energy is imparted

127
to the precursor ion through the use of 193 nm (6.4 eV) PID than is
transferred collisionally in the triple quadrupole mass spectrometer. These
data show that PID in the tandem TOF mass spectrometer provides at least
as much analytical information as CID in the triple quadrupole mass
spectrometer. Further studies of the PID process for organic compounds as a
function of laser wavelength and power are being carried out in Professor

Enke’s laboratory.
6.3 Feasibility of GC/MSIMS

The collection of complete data sets for compounds as they elute from a
gas chromatographic column is still not possible. Many of the barriers that
stood in the way of this goal have been removed, however. Particularly, we
have shown that useful product spectra can be generated at the repetition
rate of an excimer laser; Monitoring GC efﬂuents will require the
implementation of an improved instrument control and data collection
system. The current instrument control system is based on a LabWindows®
program which communicates via GPIB with a delay generator and a
transient recorder (see chapter 2). The GPIB cannot transfer data at an
adequate rate to allow GC/MS/MS. Also, delay times, which specify precursor
ion m/z, currently are entered manually.

The automated system which would allow GC/MS/MS will

continuously collect normal spectra until a compound begins to elute. The

128
spectrum of this compound will then be assessed for m/z values having
signiﬁcant intensity. Next, the control system will sequentially implement
the necessary delay times to collect a product ion spectrum for each
signiﬁcant m/z value in the normal spectrum. The control system will then
continue to check for evidence of the next elution. An integrating transient
recorder, like that developed at MSU [5-7], will be required for data
collection. While custom electronics and sophisticated programming would be
required to construct such a control system at this time, computer interfacing
and data collection electronics are both high-growth technologies, and the
building blocks of such a system may become commercially available in the

relatively near future.
6.4 TOF/TOF in other analytical applications

There are numerous potential applications of tandem TOF beyond
those which have previously been described in this thesis. Two of these,
which involve the extension of this technique to non-volatile compounds, are
described below. Besides the analytical diversity gained by extending this
instrument’s capabilities to higher mass compounds, this would allow
characterization of PID as a potential fragmentation technique for high m/z
ions.

The advantages of tandem TOF in the characterization of GC eﬂluents

are directly applicable to liquid separations. The successful coupling of LC

129

with TOF is challenging, however, since the ion sources (such as
electrospray) which are generally used to interface LC with mass
spectrometry operate in continuous rather than pulsed mode. A possible
solution to this interface problem is being developed in Professor Enke’s
laboratory. Speciﬁcally, Qinchung Ji [8, 9] has coupled an ion trap storage
device with a TOF mass spectrometer such that the storage device
accumulates continuously generated ions between extractions. This will
allow any type of ionization to be used with TOF, thereby taking advantage
of the spectral generation rate and sensitivity of TOF without sacriﬁcing
duty cycle.

Matrix-assisted laser desorption ionization (MALDI) has proven itself
as an extremely useful technique for ionizing large biomolecules or other
polymers. MALDI is often used in conjunction with TOF since TOF analyzers
have virtually unlimited mass range, are inexpensive, and are directly
compatible with pulsed ionization. MALDI is a soft ionization technique
which often yields exclusively the protonated molecule. While this is useful
in determining molecular weight, it provides little structural information.
The possibility of using tandem TOF to enhance the structural information
obtainable is therefore attractive. Preliminary studies were performed in
collaboration with Ben Gardner and Professor J. Throck Watson [10] in the
tandem TOF instrument. The photodissociation of protonated bradykinin

(m /z 1061) represents the ﬁrst successful photo-fragmentation of a peptide

130
ion in a TOF mass spectrometer. Unfortunately, the low acceleration
potential of this instrument (650 V) proved incompatible with MALDI, which
generally employs accelerations on the order of 30 kV to overcome a large
energy spread and to reduce ion loss due to radial divergence. Signal levels
were simply too low to warrant further studies. The implementation of an ion
trap storage source (as described above) would allow a more successful

coupling.
6.5 References

1 Vlasak, P. R.; Beussman, D. J .; Enke, C. G. Proceedings of the 42nd
ASMS Conference on Mass Spectrometry and Allied Topics, Chicago,
IL, May 29 - June 3, 1994, 1036.

2 Beussman, D. J. Ph. D. Thesis, Michigan State University, 1995.

3 Beussman, D. J.; Erickson, T. A; Enke, C. G. (submitted to J. Am. Soc.
Mass Spectrom.)

4. Beussman, D. J .; Erickson, T. A; Vlasak, P. R.; Enke, C. G. (in
preparation for submission to Anal. Chem.)

5 Erickson, E. D.; Enke, C. G.; Holland, J. F.; Watson, J. T. Anal. Chem.
1990, 62(10), 1079.

6 Tecklenburg, R. E., Jr.; McLane, R. D.; Grix, R.; Sweeley, C. 0.;
Allison, J.; Watson, J. T.; Holland, J. F.; Enke, C. G.; Gruner, U.; Gotz,
H.; Wollnik, H. Proceedings of the 38th ASMS Conference on Mass
Spectrometry and Allied Topics, Tucson, AZ, June 3-8, 1990.

7 Holland, J. F.; Newcombe, B.; Tecklenburg, R. E., Jr.; Davenport, M.;
Allison, J.; Watson, J. T.; Enke, C. G. Rev. Sci. Instrum. 1991, 62(1),
69.

8 Ji, Q.; Vlasak, P. R.; Holland, J. F.; Enke, C. G. Proceedings of the
42nd ASMS Conference on Mass Spectrometry and Allied Topics,
Chicago, IL, May 29 - June 3, 1994, 1042.

10

131

Ji, Q.; Davenport, M. R.; Holland, J. F.; Enke, C. G. presented at The
43rd ASMS Conference on Mass Spectrometry and Allied Tapics,
Atlanta, Georgia, May 22, 1995.

Gardner, B. D.; Vlasak, P. R.; Beussman, D. J.; Enke, C. G.;
Watson, J. T. Proceedings of the 42nd ASMS Conference on Mass
Spectrometry and Allied Topics, Chicago, IL, May 29 - June 3, 1994,
1037 .

"1111111111111111111111111115