‘3"; "1' ,-

. €

:
D;

.x “ t .
g 7..
1.2;.
1... I
.12.
'0.’ f:—

I r
”mu... 5 kamvflinrqa
firuwgwnswuamefl. ... in...
Ha, if 16%»... .9.
. aidfimmwuni
. .

333.71....

3.3;... u

l: .I 31'

huffioflysxs
..

-" I)...
3.1.12

f.

inzr.‘

.. . 3.2.3.4..
It
.

1.3:. s u... o.
. «ii

.9
. . S
:15. 09190;...
SJ... { t
22.4“ :
,.. v: .v
an
. . 4! A1.
(.21..
. . .1 I.
at,

 

t:
u...

3.
.5 vxii.

1:21.}...
.17. 3......

.\:.,

‘2‘...)

.I.1.Cifl...t.l.t ‘
A. .. v i:
.w.\ .2.£v.».\
7! ii...

39“ ‘3”

“Wm”,

.. i.

, .

. .313...“ n

5 .55.}

prmsfin.
5.1 .

‘ l

vp. .u.... 3,.

 

 

THESIS

aIs ITY LIBRARIES

I II II

MICHI IGAN STA

III IIIIIIIIIIIII II IIIIIII

I 29

 

III

 

 

 

 

 

 

 

 

 

I

 

 

This is to certify that the

dissertation entitled

GOOD VIBRATIONS

presented by

Thomas M. Halasinski

has been accepted towards fulfillment
of the requirements for

Ph.D. Chemistry

degree in

 

 

fa/ i/A/ \

 

Major professor

Date May 2, 1996

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

LIBRARY

Michigan State
University

 

PLACE IN RETURN BOX to remove this checkout from your record.

. TO AVOID FINES Mum on or Moro dd. duo.

DATE DUE DATE DUE DATE DUE

I
-

-LI
-=

m

MSU chnN'flmIltIvo MIONEqud Oppomfluylmtflwon
m 7 WW1

 

 

 

 

 

 

 

 

 

 

II
I

 

 

 

 

 

 

E

  

 

 

 

   

I
I

 

 

 

 

 

 

I

 

 

m-.._

GOOD VIBRATIONS
By
Thomas M. Halasinski

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1996

ABSTRACT
GOOD VIBRATIONS
By

Thomas M. Halasinski

We have proposed instrumentation for the accumulation and infrared
spectroscopic characterization of mass-selected ions using matrix-isolation techniques. In
the method, both mass-selected cations and anions can be deposited with an inert host on
a window held at low temperature for subsequent spectroscopic examination. Our
ultimate goal is to couple mass-selective ion sources that produce sufficient numbers of
both positively- and negatively-charged species with the appropriate matrix-isolation
hardware to allow the determination of ionic structures by vibrational spectroscopy. Upon
completion of an ion source that produces suitable currents of mass-selected CF34", we
found that these cations can be trapped in a matrix, allowing for infrared measurements to
be carried out without the concurrent deposition of negatively-charged species. We have
also discovered that the addition of carbon dioxide to the neon matrix gas during
deposition of a mass-selected cation beam results in detectable quantities of C02". It is
shown that negative charge formation is strongly dependent on the temperature of the
metallic surfaces surrounding the matrix window and the presence of condensable gases
on these surfaces.

Vibrational spectroscopic characterization of neon matrices in which mass-selected
C82“ was deposited reveals absorptions due to CSf‘ and C82". A comparison of the
results from these investigations with those from COf‘ studies reveal several interesting
aspects of the matrix deposition process. For this set of mass-selected cations, no “matrix
chemistry” products, such as the production of H20+' through charge exchange with the

mass-selected cation, and no fragmentation products of the mass-selected cations are

observed. It is also determined that C02 is a better matrix additive than C82 for
counterion generation. The results from these experiments are used to interpret the data
from investigations of mass-selected organic cations such as the tropylium (C7H7+) cation.
We also report controlled annealing studies of the matrices in which clustering of ionic

species with neutral molecules is observed.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ vii
LIST OF FIGURES ...................................................................................................... viii
Chapter 1 Introduction and Research Objectives ............................................................. 1
1. Chemistry and the Structure of Molecular Species ...................................................... 1
II. Spectroscopic Approaches to the Structure Elucidation of Molecular Ions ................. 2
A. Gas Phase Optical, Infrared, and Microwave Molecular Ion Spectroscopy ......... 3
B. Photoelectron Spectroscopy and ZEKE Spectroscopy ....................................... 4
C. Electron Spin Resonance (ESR) Spectroscopy of Ions in Solution ..................... 5
D. X-Ray Spectroscopy ......................................................................................... 6
E. Superacids ......................................................................................................... 6
F. Isolation of Transient Species in Noble Gas Matrices ......................................... 7

III. A Proposal for Infrared Spectroscopy of Mass-Selected, Matrix-Isolated

Ions ........................................................................................................................ 8
IV. Preliminary Studies of Mass-Selected, Matrix-Isolated Ion Spectroscopy ................ 9
References ..................................................................................................................... 10
Chapter 2 Mass Selection of lonic Species .................................................................... 14
1. Mass Spectrometry and the Structure of Molecular Ions ........................................ 14
11. Identification of Ionic Isomers with Mass Spectrometry ........................................ 17
H1. The Quadrupole Mass Filter .................................................................................. 19
A A Qualitative Description of the Operation of a Quadrupole Mass Filter .......... 19
B. The Stability Diagram ...................................................................................... 22
C. The Transmission/Resolution Relationship of the Quadrupole Mass Filter ........ 28

iv

D. Ion Trajectories at the Exit of the Quadrupole ................................................. 30

E. Ion Focusing at the Quadrupole Exit ............................................................... 33
IV. Modification of the Quadrupole Mass Spectrometer Vacuum Chamber ................. 36
V. Design and Construction of the Ion Optical Components ....................................... 39
A. Ion Optical Design: SIMION and Ion Beam Visualization ............................... 39
B. Post-Quadrupole Focusing .............................................................................. 40
C. Deflection of the Ions into the Matrix Region .................................................. 42
D. Improved Post-Quadrupole Ion Beam Focusing .............................................. 46
VI. The Production of Positive Ions ............................................................................ 49
VII. Mass Spectral Currents at the Matrix .................................................................... 51
References ..................................................................................................................... 58

Chapter3 Infrared Detection of Mass-Selected, Matrix-Isolated Ions and

Counterions .................................................................................................. 63
I. Initial Predictions of the Processes Occurring during Cation Depositions ............... 63
II. Mass-Selection, Matrix-Isolation of CF3+‘ ............................................................. 68
III. Counterions .......................................................................................................... 71
IV. Detection of Counterions: The Allions Experiment ................................................ 72
V. Infrared Detection of the CF3CI" Anion ................................................................ 79
VI. Infrared Detection of the C02" Anion ................................................................... 83
VII. Mechanisms for the Formation of Countercharges ................................................. 85
A. Electrons from the Ion Source ......................................................................... 87
B. Field-Assisted Extraction of Electrons from the Substrate Holder .................... 88

C. Ionic Bombardment of Gas Phase and Adsorbed
Molecules near the Matrix Region ................................................................... 92
VIII. Further Discussions of Counterion Generation .................................................... 106
A. Auger Neutralization ..................................................................................... 108

B. The Chemical Reaction of Carbon Dioxide and Metal Surfaces ...................... 111

DC Positive Countercharges ...................................................................................... 114
X. Conclusion .......................................................................................................... 115
References................... ................................................................................................ 116
Chapter 4 Investigations of Mass-Selected CS2+' Depositions with Neon .................... 120
I. Introduction ........................................................................................................ 120
11. Experimental Parameters of the Mass-Selected Cation Depositions ..................... 121
III. Results of the Mass-Selected Depositions of CSf’ and CS+‘ ............................... 123

A. cs; ............................................................................................................. 129

B. C02" ............................................................................................................. 130

C. (C02 + 02)" .................................................................................................. 130

D. C82" ............................................................................................................. 134

E. CS ............................................................................................................. 134

F. cog“- ............................................................................................................ 135

G. Other Spectral Features ................................................................................. 136
IV. Annealing of the Matrix: The Cold Diffusion Experiment .................................... 137
V. A Comparison of the COZH" and CSZH" systems ................................................ 144
VI. Observation of Emission dun'ng Cold Diffusion of Matrix Components ............... 146
References ................................................................................................................... 149
Chapter 5 Investigations of Mass-Selected C7H7+ Depositions with Neon ................... 151
I. Application of the MS/MI Technique to Organic Cations .................................... 151
II. The Tropylium Cation and Mass Spectrometry .................................................... 152
III. Generation of Mass-Selected TrOpylium Cations ................................................. 154
IV. Previous Spectroscopic Investigations of C7H7+l0 and C7H8'HO ............................ 155
V. Results of Mass-Selected, Matrix-Isolated Tropylium Cation Studies .................. 156

VI. The Origin of Acetylene in Tropylium Cation Depositions ................................... 164

VII. Reduction of the Incident Mass-Selected Ion Kinetic Energy ............................... 167
VIII. Vibrational Intensity Information of the Matrix Components ............................... 169
IX Conclusions from the Mass-Selected C7H7+ Deposition Studies .......................... 171
References ................................................................................................................... 173
Chapter 6 Conclusions and Future Directions ............................................................. 176
1. Conclusions ........................................................................................................ 176
II. Possible Modifications to the Experimental Apparatus ......................................... 178
III. Future Studies ..................................................................................................... 179

A. Interstellar Molecular Ions ............................................................................. 180

B. Matrix Chemistry .......................................................................................... 181
References ................................................................................................................... 183
Appendix A Mass Spectral Ion Currents at the Matrix Region ..................................... 185

AppendixB Calculation of the Electric Potential at a Distance from a Charged

Disk ........................................................................................................ 193

vii

 

 

I:

12

Ta

Tal

Tab

Tabll

LIST OF TABLES

Table 1.1 Measured values for the symmetric stretch (VI) of CSf’ ........................... 5

Table 1.2 Comparison of the fundamental frequencies for HZOJ" in the gas
phase with those observed in neon ............................................................ 8

Table 3.1 A list of the ions that have been successfully detected by
mass-selected, matrix-isolation spectroscopy ......................................... 112

Table 4.1 Therrnochemical data (eV) for the neutral and ionic species of CX2
(X = O, S). All thermochemical data listed and values used to
compute the BDEs were taken from reference 4 ....................... I ............ 122

Table 4.2 Tabulated vibrational wavenumbers (cm") of the spectral bands
observed in Figure 4.1 A-E ................................................................... 127

Table 4.3 Calculated vibrational frequencies and intensities for the neutral and
ionic forms of C02, C82, CO, and CS. The corresponding
frequency and intensity values of each species containing a
carbon-13 isotope are given in parentheses ........................................... 128

Table 5.1 The observed vibrational frequencies and peak absorbance
intensities of matrix-isolated C2H2- The middle column lists results
from a direct deposition of Nezcsz 100021. The column on the
right lists the frequencies observed in a C7H7+ deposition which are
assigned to Csz- (A. U. = absorbance units) ....................................... 165

Table 5.2 A comparison of the calculated vibrational intensities of several
infrared-active modes of species pertinent to MS/MI investigations.
Available experimental values are also listed. The reference which
provided each experimental value is given as a superscript next to
the value ............................................................................................... 170

viii

Figure 2.1

Figure 2.2
Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6
Figure 2.7

Figure 2.8

Figure 2.9
Figure 2.10

Figure 2.11

Figure 2.12

LIST OF FIGURES

A schematic of the processes involved in the interpretation of a

mass spectrum ........................................................................................ 14
The mass spectrum of acetone ................................................................ 15
An illustration ofa quadrupole mass filter ............................................... 20

The Mathieu stability diagram for u = x. The shaded region in the
expanded area represents the region in which the values of a and q
result in stable ion trajectories. The curves designated 30,3 and b0,3
divide the diagram into regions of stability and instability ........................ 26

The overall stability diagram for the two-dimensional quadrupole

field. See text for details of region I and II ............................................. 27
Region I ofthe Mathieu stability diagram ................................................ 29
Ion trajectory for a single ion in the quadrupole mass filter ...................... 31

The preferred mass scan line superimposed onto the Mathieu

stability diagram ..................................................................................... 34
An illustration of the front view ofthe MS/MI instrument ....................... 37
An illustration ofthe top view ofthe MS/MI instrument ......................... 38

SIMION plots of poshquadrupole ion focusing. (a) An illustration
of a 90° divergent 27 eV ion beam exiting the mass filter. (b) The
effect of the einzel lenses on the ion beam as modeled by SIMION.
All electrodes are at 0 V unless otherwise indicated ................................ 43

A SIMION model of the deflection of the mass-selected ion beam
into the matrix region. The initial ion trajectories and energies were

F12

F112

Fla

F12

I

taken from the values in Figure 2.11 at the same position. All
electrodes are at 0 V unless otherwise indicated ...................................... 44

Figure 2.13 A plot of the m/z= 69 (CF3+) ion current versus position of the
Faraday plate .......................................................................................... 45

Figure 2.14 A plot of the m/z= 69 (CF3+) ion current versus position of the
Faraday plate .......................................................................................... 48

Figure 2.15 A schematic of the power supplies used to operate the ion source of
the mass spectrometer ............................................................................. 50

Figure 2.16 An illustration of the degradation of mass resolution due to
contamination of the quadrupole rods during lengthy MS/MI
experiments. (a) The mass spectrum of acetone taken with a clean
quadrupole. (b) The mass spectrum of acetone taken after
18 hours of continuous m/z = 43 ion deposition ...................................... 53

Figure 2.17 An illustration of the effect of oil contamination of the quadrupole
mass filter on the mass spectrum of acetone; (a) the mass spectrum
of acetone taken with an oil contaminated quadrupole mass filter,
(b) the mass spectrum taken after the quadrupole was cleaned ............... 54

Figure 2.18 (a) A plot of mass-selected CS} ion current on the Faraday plate
in front of the matrix substrate and at the retracted position versus
the ion energy. (b) A plot of mass-selected C6F6+' ion current
versus the ion energy .............................................................................. 55

Figure 2.19 An illustration of the occlusion of the mass-selected ion beam by
the radiation shield surrounding the matrix substrate ............................... 57

Figure 3.1 (a) A SIMION plot illustrating the deposition of a 100 eV CF3+
ion into a +99 V matrix. (b) A SIMION plot illustrating the
deflection of a 100 eV CF; ion from a +100 V matrix. The circle

surrounding the matrix represents the grounded vacuum chamber
walls ....................................................................................................... 65

Figure 3.2 The electric potential generated by placing 1x1016 charges on the
sample window ....................................................................................... 67

Figur

1'1ng

Figur

Flgm

HEW

Figure 3.3 Infrared absorption spectra in the 1700-1640 cm'1 region of neon
matrices after 11 hours of deposition of (a) m/z = 69 (CF 3+, 20 nA)
generated from electron impact of CF3Cl, (b) m/z = 120 (no ion
current) while subjecting CF3Cl to electron impact, and (c)

m/z = 50/51 (CF2+/CF2H+, 15 nA) generated from electron impact
of CF3H .................................................................................................. 69

Figure 3.4 Transient current observed at the Faraday plate located ~1 cm from
the substrate during warming of (a) a neon matrix after a 25 hour
deposition of 20 nA of CF3+ generated from electron impact of
CF3CI and (b) an argon matrix after a 25 hour deposition of 20 nA
of CF 3+ generated from electron impact of CF3CI, followed by
2 hours of 40 nA, 10 eV electron bombardment. The final
temperatures of the argon and neon matrices were ~40 and ~20 K,
respectively ............................................................................................. 73

Figure 3.5 Schematic illustration of floating various components in the matrix
region from ground while detecting current fi'om them. This
experimental setup can be used to detect current from a grid on the
C51 window, a copper plate in place of the window, and the
movable Faraday plate ............................................................................ 75

Figure 3.6 Currents observed on the Faraday plate from a warming neon
matrix after a 27 hour deposition of 40 nA of CO}. A heater on
the window holder is turned on at t=0 seconds. At t=30
seconds, the temperature of the neon matrix was ~40 K. The
potential on the Faraday plate during collection of the signal was
-70 V with the window holder grounded ................................................. 77

Figure 3.7 SIMION plots showing the potential surface formed between the
Faraday plate and the C51 window. Equipotential lines, separated in
value by 10 V, are shown. The CsI window is represented by a
dashed line in both plots. The grounded copper grid has been
added to part b. The Faraday plate has a potential of -70 V, and all
other surfaces are at a potential of 0 V in both plots ................................ 78

Figure 3.8 Currents observed from a 1700:l NeC02 matrix after a 20 hour
deposition of 30 nA of 0,117+ generated by electron impact
ionization of toluene. Heating conditions are the same as in

Fig

Pia
,u

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

Figure 3.14

Figure 3.15

Figure 3.16

Figure 3.17

Figure 3.18

Figure 3.6. The potential on the Faraday plate during collection of
the signals was +70 V with all other elements held at 0 V ........................ 80

Infrared absorption spectrum of an argon matrix in the
960-900 cm'1 region following a 27 hour deposition of 20 nA of
CF3+ formed by electron impact ionization of CF3C1. This
spectrum was obtained by averaging 256 scan files taken with
1.0 cm’1 resolution .................................................................................. 82

Infrared absorption spectrum of a neon matrix in the
1750-1350 an1 region following a 20 hour deposition of 40 nA of
co; ..................................................................................................... 84

Infrared absorption spectrum of a 1000:] Ne2C02 matrix in the
1750-1350 cm’1 region following a 13 hour deposition of 20 nA of
CF3+ formed by electron impact ionization of CF3Cl ............................... 86

Plots of the electric potential and the electric field versus distance
from a point charge ................................................................................. 90

A scale model illustrating the relative sizes of a matrix-isolated
CF3CI molecule lying between a CF3+ ion and the wall of the

metallic window holder ........................................................................... 91

Illustration of two proposed mechanisms of counterion formation in
mass-selected, matrix-isolated cation experiments ................................... 93

Configuration 1. The experimental setup is indicated on the left,
data in the center, and interpretation on the right ..................................... 95

Configuration 2. The experimental setup is indicated on the left,
data in the center, and interpretation on the right ..................................... 96

Configuration 3. The experimental setup is indicated on the left,
data in the center, and interpretation on the right ..................................... 97

Configuration 4. The experimental setup is indicated on the left,
data in the center, and interpretation on the right ..................................... 98

xii

Figure 3.19 Configuration 5. The experimental setup is indicated on the left,
data in the center, and interpretation on the right ..................................... 99

Figure 3.20 SIMION plots illustrating the effect of floating the copper plate on
the window at +250 V on incoming 130 eV CF3+ ions .......................... 101

Figure 3.21 Infrared absorption spectrum of a 150021 NezC02 matrix in the
1750-1350 cm'1 region following an 18 hour deposition of30 nA of
C-,H-,+ generated by electron impact ionization of toluene. The
copper grid on the CsI window was floated at a potential of +80 V
during ion deposition ............................................................................ 105

Figure 3.22 SIMION plots illustrating possible flight paths of C02“ sputtered
off the radiation shield to a matrix held at a potential of +20 V. the
starting point of the anion in plot B is ~1 mm to the right from the
starting point in plot A. The electrode simulating the matrix is held
at +20 V in both plots, and in each case the anion starts with 0 eV
kinetic energy ....................................................................................... 107

Figure 3.23 Schematic diagram of possible electronic transition processes for
incident ions (X+) or excited neutral species (X') and metal
electrons em' ......................................................................................... 109

Figure 4.1 Spectra in the 1460-1080 cm'1 region of five
[ion —-> matrixzadditive] experiments ..................................................... 124

Figure 4.2 The growth of peak absorbances of several matrix components
during the deposition of the matrix which resulted in Figure 4.1A ......... 126

Figure 4.3 FTIR spectra of several regions of a neon matrix following a
24 hour [CSf‘ —9 Ne:COz 1000: l] deposition ...................................... 131

Figure 4.4 FTIR spectra of several regions of a neon matrix following a
24 hour [csf- —> Neco2 300:1] deposition ........................................ 132

Figure 4.5 (A) Relative change of the v3 C02 absorption, (b) relative change
in absorption of v3 CSf' and v3 C82", and (C) relative change in
the absorption representing the nonrotating and rotating v2 bands of
H20 at each annealing step, following a 17 hour
[13C82+' —-) Ne2COz] deposition. In each case,

xiii

F12

Fig

Fig

Fig,

%AA = {A(i K) - A(4 K)}/ A(4 K)} x 100, where A(i K) represents
the absorption of the spectral feature taken afler an annealing
excursion to a temperature of i Kelvins ................................................. 139

Figure 4.6 FTIR spectra in the 1690-1640 cm’1 region of a neon matrix at
several annealing temperatures following a 17 hour
[13CS2+' -9 Ne:COz] deposition ............................................................ 141

Figure 4.7 FTIR spectra in the 1650-1585 cm’1 region of a neon matrix at
several annealing temperatures following a 17 hour
[13CSz+' —9 NezCOZ] deposition ............................................................ 143

Figure 4.8 Visible emission detected, following a 24 hour [CSf‘ -—) NezCOz]
deposition, upon repetitive cycling of the temperature between 6
and 10 K, at a rate of~0.25 sec'l. ......................................................... 148

Figure 5.1 The mass spectrum oftoluene ........... 155

Figure 5.2 Infrared absorption spectra in the 1590-1250 cm’1 region of neon
matrices after 19 hours of deposition of (a) m/z = 91 (C7H7+,
40 nA) generated from electron impact of toluene, (b) m/z = 100
(no ion current) while subjecting toluene to electron impact, and (c)
m/z = 91 (C7H7+, 30 nA) generated from electron impact of
toluene, while the copper grid on the CsI window was floated at a
potential of +80 V during deposition (18 hours) .................................... 158

Figure 5.3 Infrared absorption spectra in the 800-700 cm'1 region of neon
matrices after 19 hours of deposition of (a) m/z=91 (C-,H—,+,
40 nA) generated from electron impact of toluene, (b) m/z = 100
(no ion current) while subjecting toluene to electron impact, and (c)
m/z = 91 (0,117+, 30 nA) generated from electron impact of
toluene, while the copper grid on the C51 window was floated at a
potential of +80 V during deposition (18 hours). These spectra
were taken with 0.5 cm’1 resolution and are the result of averaging
1024 scans ............................................................................................ 159

Figure 5.4 Infrared absorption spectra in the 2200-2000 cm‘1 region of neon
matrices afier 19 hours of deposition of (a) m/z=91 (C7H7+,
40 nA) generated from electron impact of toluene, (b) m/z = 100
(no ion current) while subjecting toluene to electron impact, and (c)

xiv

FIE

Fig

App:

APPe

m/z = 91 (C7H7+, 30 nA) generated from electron impact of
toluene, while the copper grid on the CsI window was floated at a
potential of +80 V during deposition ( 18 hours) .................................... 160

Figure 5.5 Infrared absorption spectra in the 3000-2700 cm'1 region of neon
matrices afier 19 hours of deposition of (a) m/z=91 (C7H7+,
40 nA) generated from electron impact of toluene, (b) m/z = 100
(no ion current) while subjecting toluene to electron impact, and (c)
m/z = 91 (C7H7+, 30 nA) generated from electron impact of
toluene, while the copper grid on the C51 window was floated at a
potential of +80 V during deposition (18 hours) .................................... 161

Figure 5.6 Infrared absorption spectra in the 3350-3110 cm"1 region of neon
matrices after 19 hours of deposition of (a) m/z=91 (C7H7+,
40 nA) generated from electron impact of toluene, (b) m/z = 100
(no ion current) while subjecting toluene to electron impact, and (c)
m/z = 91 (C7H7+, 30 nA) generated from electron impact of
toluene, while the copper grid on the CsI window was floated at a
potential of +80 V during deposition (18 hours) .................................... 162

Figure 5.7 Infrared absorption spectra in the 2070-1720 cm'1 region of neon
matrices after 19 hours of deposition of (a) m/z=91 (C7H7+,
40 nA) generated from electron impact of toluene, (b) m/z= 100
(no ion current) while subjecting toluene to electron impact, and (c)
m/z = 91 (C7H7+, 30 nA) generated from electron impact of
toluene, while the copper grid on the CsI window was floated at a

potential of +80 V during deposition (18 hours) .................................... 163
Appendix A] The mass spectrum of acetone .............................................................. 185
Appendix A2 The mass spectrum of carbonyl sulfide .................................................. 186
Appendix A3 The mass spectrum of methyl ethyl ketone ............................................ 187
Appendix A4 The mass spectrum of carbon disulfide .................................................. 188
Appendix A5 The mass spectrum of toluene ............................................................... 189
Appendix A6 The mass spectrum of methyl bromide .................................................. 190

Appendix A7 The mass spectrum of chlorotrifluoromethane ....................................... 191

Appendix A8 The mass spectrum of methane. The mass-selected ion source
operating parameters were not recorded for this spectrum ..................... 192

xvi

by 1

3w

dnw
"1321
Ha)
Strum
know
Unknc

a“? ret
1180 p

Chapter 1

Introduction and Research Objectives

1. Chemistry and the Structure of Molecular Species

The chemistry of molecular species is intimately related to the three-dimensional
arrangement of the chemical bonds between its atoms. Knowledge of the structure of a
molecule is paramount to the ability of describing a molecule’s reactivity and physical
properties. This idea of stereochemistry was first described independently in 1874 by
Jacobus Henricus van’t Hoff and Joseph Achille Le Bel. Van’t Hoff‘s theories were
published in extended form as La chimie dans l’espace,l in which he introduced the
concept of the tetrahedral carbon atom (which he defined as an asymmetric carbon atom).
Le Bel expanded upon Louis Pasteur’s work on the optical activity of organic compounds
by relating it to their stereochemical structure.2 Although their theories were not globally
accepted at first, they have become the foundation of modern chemistry.

The importance of the elucidation of the structures of molecular compounds has
driven the development of several analytical techniques throughout the 19003. Nuclear
magnetic resonance (NMR), infrared spectroscopy (IR), mass spectrometry (MS), and
x-ray crystallography are among the most utilized tools for determining molecular
structure. Extensive collections of data from the various techniques for compounds of
known structure have been compiled to assist in the determination of the structure of
unknown compounds. The major collections, however, only include neutral molecules.

Charged molecular species also play a vital role in chemistry. Organic radical ions

3 Molecular ions

are recognized as important intermediates in many chemical reactions.
also play important roles in plasmas, discharges, flames, and in extraterrestrial

environments, such as in the interstellar medium and planetary atmospheres.4 As with

01

i0:

slit

in t‘
and
is tl

meti

neutral molecules, one must determine the structural aspects of these ionic species to fully
understand their chemical nature. The implementation of the spectrosc0pic techniques
which can be applied to neutral molecules is not readily accomplished for ionic species.
The following section will discuss the reasons for this difficulty and briefly highlight the
current methods of obtaining spectra of molecular ions. By doing so, it is hoped to make
apparent the main shortcoming of the current methods: the inability to simultaneously
create suitable concentrations of ionic species and at the same time discriminate against all

other neutral and ionic species formed during the creation of the ion to be investigated.

II. Spectroscopic Approaches to the Structural Elucidation of Molecular Ions

There are three main obstacles which hinder the spectroscopic study of molecular
ions. First, because of the possibility of charge recombination, molecular ions are
extremely reactive as compared to neutral molecules. For this reason, ions possess very
short lifetimes if they are not isolated from ions of opposite charge. Second, the positive
charge on cationic species makes it difficult to achieve densities above 106 ions/cm3
(< 0.01 femtomole/cm3)5' to 108 ions/cm3 (< 1 femtomole/cm3),5b placing stringent
demands on the detection limit for most techniques. The third obstacle is a result of the
difficulty in generating most molecular ions. The production of organic ions is ofien
accompanied by the creation of various neutral and ionic fragments. These fragments are
difficult to separate from the species of interest, often generating ambiguity in the origin of
spectral features.

These obstacles, however, have not prevented significant spectroscopic advances
in the study of molecular ions. Techniques to investigate ionic structures in gas, liquid,
and solid phases have been developed. The foremost research proposal of this dissertation
is the redress of the main deficiency which is inherent in the current spectroscopic

methods. Since the experimental verification of this proposal relies heavily on the data

fror

tecl

old

of

Spt

by

Spt

Sc

im

16!

DH

i0]

Vii

el

5F

fi'om the current spectroscopic methods, a brief overview of several of the principal

techniques of molecular ion spectroscopy will be given.

A. Gas Phase Optical, Infrared, and Microwave Molecular Ion Spectroscopy

The emission spectra obtained from ionic species in flames or discharges is the
oldest method of spectroscopically investigating molecular ions.5b’ 6 The earliest spectra
of N24“, 02+; and CO)“ were obtained in this way.6b Electron-impact emission
spectroscopy has also been utilized to obtain spectra of molecular ions.7 The first
electronic absorption spectrum of a molecular ion was that of N2+', which was produced
by a flash discharge method in 1968.8 The initial molecular ion studies with microwave
spectroscopy in the laboratory occurred in 1974 when CO+ was detected.9 In 1977, H. A.
Schwartz reportedlo low resolution infrared absorption detection of H3O+onHZO (n = 0-6)
ions formed during pulsed radiolysis experiments. In 1980, T. Oka reported11 the high
resolution vibrational-rotational spectrum of v2 of Hf. Since these initial experiments,
numerous advances have permitted the detection of several other diatomic and polyatomic
ions. Detailed reviews of the methods that have been developed and the additional ions
that have been studied can be found elsewhere.12

When high resolution is available in the optical methods mentioned above,
vibrational information can often be obtained. In general, emission spectra can provide
ground state vibrational energy levels and absorption spectra can provide excited
electronic state vibrational energy levels. The techniques used to obtain the high densities
of ions required for detection with these methods, however, utilize harsh ionization
methods, such as electric discharges. For this reason, only diatomic and small polyatomic
ions have been studied with these methods.

Some significant techniques have been devised to simplify the assignment of the
spectra obtained with these methods. These include the development of velocity

modulation infrared laser spectroscopy by R. J. Saykally and coworkers.13 This method

2C
or)
fol
Unc

or}

makes use of the Doppler shifts of charged molecules to discriminate against the
overwhelmingly more abundant neutral species that exist in a plasma.

The multiple ionic species formed during ionization may also cause difficulties with
spectral assignments. Mass-selection of a particular ion, while discriminating amongst
other ions, results in a reduction of ion densities and precludes direct absorption
measurements and often fluorescence detection, as well. J. P. Maier and coworkers have
succeeded in measuring spectroscopic transitions of mass-selected ions in the gas phase
through an indirect two-photon photoionization approach.M A proposal to obtain the
fluorescence and absorption spectra of mass-selected ions in an ion cyclotron resonance

(ICR) ion trap has also been described.”

B. Photoelectron Spectroscopy and ZEKE Spectroscopy

The observation of vibrational fine structure in photoelectron (PE) spectra permit
the measurement of molecular ion vibrational frequencies. Likewise, photoionization mass
spectrometry (PIMS) can also provide molecular ion vibrational frequencies. These
techniques have arguably been the dominant source for such information. The theory
behind these methods and listings of the data gathered from them can be found
elsewhere.16

16" is about

The typical resolution of a conventional PE spectroscopy experiment
200 cm'1 and caution should be exercised when comparing the vibrational frequencies to
other more accurate techniques, such as emission or infrared absorption experiments. The
following table of measured values for the symmetric stretch (VI) of CSZ+° illustrates the
uncertainty in the frequencies that can be reported for molecular ions by PE spectroscopy

or PIMS.

Table 1.] Measured values for the symmetric stretch (VI) of CSf’

 

 

 

 

 

 

 

 

 

(v1) CSf' (cm'l) method year reference
595 PIMS 1979 17
615 emission 1984 18
617 emission 1976 19
618 emission 1980 20
624 PB 1968 21
633 PIMS 1978 22
660 PE 1988 23
676 PB 1980 24
677 PE 1984 25

 

 

 

The recent development of zero kinetic energy photoelectron spectroscopy (ZEKE
spectroscopy) has permitted spectral resolutions of better than 1 cm’l, often permitting
rotational analysis. Reviews of this technique are given elsewhere.26 Bondybey and
coworkers have reported a value of 616 cm"1 for the symmetric stretch of CSf‘ using
ZEKE spectroscopy.” This value is in good agreement with the values obtained through
emission experiments in the table above.

While ZEKE spectroscopy is an extremely useful method to obtain vibrational
frequencies of molecular ions, the detection of some of the vibrational modes may not be
allowed due to symmetry constraints. Furthermore, to measure the positive ion vibrational
frequencies, one must first produce the neutral species in the gas phase. Many ions do not

have stable neutral counterparts and therefore are not suitable for ZEKE analysis.

C. Electron Spin Resonance (ESR) Spectroscopy of Ions in Solution
Organic radical ions have played a key role in ESR spectroscopy. A review of
ESR spectra of radical ions in solution has recently been given.28 A proposal to optically

detect electron spin resonance of trapped, mass-selected gas phase molecular ions has also

been described.”

i01

by

C01;
acic
pen‘
0n 1
defin
tetrac
“Carbt
011 (:3
aWard,
extrem
1101 cal

ge“trill.

D. X-Ray Spectrosc0py

Although one may argue that the x-ray spectrum of a molecular species provides
irrefirtable evidence of geometrical structure in the solid phase, the difficulty in preparing
molecular ions in crystal form has severely hampered its contribution to molecular ion
structure data. However, T. Laube and coworkers have succeeded in crystallizing several
carbocations in the past decade. These ions include the 3, 5, 7-trimethyl-1-adamantyl
carbocationf'o the 1, 2, 4, 7-anti-tetramethyl-2-norbomyl cation,31 and the tert-butyl

2

cation.3 A review of the crystal structures of other carbocations can be found

elsewhere.33

E. Superacids

A particularly successful approach in the preparation and stabilization of molecular
ions for structural studies is the use of “superacids”, which has been developed principally
by George Olah. Superacids such as hydrogen fluoride-antimony pentafluoride
(HF-SbFs), are ~1018 times stronger than 100% sulfuric acid.34 In the 1960s, Olah and
coworkers discovered that stable, long-lived, molecular cations can be obtained in these
acidic systems. This preparation allows a variety of structural investigations to be
performed on the cations, including NMIL 1R, Raman, and x-ray spectroscopic studies.
On the basis of structural investigations, Olah reported in a landmark publication35 the
definition and differentiation of trivalent “classical” carbenium ions from penta- or
tetracoordinated “nonclassical” carbonium ions. He also introduced the term
“carbocation” for the cations of all carbon compounds. Based on his pioneering research
on carbocations and their role in the chemical reactions of hydrocarbons, Olah was
awarded the 1994 Nobel Prize in Chemistry. Although these studies have resulted in
extremely insightful characterizations of carbocations, the method of creating these ions is
not capable of producing a wide variety of ionic species, such as those that can be

generated by mass spectrometry.

sp=

W3

dc;
the.
thei

the I
15013
mole

studi

E Isolation of Transient Species in Noble Gas Matrices

The development of the matrix-isolation method for the investigation of unstable
species was initiated in 1954 by G. Pimentel and coworkers}6 An excellent review of the
early development of this method can be found in reference 37. During the late 19605, it
was also shown possible to use this technique to trap and detect molecular ions. In 1968,
L. Andrews reported the detection of the coulombically bound Li+ 02" species by

8 Shortly

depositing Li atoms and oxygen molecules in a growing argon matrix.3
thereafter, it was also shown possible to form molecular ions which are truly isolated from
their counterions in matrices. P. Kasai reported39 the detection of the B2116" anion in an

d40 the detection of

argon matrix with ESR, while D. E. Milligan and M. E. Jacox reporte
the C2" anion by its electronic absorption spectrum. Since these initial studies, the matrix
isolation method has been a dominant technique for spectroscopic investigations of
molecular ions, including infrared absorption spectroscopy. Reviews of the ionic species
studied and the current methods used to generate the ions can be found elsewhere.41

The advantages of this technique over all gas phase methods is the prevention of
chemical reactions of the ions and the maintenance of suitable concentrations for
spectroscopic analysis. The vibrational frequencies measured in the matrix also compare
well with gas phase measurements. In general, neon-matrix shifts are often less than 1%
and argon-matrix shifts are less than 2.5% from the gas phase values for most transient
species, including ions.42 The matrix shifts for the filndamental vibrations of H20+° are

listed below as an illustration of the agreement between values measured in the gas phase

and in solid neon:

COT

Prepog

numerc

Table 1.2 Comparison of the fundamental fi'equencies for H20+° in the gas phase with
those observed in neon.

H20+' v1 v2 v3 references
gas phase 3212.9 cm'1 1408.4 cm'1 3259.0 cm" 43, 44, 43
neon matrix 3182.7 cm'1 1401.7 crn'l 3219.5 cm" 45, 45, 45
matrix shift 0.9 % 0.5 % 1.2 %

 

 

 

 

 

 

 

It is apparent from this comparison of vibrational frequencies that the geometrical
structure of the HZO+° cation is not perturbed to any great extent when isolated in a neon
matrix.

Similar to several of the gas phase approaches to structural studies of molecular
ions, selection of the desired species from neutral precursors and other neutral and ionic
fragments is often lacking in these studies. A notable development in this area has been
made by M. Vala and coworkers. By means of a “crossed beam” configuration, Vala and
coworkers are able to scan the same sample matrix for both [R and UV/visible spectra.46
In this configuration, infrared absorptions can often be assigned to particular species by
correlating their grth during photolysis with their known electronic absorptions.

III. A Proposal for Infrared Spectroscopy of Mass-Selected, Matrix-Isolated Ions

As mentioned above, matrix-isolation spectroscopy is a major contributor to the
study of molecular ions. When both cations and anions are present in the matrix very
large concentrations of ionic species can be obtained. The low temperature of the inert
gas matrix prohibits the ions from recombination/neutralization or reactions with other
neutral species. The one major disadvantage, lack of selectivity, has formed the basis of a
proposal from this laboratory to mass-select a particular ion before its deposition into the
matrix." Generation of the ionic species outside the matrix also permits the use of the

numerous ionization methods developed in mass spectrometry. The current methods for

del
ma
Cur
sh0
durj

Crea

generating ions for matrix isolation, such as photoionization, are not capable of producing

the wide variety of ionic species that can be generated by mass spectrometric methods.

IV. Preliminary Studies of Mass-Selected, Matrix-Isolated Ion Spectroscopy

The enormous advantages inherent in mass-selected, matrix-isolation (MS/MI)
spectroscopy has motivated several research laboratories to develop such a method.“8
Before the period of work to be described in this dissertation, only electronic absorption
and laser-induced fluorescence (LIF) detection methods of the ionic species had been
accomplished. While it is possible to obtain some vibrational data of ionic species with
these methods, it was the goal of this laboratory to obtain IR absorption spectra of the
mass-selected, matrix-isolated ions.

In the initial LIF studies,” it was speculated that the low mass-selected ion
currents available with the existing apparatus was the limiting factor in the unsuccessful 1R
detection experiments. The research objective in this dissertation is the construction of a
mass-selected ion source capable of generating at least an order of magnitude more in ion
current to allow examination of the ionic species with infrared detection. It will also be
shown that the IR detection of MS/MI ions allows one to identify the counterions created

during cationic depositions. Moreover, the mechanism by which the counterions are

created is also investigated.

IQ

l0.
ll.
12.

10

References

1.

10.
ll.
12.

13.
14.

van’t Hoff, J. H. La chimie dans l ’espace; Rotterdam, 187 5 (translated into
English by Marsh, J. E. Chemistry in Space; Oxford, 1891).

Le Bel, J. A Bulletin de la Societe chimique de Paris 1874, 22, 337.

Roth, H. D. Tetrahedron 1986, 42, 6097.

(a) Smith, D.; Adams, N. G. T op. Cur. Chem. 1980, 89, 1. (b) Winnewisser, G.
Top. Cur. Chem. 1981, 99, 39.

(a) Allison, J.; Stepnowski, R. M. Anal. Chem. 1987, 59, 1072A. (b) Miller, T.
A; Bondybey, V. E. Phil. Trans. R Soc. Lond. 1982, A307, 617.

(a) Carrington, A In Molecular Ions: Geometric and Electronic Sructures;
Berkowitz, J., Groeneveld, K.-O., Eds; Plenum Press: New York, 1983; pp 1-9.
(b) Hezberg, G. Quart. Rev. Chem. Soc. 1971, 25, 201.

I-Iirota, K.; Hatada, M.; Ogawa, T. Int. J. Radiat. Phys. Chem. 1976, 8, 205.
Herzberg, G.; Lagerqvist, A. Can. J. Phys. 1968, 46, 2363.

Dixon, T. A; Woods, R. C. Phys. Rev. Lett. 1975, 34, 61.

Schwarz, H. A. J. Chem. Phys. 1977, 67, 5525.

Oka, T. Phys. Rev. Lett. 1980, 45, 531.

(a) Herzberg, G. The Spectra and Structure of Simple Free Radicals; Dover
Publ., Inc.: New York, 1971. (b) Molecular Ions: Spectroscopy, Structure, and
Chemistry; Miller, T. A, Bondybey, V. E., Eds; North-Holland: Amsterdam,
1983. (c) The Spectroscopy of Molecular Ions; Carrington, A, Thrush, B. A,
Eds; The Royal Society: London, 1988. (d) Ian and Cluster Ion Spectroscopy
and Structure; Maier, J. P., Ed.; Elsevier: Amsterdam, 1989.

Gudeman, C. S.; Saykally, R. J. Ann. Rev. Phys. Chem. 1984, 35, 387.

Danis, P. 0.; Wyttenbach, T.; Maier, J. P. J. Chem. Phys. 1988, 88, 3451.

lS.

l6.

24.

25.

26.

15.

16.

17.
18.
19.
20.
21.
22.

23.

24.

25.

26.

11

Jackson, G.; Huang, Y.; Guan, S.; Kim, H. 8.; Marshall, A. G. Proceedings of the
43rd ASMS Conference on Mass Spectrometry and Allied Topics; Atlanta, GA,
1995; p 1090.

(a) Bock, H.; Mollere, P. D. J. Chem. Ed. 1974, 51, 506. (b) Turner, D. W.;
Baker, C.; Baker, A. D.; Brundle, C. R. Handbook of Molecular Photoelectron
Spectroscopy; Wiley: New York, 1970. (c) Kimura, K.; Katsumata, S.; Achiba,
Y.; Yamazaki, T.; Iwata, S. Handbook of He] Phtotelectron Spectra of
Fundamental Organic Molecules; Halsted Press: New York, 1981. (d) Eland, J.
H. D. Photoelectron Spectroscopy; Butterworths: London, 1984. (e) Berkowitz,
J. Photoabsorption, Photoionization, and Phototelectron Spectroscopy, Academic
Press: New York, 1979.

Trott, W. M.; Blais, N. C.; Walters, E. A. J. Chem. Phys. 1979, 71, 1692.

Endoh, M.; Tsuji, M.; Nishimura, Y. Chem. Phys. Letters 1984, 109, 35.

Balfour, W. J. Can. J. Phys. 1976, 54, 1969.

Bondybey, V. B; English, I. H. J. Chem. Phys. 1980, 73, 3098.

Eland, J. H. D.; Danby, C. J. Int. .1. Mass Spectrom. Ion Phys. 1968, I, 111.

Frey, R.; Gotchev, B.; Peatman, W. B.; Pollak, H.; Schlag, E. W. Int. J. Mass
Spectrom. Ion Phys. 1978, 26, 137.

Wang, L.-S.; Reutt, J. E.; Lee, Y. T.; Shirley, D. A. J. Electron Spectrosc. Relat.
Phenom. 1988, 47, 167.

Hubin-Franskin, M.-J.; Delwiche, J.; Natalis, P.; Caprace, G. J. Electron
Spectrosc. Relat. Phenom. 1980, 18, 295.

Reineck, 1.; Wannberg, B.; Veenhuizen, H.; Nohre, C.; Maripuu, R.; Norell, K.-E.;
Mattsson, L.; Karlsson, L.; Siegbahn, K. J. Electron Spectrosc. Relat. Phenom.
1984, 34, 235.

(a) Wright, T. G.; Reiser, G. F.; Muller-Dethlefs, K. Chem. Ber. 1994, 30, 128.
(b) Schlag, E. W. Ber. Bunsenges. Phys. Chem. 1994, 98, 1389. (c) Fischer, 1.;

31.

32.

33.

34.

35.

36.

37.

39.

40.

41.

N:

27.

28.
29.

30.
31.
32.
33.
34.

35.
36.
37.

38.

39.

40.
41.

12

Lindner, R; Muller-Bethlefs, K. J. Chem. Soc. Faraday Trans. 1994, 90, 2425.
(d) Grant, E. R; White, M. G. Nature 1991, 354, 249. (e) Mflller-Dethlefs, K.;
Schlag, E. W. Ann. Rev. Phys. Chem. 1991, 42, 109.

Fischer, 1.; Lochschmidt, A; Strobel, A; Niedner-Schatteburg, G.;
Muller-Dethlefs, K.; Bondybey, V. E. Chem. Phys. Letters 1993, 202, 542.
Davies, A G. Chem. Soc. Rev. 1993, 22, 299.

Li, G.-Z.; Guan, S.; Dalal, N. 8.; Marshall, A. G. 43rd ASMS Conference on Mass
Spectrometry andAllied Topics; Atlanta, GA, 1995; p 807.

Laube, T. Angew. Chem. Int. Ed Engl. 1986, 25, 349.

Laube, T. Angew. Chem. Int. Ed. Engl. 1987, 26, 560.

Hollenstein, S.; Laube, T. J. Am. Chem. Soc. 1993, 115, 7240.

Saunders, M.; Jiménez-Vézquez, H. A. Chem. Rev. 1991, 91, 375.

Olah, G. A; Prakash, G. K. S.; Sommer, J. Superacids; “filey-Interscience: New
York, 1985.

Olah, G. A. J. Am. Chem. Soc. 1972, 94, 808.

Whittle, E.; Dows, D. A; Pimentel, G. C. J. Chem. Phys. 1954, 22, 1943.
Downs, A J .; Peake, S. C. In Molecular Spectroscopy; Chemical Society
Specialist Periodical Reports, 1973; Vol. 1, p 523.

(a) Andrews, L. J. Am. Chem. Soc. 1968, 90, 7368. (b) Andrews, L. J. Chem.
Phys. 1969, 50, 4288.

Kasai, P. H. J. Chem. Phys. 1969, 51, 1250.

Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1969, 51, 1952.

(a) Chemistry and Physics of Matrix-Isolated Species; Andrews, L., Moskovits,
M., Eds; Elsevier Science Publishing: Amsterdam, 1989. (b) Radical Ionic
Systems; Lund, A, Shiotani, M, Eds; Kluwer: Dordrecht, The Netherlands, 1991.
(c) Almond, M. J.; Downs, A. J. Spectroscopy of Matrix Isolated Species; Wiley:
New York, 1989.

47.

48.

42.
43.

44.
45.
46.

47.
48.

13

Jacox, M. E. Chem. Phys. 1994, I89, 149.

Huet, T. R.; Pursell, C. J.; Ho, W. C.; Dinelli, B. M.; Oka, T. J. Chem. Phys.
1992, 97, 5977.

Brown, P. K; Davies, P. B.; Stickland, R. J. J. Chem. Phys. 1989, 91, 3384.
Forney, D.; Jacox, M. E.; Thompson, W. E. J. Chem. Phys. 1993, 98, 841.
(a) Szczepanski, J.; Personette, W.; Vala, M. Chem. Phys. Letters 1991, I85, 324.
(b) Szczepanski, J .; Personette, W.; Pellow, R.; Chandrasekhar, T. M.; Roser, D.;
Cory, M.; Zemer, M.; Vala, M. J. Chem. Phys. 1992, 96, 35. (c) Szczepanski, J.;
Roser, D.; Personette, W.; Eyring, M.; Pellow, R.; Vala, M. J. Phys. Chem. 1992,
96, 7876. (d) Szczepanski, J.; Vala, M.; Talbi, D.; Parisel, O.; Ellinger, Y. J.
Chem. Phys. 1993, 98, 4494.

Sabo, M. 8; Allison, J.; Gilbert, J. R.; Leroi, G. E. Appl. Spectrosc. 1991, 45, 535.
(a) Forney, D.; Jakobi, M.; Maier, J. P. J. Chem. Phys. 1989, 90, 600. (b)

Zhang, X. K.; Parnis, J. M.; March, R. E. 40th ASMS Conference on Mass
Spectrometry and A [lied Topics; Washington, D. C., 1992; p 1767. (c) Knight, L.
B., Jr. in: Chemistry and Physics of Matrix-Isolated Species; Andrews, L.,
Moskovits, M., Eds; Elsevier Science Publishing: Amsterdam, 1989.

int

spt

F121

0f 11
meal
V3111:
rElati
frag"
Die .

Chapter 2

Mass Selection of Ionic Species

1. Mass Spectrometry and the Structure of Molecular Ions

The structure determination of molecular ions is not only important to the study of
intermediates in chemical reactions and the species involved in plasmas and extraterrestrial
environments, as mentioned in the previous chapter, it also plays a fundamental role in the
interpretation of mass spectral data. The processes involved in interpreting a mass

spectrum of a particular compound are shown schematically in Figure 2.1.

Structures

Relative Intensities +_ ’ Mechanisms
and m/z Values

Figure 2.1 A schematic of the processes involved in the interpretation of a mass spectrum.

In mass spectrometry, the molecular compound (analyte) is introduced into the ion source
of the mass spectrometer where it is ionized by electron impact or other appropriate
means. The molecular ions and fragment ions formed are separated by their mass/charge
value and are detected. The relative abundances of each ion formed is displayed as the
relative intensity of the detected signal at each mass/charge value. Proposed
fragmentation mechanisms are then given for acceptable structures of the ions formed.

The three elements of information, the mass spectrum, the proposed fragmentation

14

me
the

5011‘

wh.
fon

ail.

The
mas:
then
P031:
Pair

This

15

mechanisms, and the structures of the ions formed are then “pieced” together to deduce
the structure of the molecule. Thus, the structures of the ions formed within the ion
source play a crucial role in the interpretation of a mass spectrum.

The one piece of measured data (the mass spectrum), however, does not provide
what many chemists would consider definitive evidence of the correct structure of the ions
formed. The mass spectrum of a relatively simple molecule such as acetone, Figure 2.2,

will be used to illustrate this point.

 

relative intensity

 

 

 

41 114] AA‘ j_l 1 A A

1 I I I I

10 20 30 4O 5O

m/z

 

Figure 2.2 The mass spectrum of acetone.

The molecular ion of acetone formed in the ion source is represented by the peak at the
mass/charge (m/z) value of 58. These ions are formed with excess energy and a portion of
them fragment. Some of these molecular ions may undergo an inductive cleavage. The
positive charge, which is assumed to be located on the oxygen atom, attracts an electron
pair fiom a neighboring bond, and ultimately results in the fragmentation of a C—C bond.

This is illustrated below.

h)

ma

This
mec

Spec
Only

may I

16

, . to:
i? ,I, i

c +5 _. CH3+ + Hac—cz—b
/ \ / \
H3C CH3 H3C CH3

The ion represented by the peak at m/z = 15 must contain only one carbon atom and three
hydrogen atoms. The only chemically acceptable structure of this ion is the methyl cation.
The mechanism proposed above would make a very plausible process for its formation.
What is the structure of the ion represented by the most intense peak at m/z = 43?
This ion must contain one oxygen atom, one carbon atom, and three hydrogen atoms. A

mass spectrometrist would propose the following a-cleavage mechanism for its formation:

:04
"-3 . C . _5
C —’ H3 4' H3C—C_ :

/ \
H3C C H3

 

This mechanism, which produces the acetyl cation seems a plausible fragmentation
mechanism resulting in the ion represented by the peak at m/z=43. Yet, the mass
spectrum does not provide conclusive evidence that the ion really is the acetyl cation. It
only provides essentially the mass of the cation and the elemental composition.

In fact, besides the acetyl cation, there are other structures of C2H30+ that one

may consider. The acetyl cation and some of its isomers are listed below.

rise—cab: H2c=c+:—§H H3C—"=C: H2C—CH
1 2 3 4

ions 1

17

The stabilities of the gas-phase structures of these cations have been examined with ab
initio molecular orbital theory at the MP3/6-31G“ level with an added correction for

1 The acetyl cation (1) has been determined to be the

zero-point vibrational energies.
lowest energy isomer in the study. Structure (2), the 1-hydroxyvinyl cation, was
determined to lie 1.88 eV higher in energy than (1). The relative energy of structure (3)
was determined to be 2.24 eV higher in energy than (1). The next highest relative energy
isomer was determined to be (4), the oxiranyl cation, which lies at an energy of 2.53 eV
above (1).

These amounts of energy are very large when considering relative stabilities of
chemical species, yet these ions are commonly formed by 70 eV electron-impact within the
ion source of the mass spectrometer. The molecular ions which are formed have internal
energy distributions that may range from 0 to >20 eV.2 Some of the molecular ions and
fi’agment ions may contain enough internal energy to rearrange. Structures (1), (2), and

3 There is

(4) have all, in fact, been experimentally determined to exist in the gas phase.
not sufficient information in the mass spectrum to completely discount the possibility of
their formation. The fragmentation mechanism and the structure of the ions are, therefore,

only proposed entities.

II. Identification of Ionic Isomers with Mass Spectrometry

The lack of experimentally proven fragmentation mechanisms and structures of the
ions in the interpretation of mass spectra has certainly not deterred mass spectrometry
fiom becoming a powerful analytical tool. Its sharp spectral features and low detection
limits often make it a more useful qualitative technique than most other spectroscopic
techniques. Yet, the correct structures of the ions detected are still of great interest to
many mass spectrometrists.

Several mass spectrometric methods have been devised to investigate the chemical

structure of the ions produced in a mass spectrometer. Although these studies do not

18

provide information of the geometric structure of ions as most spectroscopic methods can
through unambiguous bond type, bond length, and bond angle information, they can often
difl‘erentiate between isomers by their different physical and chemical properties.

One of the most widely used techniques in identifying ionic isomers is the use of
tandem mass spectrometry (MS/MS). The mass spectrum of ions formed from the
fi'agmentation of a parent ion can be used for its identification, just as for neutral
molecules. A detailed review of MS/MS can be found elsewhere.4 In addition to the
study of ionic fragments of unimolecular dissociation processes, the neutral fragments may
also provide evidence for parent ion structure through a technique known as
neutralization-reionization mass spectrometry (NRMS).5 Other physical processes such as
charge-permutation reactions have also found extensive use in differentiating between
isomeric structures of organic ions.6

Ion/molecule reactions are of also considerable importance in the determination of
gas phase ionic structures. In these studies, it is the difference in chemical reactivities that
can differentiate ionic isomers. The use of the unique reaction of the tolyl cation with

7 is an example of this type

dimethyl ether to selectively detect it from other isomeric forms
of chemistry.

Although the mass spectrometric techniques briefly discussed above do not
provide direct information of ionic structure, the information that they do provide is often
usefiil in interpreting results from mass-selected, matrix-isolation (MS/MI) experiments.
In some instances, they are particularly USCfiJl in locating probable (formal) locations of
radical and/or charge sites. As will be shown in Chapter 5 for the C7H7+ isomers, the

information fi'om mass spectrometric experiments can be essential in the MSM studies of

some cations.

flux
bea:
utili.

inth

AA

ions t
forth i
the ab:
H. G. 1

Similar.

hl'perbo.
”along

19

III. The Quadrupole Mass Filter

As stated in the previous chapter, it is the goal of this dissertation to select only
ions of a particular m/z value to be isolated in an inert gas matrix for spectroscopic
analysis. The field of mass spectrometry is nearly 100 years old and several different types
of mass spectrometers have been developed over this period. All of them separate ions in
space by a uniquely different manner and detailed reviews of each technique can be found

8 Although other types of mass separators could possibly be utilized for the

elsewhere.
selection of ions for matrix isolation studies, only the quadrupole mass filter has been used
successfully to date for this purpose. The reasons are primarily due to its low cost and its
ability to mass separate a continuous beam of ions with low kinetic energies and high ion
fluxes. The biggest disadvantage of using the mass filter for these studies is the lack of ion
beam focusing upon exiting the quadrupole. The advantages and disadvantages of

utilizing quadrupoles for mass-selection in MS/MI experiments will be examined in detail

in the following discussion of the operating principles of the quadrupole mass filter.

A. A Qualitative Description of the Operation of a Quadrupole Mass Filter.

The possibility of using an electrodynamic quadrupole field for the separation of
ions was first recognized by Wolfgang Paul in 1953.9 A similar proposal was also put
forth independently by Richard Post10 in 1953, but was not widely recognized because of
the absence of formal publication. W. Paul shared the 1989 Nobel Prize11 in Physics (with
H. G. Dehmelt and N. F. Ramsey) for his work in isolating ions and electrons with devices
similar to the quadrupole mass filter.

The quadrupole mass filter (Figure 2.3) consists of four electrodes, ideally of
hyperbolic cross section, that are positioned in a radial array, although circular cross

sections are frequently used for economic reasons. The optimum quadrupole field can be

20

ion source quadrupole ion detector

 

 

 

 

 

 

 

 

 

llh—«t

 

Figure 2.3 An illustration of a quadrupole mass filter.

101

cur
rod
pot:
by c
the ,
illus
filter

will 1

mucl

21

established between the four rods when the circular radius of the electrodes (r) is related
to the quadrupole field radius (r0) by:12
r = 1.148ro.

Mass separation is achieved by a combination of radiofrequency (RF) and direct
current (DC) potentials with opposite polarities on the pairs of diametrically opposing
rods. The RF potential applied to one pair of the rods is 180° out of phase to the RF
potential applied to the other pair. The mass separation of the ions can be best visualized
by considering the effect of the fields on the trajectories of ions as they transverse through
the plane of the positive DC rods and the plane of the negative DC rods separately. For
illustrative purposes, only cations will be considered, although the operation of the mass
filter does not depend on the sign of the charge under consideration.

Along the plane of the positive DC rod axis, the trajectories of the lighter cations
will be most influenced by the applied RF field. These ions will tend to oscillate with a
much larger amplitude between the rods as compared to heavier ions which will not be
afi‘ected as much. Heavier cations will be focused towards the center of the axis along the
plane of the rods, and are more likely to be transmitted through the filter. Lighter cations
will be neutralized as they strike an electrode and are subsequently removed from the
analyzer region by the vacuum pumps. In this way, the positive DC rod axis will act as a
high pass mass filter. The negative DC rods will have an opposite affect on the cations.
Heavier ions will be most influenced by the negative voltage and will tend to be defocused
as they travel through the filter. Lighter ions will respond to the focusing action that
results when the positive portion of the RF field becomes larger than the static negative
potential. The negative DC rods will act as a low pass mass filter. The overall effect of
both the positive and the negative DC rods is to select only one stable mass which is
allowed to pass through the entire length of the electrodes. The mass-selected ions will be
heavy enough to avoid hitting the positive DC rods and light enough to avoid hitting the

negative DC rods. The opposite effect would apply for negative charges.

22

B. The Stability Diagram

The previous discussion of the operation of a quadrupole mass filter is only a
qualitative one, in the sense that it does not provide the details of the trajectories of ions in
a quadrupole field. To gain an appreciation of the relationship between the voltages
applied to the electrodes and the effective resolution of the device, a more quantitative
description of the motion of the ions as they pass through the filter is needed.

While the first quadrupole mass filter was constructed in the 19505, the equations
of motion which describe the trajectories of the ions through the filter were solved in 1868
by E. Mathieu.” The problem at hand in Mathieu’s time was not the trajectories of ions
through a quadrupole field, but rather the determination of the vibrational modes of a
stretched membrane having an elliptical boundary. The two different physical problems
are described by the same mathematical formulas.

Although most quadrupole mass filter electrodes are of circular cross section, the
following discussion of the equations of motion of the ions will assume a hyperbolic cross
section. For such a device, the potential of the quadrupole field between the rods can be
described by the time-dependent nature of the voltages applied to the electrodes and a

spatial factor which describes the structure of the electrodes”

<1> = [U+Vcos(a)t)]:-2;-Z:. (2.1)

U is the magnitude of the applied DC potential, V is the magnitude of the applied RF
waveform, (0 is the angular frequency (279'), t is time, x and y are defined as the distance
along the positive DC and negative DC rod coordinates (see Figure 2.3), respectively, and
r0 is the distance from the center axis (the z axis) to the surface of any electrode.

The magnitude of the electric field E, which is the change of potential with

distance, can be described along the three coordinates as:

dd) J:
E = --—— = - U V r —
x dx [ + cos(ar )] r 2 (2,2)

23

d y
0" <1)
E = -— = o
. a: (2.4)

The force F imposed on the ions is equal to the product of the magnitude of the

electric field E and the charge on the ion e:

F = eE (2.5)
If‘ = —[U+Vcos(a)t)]:—: (2,6)
F, = [U+Vcos(an)]r9,’- (27)
F. = 0. _ (2.8)

Since the potential between the electrodes is not a function of the distance along the
length of the electrodes, the electric field and the force imposed on a charge along the z
axis is 0. This results in no addition to the kinetic energy of the ions along the z axis as
they pass through the quadrupole. A 100 eV ion which enters the quadrupole will exit
with the same kinetic energy.

With the use of Newton’s law:

a
—t—x (2.9)

F=ma=nr

the mass of the ion can be substituted into formulas 2.5-8 as follows:

dzx ex

 

 

 

 

 

dt2 + mr°2[U+Vcos(a)t)]= 0 (210)
any — ey [U+Vcos(a)t)]= 0 (211)
3!2 mro2 '
dz:
. = 0 .
or (212)

Equations 2.10-2.12 describe the relationship between the position of a charged particle

within the quadmpole field and the time-dependent applied voltages to the electrodes.

24

The previous qualitative discussion of the operation of the quadrupole mass filter, in which
the influence of the potential on the positive and negative DC electrodes on the ion beam
was considered separately, is made possible because there are no cross-coordinate terms in
equations 2.10-12. These equations can be transformed into the canonical form of

Mathieu’s differential equation by the following substitution of variables:

 

 

 

(1 fi _ 4eU d _ 2eV
enea~ wzrzman q— m (2.13)
2 2
it: + i:1—[a+2q cos(a)t)]x= 0 (2,14)
52)) (02
dtz — T[a+2qcos(a)t)1y= 0 (215)
wt
defineé = —2- (2.16)

Application15 of the chain rule and the product rule allows one to write:
a" 2 u (02 a" 2u
—— = — —, hereu=xor 2.17
a t2 4 a; 2 W y ( )

Substitution of equations 2.16 and 2.17 into 2.14 or 2.15 results in the general form of the
Mathieu equation for both x and y:
d 2r
5 6

 

: + [a+2qcos(2§)]u= 0. (2.18)

It is the solutions to equation 2.18, the Mathieu equation, which are of
firndamental importance to the operation of the quadrupole mass analyzer. These
solutions, which represent the trajectories of an ion within the quadrupole, can be

determined by integrating equation 2.18, and may be written as:16

CD w
u = a'e” 2C2, e3”: + a"e”"f 2C2, cm": (2.19)

n=-m "a-”

where a’ and a” are integration constants depending on the initial conditions as the ion
enters the quadrupole, that is, the initial position no, the initial ion velocity bit/86, and the

initial phase of the RF frequency 4'0. The constants C2,, and ,u depend on the values of a

25

and q, but do not depend on the initial conditions. The solutions to equation 2.19 (the
values of u) can be classified into two types: bounded solutions and unbounded solutions.
The bounded solutions represent the stable ion trajectories which oscillate in the x-y plane
with limited amplitudes. They pass through the quadrupole without hitting the electrodes.
The unbounded solutions have a values that increase without limit as 5—r ao. These
values represent the ion trajectories which exceed r0 at some point in time along the path

of the ion through the quadrupole.

Boundary values of a and q of the stable solutions to the Mathieu equation have
been tabulated.”18 These boundary values are denoted a0, b1, a1, b2, and a plot of a
portion of the values for u = x is shown in Figure 2.4. Plots such as these are known as
Mathieu stability diagrams. The shaded regions between the boundary curves of the inset
figure represent the a and q values which provide stable ion trajectories through the
quadrupole along the x-axis. Figure 2.5 illustrates the u = y stability regions superimposed
onto the u = x stability diagram. The intersecting areas represent those values of a and q
which provide stable ion trajectories in both the x and y coordinates. If [U - Vcos(a)t)]
was chosen as the form of the time-dependent potential of the quadrupole field, then the x

and y stability regions in Figure 2.5 would be reversed.

There are several regions shown in Figure 2.5 in which ion trajectories are stable in
both the x and y coordinates. It can be envisioned that by changing the values of U and V
applied to the quadrupole, it is possible to sequentially transmit ions of different m/z values
through the mass filter (the charge of the ion will be denoted as 2 in the remainder of this
dissertation). Of the several regions in which this is possible, only the regions labeled I
and II in Figure 2.5 have been used in practice. Although the second region was originally
suggested for practical use in one of the original quadrupole field mass spectrometry

publications10 and was explored in later studies,19 only region I has been utilized in

26

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

 

 

/ / / 7/
/ x-stable/

’/

   
  

 

 

 

 

 

as be
a, a:
b2 b2
........ -20.._..........
b, at

a r

40 "

a1 b1
a0 a0
-60 "'
A l . l . l A A j A l A l A
40 -30 -20 ~10 0 10 20 30 40

Figure 2.4 The Mathieu stability diagram for u = x. The shaded region of
in the expanded area represents the region in which the values of a and q
result in stable ion trajectories. The curves designated 30.: and bmdivide
the diagram into regions of stability and instability.

27

l—-

/

1O
/ / ///////Vx-stable
/ / //

,//

y

 

/ / REGION ll

4/ REGION 1

 

 

 

 

 

\
s
‘1 ~ ~,
‘1‘ \
‘ . K
\
\ 1

s, “~\ \q \\ .
\\ \ \\
\\\ \ X ‘y-stable

‘\\;\\,\\
r \

q

\\\\~g\‘\\;\:,,§\“‘\ \ \ \ \\\

 

 

 

 

Figure 2.5

r -10

The overall stability diagram for the two-dimensional

quadrupole field. See text for details of region I and II.

28

commercial instruments. The impracticability of the higher regions of stability is due to
the higher U and Vvalues required.

An expanded view of region I is given in Figure 2.6. In principle, different masses
could be transmitted through the quadrupole by selecting a and q values which are
independent from one another. In practice, the ratio of a/q (2U/V) is held constant
throughout the operation of the quadrupole. The mass scan line superimposed onto the
Mathieu stability diagram in Figure 2.6 is a plot of the a and q values as the U and V
voltages applied to the quadrupole are scanned in a linear fashion. Since the mass of the
ion (m) is inversely proportional to both a and q (equation 2.13), the heavier masses are
represented by the a and q coordinates of the bottom left part of the line. For a particular
set of U and Vvalues, only those masses which lie along the mass line and are contained in
the x and y-stable region will be transmitted through the quadrupole. When the voltages
U and V are increased while keeping their ratio constant, the magnitude of the masses
which will be transmitted will also increase. Thus, the quadrupole will act as a mass filter
by scanning the U and Vvoltages in this way.

Figure 2.6 is also a useful illustration of how the mass resolution of the quadrupole
is dependent on the operating parameters of the quadrupole. By increasing the UW ratio,
the slope of the mass scan line will also increase. This will limit the number of masses
which will be simultaneously transmitted through the quadrupole, in effect increasing the
resolution. In a similar manner, by decreasing the UW ratio, the resolution is also
decreased. If the UW ratio is decreased by eliminating the applied DC voltage, the
quadrupole will act as a high pass mass filter. These devices, known as RF-only

quadrupoles, are USCfiJl as collision cells in triple-quadrupole mass spectrometers.”

C. The T ransmission/Resolution Relationship of the Quadrupole Mass Filter
The resolution of the quadrupole is often degraded to less than unity to attain

acceptable ion currents for use in the spectroscopic detection of ions by MS/MI.“ 22

29

 

owe
9°")
09%
(o
t9
y-unstable «A
region x-unstable
a region
x- y stable
region

 

Figure 2.6 Region 1 of the Mathieu stability diagram.

30

When the quadrupole is set at a high resolution by increasing the U/ V ratio, the number of
mass-selected ions that travel through the quadrupole is no longer limited by the ions
exiting the source but rather by the mass filter itself. This is known as the analyzer limited
condition.23 Thus, even though the U/V values are set to transmit ions of a particular
mass, all of the ions exiting the ion source with this mass may not pass through the mass
filter because of the physical boundary of the electrodes and the limitation to the area of
the quadrupole field contained inside. When this condition is present, the relationship
between the resolution (R) of the quadrupole and the transmission of the ions through it
(T) has been theoretically determined to be T = UK24 Therefore, the number of ions of a
particular mass which enter the quadrupole can be greatly increased by lowering the
resolution. At extremely low resolutions, however, the mass spectral peaks become
trapezoidal flat-topped peaks where the maximum intensity is no longer dependent on the

resolution.”

D. [on Trajectories at the Exit of the Quadrupole

The biggest disadvantage of using a quadrupole mass analyzer as the
mass-selection device is the properties of the ion beam as it exits the quadrupole. In the
MS/MI instrument previously described,21 the exit of the quadrupole is approximately
60 cm from the matrix region. It is much more difficult to focus the diverging beam from
a quadrupole as compared to the ribbon-like ion beam which is encountered in magnetic
sector instruments.26 This advantage of magnetic sectors for MS/MI is offset by the high
kinetic energies of the ions needed for separation by the magnetic fields used in these
instruments.

In quadrupole mass spectrometry, the RF frequency applied to the electrodes
causes the ions to undergo oscillations in both the x and y directions. The oscillations for

a single ion are illustrated in Figure 2.7. If one considers the ion upon exiting the

31

 

 

cox: $2: 28233. 2: E :2 295. a he >580an :3 EN 053m

 

32

quadrupole, it is easy to visualize that the x and y components of its trajectories will be
both non-zero, even if the initial values before entering the quadrupole were zero.

In addition to the contributions from the field inside the quadrupole to the x and y
components of the exiting ion trajectories, the fiinging fields created beyond the end of the
electrodes also will affect their trajectories.” At present, a complete theoretical model of
the overall trajectories ions take from their entrance to the quadrupole to points beyond
the exit is not available. The difficulty in producing such a model lies in the multitude of
variables involved. The ions emitted from the source start at different positions with
different velocity vectors. The ions also enter the quadrupole during different phases of
the RF-frequency applied to the electrodes. Moreover, as the ions are in the fiinging fields
of the quadrupole, the assumption of mutually independent motion in the x and y
directions (equations 2.10-12) is no longer truezm) The use of phase-space dynamics has
been useful in confronting the complexities in the calculations of ion displacements in
quadrupole fields?" “ *1

Experimental studies of ion trajectories at the quadrupole exit are also scarce. In
some ways, this is due to the detection systems used with quadrupoles. The common
electron-multiplier detector is held at a large negative potential and the ion collection
efficiency is very high. The paths taken by the exiting ions is not of a great concern to
most mass spectrometrists. Although quadrupole mass analyzers had been in existence for
30 years, it was not until 1978 that the cross shape cross-section of the ion beam at the
exit of the quadrupole had actually been directly observed (although the cross-shaped
patterns commonly formed by the ion beam on exit apertures after extended periods of use
had alluded to this pattern).28 Further studies have supported this observation.29

The divergence of the cross-shaped beam has also received little attention. In this
laboratory, an Ar+ ion beam exiting a quadrupole has been experimentally measured to
diverge up to ~68.9° along the -DC rod pair axis.29c These results also show that the

divergence increases with the mass of the selected ion. The effect of initial ion energy and

33

the resolution setting of the quadrupole on the extent of divergence has yet to be studied.
Clearly, more information about the nature of ion trajectories upon exiting the quadrupole
is needed to design an ion optical system which transports ions from the quadrupole exit

to the matrix region with little to no loss in ion current.

E. [on Focusing at the Quadrupole Exit

Though no device presently exists which can cleanly focus the entire ion beam
exiting the quadrupole, there have been a few ion optical techniques which have been
developed to confront this problem. These devices attempt to reduce the effects of the
fiinge fields of the quadrupole. Although the major application for these devices is in
improving transmission of the ions as they enter the quadrupole, the approach can be
extended to the exiting ions.

The physical basis behind these devices can be illustrated with the Mathieu stability
diagram. A stability diagram, similar to Figure 2.6 is shown in Figure 2.8. As explained
earlier, the ions which experience a and q values (U and V amplitudes) represented by the
area of the triangular region in the diagram undergo stable ion trajectories in the
quadrupole field. Those ions whose masses are lower or greater than the set mass will
undergo unstable oscillations and hit one of the electrodes. The ions which are desired to
be transmitted through the quadrupole experience lower electric fields as they travel
through the region immediately before and after the quadrupole than they do inside the
quadrupole. Most studies of the fringe fields assume the fields drop linearly in potential
with distance from the entrance, although a more realistic model has been proposed?"
These lower fields experienced by the ion place the desired mass outside the stability
region. In particular, it is the lower RF fringe field, represented by a lower value of V,
which places the desired ion outside of the stability region in Figure 2.8. That is, the

selected m/z value now lies to the left of the stability region. Therefore, by placing

additional electrodes (a “pre-filter”) at the entrance and exit of the quadrupole, upon

34

1 preferred mass scan line

y-unstable
region I
i x-unstable
, region
, ’ x-y stable
, -— ’ region

Cl

 

Figure 2.8 The preferred mass scan line superimposed onto the Mathieu
stability diagram.

35

which only the full RF potentials are applied, the trajectories of the ions will no longer be
unstable. A plot of the a and q values as the ions travel through the pre-filter to the
quadrupole may take on the shape of the curve denoted as the “preferred mass-scan” line
shown in Figure 2.8.30

This mode of operation is termed a “delayed DC ramp” and the device is often
referred to as a Brubaker lens after the originator of the device, Wilson M. Brubaker.30
Brubaker has also shown that a pre-filter utilizing a DC potential of an opposing polarity
relative to its pairing rod can cause the DC field in the vicinity of the ion entrance aperture

to approach zero.31

Wade L. Fite has patented a leaky dielectric material which appears
to RF fields as a dielectric but to DC fields as a conductor.32 This material is placed
between the poles of the mass filter to achieve the same effect as the delayed DC ramp
pre-filter. This device is known commercially as an ELFS (Extranuclear Laboratories
Field Separation) lens. The effect of RF-only pre-filters on the performance of
quadrupoles as mass analyzers has been studied in detail.33

As will be discussed later in this chapter, an electrostatic octopole lens28 was
ultimately utilized for the focusing of the ions after the mass filter. While other ion optical
devices may have proved beneficial to the focusing of the ion beam, such as the RF-only
filter, the decision to rely on the electrostatic octopole was based on two postulates. One
was that it would provide adequate focusing action on the ion beam as demonstrated in
reference 28. The second was that it could possibly act similarly to the RF-only pre-filter,
by providing electric fields to the fringe field area of opposite sign to the fiinge fields and
thereby reducing their effect. Electrostatic octopoles have been used in several

34 As will be shown in the

experimental instruments to focus ions exiting mass filters.
remainder of this dissertation, the electrostatic octopole and additional focusing elements
have proven to guide a large portion of the mass-selected ions from the exit of the
quadrupole to the matrix region. It will also be shown, that for some ions, the resulting

mass-selected ion current is large enough to detect the matrix-isolated ions with infrared

36

spectroscopy. However, if further generations of MS/MI instruments utilize quadrupoles

as the mass selection device, experimentation with other ion optical systems is suggested.

IV. Modification of the Quadrupole Mass Spectrometer Vacuum Chamber

As mentioned in Chapter 1, the first generation mass-selected ion source proved to
generate insuflicient ion currents for infrared detection of several cations by MS/MI.35
These cations included CSf‘, CHZNH2+, and NOL Following those initial studies, a
commercial Finnigan 3200 mass spectrometer was donated to this laboratory by the
Michigan State Police. The chemical ionization (CI) source capability of the Finnigan
3200, and the associated vacuum pumping capacity of this instrument, promised access to
much higher ion currents.

The modified Finnigan 3200 vacuum housing and its coupling to the previously
described matrix chamber housing35 is shown in Figure 2.9 (front view) and Figure 2.10
(top view). Figure 2.10 displays the removal of the 3200 detector and the placement of
the supports upon which the mass filter is secured in the vacuum chamber. To insure low
background pressures when introducing gaseous samples into the ions source region, the
solid probe inlet teflon o-ring assembly was replaced with a swagelock connection
assembly.

The ion source housing is pumped by a Varian VHS-4 diffusion pump (maximum
speed = 1,200 US [air]; maximum throughput = 2.5 Torr-lls; ultimate pressure
<5x10'9 Torr). The analyzer region is pumped by a Varian HS-2 diffusion pump with a
Varian 325 Cryotrap (maximum speed = 175 US [air]; maximum throughput =
0.55 Torr-Vs; ultimate pressure <5x10'9 Torr). Both diffusion pumps are backed by a
single Alcatel model ZM2012A two-stage rough pump (310 l/min). Both diffusion pumps
are filled with Dow Corning 705 diffusion pump fluid. The cryotrap is filled with liquid

N2 at least every 2 hours while the mass spectrometer is open to the matrix chamber.

37

4:08:35 292 05 no 33> 28¢ 05 mo ecu—98.: :< ad 2sz

i a .fi

 

 

 

 

 

 

\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

aEaa 1%
c2256
Logo—cm

9:3
.t 533:6
858 :2

 

 

 

 

 

 

 

conEmco
Easom>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

xEmE

\

.Emobo

 

 

 

 

 

9.9 6.5 mom
2% 29:3

.onEmco E339
ago 896on
mde

38

4:08:55 952 2: .«o 33> a3 05 mo guess—z :< 2 .N 25E

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

c2023
.1. \._.0_)_
t \. 29a
1. \cofimcmu mam. .350 min. 6.5
so: $353.68 283690 89
too 8.:0m111.\.l.\v\ 9
CO. DEN / .\
\ / \ l ‘ q=1
> Ema . . _ . . . _.
mumcmullllv ..................... fllllril...
o5m>oE \\/ __ . . T v... . l t J. 7 E
goes \4/

 

 

cozmfim.
new Swamps...“ / Esteem 1
xEmE

mouma mam. _o~c_o c.8200 8:: 850m 383

cozoozou 28233.38 39: co. coastfi

a2 new to. . . .
conEmco

xEmE

39

V. Design and Construction of the Ion Optical Components

As can be seen fi'om Figures 2.9-10, the vacuum diffusion pumps, which are large
in size, precluded the placement of the quadrupole mass filter inside the matrix vacuum
chamber as before. To guard against the diffusion of oil vapor from the diffilsion/rough
pumps into the cryopumped chamber during the initial pump down, it was necessary to
place a high-vacuum gate valve between the Finnigan and the matrix chamber. The exit of
the mass filter, as a result, is situated ~60 cm from the matrix region. Furthermore, the
area of the matrix substrate of the new Heliplex system21 is ~43% smaller than the matrix
substrate area used in the previous studies.35 These modifications place high demands on

the ion beam focusing components.

A. Ion Optical Design: SIMI ON and Ion Beam Visualization

The ion optics used for the focusing and deflection of the ions exiting the
quadrupole to the matrix were first modeled with SIMION 5.0,36 an electrostatic ion
optics computer program. Briefly, the trajectories of charged particles are computed in
SIMION via a modified fourth-order Runge-Kutta method. The potential and voltage
gradients at a specific point in the model are obtained by trapezoidal methods from linearly
interpolated potentials of neighboring points. In this way, the electric field near electrodes
and grids are computed. More detailed explanations of the methods used in computing
the electric field and ion trajectories can be found elsewhere.”

Although modeling programs such as SIMION 5.0 offer valuable insights into ion
trajectories for cylindrically-symmetric electrostatic situations, their use is of limited value
for non-cylindrically-symmetric time dependent situations like those found in quadrupoles.
Three-dimensional ion optics programs do existf"8 however, their high cost has made them
unavailable for general use. As will be explained, much of the mass-selected ion current is
lost immediately after the quadrupole in the present ion optical system. For this reason,

only those cations whose mass spectral peaks are relatively abundant have been considered

40

for IR investigation by MS/MI. If quadrupole mass filters continue to be utilized for these
studies in the future, the design of a more efficient ion optical system through
three-dimensional modeling is a necessity.

In conjunction with the computer modeling of ion trajectories, ion beam
visualization techniques were also performed. The technique of visually monitoring the
profile of the ion beam directly, with the use of a channel electron multiplier array
(CEMA) coupled to a phosphorescent screen, has been described previously?” While
useful for the visualization of the overall beam profile, it has been found in practice to be
difficult to visually monitor the relative intensities within the cross-sectional profile. The
profiles of the ion beams exiting the mass filter in this laboratory exhibit equal intensities
throughout the profile of the cross shaped pattern, regardless of the potentials applied to
the beam Visualizer. Other methods of monitoring real-time images of low energy ion

beams may prove to show more intensity variations within the profile.39

B. Post-Quadrupole Focusing

Studies of the divergence of the mass-selected ion beam exiting the Finnigan 3200
mass filter showed approximately the same degree of beam divergence as the mass filter
used in the previous apparatus. By using the same ion beam visualization technique as

29c

before, the following divergence angles were measured:

 

 

divergence along the divergence along the
m/z +d.c. rod axis (0) -d.c. rod axis (0)
0-100 scan 58.4 79.9
18 53.9 63 .0
28 73.1 74.0
43 59.8 85.0

 

 

 

41

While other methods for measuring the angle of divergence of low energy ion beams may
be more accurate,39 the large angles measured here seem reasonable values considering
the inherent ion oscillations and fiinge fields involved with the quadrupole.

Initially, it was thought sufficient to utilize an einzel type accelerate/decelerate
aperture lens system followed by a set of deflection plates to focus and steer the ion beam
to the matrix region. Briefly, an einzel lens consists of three aperture electrodes. The first
and third electrodes are normally held at the same potential (usually earth ground), and a
separate bias is applied to the middle electrode. If cations are to be focused, the middle
electrode is held at a negative potential for the accelerate/decelerate mode. As the ions
are accelerated in the region between the first two electrodes, they generally experience a
focusing effect. As they pass through the region between the second and third electrodes
they are decelerated and undergo a slight defocusing. With adjustment of the magnitude
of the potential applied to the middle electrode, an overall focusing effect can be created.
The term einzel is derived from the German word for single, since only one voltage, that
between the middle and outer electrodes, defines the system. For this reason, the kinetic
energy of the entering ion beam is conserved upon exiting the lens. A more detailed
review of einzel lens systems can be found elsewhere.37b

A decision was also made to place an aperture between the diffusion pumped
analyzer region and the cryopumped matrix chamber to limit the diffusion of pump oil
onto the cold surfaces of the cryostat and the cryopump. It was proposed to use two
einzel lenses for the focusing of the ion beam before and after the aperture. Since the
“best” modeled position of the second einzel lens coincided with the most convenient
place for an aperture in the vacuum chamber, the decision was made to use the first
electrode of the second einzel lens as the aperture.

Due to the limitation to the number of points in the potential array (16,000) in

SIMION, it was decided to model the post-quadrupole region and the final deflection

region in separate steps to maintain accuracy in the calculated electric field. Figure 2.11

42

displays the SIMION model of the first proposed post-quadrupole einzel lens system. A
SIMION plot with no electrodes is included to allow visualization of the original ion
beam. The angle of divergence in the unfocused ion beam is 90°. Ion kinetic energies
were chosen as 27 eV, the maximum energy available with the original Finnigan 3200

mass spectrometer.

C. Deflection of the Ions into the Matrix Region

Figure 2.12 displays the continuation of the ion trajectories calculated by SIMION
from Figure 2.11b into the ion deflection region. Only the left and right deflection plates
were modeled with SIMION. Although top and bottom deflection plates were also
constructed, the vertical alignment of the quadrupole and the matrix substrate suggested
that their presence was not critical. A three-electrode electrostatic deflection plate system
for low-energy charged beams has been described“0 and may be considered for future
modifications.

The actual mass-selected current at the matrix substrate is measured by placing the
movable Faraday plate (see Figure 2.10) in front of the matrix. The diameter of the
Faraday plate used in these studies is the same as the diameter of the matrix substrate
(1.9 cm). When the plate is directly in front of the matrix window, it lies ~2 mm in front
of the radiation shield. The change in the measured ion beam current on the Faraday plate
as it is moved across the matrix region provides information regarding the degree to which
the ion beam is focused onto the matrix region. If the ion beam is tightly focused onto the
substrate, the Faraday plate should only register high current readings when it is placed
directly in front of the matrix. Small to zero currents are expected when the Faraday plate
is far away from the matrix.

Ion optics were constructed according to the SIMION models shown in
Figure 2.11-12. The actual Faraday plate readings of mass-selected CF 3+ current, formed

from CF3Cl, as the plate was moved across the matrix region is shown in Figure 2.13.

43

.3822: 3.3550 «mo—c: 83> o “a 2n 8.5.38.0 =<
.2925 .3 v0.39: 3 Enos :2 05 co nonco— _o~=_o 2: no .oobo 2:.
Q: :25 m8:— ofi mixes Enos :2 >0 R “593% com a .3 5:233:
.2 3 sense :2 28288.28 co sea 222m :a some

 

 

 

 

 

 

 

 

 

 

 

> com-

 

 

 

 

 

.1—

 

 

 

.82865 832050 3.2:: 36> o 3 oh 83.820
=< dogma 2:3 05 «a ZN 053..— E 3:53 82? 05 Sec coo—S
203 $3.05 23 $138.3... :2 355 BE. :2on x59: 05 SE Ewen
:2 vogue—883E 2: do 5:853 2: mo :69: 2925 < «mm £sz

a Bag 34.3.3.3

> mNm+ leash.

\.\trrhllllrrt r

  

45

 

20
18 ~

16-

1 2 '1 .-.-O-O‘.‘.‘O~.‘._.‘.

CF,+ Ion Current (nA)

 

 

 

o i I ' U ' I ' fi fi I ‘
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Faraday plate position (inches)

I '

Mass-Selected Ion Source Operating Parameters

ion energy = 180 eV post-quad einzel lens = -2090 V
total filament emission = 8.30 mA post-aperture einzel lens = -289.6 V
ion source pressure = 16 mTorr left plate = -15.0 V

chamber pressure = 2.9x10" Torr right plate = +16] V

ion source collector = +351 V top plate = -0.5 V

ion source extractor = +9.3 V bottom plate = -l .9 V

ion source lens = -78 V

Figure 2.13 A plot of the m/z = 69 (CFJ) ion current versus position of
the Faraday plate.

46

The potentials on the various optical elements are denoted below the plot. These
potentials were chosen on the basis of the maximum possible current that could be
observed at the Faraday plate when placed in front of the matrix substrate. Although the
SIMION model has predicted a relatively focused ion beam, clearly, it could not be
experimentally realized with this particular ion optical system. Further adjustment of the

various potentials did little to improve the extent of focusing.

D. Improved Post-Quadrupole Ion Beam Focusing

In addition to the poor ion beam focusing at the matrix region, the magnitude of
the ion current at the substrate was disappointing. Measurements of the current
immediately after the quadrupole under similar operating conditions .. show that
200-300 M of mass-selected ion current can be generated. While 10 nA is approximately
an order of magnitude larger than was possible with the modified RGAs, it is evident that
a large percentage of the mass-selected beam is lost before it reaches the matrix region.

The experimental verification of the region in which most of the ion current was
lost was made difficult by the inability to continuously measure the magnitude of the
current along the ion optical path. Movement of the Faraday plate required breaking
vacuum, a time consuming procedure owing to the absence of gate valves between the
diffusion pumps and the chamber. Often, the maximum current at a particular position
would vary greatly among several measurements. These inconsistencies were mostly due
to the oil deposits which formed on the ion optical electrodes with each vacuum
break/pumpdown. The oil deposits, along with the further contamination of metal
surfaces due to ion bombardment, degrade the performance of each electrode, particularly
those of the mass filter.

It was possible, however, to measure ion currents of ~100 nA at the region of the
entrance to the post-aperture einzel lens. The results of attempts to observe the intensity

profile of the ion beam at this region were inconclusive. At positions of ~2 inches past the

47

post-quadrupole einzel lens, the entire phosphorescent screen of the CEMA assembly was
activated, showing no intensity variations, regardless of the potentials applied to the
CEMA Although conclusive evidence could not be obtained as to the exact reasons for
the difference between the SIMION models and experimental observations, it was
assumed that the major cause of the discrepancies was due to the inability to recreate with
SMON the potentials involved in the fiinge fields at the exit of the quadrupole.

Placement of a grounded aperture immediately after the quadrupole to reduce the
fiinge fields resulted in no noticeable improvement of ion beam focusing. The presence of
the gate valve and an oil diffusion limiting aperture between the quadrupole and the matrix
precluded the use of continuous electrostatic ion beam guides."l Two other options were
considered, the use of an RF-only filter and the electrostatic octopole. As previously
discussed, the decision was made to construct an octopole to provide both a focusing
action on the x and y oscillating ion beam exiting the quadrupole and also produce electric
fields that would reduce the effect of the fringe fields.

The electrostatic octopole was constructed as closely as possible to the dimensions
given in reference 28. The symmetry of the three-dimensional quadrupole fiinge field and
octopole field precluded its modeling with SIMION. Thus the optimal position of the
post-quadrupole einzel lens was speculated to be as close to the end of the octopole as
possible. Although beam visualization experiments showed only the same images found
without the octopole (the entire screen was illuminated with equal intensity), the current
measurements in the matrix region were strikingly different. Figure 2.14 plots the
mass-selected CF3+ ion current as the Faraday plate moves across the matrix region.
Clearly, the octopole is providing a more tightly focused beam, as evident by the
dramatically smaller currents measured with the retracted Faraday plate, and the greater

currents measured in front of the matrix substrate.

48

 

 

 

 

30 'l
4 ./.
/'/
25 7 (0
A i
g /
.. 20« /
5 .
l: r
:1
o ./
= /
.9. .
+ n /
LII! 4 C
O
/
5 1 0
o"/
4 /.l
0 ‘ rO'.‘.’.’.
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Faraday plate position finches)

Mass-Selected Ion Source Operating Parameters

ion energy = 101 eV post-quad einzel lens = -690 V
total filament emission = 7.58 mA post-aperture einzel lens = -76.6 V
ion source pressure = 10 mTorr left plate = -25.5 V

chamber pressure = 1.3x10" Torr right plate = +320 V

ion source collector = +328 V top plate = +7.1 V

ion source extractor = +354 V bottom plate = +3.7 V

ion source lens = -45 V

Figure 2.14 A plot of the m/z = 69 (CF,+) ion current versus position of
the Faraday plate.

49

VI. The Production of Positive Ions

All the cations generated for the MSM experiments to be discussed have been
produced with the commercial Finnigan 3200 electron impact (E1) ion source. On
average, 10-50 mTorr of neutral precursor have been utilized to increase ion production.
While higher pressures may be utilized, the ion source filament life is drastically reduced.
Experiments with the more enclosed chemical ionization (CI) ion source have shown that
it is not capable of producing the currents available with the E1 source. Presumably, the
smaller ion source apertures reduce both the magnitude of the electrons available inside
the ion volume and also the number of ions that are transmitted to the quadrupole mass
filter. Initially, the filaments used for electron generation were the stainless steel body
filaments (Finnigan part no. 30004-60020) with 0.007” dia. Rhenium wire. .These were
later replaced by a ceramic body filament (Scientific Instrument Services, Inc.
part no. FF3200) with 0.0045” x 0.009” Rhenium ribbon.

Although the design and placement of the ion optics in the ion source were left
unmodified, it was determined that the Finnigan 3200 ion source power supplies were not
sufficient for generating high ion currents. Figure 2.15 displays the power supply
configuration used for most of the MSM experiments. The ability to regulate the total
filament emission has not been included in the present design. The lack of regulation
requires the filament heater current to be manually decreased over the course of long ion
depositions to keep the filament emission constant and avoid filament burnout. Although
long ion depositions are possible under present conditions, it would be advantageous to
include filament regulation42 in future experiments. Additional commercial and home-built
voltage power supplies were utilized for the ion optics of the ion source and the
post-quadrupole optics as well.

When the ion kinetic energy is selected to be 130 eV by applying +130 V to the
ion volume, it has been determined that the electron energy which generates the maximum

mass-selected ion current is also in the 100-130 eV range. For the experiments to be

50

electron collector \

ion volume :—]—
\

6 HP 6516A DC Power +
Supply

 

 

i

 

 

 

 

l'_'_

e
e
filament ——_)?1

L.

'6 HP 6281A DC Power 1
Supply T

 

 

 

 

 

 

 

 

 

 

I

do

Keithley 169
multimeter

 

 

 

 

Figure 2.15 A schematic diagram of the power supplies used to operate
the ion source of the mass spectrometer.

51

discussed, the electron energy was kept the same as the ion kinetic energy. This is
brought about by keeping the ion source filament at ground potential and floating the ion
volume at 130 eV. Modifications to this setup to include variable electron energies at
constant ion volume potentials may be considered for filture experiments. The Finnigan

3200 quadrupole electronics were left unmodified.

VII. Mass Spectral Ion Currents at the Matrix

As discussed earlier, the transmission of ions through the quadrupole increases as
the resolution decreases. The resolution in the MSM experiments performed here has
been degraded to less than unity to achieve acceptable mass-selected ion currents for
infi'ared detection. The low resolution mass spectra of acetone, OCS, methyl ethyl ketone,
C82, toluene, CH3Br, and CF 3C1, and CH4 and the experimental parameters used to obtain
them are shown in Appendix A. These spectra were taken with scan speeds of 1 sec/amu
at the Faraday plate in front of the matrix region. The current from the Faraday plate was
converted to a voltage signal with a Keithley 610C picoammeter and digitized with
MacADIOS 411 data acquisition hardware and software.

Comparison of the spectra shown in Appendix A with the literature mass spectra
reveals that only in the studies of CH; formed from CH4 and C7H7+ formed from toluene
is there a significant current of ion intensity of a nearby m/z value. The significance of the
added ion intensity of C7H8+' for the C7H7+ investigation will be discussed in Chapter 5.
It should also be noted that the relative ion intensities in each mass spectrum shown do not
differ significantly from the literature relative intensities. The high neutral precursor gas
pressures used in the ion source, required to obtain such large ion currents, may cause
significant deviations in the relative intensities for other molecules not studied here.

The contamination of the quadrupole electrodes, as a result of the high ion source
pressures and ion currents used, degrades the performance of the quadrupole during

lengthy experiments. An illustration of the effect on the resolution of the quadrupole is

52

shown in Figure 2.16. Figure 2.16a displays the mass spectrum of acetone at the
beginning of an m/z = 43 ion deposition experiment. Figure 2.16b displays the degraded
resolution after 18 hours of continuous ion deposition. Although the lower resolution has
not impeded long depositions of the ions studied to date, it may be of some concern for
future studies. To reduce this problem without reduction of the ion currents, the spacing
between the quadrupole rods would have to be increased, along with appropriate
modifications to the potentials applied to the rods.

Diffusion and rough pump oil contamination of the quadrupole rods also lowers
the performance of the quadrupole, as shown in Figure 2.17. Figure 2.17a displays the
mass spectrum of acetone taken after a ~1 minute power outage to the instrument, during
which the diffilsion and backing pumps were turned off. Other than contamination due to
the oil vapor, which diffused into the analyzer region, the quadrupole was clean.
Figure 2.17b displays the spectrum after the oil was removed from the quadrupole. To
maintain optimal performance of the mass filter and ion optical components, the mass
spectrometer vacuum chamber is actively pumped for only 1 or 2 days before the start of
an ion deposition experiment. The advantage of extended periods of vacuum pumping to
achieve low background pressures is offset by oil contamination of the mass filter, in this
particular mass spectrometer.

The kinetic energy of the mass-selected ions, used to produce the mass spectra
shown in Appendix A and used in all the experiments to be discussed here, have all been
~l30 eV. This relatively high energy is determined by the positive potential applied to the
ion volume element of the ion source, and is utilized to attain large ion currents. As
Figure,2.18a-b displays, the maximum ion current is achieved over a relatively narrow
range of ion energy. It is not clear why such high energies are needed for maximum
current. As explained in Section III of this chapter, the oscillations that the ions undergo

as they traverse through the quadrupole is fiindamental to the mass-selection process. The

53

 

35

eo- a

20"

 

15”

Ion Current (nA)

 

 

 

 

m/z —->

 

35

25"

20'

Ion Current (nA)

q

 

 

 

m/z ——->

Figure 2.16 An illustration of the degradation of mass resolution due to
contamination of the quadrupole rods during lengthy MS/MI experiments.
(a) The mass spectrum of acetone taken with a clean quadrupole. (b) The
mass spectnlm of acetone taken after 18 hours of continuous m/z = 43 ion
deposition.

54

 

35

25'

20’

15'

Ion Current (nA)

10'

j

 

 

 

m/z ——>

 

35
b

25"

20- V

15”

1.. (l /

 

 

 

Ion Current (nA)

 

N

 

 

 

m/z —->

Figure 2.17 An illustration of the effect of oil contamination of the
quadrupole mass filter on the mass spectrum of acetone; (a) the mass
spectrum of acetone taken with an oil contaminated quadrupole mass filter,
(b) the mass spectrum taken after the quadrupole was cleaned.

55

 

 

 

 

 

 

 

 

 

80
Ion current A
measured on the 0 ' o '
A 60 ’ Faraday plate in / \
E front of the matrix '
V * substrate \
E o
a)
s . \
0 4o .
e \
"" O
4-
N
U) 0
0 Ion current measured
20 ’ on the retracted
Faraday plate ’0 o
r 0'0
0/. O 0 CI
9' o
0 M 1 1 1 . 1 1 W000 1 A 1 1 .
0 20 40 60 80 100 120 140 160 180
Ion Energy (eV)
35
b
3° " Ion current \
l measured on the .
A Faraday plate in
E 25 _ front of the matrix
V substrate
78 . °
': 20 “
a
o /
5 15~ -
+
0
u /
0" 10 - o
. /
O
5 ~ ’.
l ,0’.
. I
o 4. 1 x L 4 1 A l A L n l M J
o 20 40 60 80 100 120 140
Ion Energy (eV)

Figure 2.18 (a) A plot of mass-selected C82+ ion current on the Faraday
plate in front of the matrix substrate and at the retracted position versus the
ion energy. (b) A plot of mass-selected CfiF6 ion current versus the ion

energy.

56

faster the ions travel through the quadrupole, the fewer oscillations will be undertaken.
Thus, the resolution decreases and the transmission will increase.

Another factor that may play a part the ion energy/maximum current relationship is
the change of focusing with higher energies. It might be rationalized that faster ions spend
less time in the fiinge fields of the quadrupole and diverge to a lesser extent. However, as
Figure 2.18a shows, the ion beam profile at the matrix increases with ion energy. Since all
potentials applied to the ion optics were unchanged as the measurements in Figure 2.18
were made, further studies need to be undertaken to establish any relationship between the
ion beam focusing and ion energy. The effect of this high ion kinetic energy on the
deposition of mass-selected ions into a growing matrix is discussed in Chapter 3.

The position of the Faraday plate for current measurements has been determined
by placing an additional Faraday plate in place of the matrix substrate and measuring the
ion currents at both plates as the one in front of the matrix is moved across the matrix
region, as shown in Figure 2.19. When the current at the matrix substrate drops to zero,
the Faraday plate in front of the matrix has completely eclipsed the matrix. From
Figure 2.19, it is determined that the optimal position is at ~1.37 inches from the retracted
position. Measurements such as this should be occasionally performed to make certain
that the ion beam is being correctly focused onto the matrix region during ion depositions.

A further observation from Figure 2.19 is that a portion of the ion beam is blocked
by the radiation shield. It can be calculated from this particular plot that ~20% of the
incoming ion beam is blocked by the radiation shield. Similar measurements for other ion
beam experiments have resulted in up to ~3 8% blockage. The intensity of the measured
ion current should be reduced appropriately when considering the actual current that
reaches the matrix. As a final note, the two plots in Figure 2.19 are not symmetrical since
the support arm of the Faraday plate also blocks some of the ion beam as the plate is

moved past the matrix region.

57

 

 

 

 

 

12
1 Ion current measured
on the Faraday plate . 0 o . .
10 _ in front of the matrix \ . 9 0 .
substrate ./ ‘0‘
/ .\
2 l /. .\
< O O . .\
5 8Tb 0000000000000 / .\.
'10-! \ C \
cc) 0\ / '\
h 0 . O
s. )( \
a co
0 6 ‘ ./ \ Ion current measured 7
C / 0 on the Faraday plate
2 . \ in place of the matrix
+ ,/ 0\ substrate
u? 4 1 ./
/ O
0 o \ / 00
/'l 0 CI
0 \ o
. I
2 ‘ o . . O /O O
O . . \ xo
0 0 o o
q \ ,0/
O\ O
0 ' 1 ' 1 ' 1 ‘ I 1 r r 1 0090 1 ' I '
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Faraday Plate Position (inches)

Figure 2.1.9 .An illustration of the occlusion of the mass-selected ion beam
by the radratron shield surrounding the matrix substrate.

References

1. Nobes, R. H.; Bouma, W. J.; Radom, L. J. Am. Chem. Soc. 1983, 105, 309.

2. McLafi‘erty, F. W. Interpretation of Mass Spectra, 3rd ed.; University Science
Books: CA, 1980, p 103.

3. Burgers, P. C.; Holmes, J. L.; Szulejko, J. E.; Mommers, A. A. and Terlouw, J. K.
Org. Mass Spectrom. 1983, 18, 254.

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

5. (a) Holmes, J. L. Mass Spectrom. Rev. 1989, 8, 513. (b) Wesdemiotis, C.;
McLafl‘erty, F. W. Chem. Rev. 1987, 87, 485. (c) Terlouw, J. K.; Schwarz, H.
Angew. Chem. Int. Ed. Engl. 1987, 26, 805.

6. (a) Levsen, K.; Schwarz, H. Mass Spectrom. Rev. 1983, 2, 77. (b) Kingston, E.
E.; Beynon, J. H.; Ast, T.; Flammang, R.; Maquestiau, A. Org. Mass Spectrom.
1985, 20, 546.

7. Heath, T. 6.; Allison, J.; Watson, J. T. J. Amer. Soc. Mass Spectrom. 1991, 2,
270.

8. Skoog, D. A. Principles of Instrumental Anabrsis, 3rd ed.; Saunders College
Publishing: New York, 1985.

9. (a) Paul, W.; Steinwedel, H. Z. Naturforsch., T eil A 1953, 8, 448. (b) Paul, W.;
Steinwedel, H. German Patent No. 944, 900, June 28, 1956. (c) Paul, W.;
Reinhard, H. P.; von Zahn, U. Z. Phys. 1958, 152, 143. (d) Paul, W.; Steinwedel,
H. 1960, US. Pat. 2 939 952.

10. Post, R. F. Univ. of California Rad Lab. Rept. 1953, UCRL-2209.

11. Paul, W. Angew. Chem. Int. Ed. Engl. 1990, 29, 739.

12. Dayton, 1. E.; Shoemaker, F. C.; Mozley, R. F. Rev. Sci. Instrum. 1954, 25, 485.

58

13.

14.

15.

16.

17.
18.

19.

20.
21.

22.
23.
24.
25.
26.

27.

59

Mathieu, E. Jour. de Math. Pures et A ppliquées (Jour. de Liouville) 1868, 13,
137.

Quadrupole Mass Spectrometry and Its Applications; Dawson, P. H., Ed.;
Elsevier: Amsterdam, 1976.

(a) Campana, J. E. Int. J. Mass Spectrom. Ion Phys. 1980, 33, 101. (b) Miller, P.
E.; Denton, M. B. J. Chem. Ed. 1986, 63, 617.

McLachlan, N. W. Iheory and Applications of the Mathieu F uctions; Oxford
University Press: Oxford, 1947.

Ince, E. L. Proc. Royal Soc., Ed. 1925, 46, 20.

Computation Laboratory, National Bureau of Standards, Tables Relating to
Mathieu Functions, Columbia Univ. Press: New York, 1951.

(a) Pedder, R. E.; Schaefl‘er, R. A. Proceedings of the 43rd ASMS Conference on
Mass Spectrometry; Atlanta, GA, 1995; p 1069. (b) Hiroki, S.; Abe, T.;
Murakami, Y. Int. J. Mass Spectrom. Ion Phys. 1994, 136, 85. (c) Hiroki, S.;
Abe, T.; Murakami, Y. Rev. Sci. Instrum. 1991, 62, 2121. (d) Dawson, P. H. J.
Vac. Sci. Technol. 1974, 11, 1151.

Yost, R. A.; Enke, C. G. J. Amer. Chem. Soc. 1978, 100, 2274.

Halasinski, T. M.; Godbout, J. T.; Allison, J.; Leroi, G. E. J Phys. Chem. 1994,
98, 3930.

Forney, D.; Freivogel, P.; Fulara, J.; Maier, J. P. J. Chem. Phys. 1995, 102, 1510.
Dawson, P. H. Mass Spectrom. Rev. 1986, 5, 1.

Dawson, P. H. Int. J. Mass Spectrom. Ion Processes 1990, 100, 41.

Dawson, P. H.; Whetten, N. R. Dynam. Mass Spectrom. 1971, 2, l.

O’Connor, P. J.; Leroi, G. 13.; Allison, J. J. Am. Soc. Mass Spectrom. 1991, 2,
322.

(a) Titov, V. Int. J. Mass Spectrom. Ion Phys. 1995, 141, 13-65. (h) Jarvis, K.
13.; Gray, A. L.; Houk, R. S. Handbook of Inductively Coupled Plasma Mass

28.

29.

30.
31.
32.

33.

6O

Spectrometry; Chapman and Hall: New York, 1992, p 42-44. (c) McIntosh, B. J .;
Hunter, K. L. Int. J. Mass Spectrom. Ion Phys. 1989, 87, 165. ((1) Hunter, K. L.;
McIntosh, B. J. Int. J. Mass Spectrom. Ion Phys. 1989, 87, 157. (e) Dawson, P.
H. Int. J. Mass Spectrom. Ion Phys. 1980, 36, 353. (f) Dawson, P. H.; Meunier,
M.; Tam, W.-C. Adv. Mass Spectrom. 1979, 88, 1629. (g) Dawson, P. H. Int. J.
Mass Spectrom. Ion Phys. 1977, 25, 375. (h) Dawson, P. H. Int. J. Mass
Spectrom. Ion Phys. 1975, 17, 447. (i) Dawson, P. H. Int. J. Mass Spectrom. Ion
Phys. 1975, I 7, 423. (j) Dawson, P. H. Int. J. Mass Spectrom. Ion Phys. 1974,
I4, 317. Brinkmann, U. Int. J. Mass Spectrom. Ion Phys. 1972, 9, 161. (l)
Dawson, P. H. J. Vac. Sci. Technol. 1972, 9, 487. (m) Dawson, P. H. Int. J.
Mass Spectrom. Ion Phys. 1971, 6, 33. (n) Brubaker, W. M. Proceedings of the
Fifth International Instruments and Measurements Conference, Stockholm, 1960;
Academic Press Inc.: New York, 1961, p 305.

Birkinshaw, K.; Hirst, D. M.; Jarrold, M. F. J. Phys. E: Sci. Instrum. 1978, 11,
1037.

(a) Weaver, H. 13.; Mathers, G. E. Dynamic Mass Spectrometry (D. Price and J.
F. J. Todd, eds), vol. 5, p41, Heyden, London, 1978. (b) Campana, J. E.; Jurs, P.
C. Int. J. Mass Spectrom. Ion Phys. 1980, 33, 119. (0) Gilbert, J. R.; Leroi, G.
13.; Allison, J. Int. J. Mass Spectrom. Ion Phys. 1991, 107, 247. ((1) Short, R. T.;
Grimm, C. C.; Todd, P. J. J. Am. Soc. Mass Spectrom. 1991, 2, 226. (e) Kane, T.
13.; Angelico, V. J.; Wysocki, V. H. Anal. Chem. 1995, 67, 1019.

Brubaker, W. M. Adv. Mass Spectrom. 1968, 4, 293.

Brubaker, W. M. J. Vac. Sci. T echnol. 1973, 10, 291.

(a) Fite, W. L. Rev. Sci. Instrum. 1976, 47, 326. (b) Ketkar, S. N.; Fite, W. L.
Rev. Sci. Instrum. 1988, 59, 987.

(a) Simion, M.; Bohatka, S.; Trajber, Cs.; Futo, 1. Rapid Commun. Mass
Spectrom. 1995, 9, 629. (b) Trajber, Cs.; Simion, M.; Bohétka, S.; Futb, I.

34.

35.
36.

37.

38.

39.

40.

61

Vacuum 1993, 44, 653. (c) Trajber, Cs.; Simion, M.; Bohatka, S. Rapid Commun.
Mass Spectrom. 1992, 6, 459. (d) Trajber, Cs.; Simion, M.; Csatlos, M. Meas.
Sci. Technol. 1991, 2, 785.

(a) Jarrold, M. F.; Bower, J. E.; Kraus, J. S. J. Chem. Phys. 1987, 86, 3876. (b)
Shul, R. J .; Upschulte, B. L.; Passarella, R.; Keesee, R. G.; Castleman, Jr., A W.
J. Phys. Chem. 1987, 91, 2556. (c) Jarrold, M. F.; Birkinshaw, K.; Hirst, D. M.
Mo]. Phys. 1980, 39, 787. (d) Dawson, P. H.; Meunier, M.; Tam, W.-C. Adv.
Mass Spectrom. 1979, 88, 1629.

Sabo, M. 8; Allison, J.; Gilbert, J. R.; Leroi, G. E. App]. Spectrosc. 1991, 45, 535.
Dahl, D. A.; Delmore, J. E. SIMION PC/PSZ Version 5.00, Idaho National
Engineering Laboratory, 1991.

(a) Dahl, D. A.; Delmore, J. E.; Appelhans, A. D. Rev. Sci. Instrum. 1990, 61,
607. (b) Gilbert, J. R. Ph.D. Dissertation, Michigan State University, 1990. (c)
Chisholm, T.; Stark, A. M. J. Phys. D: Appl. Phys. 1970, 3, 1717.

(a) CPO-3D, Integrated Sensors Ltd., PO. Box 88, Sackville Street, Manchester
M60 lQD, UK. (b) E3 1.0, Field Precision, PO. Box 13595, Albuquerque, NM
87192, USA. (c) MacSimion 3D, McGilvery, D. C.; Morrison, R. J. S.
Proceedings of the 43rd ASMS Conference on Mass Spectrometry; Atlanta, GA,
1995; p 718. (d) SIMION 3D version 6.0, Dahl, D. A. Proceedings of the 43rd
ASA/[S Conference on Mass Spectrometry; Atlanta, GA, 1995; p 717.

(a) Yasuike, K.; Miyamoto, S.; Nakai, 8. Rev. Sci. Instrum. 1996, 67, 437. (b)
Pai, J. R.; Venkatramani, N. Rev. Sci. Instrum. 1992, 63, 5234. (c) Maniv, S.
Rev. Sci. Instrum. 1991, 62, 1179. ((1) Brown, H. L.; Cross, R. H. Rev. Sci.
Instrum. 1976, 47, 887. (e) Coupland, J. R.; Green, T. 8.; Hammond, D. P.;
Riviere, A. C. Rev. Sci. Instrum. 1973, 44, 1258.

Grisenti, R.; Zecca, A. Rev. Sci. Instrum. 1983, 54, 505.

41.

42.

62

Limbach, P. A.; Marshall, A. G.; Wang, M. Int. J. Mass Spectrom. Ion Processes
1993, 125, 135.

(a) Yinon, J.; Ganz, J. Rev. Sci. Instrum. 1975, 46, 1707. (b) Chapman, R. Rev.
Sci. Instrum. 1972, 43, 1536. (c) Propst, F. M.; Tomaschke, H.; Skaperdas, D.
Rev. Sci. Instrum. 1963, 34, 312. (d) Skaperdas, D.; Tomaschke, H. Rev. Sci.
Instrum. 1961, 32, 1261.

Chapter 3

Infrared Detection of Mass-Selected, Matrix-Isolated Ions and Counterions

1. Initial Predictions of the Processes Occurring during Cation Depositions

As mentioned in Chapter 1, much of the reason why very few ionic structures of
molecular species have been experimentally determined relative to the number of neutral
molecules is due to the charge on the ion. Achievable densities of ions of the same charge
are far too low for the majority of the structural-detennination methods used by chemists.
This is precisely the reason why a mass-selected ion source of negative ions was included
in the original proposal for this research project. The presence of ions of both charge
would mutually negate the coulombic repulsion between ions of like charge and allow
detectable densities to be achieved.

However, the first successful MS/MI experiment by Maier et a].1 was performed
with only a mass-selected ion source of cations. No anions were intentionally introduced
into the matrix. The sensitivity of the optical absorption detection apparatus used
precluded the possibility of only positive ions being present. The first efforts in this
laboratoryr2 have also shown that continued deposition of only cations can allow for
detectable concentrations to be achieved. Clearly, the processes involved during isolation
of the cations in the matrix were not fully understood in these experiments.

To gain a more complete understanding of the processes occurring during ion
deposition, it is usefiil to quantitatively consider the electric fields generated by only the
matrix-isolated cationic species and their effect on the incoming mass-selected ions. For
illustrative purposes, a mass-selected ion beam of CF3+ with 100 eV of translational

kinetic energy will be considered. To be sure, the potential of the matrix in which the

63

64

CF; ions are being deposited can never attain a potential above +100 V. Once +100 V is
reached, the ion beam will be deflected from the matrix region, preventing further
deposition. This is plainly illustrated in a SIMION plot in Figure 3.1. Figure 3.1(a)
displays a +100 V CF3+ ion being deposited into a matrix containing +99 V. The large
circle around the matrix represents the wall of the experimental vacuum chamber which is
held at ground potential. The angle of the flight path of the ion to the matrix is similar to
the actual angle of mass-selected ions used in MS/MI experiments. Figure 3.1(b) confirms
the fact that a matrix containing +100 V of charge will deflect the next incoming CF3+ ion
from the matrix region. The actual potential of the matrix during experiments plays an
important role in the mass-selected ion deposition and will be discussed in fiirther detail
later.

Would the number of matrix-isolated ions that would create an electric field of
+100 V be within the detectable limits of the present apparatus or most other detection
methods available for matrix-isolation spectrosc0py? To answer this question, one can
formulate a mathematical relationship between the number of cations in the matrix and the
resulting electric field. For a disk of radius R containing a surface charge density of O’
(# of ions/m2), the potential V of the matrix at some point x along the axis of the disk can

be described by

V: —0—--\/x2+R2—x] (31)
28° '

where so is the vacuum permittivity constant. A derivation for this formula is given in
Appendix B.

An important assumption for this relationship is that the charge density is
contained on a 2-dimensional surface, the fact that the actual matrix will have a typical
thickness on the order of hundreds of micrometers3 is not taken into account. The electric
potential can be assumed to decrease if the same number of ions on the two-dimensional

surface were allowed to move within a third dimension (the thickness of the matrix).

65

 

Figure 3.1 (a) A SIMION plot illustrating the deposition of a 100 eV
CF,+ ion into a +99 V matrix. (b) A SIMION plot illustrating the
deflection of a 100 eV CF; ion from a +100 V matrix. The circle
surrounding the matrix represents the grounded vacuum chamber walls.

66

Therefore, for a given matrix potential V, equation 3.1 represents a lower limit to the
number of ions within the matrix. Additionally, the formula assumes the ions are in a
vacuum. The presence of mainly Ne atoms would presumably allow for a decrease in the
potential of the matrix (the permittivity constant for a neon matrix is larger than that for a
vacuum), although it can be considered negligible. These approximations, however, do
not preclude useful estimations.

The sample window used during all the experiments reported here has a diameter
of ~1.9 cm. Assuming the matrix is contained within this area, how many ions then will be
contained in matrix containing +100 V of potential (at x = 0)? The corresponding surface
charge density 0' would be 1.86 x 10’7 C/mz. The matrix would then contain 5.3 x 10'11 C
of charge, which translates to 3.3 x 108 ions or ~1 femtomole. These theoretical values do
seem to correspond to some experimental observations of the interaction of 50 nA cation
beams with metallic surfaces. It has been observed from the CEMA ion beam
visualization experiments discussed in Chapter 2 that the ion currents generated from the
mass-selected ion source can charge up metallic surfaces which have not been grounded.
This is observed as a deflection of the ion beam image on the CEMA from the position if
the surface was grounded. The fact that the beam is observed to shift by eye with
essentially no time lag from the introduction of the ion beam to the surface, and
considering that it would take a 50 nA beam ~1 ms to introduce 3.3 x 108 ions, gives
credence to equation 3.1.

Clearly, the present apparatus cannot detect a femtomole of ions. In fact, previous
estimations (using the C——N stretch of acetonitrile in solution) on a similar instrument" 4
show that ~1 x 10‘6 or ~10 nmole of absorbers may be required for reliable detection.
What, then, would the potential in the vacuum chamber be if 1 x 1016 ions were placed on
the sample substrate? A plot of the electric field V vs. a distance x along the center axis of

the sample substrate upon which 1 x 1016 ions (1.6 x 10'4 C or 5.7 x 10'1 C/mz) have been

placed is shown in Figure 3.2.

 

 

67

 

3.0x10 9

2.5x10 9 -

2.0x10 9 -

1.5x109‘

Electric potential (N - m/C)

1.0x1o9-

5.0x10 8- x

0.0 ' j fif I ' l ' 1
0.0 0.1 0.2 0.3 0.4 0.5

 

 

 

Distance from sample window (m)

Figure 3.2 The electric potential generated by placing 1 x 10'6 charges on
the sample window.

68

According to Figure 3.2, at a distance of 8 inches (~0.2 m) from the sample
window (the chamber diameter is 8 inches), the electric field would be ~7 x 10‘5 V. This
large an electric potential simply could not be formed from a 100 eV ion beam, even
considering the assumptions made in the derivation of equation 3.1. The remainder of this
chapter will show that, in fact, it is apparently possible to deposit only cationic species into
the growing neon matrix and detect them with infrared spectroscopy. A mechanism which

accounts for this experimental result will be proposed and tested.

II. Mass-Selection, Matrix-Isolation of CF;

The CF3+ cation was chosen as a “test” ion due to the large absorptivity coefficient
expected for species with polar C—F bonds. This can be shown by the trend in the
experimental absorption strengths for HI, HBr, HCl, and HF which have been measured to
be 009,5 8.8,6 19.8,7 and 922.88 (km/mole), respectively. Furthermore, the most abundant
cation in the mass spectrum of CF3C1 is CF3+; ion currents on the order of 20-40 nA can
be routinely obtained with the present apparatus.

Co-deposition of 15-25 nA beams of CF3+ from CF 3C1, CF 3Br, or CF 3H and neon
for up to 25 hours onto the cryogenic substrate resulted9 in the appearance of two new
infrared absorptions, at 1670 cm'1 and 1651 cm']. Only these bands, shown for generation
of CF; fiom CF3C1 in Figure 3.3 trace a, among absorptions due to matrix-isolated H20,
C02, and precursor molecules, have intensities that correlate with the integrated ion
current. (The most prominent of these background absorptions are those of the precursor,
CF3C1. After 11 hours of deposition, the absorbance of the most intense of these
transitions, v, at 1104 cm], 10 was 0.27.) The 1670 cm'1 and 1651 cm'1 peaks appeared
both with, and in the absence of, C0,, mixed with the neon as a potential electron

1 peak to the antisymmetric stretch (v3) of

scavenger. We assign the 1670 cm'
mass-selected, matrix-isolated CF3+. The identity of the 1651 cm“1 feature was originally

not certain. The frequency did not correlate with previously identified fragments of

Absorbance

69

 

 

 

 

10.002 AM.
a
b
MW
1700 t 1680 . 16.60 T 1640

Wavenumbers (cm'1)

Figure 3.3 Infrared absorption spectra in the 1700-1640 cm‘I region of
neon matrices after 11 hours of deposition of (a) m/z = 69 (CF,‘, 20 nA)
generated from electron impact of CF,Cl, (b) m/z = 120 (no ion current)
while subjecting CEO to electron impact, and (c) m/z = 50/51
(CFf/CFzH', 15 nA) generated from electron impact of CF3H.

70

CF3CI, and the band was not observed when argon matrices are employed. Our original
suggestion was that it might arise from CF 3+ in a different neon matrix environment,
although 19 cm'1 is unusually large for a site splitting. A later publication by Jacox and
coworkers11 in which CF3+ in neon was observed with infrared absorptions on the order
of tenths of absorption units assigned 11 site splittings to the asymmetric stretch, including
one at 1650.6 cm'l.

Control experiments demonstrate that the 1670 cm'1 and 1651 cm'1 absorptions
are not due to neutral species generated during the ionization process. When the mass
spectrometer was set to select m/z 120, but all other experimental parameters were
retained, no ions were transmitted for deposition in the matrix, yet the flux of neutrals
from the ion source remained. During such experiments, the peaks at 1670 cm'1 and
1651 cm‘1 did not appear (Figure 3.3, trace b), but the rest of the spectrum was
unchanged. To determine if these features were due not to mass-selected CF3+, but to a
secondary species formed upon collisions of the ions with solid surfaces, gas phase
species, and/or the growing matrix, mass-selected CF2H+ (m/z 51) [along with a small
amount of CF2+' (m/z 50)] was deposited into the neon matrix. As illustrated in trace c of
Figure 3.3, the new peaks were again absent. (It should be noted that bands attributable
to CFZH+ or CF; were not observed, but based on previous assignments12 the CFZH”
absorptions would have been obscured by background water absorptions, and the amount
of CF2+' deposited was too small to observe in this experiment.)

The position of the 1670 cm'1 band is consistent with a previous assignment to V3

13 1

of CF3+ matrix-isolated in neon. In that study, split absorptions at 1670 cm' and
1664 cm'1 were reported; the feature at 1664 cm'1 was not observed in our study and was
not reported in a later publication11 by Jacox and coworkers. The antisymmetric stretch
of CF; also has been reported in the range 1663-1667 cm'1 in argon matrices.“ 14' 15
When we deposited mass-selected CF3+ in an argon matrix at 5 K, an absorption at

1667 cm'1 with an apparent shoulder at 1665 cm'1 was observed.

71

Ifthe CF3+ ions are partially neutralized during deposition, spectral features due to
CF 3' should be observed. An extremely weak absorption was observed in neon matrices at
1254 cm], which agrees with previous assignmentle for CF3' in neon. The low intensity
of this feature (absorbance = 0.007), indicates that only barely detectable quantities of
CF3' are present in the matrix. Another possible source of CF3' is sputtering of neutral
fragments of CF3Cl which are adsorbed on the heat shield. This possibility will be
discussed in greater detail in Chapter4 of this dissertation. Features due to other
neutralization/fragmentation products, such as CF;16 and CF' 16 were not observed.

In summary then, these experiments show that it is possible to mass-select ions
from a quadrupole mass filter, deposit them into a growing neon matrix, and detect them
with FTIR spectroscopy. It has also been shown that direct deposition of the
mass-selected ions is the only source of the mass-selected ions in the matrix. Control
experiments clearly indicate that dissociation of the neutral precursors which have found
their way to the matrix are not the source of the absorptions assignable to the
mass-selected ion. These absorptions also grow in intensity over several hours of
continuous ion deposition. Negative species, either anions or electrons, must be present in

the matrix to account for these observations.

11]. Counterions

Obviously counterions must be present, yet their identity and mechanism of
formation were not known in these initial experiments. Often, counterions in
non-mass-selected, matrix-isolation (non-MS/MI) experiments also are not identified.
This is frequently the case when electronic spectroscopy is used for detection. Electronic
absorption spectra of negative ions are often unstructured, making their identification
diflicult.l In fact, all the reported MS/MI spectra to date from the J. P. Maier laboratory
in which cations were selected fail to report the spectral identity of anionic species (a

review of Maier’s work up to 1992 in this area can be found in reference 17). It is also

72

not uncommon to find reports of infrared spectra of non-mass-selected cations in which
anionic species are not identified. Examples of these include the phenanthrene cationm
(CMHIOP), the butadiene cation,19 and the anthracene cation.20 A matrix isolation study
using infrared spectroscopy in which the only ionic species identified was AlH4' has also
been reported.21

The identity of negative counterions in non-MS/MI experiments, however, can be
often inferred by the presence of neutral fragments of precursor molecules which have
undergone dissociative electron attachment. This is the case when electron scavengers
such as CC14 (~0. 1%) are added to the matrix22 to increase the cation concentration. The
presence of Cl’ (which is not easily detectable by optical techniques) can be deduced fi'om
observations of the growth of bands due to Cle and CCl3' during matrix growth.
CH2C12 has also been used as an electron scavenger.19 The use of its isotope CD2C12

allows bands due to the electron scavenger to be identified and excluded from assignment

to the cation which is under study.

IV. Detection of Counterions: The Allions Experiment

A major reason why mass spectrometry holds such a prominent position in
analytical chemistry is due to its sensitivity. It is much more sensitive to detect the
presence of a charged species by measuring a current due to its neutralization on a metallic
surface than it is to detect it by its absorption of photons. In fact, the first clear
observations of negative species in cation MS/MI were detected in this manner.9

When the Faraday plate, installed to monitor the impinging ion flux, was placed
directly in front of the cryostat window as neon matrices containing mass-selected cations
were warmed to about 20 K, current was measured. This procedure has been
affectionately dubbed the “Allions experiment” in our laboratory. As shown in trace a of

Figure 3.4, very small (picoampere) transient signals due to both positive and negative

species entrained in the vaporizing matrix were measured when the neon matrix was

73

 

 

 

 

 

 

 

20 .. a
a 0*
O.
:-20
5
t -40 <
3
2 -60
2
Q
5‘ 20
R b
E 0‘
IL

-20 ,

5 15 25 35

Time (seconds)

Figure 3.4 Transient current observed at the Faraday plate located ~1 cm
from the substrate during warming of (a) a neon matrix after a 25 hour
deposition of 20 nA CF, generated from electron impact of CEO and (b)
an argon matrix after a 25 hour deposition of 20 nA CF,+ generated from
electron impact of CF,Cl, followed by 2 hours of 40 nA, 10 eV electron
bombardment. The final temperatures of the argon and neon matrices were
~40 and ~20 K, respectively.

74

warmed. For the control experiments, where the peaks attributable to CF3+ were absent,
no such transient ion signals were observed. These results indicate that
negatively-charged species are being produced and matrix isolated as a result of the
deposition of positive ions. This experiment also may provide some insight into the
composition of the matrix. If the matrix were to have a homogeneous distribution of
positive and negative charge carriers, they would either recombine and be neutralized, or
their signals would cancel. The presence of transient currents of both sign may indicate
that the counterions are being formed in discrete time periods, rather than continuously,
suggesting some sort of layering in the matrix. When an argon matrix containing CF;
was subjected to two hours of 40 nA, 10 eV electron bombardment (the CF; absorption
was not reduced afler this electron bombardment), the transient signal shown in trace b of
Figure 3.4 was measured as the substrate was warmed. In this case, desorbed negative
charge carriers clearly dominate. This experiment suggests that the matrix can tolerate
some charge imbalance. Perhaps during positive ion deposition, counterions are formed
only after the potential in the matrix is sufficiently high to activate the counterion
generation process.

Positive ion detectors in mass spectrometry are usually held at a large negative
potential to increase detection limits. An accelerated cation can eject more than one
electron, producing an amplification in the signal current. The large negative potential
also improves the cation collection efficiency of the detector. With this in mind, we have
experimented with improvements to the Allions technique by isolating the window holder
and floating it and the Faraday plate at various potentials during warming of the matrix.”
Figure 3.5 displays a schematic of the details of this experimental setup. It was hoped that
the potentials would increase the number of ions detected, and also provide the means to
separate the positive and negative ions.

Unexpected results are' sometimes obtained when ions are collected in this manner

as the matrix is warmed. This is particularly true when the Faraday plate is biased, and the

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

output
recorder
O
inverting common
analo
outp
Keithley 480
'coammeter
input p '
'r
: O - : . .
: 300v L__Insulatlng
. battery .’ layer
L—: + :
I“
L__
vacuum
<—— chamber
cryostat wall
threaded . .
insulating 1 radiation
' I ShiGld
fog. / III :II I
sapp ire , 'l' ' I, : Farada late
spacer ' " (movgb'ie)
window
holder grid on Csl
or copper plate

Figure 3.5 Schematic illustration of floating various components of the
matrix region from ground while detecting current from them. This
experimental setup can be used to detect current from a grid on the CsI
window, a copper plate in place of the window, and the movable Faraday
plate.

nit

bia.

ch:

dyr
ma
tha
on
fie]
like
de:
v.11

5111

p01
of -

the

cle

Sur

deg
Win

and

76

window holder is held at ground. An example, obtained when the Faraday plate was
biased at -70 V with respect to the grounded window holder, is shown in Figure 3.6.
Although one would expect only positive ions to be collected, both positive and negative
charges clearly strike the Faraday plate.

Several points are relevant in considering such results. As the matrix warms, the
dynamics for released, charged species are complex. Warming may occur first at the
matrix/window interface. It is unclear how inhomogeneous heating, with a neon matrix
that rapidly vaporizes, influences the emanating ions. There are a variety of forces that act
. on ions which determine their trajectories as the matrix vaporizes. The first is applied
fields, which we create to separate ions of opposite charge. However, ion/ton interactions
likely create stronger and more complex fields. The many collisions of the ions with the
desorbing neon must also be considered. Further, the distribution of anions and cations
within the matrix and across thecold window need not be homogeneous. Thus, it is not
surprising that, even when a field is established to collect cations on the Faraday plate,
some anions reach the collector as well. Also, while an experiment with the window
holder at ground and the Faraday plate at -70 V may be assumed to create conditions to
collect only cations at the Faraday plate, all points across the salt window are not at a
potential of 0 V. This situation is illustrated in Figure 3.7a, which shows a SIMION plot
of the potential gradient formed between the elements that create it, the Faraday plate and
the other metallic surfaces. The window location is indicated by the dotted line in the
figure. The 11.1" -70 V is not realized by ions originating at the window, and the field is
clearly not linear between the salt window and the collector. The complex potential
surface on which desorbed ions move is not that which we hoped to form.

To create a more ideal electrostatic situation for detection and possibly sorting of
desorbed ions, we placed a metallic grid (70% transmission) across the face of the Csl
window. A SIMION plot of the potential gradient formed, with the grounded grid present
and with -70 V applied to the Faraday plate, is shown in Figure 3.7b. The potential

77

 

I

1 000

500 - ll

Current (pA)

_ I
i

 

 

.1000 l 4 I L l L l 1 l J J A l
0 5 10 15 20 25 30

Time (seconds)

 

Figure 3.6 Currents observed on the Farady plate from a warming neon
matrix after a 27 hour deposition of 40 nA of CO{. A heater on the
window holder is turned on at t = 0 seconds. At t = 30 seconds, the
temperature of the neon matrix was ~40 K. The potential on the Faraday

plate during collection of the signal was -70 V, and the window holder was
grounded.

78

 

-10V -60V

-1V

 

 

 

[IIXIIIIIIIU IIIII
fill!
‘1
Ni
ll
ii
I:

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-10 V -60 V

Figure 3.7 SIMION plots showing the potential surface formed between
the Faraday plate and the Csl window. Equipotential lines, separated in
value by 10 V, are shown. The CsI window rs represented by a dashed line
in both plots. The grounded copper grid has been added to part b. The
Faraday plate has a potential of -70 V, and all other surfaces are at a
potential of 0 V in both plots.

79

surface is now considerably more linear between the sample window/grid and the Faraday
plate. Figure 3.8 shows the currents observed on both surfaces with the Faraday plate
held at a potential of +70 V as a matrix is warmed. In this plot it is clear that negative
signals dominate at the Faraday plate and positive signals at the grid collector on the
sample window, suggesting that desorbed ion motion is influenced primarily by the applied
field in this case.

The results of all configurations of these experiments show that extensive
cation/anion recombination takes place, and that the signals measured are only a very small
fraction of the total charge present in the matrix before warming. A deposition of 20 nA
of cations for 5 hours corresponds to 2x1015 ions in the matrix (assuming all ions are
isolated). A 200 nA ion signal for 1 second, similar to those shown in Figure 3.8,
corresponds to 1x1012 ions. Clearly, most matrix-isolated ions are neutralized or impinge
on surfaces other than those used here as collectors upon warming. Also, because the
magnitudes of the signals vary between experiments, quantitative information cannot be
extracted from them. Such electrical measurements provide a means for quickly detecting
the presence of cationic and anionic species, of particular importance in experiments
where this cannot be done spectroscopically. The technique is useful for optimizing the
experimental system since emitted ion currents can be obtained after <2 hours of ion
deposition, whereas longer times are required when spectrosc0pic detection is the only

tool available for determining matrix composition.

V. Infrared Detection of the CFJCI" Anion

While coulombic detection of negatively-charged species has a much greater
detection sensitivity than most spectroscopic methods, it completely lacks qualitative
information (although it may be possible to mass-select the ionic species as they are
released from the matrix with mass spectrometry). The first spectral evidence of a

counterion in the CF3+ MS/MI experiments was the observation of two infi'ared

80

 

 

 

 

 

 

 

 

 

 

I I F fl * I I I j
50 l- d
r 1
40 ' Current measure-d ‘
~ at the grid and wrndow
30 _ holder on the _
sample wrndow
20 l- u
2 10 ~ -
5 O l- J L_A— 2.... W -
q—o
c
a o. -
L W’ .
3 . if .
0 10 -
Current measured
at the Faraday
-20 - plate ~
1
-30 - -
.40 - l
.l
-50 1 . 1 A 1 4 1 I I L l

0 5 10 15 20 l 25 30
Time (seconds)

Figure 3.8 Currents observed from a 170021 Ne:CO2 matrix after a 20
hour deposition of 30 nA of CH] generated by electron impact ionization
of toluene. Heating conditions are the same as in Figure 3.6. The
potential on the Faraday plate during collection of the signals was +70 V
with all other elements held at 0 V.

81

absorptions at 932.2 and 937.5 cm”1 in an argon matrix, shown in Figure 3.9. These bands
correspond closely to those previously assigned“ to CF3Cl" in argon. The third band
assigned to this species at 666 cm‘1 could not be observed in this experiment due to the
wavenumber limitation of the spectral window by the Can windows on the vacuum
chamber (these were later replaced by KBr to extend the spectral window). Presumably,
this species is formed by electron attachment of CF 3C1, which is present in the matrix in
abundance due to diffusion of neutral precursors from the ion source. Thus, CF3Cl may
act as an electron scavenger as discussed previously. The corresponding peaks are not
readily observed in our experiments with neon matrices. It must also be noted that a
vibrational mode was not assigned to the bands in reference 24. The bands were assigned
to the molecular anion solely on the basis of observations of photolysis behavior of the
bands and the related bands in CF 31 and CF3Br.

It is clear where the CF 3C1 originated in this experiment. Observation of CF3Cl"
anion does not by itself identify the origin of the electron. Some experimental results fi'om
other laboratories on the formation of CF3Cl" may provide clues to its formation in the
present experiment. The C—Cl bond in this radical anion is extremely labile. The spin
density distributions calculated from ESR parameters” of CF3Cl" (formed in a
tetramethylsilane glass following 7 irradiation) suggest that the unpaired electron resides in
an 01(0“) anitibonding orbital which is composed largely of the p orbitals from the carbon
and the chlorine atom which lie along the C3,, symmetry axis of the radical anion. This
observation was substantiated by the growth of an ESR signal due to the CF; radical
during the decay of the CF3Cl" spectrum in neopentane above 100 K. In the gas phase, no
negative parent molecule of CF3CI could be observed following low-energy (0-10 eV)
impact with near monoenergetic electrons.26 However, electron attachment to CF3Cl
clusters show that CF30" is generated, among various products, in its relaxed

configuration which is not accessible in electron capture by the isolated molecule.”

82

 

01KB-

 

01M!“

 

DIM 4

0.00 - [’W/h

I ‘ I ' I ' I ' I ' I ' I
960 950 940 930 920 910 900

Wavenumbers

Absorbance

 

 

 

 

 

 

Figure 3. 9 Infrared absorption spectrum of an argon matrix in the
960-900 cm'| region following a 27 hour deposition of 20 nA of CF,

formed by electron impact ionization of CF, Cl. This spectrum was
obtained by averaging 256 scan files taken with 1. 0 cm lresolution.

83

The observation of CF3Cl" in the present argon MS/MI experiment and the failure
to detect it in neon may be due to the nature of the C—Cl bond in CF3Cl" and the way in
which the noble gas forms a matrix site around the anion. Perhaps as the smaller neon
matrix site forms around the CF3Cl" anion, the weak C—Cl bond is severed by the
packing forces of the noble gas. The formation of the larger argon site may simply “cage”
the anion, locking the CF3' radical and the Cl' anion together. The weakness of the C—Cl
bond might also suggest that this anion must be formed on the matrix. If the molecular
anion were formed outside the matrix region, the force of the impact upon entering the
growing matrix would readily induce dissociation.

If the counterion in the neon experiments is Cl’, it may prove possible to detect the
presence of this species by infrared spectroscopy by providing an abundance of species
within the matrix with which the Cl' may form a complex. The observation of a shift in a
fundamental vibration of the molecular species could provide spectral evidence of the
presence of an atomic anion. For example, it is known that complexation of CI' to
acetonitrile leads to the development of a low frequency component in the C——H

stretching band.28

VI. Infrared Detection of the C02" Anion

The success of MS/MI spectrOSCOpy with only cation deposition and any rigorous
study of the concurrent counterion formation relies on the ability to readily detect a
molecular anion with the use of infrared spectroscopy. This was achieved fortuitously
during the study of the MS/Ml of other cations whose vibrations have been previously
assigned. One of the chosen cations for mass-selection was COf’. A 20 hour deposition
of ~40 nA of C02+' resulted23 in the observation of a band at 1422.2 cm'1 (Figure 3.10),
which had been previously assigned to asymmetric stretch of the cation.29 Additional

l

absorptions were also observed at 1658 cm'1 and 1665 cm' . These have been previously

assigned to the C02" anion.29

ABSORBANCE

84

 

 

 

 

 

 

 

1300

CO'

0.03- a N
(102'

- , HO
0.01 (C02): ‘ I i

. 002’

Jr
(100'
1800 1700 1600 l 1500 1400
WAVENUMBERS
Figure 3.10 Infrared absorption spectrum of a neon matrix in the

1750-1350 cm" region following a 20 hour deposition of 40 nA of COL

85

Since C02)" was the mass-selected cation in this experiment, it was speculated that
the COf‘ cation had a direct parentage to C02" through charge inversion or another as
yet determined mechanism. To investigate this possibility, C133+ was mass-selected and
deposited into a growing neon matrix which was doped with C02 in a 1000:] Ne:COz
ratio. The result of this experiment is shown in Figure 3.11. Signals representing C02”
are clearly present, showing that a process is occurring that results in their formation, but
that they are not formed directly from the incoming cation. Unlike the CF3Cl" anion, this
anion could be readily observed in neon matrices. It is interesting to note that, in ESR
studies of matrix-isolated cations, Knight30 also detects a signal representing C02". The
observation of C02" in MS/MI experiments in which cations were selected offers two
advantages. It enhances the positive ion accumulation efiiciencies; detectable signals can
be achieved in noticeably shorter times when C02 is present. Moreover, the combination
band (v1 +v4)13 of CF3+ is now detectable at 1625.7 cm’1 in Figure 3.11. Future sections
of this dissertation will evaluate the use of this matrix additive in MS/MI experiments.
The next section describes the use of this readily observable counterion to investigate the

mechanism for its formation.

VII. Mechanisms for the Formation of Countercharges

Since the beginning of MS/MI spectroscopy, there have been several proposed
mechanisms for the formation of counterions in experiments where only cations are
selected for deposition. The first report1 acknowledged the requirement of counterions
and proposed that the anions in the matrix were “probably produced by dissociative
electron attachment to trace impurities present in the matrix.” The authors further
proposed that the “electrons stem either from the ion source or by ion bombardment of
metal surfaces.” The first publication of cation MS/MI spectroscopy from this laboratory’
furnished a third mechanism for the generation of electrons: “The copper support for the

substrate is held at electrical ground; local fields due to ions in the matrix might result in

86

 

 

 

 

 

 

 

0.025 ~
CO,’
I N
0.020 -
w b
g 0.015 - (ӣ0
CD +
g 0.010 ~ CF’
‘é’ t \ or:
0.005 ’ (003); L) “,0
0.000 -
'°'°°51000 L 1700 I 1600 A 1500 A 1400 A 1300
WAVENUMBERS

Figure 3.11 Infrared absorption spectrum of a 1000:1 Ne:CO, matrix in
the 1750-1350 cm'l region following a 13 hour deposition of 20 nA of CF,’
formed by electron impact ionization of CF,Cl.

87

extraction of electrons from the metal.” Later, another mechanism was proposed from
this laboratory:23 “the incoming ions have sufficient kinetic energies to ionize gas phase
species in front of the growing matrix, forming positive and negative charge carriers.” An
additional mechanism was proposed in which the CsI substrate, upon which the matrix

31 While we have ultimately

lies, provides the matrix with a source of I‘ anions.
proposed” that one of these mechanisms dominates in the generation of counterions, it is
useful to discuss both the theoretical and experimental observations from which each
mechanism was discounted or verified. Only by considering in full all possibilities, even

those that have been since experimentally invalidated, does one gain a complete

understanding of the actual processes occurring during MS/MI ion deposition.

A. Electrons from the Ion Source.

Early versions of MS/MI instruments of this laboratory and that of J. P. Maier’s
laboratory placed the mass-selected ion source relatively close to the matrix region. The
proximity of the growing matrix to the ion source filament, from which emanates electron
currents on the order of milliamperes, makes it a possibility for the origin of the negative
charges. The large amount of ~70 eV electrons made this a plausible mechanism, even
though the electric potentials on the few ion optics between the ion source and the matrix
were set to focus only positive charges. However, later generations of MSM
instrumentation from this laboratory,9 and that of J. P. Maier’s," utilized ion sources far
removed from the matrix region. This was mostly due to the addition of vacuum pumps
to the ion source to remove the large amount of neutral precursors needed to obtain high
mass-selected ion currents. The ion source in both cases was additionally offset at an
angle from the axis normal to the surface of the matrix substrate to further reduce the
amount of neutrals in the matrix (Maier has utilized a 90° bending quadrupole in his latest
version"). The spectra acquired following these modifications show no decrease in

spectroscopic signals due to matrix-isolated cations. All efforts to measure negative

88

currents fi'om the ion source (with the ion source filament on, but no sample gas present)
at the Faraday plate in front of the matrix in this laboratory have failed. This mechanism

has thus been discounted.

B. F ield-Assisted Extraction of Electrons from the Substrate Holder

The next mechanism of counterion generation to be considered is the matrix
electric field assisted extraction of electrons from the grounded substrate holder. This
mechanism is a “condensed phase” mechanism, in which accumulating positive charge in
the matrix creates a field which is sufficient to extract electrons from the grounded metal
window holder (with which the matrix is in contact). Thus, the primary negative charge is
an electron, which migrates through the matrix and is captured by some molecular species
capable of forming a stable anion.

Clearly, very high electric fields are needed for a cation to extract an electron fi'om
a metal surface without making direct contact with the surface. As mentioned previously,
however, the potential of the matrix will never go above the kinetic energy of the
mass-selected ion beam. If a 100 eV ion beam is selected, then the matrix cannot hold
more than 100 V of overall charge. Since the matrix window has a diameter of ~2 cm
across, this translates to an electric field of ~5000 V/m. If the overall potential of the
matrix is insufficient to extract an electron, cations isolated close to the surface may be
able to generate fields of much higher magnitudes. It is known from field ionization mass
spectrometry that fields on the order of 109 V/m are needed to ionize gas phase molecules
and atoms.32 The potential well for an electron of a molecule in such a high field becomes
distorted, allowing the electron to escape by undergoing a “tunneling” process. While the
use of this electric field value in a comparison of the field needed to cause a similar
tunneling process between a cation and a metal surface is far from a rigorous one
(although the average metal work function and molecular ionization energy differ by only

an approximate factor of 2), it does provide a “ballpark” value in which one may begin to

89

consider the mechanism in question. With this starting point, the required distance of the
cation to the metal surface to generate the appropriate electric field can be approximated
by calculating the potential and the field generated by a point charge with the following

equations:

 

V: 1 1 and E=
47reor

K

r

As Figure 3.12 illustrates, the electric potential of a point charge at 1.18x10'9m is
relatively low, 1.22 V. The electric field, however, which is the change in potential with
distance, is very high. At 1.18x10'9 m, the electric field is 1.03x109 (N/C). Therefore,
from the earlier estimation, a cation should lie within 1.18x10’9 m (11.8 A) of the metal
surface to extract an electron from the surface. Despite all the approximations that were
used for the estimation of this distance, it agrees well with a quantum-mechanical
calculation33 which concluded that the probability for neutralization of a 25 eV I-I+ ion
reaches unity at a distance of 3.5 A from a metal surface.

If this mechanism is the dominant process by which the matrix containing ions
maintains neutrality, not only must the cation extract an electron from the metal surface
while it is isolated in the matrix, it must somehow keep the electron from neutralizing
itself. Since CF3Cl" was observed in an argon matrix, and presumably Cl' is present in the
neon experiments, it might be concluded that the CF3Cl is the dominant electron
scavenger in the matrix when CF3C| is the neutral precursor of the mass-selected cation.
With Figure 3.13 (an approximate scale model using actual atomic radii) as a guide, the
possibility that CF3CI, which is present in the matrix in abundance when used as a neutral
precursor, could “fit” in a 10 A space between a CF 3+ ion and a metal surface is clear.

Considering Figure 3.13, several possible shortcomings of this mechanism can be
envisioned. The electric field experienced by the CF3Cl is larger than the field at the

surface due to the cation. If this field is high enough to extract an electron from the

surface, it probably is also high enough to extract a negative charge from a CF3Cl" anion.

90

 

1.4"

1.2“

Electric Potential (Volts)

1'11 \\

 

 

 

I '

10x109 1.1::10'9 1.2x10‘9 1.:ix10e 1.4x10”
r(m)

 

1.4x109 ‘
l
1.3m!9 ‘

4

1.2x10’ ‘ \\

‘ \
1.1x10’ 1 ‘\
1 \‘

1.0mm9 ‘ ‘\

Electric Field (N/C)

9.011105 4 \
1 \

amid“ ‘

d

 

 

 

9 fi I 1 I 9 Y 9 V 9
1.0x10 1.1x10’° 1.2x10’ 1.3x10' 1.4x10'
7 (m)

Figure 3.12 Plots of the electric potential and the electric field versus
distance from a point charge.

91

 

 

 

[(0. c0. 0,

Figure 3.13

 

A scale model illustrating the relative sizes of a

matrix-isolated CF,C1 molecule lying between a CF; ion and the wall of
the metallic window holder.

Ifi

fur

ho

1C

C)!

ti.

pi

92

If it is not, the layer of negative charges that would build up over time should prevent
further extraction by other cations. For these reasons, the probability that field assisted
extraction of electrons from the substrate holder does not seem a highly plausible one.
Although this mechanism is difficult to evaluate theoretically, it is relatively easy to
experimentally test. To investigate this mechanism, we electrically isolated the window
holder, thus terminating the connection to its source of electrons (ground). If only cations
are accumulated in a growing matrix, and the impinging cations have kinetic energies of
100 eV, the potential of the matrix cannot exceed +100 V. This will be established within
the first second of deposition, as mentioned previously. If this potential is sufficient to
extract electrons (or anionic impurities) from the grounded window holder, then isolating
the window holder fi'om ground should prohibit the continuous deposition of cations over
periods of several hours. Experiments performed with the isolated window holder, with
all other parameters the same, resulted in no change in the infi'ared absorption spectrum.
Both cationic and anionic species were observed with absorption intensities equivalent to
the earlier experiments in which the window holder was not isolated. This mechanism was

therefore eliminated from further consideration.

C. Ionic Bombardment of Gas Phase and Adsorbed Molecules near the Matrix Region
The last two possible mechanisms which we considered are similar in nature. They
will be referred to as mechanism 1 and mechanism 2 and are illustrated in Figure 3.14.
Mechanisml is a “gas phase” mechanism. The incoming ions have sufficient kinetic
energies to ionize gas phase species (exemplified by C02 in Figure 3.14) in front of the
growing matrix, forming positive and negative charge carriers. The negative charge
carriers, which may be electrons or negative ions depending on the molecules that are
present in this region, are attracted toward the positive potential of the matrix. The
nascent cation is repelled by the same field. Mechanism 2 is a “surface bombardment”

mechanism. Similar to mechanism 1, incoming ions collisionally generate gas phase

93

 

 

 

 

 

radiation shield window holder
Csl window
|C °°’®C°®CO@00.®C°-®°°/:
CO\ I; \\ 6 co, co, co, co
091% CO,;\/ I
G) G)

® ®

Figure 3.14 Illustration of two proposed mechanisms of counterion
formation in mass-selected, matrix-isolated cation expenments.

do
this
sin:
col

Fig

spe

10

an

th

lie

94

electrons or negative ions which are subsequently attracted to the matrix. However, in
this case the collisions are with metal surfaces. This mechanism offers many possibilities,
since the surfaces close to the growing matrix are not only metallic, but are sufficiently
cold such that a variety of “background” gases in the vacuum system (again exemplified in
Figure 3.14 by C02) can be condensed onto them.

To investigate whether mechanism 1 or 2 is the cause of negatively-charged
species being deposited into the matrix, the CsI window was replaced with a copper plate
to measure current that may impinge on the location of the salt window in these
experiments. Similar to the grid used in the Allions experiment, the copper plate can be
grounded or biased while being used as a Faraday plate. See Figure 3.5 for the details of
this experimental setup. Several experimental configurations were investigated, and five
are shown as Configurations 1-5 in Figure 3.15-19. In each case, the experimental
configuration is shown on the left, the results of the experiment are shown in the middle,
and our interpretation of the results is shown on the right. For each pair of data traces,
the top trace shows current measured at the copper plate that was substituted for the Csl
window, and the bottom trace is current measured at the retracted Faraday plate, which
lies approximately 1 inch to the right of the copper plate, outside of the radiation shield.
Each pair of traces was obtained simultaneously. A manual gate valve was opened/closed
to start/stop the incoming CSf‘ ion beam. The ion beam was presented for two one-
minute periods (approximately 30-90 seconds and 150-210 seconds into the measurement)
during the displayed time period. Matrix gas (NezCOz 1000:1) was present in all
experiments.

When all surfaces are at room temperature, configuration 1, the incoming cation
beam has a current of approximately 30 nA, as shown in trace 1a. The retracted Faraday
plate, which sits to the side of the radiation shield, collects approximately 1 nA of cation
current (trace 1b), demonstrating that not all of the incoming ions strike the surface on

which the spectroscopically-observable matrix is formed. The origin of the spike observed

95

.25. 05 5 59329.35 can .333 2: E 3% o—
2: 5 336%.: m_ 958 3:08:36 2:. ._ grandma—80 m2“ 2 E

Amocooomv oEF
o3 om? om om

(choH ii iii o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

<5H
W h
IWI hh

E

38H

05

.Ewt 05 co 5:80.985 van .0500 05 E 83 d0.
05 co 380:2: 3 0200 3:08.530 05. .N cow—23:50 Ed Semi

Amocooomv 0EF

omw

om

om

 

 

 

 

oEF

 

 

rueung

if
\a

N

97

(EH

kw
bum
I

 

<cmH

.Ewt 05 co 5358925 use .0300 05 E 3% d0—
05 co 0030:05 £ 958 3:08.598 05. .m cozesmucou 2d Ezmfi

Ame-600$ 05....

05 cm? cm

on

 

v ' I

 

 

 

 

 

 

 

 

 

 

 

rueuno

0i .2.
milm

      

  

.23. 05 :o 5:505:85 :5 .._8:00 05 :m 834M0—
05 :o 333:5 3 9:3. 3:06.598 05. .v :owaSmcgU ”—6 0: E

3:583 06:.

SN 8..

cm

cm

 

«5H airiérlilf.

 

 

 

98

 

a:

 

 

 

 

a... .0
Ti
helm...

oEF

 

 

 

tueunc)

5.0/th 0. mm“ d2
E

\aL.

v

 

rueuno

CU

99

game 05 :o 5:80.925 :5 .:2:00 05 :_ 80: d0.
05 :o Baofifi 2 9:00 3:25.096 2F .m :ozfiewcgo 0:“ 2:3..—

Amucooowv 0E5.
o 5. om F cm on

 

 

 

 

 

 

 

 

 

100

in trace 1-4b when the gate valve is closed is uncertain, perhaps it is due to a change in the
focal properties of the mass-selected ion beam as the gate valve blocks the ion beam.

Are negative charges being formed in this experiment, when gases are present and
all surfaces are at room temperature? To test this, configuration 2 is used. The copper
plate is biased at a positive potential of +250 V. This design is preferable for detecting
any negative current formed for the following reason. In configuration 1, trace 1a could
represent the sum of positive and negative ions striking the copper plate. With a +250 V
bias on the copper plate, incoming cations are repelled and only negative charges are
collected. A SIMION plot, figure 3.20, illustrates the effect of the +250 V copper plate
on the incoming 130 eV positive ions. Although they no longer strike the copper plate, a
portion of the ion beam is still incident on the radiation shield. Trace 2a shows that a
small negative current is formed when a positive ion beam is admitted into the
experimental chamber. We believe that the signal shown in this trace represents negative
charges which are formed by ion bombardment of nearby metal surfaces, on which
molecules are adsorbed. These negative charge carriers are attracted to the positively-
biased copper plate. Note that the fraction of incoming positive ions collected by the
Faraday plate in configuration 2 is slightly larger than in configuration] due to the
positive potential applied to the copper plate in the former case.

If mechanism 2 is operative, then cooling the radiation shield should have a
dramatic effect on such measurements. This situation is created using configuration 3.
The cryostat radiation shield can be cooled to a temperature of ~40 K, and, via a heater on
the window holder block, the copper plate can be maintained at approximately 225 K.
Under such conditions, with Ne and C02 present, C02 should accumulate on the radiation
shield, leading to the formation of a larger negative current upon cation bombardment.
When the radiation shield is cooled, and the copper plate held at ground, the data shown
in traces 3a and 3b are collected. As in configuration 1, incoming cations are detected.

The negative deflections which are superimposed on the cation signal in traces 3a and 3b

101

0 volts

 

130 eV CF,+ ions

ccccc

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

Figure 3.20 SMON plots illustrating the effect of floating the copper
plate on the window at +250 V on incoming 130 eV CF,+ ions.

102

are due to negatively-charged species striking the copper plate and the Faraday plate.
Several aspects of the data warrant comment. First, we have evaluated many experiments
that employ configuration 3 and the negative deflections detected at the copper plate, such
as those clearly seen in the 150-210 second time window of trace 3a, correlate with
identical features in the current detected at the Faraday plate, 3b. This suggests that ion
bombardment of the cold surface results in ejection of negative charges into the gas phase.
They then impinge on a variety of grounded surfaces, and thus show a specific time
correlation. Second, it is unexpected that the negative charge generation is not
continuous. This remains the most surprising aspect of this data set. Finally, the negative
current deviations are relatively small compared to the cation signal. If this is the case,
then the process of ion bombardment of cold surfaces on which C02 is condensed is not
yielding an anionic current sufficient to balance the charge of the incoming cationic
current. However, in configuration 3 the copper plate is held at a potential of 0 V; during
the experiment it is likely that the matrix holds a net positive charge.

To Optimize our ability to measure negative charges from the cation bombardment
of the cold radiation shield, configuration 4 is used. With a positive potential on the
copper plate, as in configuration 2, incoming cations are repelled from this plate and a
field is created which will collect most of the negative charges that are formed. As
expected, the grounded Faraday plate shows a simple cation signal in trace 4b. However,
the temporal and signal intensity aspects of the data shown in 4a are surprising. Negative
charge generation under these conditions is very different fi'om that detected in
configuration 2. Negative charge is not detected promptly upon cation bombardment, and
does not cease promptly when the incoming ion beam is blocked by the gate valve. Also,
negative charge currents larger than expected are collected. At present this is not firlly
understood. The total cationic current striking the entire radiation shield is at most
200 nA, the observations require that more than one negative charge carrier is generated

per incoming cation. This would suggest a surprisingly efficient utility of the cation

kine
ion”
on t
the

surf
mea
trac

due

neg
enh

bea

obt

CUI‘

coc
00p
resr
met
in 1
Pas

COn

iSol
903

Sun

103

kinetic energy. Unfortunately, we are working in an energy regime in which few relevant
ion/surface interaction experiments have been performed. Considering that anions impinge
on the biased surface at energies of ~250 eV in this configuration, it is possible that part of
the total negative current measured in 4a includes anions losing two electrons each at the
surface and leaving as cationic species.“ That is, the current sensed may not be a simple
measure of the primary negative charges formed. The difference between the data in
traces 2a and 4a may also suggest that part of the current detected in 4a is field assisted,
due to the positive potential on the copper plate.

However, the important observation from configuration 4 is that the number of
negative charges that can be produced per second, induced by cation bombardment, is
enhanced when the radiation shield is cooled. Since we know that a portion of the ion
beam strikes this copper surface, and that C02 condenses on the surface, we present this
as strong support for mechanism 2. It should be noted that results similar to these are
obtained when the C51 window is present and the window holder is used to measure ionic
currents, and that the results are independent of the identity of the mass-selected cation.

Configuration 5 is similar to configuration 3, except that the copper plate was
cooled to a temperature of ~5 K. The fact that both neon and C02 can freeze on the
copper plate, on which we hope to measure current, complicates the experiment. The
results are, however, interesting. Since C02 can adsorb onto the surface, according to
mechanism 2, negative charges can be extracted from it. We attribute the positive peaks
in trace 5a to negative charges leaving the copper surface. These negative species can
pass through space and some may be detected on the Faraday plate (trace 5b). As in
configuration 3, the time periods of the bursts of the negative charges correlate.

Insights regarding mechanisms 1 and 2 can also be obtained by using an electrically
isolated metallic grid on the Csl window (in place of the copper plate) and applying to it a
positive potential of +80 V during the course of a matrix deposition. All surrounding

surfaces are at ground. This experiment was performed with a positive potential on the

wit
F12
res:
spe
3P1
hat
ma

m0

acc
eff
acr
me
wit
spe
010
$0
wt.

sin

ma
the
ab:
Wit
ma

Wit

104

window grid in an attempt to decelerate incoming cations. The results were unexpected.
Figure 3.21 shows the results of a C7H7+ deposition similar to the deposition for which the
results are shown in Figure 3.11, except for the potential applied to the grid. The
spectrum clearly changed - a signal representing C02“ is now present. We attribute the
appearance of C02+' to ionization of neutral C02 molecules in the matrix by electrons that
have been formed by cation bombardment of surfaces and are accelerated towards the
matrix by the potential applied to the grid. If mechanism 1 is operative, electrons would
most probably be formed immediately in front of the growing matrix where the density of
gas phase particles is greatest. An electron formed in this area would likely not be
accelerated to a kinetic energy required to ionize other neutral species in the matrix so
efficiently. Only in mechanism 2 would the electrons experience the full 80 V
acceleration, allowing them to ionize neutral species in the growing matrix. Regarding
mechanism 2, large (~30%) secondary electron emission from metal meshes bombarded
with anions has been recently reported.35 Such processes are of concern to mass
spectrometrists who work with negative ions, since ion/surface interactions can yield
electrons which are detected along with incoming anions, making signals artificially large.
Schlag and coworkers” attribute the electron origin mainly to the adsorbed residual gases,
which is consistent with the temperature dependence they observed. It seems likely that a
similar mechanism for electron emission occurs to some extent in our experiments.

Other evaluations of mechanism 1 suggest that it is n_ot the process by which
positive ions are created and deposited into the matrix. Experiments performed with
mass-selected ion beams of CF; fail to produce absorptions due to isolated CF; ions in
the matrix. Only when ion beams of CF3+ are mass-selected and matrix-isolated are
absorptions due to CF 3+ observed.9 This is reasonable; even if 100 eV CFZP ions collided
with neutral, gas phase precursor CF3Cl molecules to form CF3+ and Cl', the CF; in the
matrix would not create a field that would attract gaseous, thermal CF; ions toward the

window. Only the ions in the mass-selected ion beam have sufficient energy to overcome

ABSORBANCE

105

 

 

 

 

 

 

0.030 - .

, CO. N
0.025 -
0.020 -
0.015 - H20

/J
0.010 - _
(C02):
+
0.005 - \4 ~ 002
M J,

0.000 ~

' A l L l a l A l a

1800 1700 1600 1500 1400 1300

WAVENUMBERS

Figure 3. 21 Infrared absorption spectrum of a 1500: 1 Ne. CO, matrix in
the 1750-1350 cm'l region following an 18 hour deposition of 30 nA of
C,H., generated by electron impact ionization of toluene. The copper grid
on the Csl window was floated at a potential of +80 V during ion
deposition.

106

the slight positive potential of the matrix and become isolated. Therefore, when the
window grid is positively biased and we detect COf‘, this cation is not being deposited
into the matrix, but is formed in the matrix upon electron bombardment.

In mechanism 2, C02" must follow a path from the radiation shield to the matrix.
For some C02” ions, the time of flight to the matrix may exceed the lifetime of the ion
(20-90 [19,3657 and it may dissociate into neutral C02 and electrons. This possibility was
explored with SIMION. In the SIMION plot shown in Figure 3.22, the electrode which
represents the matrix was set at a potential of +20 V in both plots. Although the actual
potential of the matrix is not known, this potential was chosen to produce the longest
possible time of flight of the ion, ensuring a stringent test of whether flight times shorter
than the C02" lifetime are possible. The maximum matrix potential of 130 eV was
expected to yield shorter ion time of flights. In plot A, an anion with a molecular mass of
44 amu was placed on the radiation shield with 0 eV of kinetic energy. The total time of
flight of this anion was ~3.2 ms. In plot B, the anion was placed ~1 mm to the right of the
previous position. The time of flight for this ion was ~3.5 us. Clearly, the time of flight
of an anion produced on the radiation shield to the matrix depends greatly on the starting
position. It can be concluded from these plots, however, that C02" formed close to the
opening of the radiation shield has a time of flight to the matrix which is shorter than the
reported lifetime of the anion. It is also possible that in the longer flights, the electrons
produced from dissociation of the anion bombard the matrix. The CO}, which has been
attributed to bombardment of the matrix by electrons generated directly from the metal
surface, might also be formed by bombardment of the matrix by electrons that have

dissociated from C02" undergoing long flight paths to the matrix.

VIII. Further Discussions of Counterion Generation
Explicitly, we suggest that both electrons and anions leave the cold radiation shield

upon cation bombardment. Although there have been numerous observations in surface

107

 

 

 

 

time of flight = ~3.2 ms

ion trajectory

 

 

 

. .\ time offlight = ~3.5 us

Figure 3.22 SIMION plots illustrating possible flight paths of CO,’
sputtered off the radiation shield to a matrix held at a potential of +20 V.
The starting point of the anion in plot B is ~1 mm to the right from the
starting point in plot A. The electrode simulating the matnx is held at
+20 V in both plots, and in each case the anion starts with 0 eV kinetic

energy.

108

science laboratories concerning the ejection of more than one negative charge per incident
cation on a metal surface, this phenomenon is still relatively new in mass spectrometry.
Since this process is of fundamental importance to MS/MI spectroscopy in which only
cations are selected, these results will be discussed in detail to gain further insight into the

processes that may be occurring here.

A. Auger Neutralization

Prior to the results presented above, arguments against mechanism 2 utilized the
tenet that a 100 eV cation should be neutralized upon impact with a conducting surface.
An electron fi'om ground, attracted by the potential of the cation, would simply tunnel
through the barrier of the metal surface and neutralize the cation. Only when the kinetic
energy of the cation reached above ~1000 eV would neutralization be accompanied by the
further ejection of electrons. The only process by which it might be possible to extract
more than one electron per cation was thought to be kinetic ejection.38 The nature of the
ion does not play a role in this process, only the kinetic energy of the ion determines the
number of electrons ejected from the surface.

Potential ejection, however, is also known to occur. This process involves the
electronic interaction between an incoming ion and the conduction electrons of the metal
while the ion is still outside the metal surface. In this case, it is the nature of the ion which
determines the efficiency of electron ejection and not its kinetic energy. In this process,
one electron from the metal tunnels through the barrier between the surface and ion and
drops into the ground state. In a radiationless process, a second electron accepts the
energy released in the downward transition of the first electron. This second electron may
then leave the solid if properly directed. This process is known as Auger neutralization
(AN) and is schematically illustrated in Figure 3.23. It is also possible for a metastable
species to eject an electron from the surface. This process is known as Auger

de-excitation (AD) and is illustrated in Figure 3.23 as well.

109

(RN)

 

 

resonance
neutralization
+ — * :-
X + nem 4 V X + (”-1)em
resonance
ionization
(RI)
auger auger
neutralization de-excitation

(AN) (AD)

x + (n-2)e;,+e'

 

 

 

 

 

 

 

 

 

 

 

 

A A
e; e;
7r
V//// E//// 28
VII/Xe- V///\
VII/l '“ /////
///// /////
///// {ll/l
///// /////
///// r ///// .15
///// /////
V j j :: ion potential energy / ’ ’ ’: ion potential energy
surface surface
metal surface metal surface

AN RN + AD

Figure 3.23 Schematic diagram of possible electronic transition processes
for incident ions (X ) or excited species (X‘) and metal electrons e;,.

110

The maximum kinetic energy (Ekin,max) of the ejected electron during potential
ejection is given by
E = E, — 209

kin-lax
where E,- is the ionization energy of the incident ion and (a is the work function of the
metal.39 In general, E,- is reduced by 1-2 eV in front of a surface, which is primarily a

0 Information concerning the

consequence of the operation of image force effects.4
electronic state densities in solids can be extracted from the kinetic-energy distributions of
the ejected electrons. This technique has formed the basis of Ion-Neutralization
Spectroscopy (IN 8)."1

It seems plausible that the physical processes behind AN may play a role in the
formation of counterions in MS/MI. However, the total yield of electrons per incident ion
is quite low. In a study of noble gas ions on a tungsten surface (¢= 4.5 eV),42 no greater
than 30% of the incident ions are converted to electrons. Xenon, which has an ionization
energy of 12.1 eV resulted in the lowest conversion percentage (~2%). The fact that
these studies utilized atomically clean metal surfaces can be assumed to drastically reduce
the number of Auger electrons produced in comparison to the present apparatus, in which
the 00pper used is far from atomically clean. Presumably, the conversion factor for
incident cation to countercharge in MS/MI experiments is much higher than those
obtained in AN experiments.

Further consideration of the cations studied with MS/MI suggest that AN is not
the dominant mechanism by which countercharges are created. The metal surfaces
involved in the present apparatus are the 00pper radiation shield and the nickel-plated
window holder. The radiation shield used in the Maier laboratory is presumably similar,
and the substrate upon which the matrix is grown is rhodium-coated copper." The work

fimctions of Cu, Ni, and Rh are 4.4, 4.5, and 4.75 eV, respectively. If AN is the sole

process by which countercharges are formed, then the cations that can be continuously

111

deposited in MS/MI must have ionization energies which are at least twice that of the
metal surfaces used. Table 3.1 lists the cations that have been detected by MS/MI to date.
The majority have ionization energies well over twice the work function of copper. In
1993, however, Maier reported the MSM of C7O+ and C60+, which have ionization
energies of 7.3 and 7.58 eV, respectively. The C6H' experiment, in which C6" was
detected, clearly shows that AN cannot be a dominant mechanism of countercharge
generation in his experiment. Additional studies in this laboratory in which the radiation
shield was fabricated with platinum (¢= 5.3 eV) showed no apparent decrease in the
infrared absorption intensities due to the mass-selected cation CF3+, as might be expected
with the increase in work function of the metal. Other mechanisms of countercharge

formation must account for the production of counterions in MS/MI.

B. The Chemical Reaction of Carbon Dioxide and Metal Surfaces

While the countercharges have been determined to be formed on the radiation
shield during MS/MI experiments, the mechanism by which they are formed has yet to be
established. Auger neutralization, which involves cations incident on a clean metal
surface, as discussed previously, is probably not the dominant mechanism by which
countercharges are formed. There must be more to the mechanism than simply the
interaction of a cation and a metal surface. The observation that the magnitude of
negative charges depends greatly on the temperature of the radiation shield and the
presence of condensable gases such as C02 must also be accounted for by the mechanism.
At present, however, a detailed experimental study of the processes that occur as C02 is
adsorbed on the radiation shield and the effect of the incident mass-selected ion beam is
not possible in this laboratory. There does exist, however, a large amount of work
devoted to the reductive activation of C02 because of the possibilities of its use in energy

storage or as a C1 building block in organic chemistry. Several laboratories have

112

matrix-isolation spectroscopy.

Table 3.1 A list of the ions that have been successfully detected by mass-selected,

 

Ionization

Ion Kinetic

Ion

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ion Year Reference
Energy (eV) Energy (eV) Current (nA)

HC4H* 10.5 150 2 1988 [1a]
C; 12.1 150 5 1988 [1a]
NCZN” 13.4 150 20 1989 [lb]
C; 12.1 150 5 1989 [lb]
ClC2C1+ 10.1 150 100 1989 [lb]
c; 12.1 125 20 1990 [43]
C1C,C1+ 10.1 150 100 1991 [44]
C; 12.1 150 5 1991 [44]
N; 15.6 _ 1000 1991 [44]
HC,,H+ 10.5 150 100 1991 [44]
cs,+ 10.1 10 1 1991 [2]
HC4H+ 10.5 50 100 1992 [17]
Cm“ 7.3 _. 10 1993 [45]
C60+ 7.6 _ 10-20 1993 [46]
3h, — — 3-7 1993 [47]
CF,+ 8.9 130 20 1994 [9]
HC2nH+ "=34! _ 50 25 1994 [48]
"0261*1+ "'2-7 _ 50 10 1994 [48]
"C2110“ "'26 _ 50 10 1995 [49]
NCCanN+"*'-5 _ 50 10 1995 [49]
C6‘ _ 50 10 1995 [50]
Can'n'“ — 50 10 1995 [51]

 

113

demonstrated that activation of C02 can occur as it binds to transition metal
complexessz’ 53

The exact nature of the processes that occur during adsorption of C02 on a copper
surface and the effect of cationic bombardment have yet to be clarified. A variety of
studies of C02 adsorption on metal surfaces have been undertaken using several different
experimental techniques. These techniques include FTIR,“ SERS,” 'I'PD,56 XPS,” and
various electron spectroscOpies such as HREELS58 and VEELS”. The results show that
C02 undergoes a “chemical reaction” with various metal surfaces; i.e. the C02 which is
linear in the gas phase, becomes bent and negatively charged as it adsorbs on the surface.
This observation has been deduced from an increase in the work function of the surface
upon addition of the adsorbate, vibrational shifts of the adsorbed species, and changes in
the binding energies of the adsorbates with temperature of the surface. It has also been
concluded that the most probable orientation of bent C02 and the metal surface is a C2,,
geometry where the two oxygen atoms are chemisorbed onto the metal surface while the
carbon atom lies above the surface plane.58b"

Several differences between the conditions of the surface used in the studies
mentioned previously and in those used in MS/MI experiments suggest possible ways to
improve the anion yield. In addition to copper, Ni, Fe, Pt, Re, and Rh have been used in
C02 adsorbate studies. If the C02" production efficiency in MS/MI experiments depends
on the orientation of the C02 on the metal surface, perhaps a metal other than copper
might provide a larger yield of anions. Furthermore, several of the studies utilized alkali
metal and alkaline earth metal covered surfaces. The lowering of the work function of the
metal surface might increase charge transfer. Molecular oxygen has also been used to
facilitate co,” formations“ ‘0

Experiments which investigate the products of the reaction of low energy

(5-500 eV) ion beams and metal surfaces may also provide insights into countercharge

generation. SERS has been used to tentatively assign surface carbonate anions on silver

114

61

electrodes, following bombardment of mass-selected carbon cluster cations. The very

first observation of C02” was during a study of the negative ions produced by

bombardment of a metal surface with Hg+ ions.62 C02 ions have also been observed to
be formed at metallic grid surfaces in a flux of excited Cs atoms and ions.“3 COf' and

C02 molecular beams have also been shown to directly form C02" upon scattering from

‘4 Harpooning reactions, which are well known in the gas phase“, have

66

metal surfaces.
been used to describe the charge transfer between the cation and the metal surface.
Presumably, negatively charged, bent C02 molecules are intermediates in these processes
just as they are in the studies in which C02 is adsorbed on metal surfaces at low
temperatures.

What is the role of the incident ion beam in the production of C02"? It has been
noted that through a process known as Auger neutralization, an incident cation may be
responsible for the extraction of two negative charges from the metal surface. Yet, the
surface science studies just mentioned show that a neutral adsorbate may also extract a
negative charge. It is possible that the ion beam which is incident on the surface merely
sputters off the already formed C02 anion. Perhaps a beam of neutral species, such as a
neon, might also sputter oft‘ C02". It might prove possible to significantly increase the

anion yield in this way, and this process is worthy of experimental investigation.

IX. Positive Countercharges

In our laboratory, C02" has been detected in matrix isolation experiments
involving a variety of mass-selected cations. These ions include both even electron cations
(CH,*, CH, C7H7“, 021130“) and radical cations (cog; csz“, 0,176+). J. P. Maier and
coworkers have also successfully detected mass-selected, matrix-isolated ions, mainly by
electronic absorption spectroscopy.17 Although they do not intentionally add CO; to their
neon matrix gas, a counter ion must exist in their experiments and presumably it is formed

by a mechanism similar to that proposed here. Does a mechanism exist whereby mass-

115

selected negative ions form positive counter ions? Experiments involving the mass-
selected anions C,,H,,,', where n=2-12 and m<3, have been performed by Maier’s group."
A prominent peak was observed at 608 nm in an experiment involving mass-selected
CGHM'. This spectroscopic transition was originally attributed to C6H'; later theoretical
calculations, however, have shown the most likely candidate to be the C6" anions9
Presumably, the positive counter charges in this experiment are also formed by the
interaction of the mass-selected ion beam with metallic surfaces nearby the growing
matrix. In fact, positive ions have been detected from collisions of negative ions with
metal surfaces in other experiments.34 The extent to which these mechanisms may be

utilized to produce countercharges in fiiture MS/MI experiments in which only anions are

selected remains to be seen.

X. Conclusion

The initial proposal for MS/MI suggested that both a mass-selected cation beam
and a mass-selected anion beam should be utilized. It has proved difficult to construct a
mass-selected anion source capable of producing currents of equal magnitude to those for
mass-selected cations. The results presented above show that it is possible to produce
large quantities of infrared-detectable anions by allowing a portion of the mass-selected
cation beam to bombard the cold radiation shield and other surfaces upon which C02 has
been adsorbed. The utility of this technique of countercharge production in MS/MI will

be explored in the remaining chapters of this dissertation.

116

References

1.

10.

11.
12.
13.

14.
15.
16.
17.
18.
19.
20.

(a) Forney, D.; Jakobi, M.; Maier, J. P Faraday Discuss. Chem. Soc. 1988, 86,
95. (b) Forney, D.; Jakobi, M.; Maier, J. P. J. Chem. Phys. 1989, 90, 600.
Sabo, M. 8.; Allison, J.; Gilbert, J. R.; Leroi, G. E. Appl. Spectrosc. 1991, 45, 535.
Rossetti, R.; Brus, L. E. Rev. Sci. Instrum. 1980, 51, 467.

Sabo, M. S. Ph.D. Dissertation, Michigan State University, 1991.

Ameer, G.; Benesch, W. J. Chem. Phys. 1962, 37, 2699.

Babrov, H. J. J. Chem. Phys. 1964, 40, 831.

Lin, C. L.; Niple, E.; Shaw, J. H.; Uselman, W. M.; Calvert, J. G. J. Quant.
Spectrosc. Radiat. Transfer 1978, 20, 581.

Wickliffe, M.; Rollefson, R. J. Chem. Phys. 1979, 70, 1371.

Halasinski, T. M.; Godbout, J. T.; Allison, J.; Leroi, G. E. J. Phys. Chem. 1994,
98, 3930.

Burger, H.; Burczyk, K.; Bielefeldt, D.; Willner, H.; Ruoff, A.; Molt, K.,
Spectrochim. Acta, 1979, 35A, 875.

Forney, D.; Jacox, M. E; Irikura, K. K. J. Chem. Phys. 1994, 101, 8290.
Kelsall, B. J .; Andrews, L., J. Phys. Chem. 1981, 85, 2938.

Andrews, L. In Chemistry and Physics of Matrix-Isolated Species; Andrews, L.,
Moskovits, M., Eds; Elsevier Science Publishing: Amsterdam, 1989; p 15.
Prochaska, F. T.; Andrews, L., J. Am. Chem. Soc. 1978, 100, 2102.

Jacox, M. F.., Chem. Phys. 1984, 83, 171.

Milligan, D. E.; Jacox, M. E., J. Chem. Phys. 1968, 48, 2265.

Maier, J. P. Mass Spectrom. Rev. 1992, 11, 119.

Salama, F.; Joblin, C.; Allamandola, L. J. J. Chem. Phys. 1994, 101, 10252.
Tang, Weilin; Zhang, X.-L.; Bally, T. J. Phys. Chem. 1993, 97, 4373.
Szczepanski, J .; Vala, M.; Dahbia, T.; Parisel, 0.; Ellinger, Y. J. Chem. Phys.
1993, 98, 4494.

21.
22.

23.

24.
25.

26.
27.
28.
29.
30.

31.
32.

33.
34.

35.

36.
37.
38.

117

Pullumbi, P.; Bouteiller, Y.; Manceron, L. J. Chem. Phys. 1994, 101, 3610.
Szczepanski, J.; Roser, D.; Personette, W.; Eyring, M.; Pellow, R.; Vala, M. J.
Phys. Chem. 1992, 96, 7876.

Godbout, J. T.; Halasinski, T. M.; Leroi, G. E.; Allison, J. J. Phys. Chem. 1996,
100, 2892.

Prochaska, F. T.; Andrews, L. J. Am. Chem. Soc. 1978, 100, 2102.

Hasegawa, A.; Shiotani, M.; Williams, F. Faraday Discuss. Chem. Soc. 1977, 63,
157.

Illenberger, E; Scheunemann, H.-U.; Baumgértel Chem. Phys. 1979, 37, 21.
Kiihn, A.; Illenberger, E. J. Chem. Phys. 1990, 93, 357.

Jayaraj, A. F.; Surjit, S. J. Mol. Struct. 1994, 327, 107.

Jacox, M. E.; Thompson, W. E. J. Chem. Phys. 1989, 91, 1410.

Knight, L. B., Jr. In Radical Ionic Systems; Lund, A., Shiotani, M., Eds; Kluwer:
Dordrecht, The Netherlands, 1991; p 83.

Robert Corn, University of Wisconsin, private communication.

Beckey, H. D. Principles of Field Ionization and Field Desorption Mass
Spectrometry; Pergamon Press: London, 1977.

Hagstrum, H. D. Phys. Rev. 1953, 89, 244.

Beir, M. E.; Vincenti, M.; Cooks, R. G. Rapid Commun. Mass Spectrom. 1987, 1,
92.

Fraunberg, W.-D. v.; Drechsler, G.; Bassmann, C.; Boesl, U.; Schlag, E. W. Int. J.
Mass Spectrom. Ion Processes 1994, 133, 211.

Cooper, C. D.; Compton, R. N. Chem. Phys. Lett. 1972, 14, 29.
Compton, R. N.; Reinhardt, P. W.; Cooper, C. D. J. Chem. Phys. 1975, 63, 3821.
Medved, D. B.; Strausser, Y. E. In Advances in Electronics and Electron Physics;

Marton, L., Ed; Academic Press: New York, 1965; Vol. 21, p 101.

39.

40.
41.

42.
43.
44.
45.
46.
47.
48.

49.
50.

51.

52.
53.
54.
55.
56.

118

Ertl, G.; Kuppers, J. Low Energy Electrons and Surface Chemistry; VCH
Publishers: Deerfield Beach, FL, 1985; Chapter 5.

Hagstrum, H. D. Phys. Rev. 1954, 96, 336.

(a) Hagstrum, H. D. Phys. Rev. 1966, 150, 495 (b) Hagstrum, H. D. Science,
1972, 178, 275.

Hagstrum, H. D. Phys. Rev. 1954, 96, 325.

Maier, J. P. Int. Rev. Phys. Chem. 1990, 9, 281.

Maier, J. P. Int. J. Mass Spectrom. Ion Processes 1991, 104, l.

Fulara, J.; Jakobi, M.; Maier, J. P. Chem. Phys. Letters 1993, 206, 203.

Fulara, J .; Jakobi, M.; Maier, J. P. Chem. Phys. Letters 1993, 211, 227.

Fulara, J.; Lessen, D.; Freivogel, P.; Maier, J. P. Nature 1993, 366, 439.
Freivogel, P.; Fulara, J .; Lesson, D.; Forney, D.; Maier, J. P. Chem. Phys. 1994,
189, 335.

Forney, D.; Freivogel, P.; Fulara, J .; Maier, J. P. J. Chem. Phys. 1995, 102, 1510.
Forney, D.; Fulara, J.; Freivogel, P.; Jakobi, M.; Lessen, D.; Maier, J. P. J. Chem.
Phys. 1995, 103, 48.

Freivogel, P.; Fulara, J .; Jakobi, M.; Forney, D.; Maier, J. P. J. Chem. Phys. 1995,
103, 54.

Amatore, C.; Jutand, A. J. Am. Chem. Soc. 1991, 113, 2819.

Kolomnikov, I. S.; Grigoryan, M. Kh. Russ. Chem. Rev. 1978, 4 7, 334.

Clarke, D. 8.; Suzuki, 1.; Bell, A. T. J. Catal. 1993, 142, 27.

Akemann, W.; Otto, A. Surface Sci. 1993, 287/288, 104.

(a) Schneider, T.; Hirschwald, W. Catal. Lett. 1992, 14, 197. (b) Rasmussen, P.
B.; Taylor, P. A.; Chorkendorff, 1. Surface Sci. 1992, 269/270, 352. (c) Millar,
G. J.; Rochester, C. H.; Howe, C.; Waugh, K. C. Mo]. Phys. 1991, 76, 833. (d)
Hadden, R. A.; Vandervell, H. D.; Waugh, K. C.; Webb, G. Catal. Lett. 1988, 1,
27.

57.

58.

59.
60.

61.

62.

63.

65.
66.

67.

119

Copperthwaite, R. G.; Davies, P. R.; Morris, M. A.; Roberts, M. W.; Ryder, R A
Catal. Lett. 1988, 1, 11.

(a) Solyrnosi, F.; Klivényi, G. Catal. Lett. 1993, 22, 337. (b) Linder, H.;
Rupprecht, D.; Hammer, L.; Muller, K. J. Electron Spectrosc. Related Phenomena
1987, 44, 141. (c) Bartos, B.; Freund, H.-J.; Kuhlenbeck, H.; Neumann, M.;
Linder, H.; Muller, K. Surface Sci. 1987, 179, 59.

Carley, A. F.; Roberts, M. W.; Strutt, A. J. Catal. Lett. 1994, 29, 169.

Chinchen, G. C.; Plant, 0; Spencer, M. S.; Whan, D. A. Surface Sci. 1987, 184,
L370.

Haslett, T. L.; Fedrigo, S.; Moskovits, M. J. Chem. Phys. 1995, 103, 7815.

(a) Arnot, F. L.; Beckett, C. Proc. Roy. Soc. (London), 1938, A168, 103. (b)
Smith, R. A. Proc. Roy. Soc. (London), 1938, A168, 19. (c) Sloane, R H.; Press,
R. Proc. Roy. Soc. (London), 1938, A168, 284.

Lundin, J .; Holmlid, L. Int. J. Mass Spectrom. Ion Processes 1992, 114, 137.

(a) Schlatholter, T.; Heiland, W. Surface Sci. 1995, 323, 207. (b) Schmidt, K.;
Franke, H.; Schlatholter, T.; Hefner, C.; Ntirmann, A.; Heiland, W. Surface Sci.
1994, 301, 326. (c) Schubert, S; Imke, U.; Heiland, W. Vacuum 1990, 41, 252.
(d) Schubert, S.; Imke, U.; Heiland, W.; Snowden, K. J .; Reijnen, P. H.; Kleyn, A.
W. Surface Sci. 1988, 205, L793.

Herschbach, D. R. Adv. Chem. Phys. 1966, 10, 319.

(a) Haochang, P.; Horn, T. C. M.; Kleyn, A. W. J. Electron Spectrosc. Related
Phenomena 1987, 45, 361. (b) Haochang, P.; Horn, T. C. M.; Kleyn, A. W.
Phys. Rev. Lett. 1986, 57, 3035.

(a) Feher, M.; Maier, J. P. Chem. Phys. Letters 1994, 227, 371. (b) Schmatz, S.;
and Botschwina, P. Chem. Phys. Letters 1995, 235, 5.

Chapter 4

Investigations of Mass-Selected CS2+ Depositions with Neon

I. Introduction

Three aspects of the previous investigations of mass-selected cations in neon
matrices make the study of CSf‘ a logical next step. First, while earlier attempts from this
laboratory using mass-selected ion beams of smaller currents and lower kinetic energies
were successful in accumulating sufficient numbers of CS} ions for laser-induced
fluorescence detection, they were unsuccessful in continued ion accumulation .for infrared

1 Since the current experimental setup includes an ion source which produces

detection.
ion currents more than an order of magnitude greater, the decision was made to return to
this particular challenge.

Second, through experiments involving relatively low mass ions, we hope to
sufficiently understand the processes which occur during ion deposition such that data for
larger ions can be understood and interpreted. The comparison of results for two similar
species, COf' and CSf‘, may reveal aspects that require further elucidation. There are
currently two extreme views of this experiment, in which ions generated with kinetic
energies above 100 eV are deposited into a growing neon matrix. The first, the ideal,
would be that the low temperature matrix serves a simple purpose - the capture and
isolation of incoming ions. Another extreme is that the system is a reactive one, in which
ionic and neutral species in the outermost layers of the matrix can react. Reaction
pathways of particular interest include collision-induced dissociation (CID) and cation

neutralization processes. To understand which view is closer to reality, the decision was

made to compare the CS2 system with the C02 system.

120

121

The thermochemical data2 in Table 4.1 suggest that deposition of CSf' ions could
yield very different results than COf’. While these two triatomic molecules are obviously
similar, they difi‘er in ways that could be critical in this experiment. The ionization energy
(IE) of C02 is more than 3 eV higher than that for C82. One may expect COf' to
undergo charge transfer with matrix components such as H20 (IE = 12.6 eV),2 whereas
CS2+' would not do so. In general, the neutral and ionic forms of C82 are characterized by
bond dissociation energies (BDEs) that are lower than the corresponding BDEs for
COz-related species. Thus, if CID occurs, it may be more extensive for CS2+°. It is
advantageous to explore possibilities such as CID in experiments with small triatomic ions
rather than with larger organic ions, since the number of possible fragments from the
triatomic is small, and their vibrational spectra are usually understood.

The third reason why C82 was selected is related to counterion generation in these
experiments. While C02 as a matrix additive is exceedingly usefirl, the linear molecule has
a negative electron affinity. Thus, one goal in this work is to evaluate CS2 as a matrix
additive, since, as Table 4.1 indicates, this triatomic species has a positive electron affinity

and may thus be a more useful additive for counterion generation.

11. Experimental Parameters of the Mass-Selected Cation Depositions

CS2+' and CSP mass-selected ion currents of ~50 nA and ~10 nA, respectively,
were used, and the ions were generated with kinetic energies of 130 eV. The resolution of
the quadrupole mass spectrometer was degraded to less than unity to increase
transmission of the mass-selected ions. Experiments involving depositions of CS2“ used
naturally occurring C82, which contains 1.1% 13CS2; however, the amount of ”CS?
deposited during CSf‘ depositions is negligible. The matrix gas was pre-mixed with C02
or C82. The Nezadditive ratio was always at least 1000:] and the gas was deposited onto
the window at a rate of 0.5 mmole/hr. FTIR spectra were recorded by averaging 256 scan

1

files with 1cm“ resolution. The spectra taken before the deposition were used as

122

Table 4.1. Thermochemical data (eV) for the neutral and ionic species of CX2 (X = O, S).
All thermochemical data listed and values used to compute the BDEs were taken from
reference 4.

 

 

 

 

 

 

x=o x=s
neutral
EA -0.6 +0.5
IE 13.8 10.1
BDE(CX—X) 5.5 4.4
cation
BDE(CX—-X+) 5.4 4.7
BDE(CX+—X) 5.8 5.8
anion
BDE(CX—X‘) 3.5 2.9
BDE(CX‘—X) 8.6 4.7

 

 

123

background for correction of the absorption after matrix formation. The %-AA values
displayed in Figure 4.5 were obtained from baseline corrected spectra. Small uncertainties
in these values evolve from manual selection of the baseline for each spectrum. The
corrected spectra are obtained by calculating a least squares fit polynomial curve through
these selected baseline regions.

Ab initio calculations were performed using the 6-311+G"' basis set as
implemented in the GAUSSIAN 92 program3 to assist in the assignment of the observed
spectral features. These calculations were carried out at the restricted closed-shell
Hartree—Fock (RHF) and unrestricted Open-shell Hartree-Fock (UHF) levels for the
neutral and ionic species, respectively, of C52, CS, C02, and CO.

The sample window temperature was measured with a Lakeshore Cryotronics 330
controller and a DT-470 Si diode sensor. The controller was also utilized to warm the
sample window for annealing studies. In the cold diffusion experiments, the temperature
of the matrix was initially ramped from 4 K to 6 K in approximately 60-70 seconds and
immediately cooled down to 4 K. A spectrum of the matrix was then acquired at 4 K to
analyze the matrix components. This process was repeated, successively ramping to
temperatures of 6, 8, 10, and 12 K, with FTIR spectra taken after the return from each

high temperature excursion.

III. Results of the Mass-Selected Depositions of CS2+' and CS+'

Figure 4.1 displays the spectra in the 1460-1080 cm'1 region from five experiments
in which mass-selected ions were deposited into neon matrices doped with either C02,
CS2, or 13C82. This region has been selected from the entire spectral region scanned
(4800-875 cm") because it contains the spectral features which are most pertinent to the
deposited cations and the counterions. Other absorptions, such as those due to neutral

C02 and C82, are in the regions not shown.

124

 

 

 

 

Des; —» Nezco,j v "cs v, "cs; v, “cs;
I : : ii ‘\1 :4 A
9 ii % v. "CS;

 

 

 

 

E’cs; —- Ne:"CS,

 

 

 

ECS; —> NGZuCSZ ]

 

 

 

 

 

 

 

10.002 A.U.
WWW E
1500 ‘ 1400 A 1300 l 1200 ‘ 1100 A
WAVENUMBERS

Figure 4.1 Spectra in the 1460-1080 cm'I region of five
[ion—)matrixzadditive] experiments.

net
der
[10
v.11
val

1118

Pi
cc
ne
di
fe

in

SF

125

Figure 4.1A displays the results following a 24 hour deposition of CS2+° into a
neon matrix doped with C02. The set of experimental variables for this experiment is
denoted as [CSf’ —> Ne2C02]. Other experiments will be designated in the same way,
[ion —) matrixzadditive]. Unless otherwise indicated, the neutral and ionic species of CS2
will refer to the unlabeled form. Figure 4.2 displays the growth of the peak absorbance
values of several matrix components during the formation of the [CSf' -)Ne:COz]
matrix whose spectrum is displayed in Figure 4.1A. These growth curves of matrix
components are typical of all experiments involving the depositions of mass-selected
cations in neon matrices d0ped with C02. As is often done in infrared studies, the
experiment was repeated with an isotopomer of the ionic analyte, 13CSZ+1 Figure 4.1B
displays the results of a 17 hour [13CSZ+' —-) NezCOZ] experiment. Experiments were also
performed in which the neon gas was doped with C82 or l3C82 during depositions of
CSf‘. Figure 4.1C displays the results of a 23 hour [CSf' -—> NezCSZ] experiment, and
Figure 4.1D displays the results of a 21 hour [CSf‘ ->Ne:l3CS2] experiment. As a
control experiment, and also to attempt to establish the vibrational frequency of CS“ in
neon, a 28 hour [CS+' —-) Ne:C02] experiment was performed; the resulting spectrum is
displayed in Figure 4.1E. A summary of the vibrational frequencies of the spectral
features observed in these experiments is given in Table 4.2. During control experiments
in which all parameters are held constant, except that the quadrupole mass filter is set at a
We. value for which the ion current is zero or negligible, only the bands due to the neutral
species of C82, C02, and H20 are observed. For the systems studied to date, neutral
fragments diffusing from the ion source have not been observed in the matrix spectra.

Ab initio calculations of the vibrational frequencies and the corresponding
absorption intensities of both neutral and ionic species were canied out to assist in the
assignment of the observed spectral bands. The results for the asymmetric stretch of the
neutral and ionic CS2 and C02, along with the vibrational frequencies of the neutral and

ionic CS and C0, are listed in Table 4.3. Although higher level calculations on neutral and

Absorbance Absorbance

Absorbance

126

 

 

 

 

 

 

 

 

 

 

 

 

2.000 __.___, ,
1.500 /. -”" vtCO.
/'/

1.000 _ ./-
0.500 —./ v,Cs,
0.000 '_._.u—-.——..—--—.--:--:—".—"T'_. .

5 10 15 20 25
Time of Deposition (hr)
0.040 .
0.030 v2 H,O "rotating" /,/'/
. ./,
0.020 _ /././ ,... .
0'010 _ /. ::.-""". . v,H,O "nonrotating"
4"?" - . - 1 - t - .
0000 5 10 15 20 25
Time of Deposition (hr)

0.010
0.008 , /o/°"'° °
0 006 V3 C0; anion o/O v3 C82 anion (g)

' - /c/°’ v, CS, cation (o)
0.004 _ /.,/° o__,,..c c
0.002 /./;:g:gfgf::::,_,_,,o o vCS
0000 5 10 15 20 25

Time of Deposition (hr)

Figure 4.2 The growth of peak absorbances of several matrix components
during the deposrtion of the matrix which resulted in Figure 4.1A.

Table

II

127

Table 4.2. Tabulated vibrational wavenumbers (cm") of the spectral bands observed in

 

 

 

Figure 4.1 A-E.
1 A 1 B 1 C 1 D 1 E Assignment
1104.4 1104.4 _
1122.2 1122.5 v3 13CSZ-.
1159.4 1159.4 1159.4 (,3 ”cs;-
1167.6 1167.6 v3 ”(352+-
1207.1 1207.1 1207.1 v3 ”cs;
1211.5 1210.5 _
1224.0 _
1237.5 1237.5 l3cs
1256.8 1256.8 160212C”1602-.
1273.7 1273.2 1272.7 12CS
1308.9 —
1340.9 a
1386.0 3
1422.2 14222 V3 izcoz.
1445.3 _

 

 

a See discussion for possible assignments.

Table
form:
each

128

Table 4.3 Calculated vibrational frequencies and intensities for the neutral and ionic
forms of C02, C32, CO and CS. The corresponding frequency and intensity values of
each species containing a carbon-13 isotope are given in parentheses.

 

Triatomic Species

V3 (9111.1)

intensity (km/mole)

 

co;
co2
co;-

933.4 (906.9)‘
2575.7 (2502.4)
1857.8 (1807.2)

231 (223)
1160 (1095)
1142 (1091)

 

Diatomic Species

v (cm'l)

intensity (km/mole)

 

CO+'
CO
CO"

 

2412.8 (2359.0)
2432.0 (2377.8)
1439.3 (1407.2)

 

8 (6)
163 (156)
2543 (2454)

 

 

 

 

 

Triatomic Species v3 (cm’l) intensity (km/mole)
cs; 1156.2(11180) 362 (333)
C82 1582.1 (1529.8) 1567 (1466)
C82" 1185.2(1146.7) 990 (933)

Diatomic Species v (cm'l) intensity (km/mole)
CS+' 1260.1 (1224.2) 2 (2)
CS 143l.5(1390.8) 155 (146)
CS" 1019.0 (990.0) 109 (107)

 

 

 

a See discussion.

ionic
inten
perfi

00m

expe
inte:
spet
the
not
10 l

T851

0b:
the
dif
shi

1h:
mi

01

Fi

129

ionic C82 and C02 have been previously published,4 few report computed spectral
intensities and all lack vibrational information for C82". Thus, the decision was made to
perform these calculations for the species of interest at the same computational level for
comparative purposes.

As Table 4.3 illustrates, vibrational intensities for the species involved in this
experiment can vary widely. In particular, the cationic species show relatively small
intensities. This decrease in the infrared intensity from the vibrational mode of a neutral
species to the corresponding cation has been discussed by Chin and Person.“ Thus, with
the presently available mass-selected ion currents, ions with low infrared intensities may
not be spectroscopically observed, although they may be present in amounts comparable
to neutral and anionic matrix guests. This may also explain the occasional experimental
result in which infrared absorptions due to cationic matrix components are not observed.

The calculated results listed in Table 4.3 compare well with the experimentally
observed values. The value for v3 of C82 is calculated as 1582.1 cm'l; it is observed in
the gas phase5 at 1535.4 cm'1 and in a neon matrix at 1533.6 cm‘l. This is typical of the
difference between computed frequencies and measured values. The calculated fractional
shift upon isotopic substitution of the carbon and the observed fractional shift for v3 of
C82 are both 0.967.

The discussion that follows will focus on the trends and correlations observed in
the present experiments. It should be noted that many questions remain, in part due to the
many experimental variables. While the ion energy and matrix gas flow can be controlled,
other important variables such as the ion beam focusing properties, ion current, and the

composition of the condensed gases on the cold metal surfaces vary between experiments.

A. CSz+°
Of the several peaks observed in the [CSf‘ —) NezCOz] FTIR spectrum,
Figure 4.1A, the absorption band at 1207.1 cm'1 can be readily assigned to v3 of CSf‘.

phot
CXpl
121
exp
for
exp
333

0111

ll
11

Va

CC

d:

130

This agrees well with emission studies of CSf‘ formed in neon matrices through
photoionization,‘S where an emission band at 2418 cm'1 was assigned to 2V3. A similar
experiment using infrared spectroscopy7 also tentatively identified v3 of C824“ at
1211 cm'l. An absorption band is observed at 1167.6 cm'1 in the [13CSZ+° —)Ne:C02]
experiment, Figure 4.1B, which is consistent with the assignment6 of 2340 cm'1 for 2V3
for l3C82“. It should be noted that v1 and v2 of CSf’ are not observed in these
experiments. The bending mode v2, which has been observed in the gas phase at
333 cm'l,8 is at too low a frequency to be observed in the spectral window provided by
our present apparatus, and V] is not infrared-active.

Previous calculations of v3 for CSf‘ compare well with the computed result of
1156.2 cm’1 listed in Table 4.3. Use of the 6-31G“ basis set4c results in a value of
1152 cm]. A complete active space self-consistent field (CASSCF) approach"f predicts a
value of 1195.0 cm]. The observed isotopic fractional shifts also compare well with the

calculated results; both are 0.967.

B. CO,"

As discussed in the previous chapter, C02 can be used as an additive to increase
counterion production. Absorption bands at 1658.5 cm'1 and 1665.7 cm'l, due to C02"
and (C02)2",9 respectively, were observed in all experiments which involved ion

depositions into neon matrices doped with C02.

C. (C0; + 0;)"

Absorption bands at 1256.8 cm’1 and 1895.8 cm“1 are observed when C02 is used
as the matrix additive, and the mass-selected ion current is high. Figure 4.3 displays these
bands, which were detected after a 24 hour [CSf' -—> Ne2C02 100021] matrix deposition.
These bands have been previously assigned to v2 and v1, respectively, of (C02 + 02)"

(which may be represented as CO4") isolated in a neon matrix.lo Three mechanisms for

131

 

[0.002 A. U.

W
l j T I 1 l

l ' I ' I ' l ‘
2180 2170 2160 2150 2140 2130 2120 2110
Wavenumbers

 

 

 

 

/\ [0.002 A. U.

 

 

 

1
1910 1900 1890 1880 1870 1860 1850 1840
Wavenumbers

 

[0.002 A. U.

/\

\
We. /\~"“v' (“JV 4" \JWf W
1 v
V V V V V

I 1 T ' T T I I l I
1300 1290 1280 1270 1260 1250 1240 1230
Wavenumbers

 

 

 

Figure 4.3 FTIR spectra of several regions of a neon matrix following a
24 hour [CSf -> Ne:CO2 1000:1] deposition.

132

 

[0.002 A. U.

 

 

 

I ' T ' I ' I ' I ' I 1 I ‘ I
2180 2170 2160 2150 2140 2130 2120 2110
Wavenumbers

 

[0.002 A. U.

 

 

 

I a I ' I ' I ' I fl I 1 I ' I

1910 1900 1890 1880 1870 1860 1850 1840

 

 

 

 

Wavenumbers
[0.002 A. U.
.’ A/Afl‘l".
x/ f. [/N/ .‘1 4‘
W. \ .V. U fVMW
1300 1290 1280 1270 1260 1250 1240 1230
Wavenumbers

Figure 4.4 FTIR spectra of several regions of a neon matrix following a
24 hour [C82+ —) Ne2CO2 300:1] deposition.

gene:
radia
wind
010’!
pres
ch81
ther

dire

ion

sinl

811‘

18

ll

133

generation of CO4" in the matrix are reasonable. First, C02 and 02 adsorbed on the
radiation shield might form CO4" directly upon ion bombardment, and accumulate on the
window as does C02". This is certainly viable, since condensable components such as
oxygen will be most concentrated on the radiation shield, and oxygen molecules are
present in these experiments as a constituent of the background gas present in the vacuum
chamber. Since we know C02" is formed and deposited into the matrix, it could combine
there with 02 to form CO4" via “matrix chemistry”. It is difficult to assess this possibility
directly, since the amount of 02 present in the matrix cannot be measured by infrared
spectroscopy. A third possibility is that 02" may desorb from the radiation shield upon
ion bombardment and react with C02 in the matrix. This is also a reasonable mechanism,
since neutral C02 is present in a concentration that is high relative to other matrix
components. It remains to be determined whether deposited ions find a reactive
environment in the growing matrix, or if they are simply trapped and isolated under the
conditions used.

A band at 1865.1 cm“1 with a shoulder at 1862.4 cm'1 has also been attributed to
CO4" isolated in a neon matrix.lo Absorption bands at 1852.9 cm'l, 1855.3 cm'l,
1862.5 cm’l, and 1864.9 cm'1 also appear in our experiments where C02 is present as a
matrix gas additive. These bands are also displayed in Figure 4.3. The 1852.9 cm"1 and
1855.3 cm'1 bands were not observed in reference 10. Figure 4.4 displays the same bands
attributed to CO4" which were detected afier a 24 hour [CSf‘ —) Ne2C02 300:1] matrix
deposition. Comparison of Figures 4.3-4 reveal that when the 1256.8 cm'1 and
1895.8 cm‘1 bands are relatively intense, the two doublets in the 1855 cm'1 region have
decreased in intensity. It is possible that these two doublet bands arise from difl‘erent
CO4" geometries. This would explain their variable intensities with respect to the
1256.8 cm'1 and 1895.8 cm’1 absorptions over the course of several experiments in which
deposition parameters and the resulting dominant mechanism for CO4" formation may

change. The intensity of an unassigned band at 2146.6 cm‘1 appears to follow the

bel

134

behavior of the two doublets, as also revealed in a comparison of Figures 4.3-4. This band
may be due to a complex of CO with some as yet identified species since it is slightly blue
shified from the CO absorption at 2141.7 cm]. It cannot be discerned whether the
behavior of these bands can be attributed to the level of C02 present in the matrix region.
Other experimental variables, such as the mass-selected ion current, also vary between

experiments and may contribute to the overall behavior.

D. CS 2"

An absorption band at 1159.4 cm'1 is present in experiments in which C82 is
present in the matrix region, due either to diffusion of neutral precursors from the ion
source or to direct doping of the neon matrix gas with C82, although it is not observed in
the [CSf‘ —-> Nezl3CSZ] experiment. This band shifts to 1122.2 cm'1 when 13C82 is used
as either the neutral precursor (Figure 4.1B) or the matrix additive (Figure 4.1D). Based
upon the ab initio calculations summarized in Table 4.3, the 1159.4 cm’1 and 1122.2 cm“1
bands are assigned to v3 of CS2" and 13C52", respectively. The calculated fractional shift
for v3 of CS2" upon isotopic substitution of the carbon and the observed fractional shift of
these bands are both 0.968. Presumably, CS2" is formed from mass-selected ion
bombardment of CS2 that has been frozen on the heat shield or other cold metal surfaces,

similar to the proposed mechanism for C02" formation.

E. CS

An absorption band at 1273.7 cm’1 is observed in all experiments in which CS2 is
present in the matrix region. This absorption shifts to 1237.5 cm‘1 when 13C82 is used as
the matrix gas additive. These bands are consistent with previous assignments of the
vibrational fi'equencies of CS in argon matrices and in the gas phase. The vibrational
frequency of CS has been reported at 1275.1 cm’1 and 1272.2 cm”1 in an argon matrixn

and in the gas phase,12 respectively. Likewise, the vibrational frequency of 13CS has been

repc

resp

pre
reh
0f
b0]
the

III
fie
fir

5P

CC

135

reported at 1239.1 cm'1 and 1236.3 cm'1 in an argon matrix11 and in the gas phase,12

respectively.

F. COX“

A weak absorption band due to COZP is observed at 1422.2 cm’1 when C02 is
present in abundance and large mass-selected ion currents have been deposited for
relatively long periods of time. The appearance of CO; has been attributed to ionization
of neutral C02 molecules in the matrix by electrons that have been formed by cation
bombardment of surfaces and are accelerated towards the matrix by the potential due to
the accumulated ions present. This has been discussed in the previous chapter.

The computed value for v3 in Table 4.3 is anomalously low for this cation.
UHF/6-311+G" calculations using the GAUSSIAN 92 program result in four vibrational
frequencies for the COf‘ cation. These are 446.4, 626.6, 933.4, and 1363.6 cm'l. The
first two values are the bending (v2) vibrational frequencies. The non-degeneracy of these
two values is a typical result for UHF calculations involving linear, open-shelled, triatomic
species.13 The ordering of the fundamental frequencies for linear symmetric triatomics is
commonly v2 < v1 < v3, as is true for the calculated results for CSf‘, C82, and C02. One
might be tempted to assign the 933.4 and 1363.6 cm"1 frequencies as v, and v3,
respectively. However, the calculated atomic displacements and IR and Raman intensities
clearly show that the higher frequency vibration (1363.6 cm!) is due to a symmetric
stretch motion (v1) and that the lower frequency (933.4 cm’l) belongs to an asymmetric
stretch motion (v3). This calculated v3 frequency is anomalously low as compared to the

l. The reason for this discrepancy may be

gas phase experimental valueH of 1423.1 cm'
similar in nature to the restricted Hartree—Fock instability attributed” to the potential

energy function for COf‘.

136

G. Other Spectral Features

When C82 is substituted for C02 as the matrix additive, an absorption band
appears at 1386.0 cm], as shown in Figure 4.1C. This band shifis to 1340.9 cm’1 upon
isotopic substitution of the carbon atom, as shown in Figure 4.1D. A possible assignment
for this band is CS+', as its vibrational frequency has been estimated16 at 1376.6 cm'1 in
the gas phase. The calculated fractional carbon-13 shift of 0.972 is comparable to the
observed shifi of 0.967 .

However, it is more likely that the species responsible for the absorption at
1386.0 cm'1 may be neutral or charged polymers of carbon and sulfur atoms. CS2
adsorbed on the radiation shield and other cold surfaces should be expected to polymerize
to a much greater extent than C02. Matrix-isolated polymers of CS2 may originate from
the efl'ect of ions striking the adsorbed C82. The absorption at 1386.0 cm'1 does not
appear in experiments in which Ne:CS2 100021 mixtures are deposited in the absence of a
mass-selected ion beam and therefore is not due to polymers of C82 formed in the gas
phase. Photopolymerization studies of CS2 have been shown to yield (C82)x polymers
with a strong absorption centered around 1410 cm'l." This absorption shifts to
1364 cm“1 in (13C82)x." The reported fractional shift of the polymer and the observed
fractional shift in our experiments are both 0.967. Moreover, the large observed intensity
of the bands make it unlikely to be due to CS“, which has a calculated absorbance
coefficient of only 2 km/mole (Table 4.3).

In support of this conclusion, when a current of 10 nA of CS+', formed fi'om C82,
is deposited for 28 hours [CS+' —> NezCOZ], no absorbances other than those previously
assigned to C82" and C02" are observed (Figure 4.1E). Presumably, the cation present in
this experiment is CS+' and it is not observed owing to its low extinction coefficient.
Therefore, it is proposed that the absorption at 1386.0 cm'1 is due to oligomeric forms of
C82. In the previous chapter, a mechanism was discussed whereby anionic species,

formed on metal surfaces, find their way to the matrix due to their attraction to the

posi‘

sput

rem
122
rep
ma
C01
p0
1111

83

tl'

137

positive potential of the matrix. Here, it appears that oligomeric (C82)x species may be
sputtered off such surfaces and diffuse to the matrix as well.

Several absorption bands are frequently observed in these experiments which
remain unassigned. The most prominent of these bands are at 1104.4 cm'l, 1210.5 cm],
1224.0 cm 1, 1308.9 cm'l, 1445.3 cm], and 2146.6 cm'l. The origin(s) of the low
reproducibility for these bands are unknown. Of all the experimental parameters, the
mass-selected ion current and ion beam focusing are the most difficult to maintain over the
course of several different experiments. As discussed in the previous chapter, the matrix
potential and ion trapping efiiciencies are all intimately related to these parameters. With
the small signal/noise ratios encountered in these experiments, it is not surprising that
some features of the spectra obtained may be difficult to reproduce.

Possible candidates for these unassigned absorptions are species containing some
combination of carbon, oxygen, and/or sulfur atoms. Absorptions which have been
assigned to species such as $0, $02, $20, OCS, C252, and C352 do not match well with
these spectral features. Other candidates that have been considered are complexes of

matrix guests with N2 or H20, due to their high abundance in the background gas. H28,

HCS+, and HzCS do not have absorptions at the listed frequencies, as well.

IV. Annealing of the Matrix: The Cold Diffusion Experiment

Photolysis (photodetachment of electrons from matrix-isolated anions) and
annealing of the matrix are two common methods used in matrix-isolation spectroscopy to
provide additional information on the origin of spectral features. When ions are under
study, the growth or disappearance of a spectral feature during photolysis may be related
to neutralization of ionic species. Annealing experiments are most often used to reduce
matrix site distributions to simplify the spectrum. The experiments presented here were
performed at matrix temperatures where diffusion of the guest species appears to occur

without significant losses due to vaporization, similar to the work of Andrews et al.18 in

Spt

1111
ha

138

which polymers of HF were formed during neon matrix annealing experiments.
Obviously, in experiments involving ionic species, infrared spectral features could
represent charged or uncharged matrix components. If the temperature is raised and
trapped species begin to diffuse, charged species would be expected to recombine at the
highest rate. Thus, the time dependence of peaks in annealing experiments might provide
a signature for the presence or absence of charge on molecules responsible for various
spectral features. This possibility is evaluated here.

Annealing experiments were performed afier both the mass-selected ion beam and
the matrix gas deposition were stopped to observe the behavior of the intensity of each
band after matrix excursions to temperatures above 4 K. The 4 K matrix was raised to
6 K over a period of ~1 minute. Once 6 K was reached, the matrix was quickly cooled to
4 K and a spectrum was obtained. This process was then repeated at 8 K, 10 K, and 12 K.
Thus, by incrementally accessing higher temperatures, conditions could be gradually
achieved at which diffusion occurs for a short period of time without considerable loss of
the guest and host due to vaporization. This is reflected in absorptions for “unreactive”
matrix guests like C02 where, as shown in Figure 4.5A, the %-AA due to the asymmetric
stretch, normalized to the absorption in the original 4 K spectrum, never varies by more
than 10% over the course of the annealing experiments.

One reason for this experiment was to test the proposal that ionic species would
diffuse and recombine more quickly than neutral species as molecular mobilities increase,
possibly providing clues to the charge state of the species responsible for a particular
spectral feature. With a few exceptions, all bands decreased in intensity at approximately
the same rate with increasing matrix temperature. The %-AA of the peak absorptions due
to (v3) CS; and (v3) C82" through the annealing experiments are shown in Figure 4.5B.
These changes in %-AAs are typical of reactive matrix components that are present at low

levels.

139

 

120 f A
110 f C02

100 f = \ __________._—o

90 *- —'

80 r
70 :

 

 

120 f B
110 r CS; '

100 -
90 L
80 } cs;
70 r

°/0-AA

 

120 r nonrotating - H20 C
110 r

4

100 ’ = =
90 r .

70

ff

 

J A L 4 k A 1 A

 

 

6 8 10 12
Temperature (K)

Figure 4.5 (A) Relative change of the v3 CO2 absorption, (B) relative
change in absorption of v3 CS," and V3 C5,", and (C) relative change in the
absorption representing the nonrotating and rotating v2 bands of H,O at
each annealing step, following a 17 hour [”CS," -+ Ne:CO,] deposition. In
each case, %AA={A(i K) - A(4 K)} / A(4 K)} x 100, where A(i K)
represents the absorption of the spectral feature taken after an annealing
excursion to a temperature of i Kelvins.

140

One of the exceptions was observed in the C02 anionic monomer and dimer region
of the spectra. An example is displayed in Figure 4.6, taken from a matrix produced from
a 17 hour [13CS2+' -+Ne:C02] deposition. The fiont spectrum of Figure 4.6, which is
labeled 4 K was acquired before annealing. Each succeeding trace is labeled with the
highest matrix temperature reached in that particular annealing step. In these spectra, the
C02" absorption at 1658.5 cm'1 decreases in intensity with each annealing step. The rate
at which this band decreases in intensity is comparable to the behavior of the majority of
bands observed in the entire spectrum obtained. The (C02)2" absorption at 1658.5 cm'l,
however, increases in intensity with each annealing step. A new absorption at 1670.6 cm'1
continues to increase in intensity through all annealing steps shown. These three spectral

features likely demonstrate clustering chemistry in the matrix:
_ co.2 _ co, -
C02 ° —') (C02)2 ' —9 (C02)3 ..

On the basis of this proposal, the 1670.6 cm‘1 absorption is assigned to (COZ)3". Jacox
and Thompson also observed9 a spectral feature at 1670.2 cm‘1 in their C02” study and
suggested that the absorption was due to a charged molecular aggregate of C02. The
(CO;);," assignment seems reasonable since C02 is presumably the species with the
highest matrix concentration other than neon. While C02" ions likely recombine with
cations at some point, the cation concentration is much lower than the neutral C02
concentration, so diffusing C02" anions encounter C02 molecules more often than
counterions. More generally, this demonstrates that one can deposit ions and neutrals and
have them react in the matrix through controlled annealing.

The absorptions at 1386.0 cm’] [CSf' —>Ne:CS2] and 1340.9 cm’1
[CSf’ -> Ne:l3CSZ] have also been observed to initially increase in intensity during the
annealing process. This observation reinforces the conclusion that these absorptions are

not due to CS“ and 13CS+', respectively, since it is unlikely that an ionic monomer would

.528qu
160qu T amu”; So: 2 a 9.36:8 8.382.an wE—aoccu .826.“
3 x52. :8: a mo comm”: ..Eo 9.2.82 05 E «58% 5.: WV. 953..—

 

141

 

 

 

 

 

 

 

 

wmwmznzmzss
ave. one. one. came anew cam.
v. v - 8o...
2 v. a
(Will/V) _ x we a
1 33
\ . 1 Sod
‘ - a:

 

 

 

 

 

 

EONVSBOSBV

incr
cm
of 1

ads

Tl

IE

142

increase in abundance during the annealing process. The observation is taken as further
evidence that the species responsible for these absorptions are probably oligomeric forms
of CS2. While (C82)x may originate from the effect of the mass-selected ion beam striking
adsorbed C82 on metal surfaces, the polymers might increase in abundance during the
annealing process due to a reaction involving charged CS2 ions with C82 in the matrix.
Thus, while C02" may form oligomers upon annealing, the more chemically reactive
anionic and/or cationic forms of C82 may exhibit polymerization chemistry, as does CS,
the photolysis product of C82.19

The only other exception observed in the annealing experiments was the (v2) H20
region of the spectrum. Figure 4.7 displays the 1650 cm’1-1585 cm'l region of the same
spectra shown in Figure 4.6. The absorptions at 1595.8 cm'1 and 1631.0 cm'1 have been
previously assigned20 to (00.0 4— 00’0) v2 and (11,1 (— 00,0) v2 of H20, respectively. These
will be referred to as the nonrotating and rotating bands of H20. As Figure 4.7 shows, the
rotating water band slightly decreases in intensity and the nonrotating component
increases in intensity during the annealing process. To assist in describing how the peak
absorptions at 1595.8 cm'1 and 1631.0 cm'1 change during the course of the annealing
steps, the °/o-AAs are provided in Figure 4.5C. Their matrix-temperature dependence
clearly differs from the other spectral features. Ion-dipole and dipole-dipole interactions
should become important as the mobilities in the matrix of ionic and polar species increase,
as the matrix is heated. The abundance of nonrotating H20 in the matrix should then be
expected to increase as observed due to these “clustering” interactions. The shoulder at
1599.2 cm'l, which has been previously assigned20 to (1120);, is also observed to increase
during the annealing process. This is further evidence for the “clustering” process. We
note that this behavior has not been observed in previous photoionization experiments,” 21
where the rotating H20 band increased slightly in intensity upon annealing. That behavior
was attributed to the decreasing electric field as ions in the matrix are neutralized upon

annealing. The reasons for the different observations in the v2 H20 region in these two

143

cam?
-

.coEmoaov

@00qu T tamoau So: 2 a 336:8 4.238388 mam—855 .838
“a 3:38 :8: a .8 SEE 7:6 mam—-32 05 E 880% 5.: 5. 2:3"—

oowr

p

P

wmwm232w><>>

crow

P

L

one? One? over ommp

-

P p p P _

 

 

 

 

 

 

 

 

a,
a

x m

w.
Xm
v.9.
XNF

 

 

 

 

 

 

 

 

 

1

AL

1

1

l

 

OOOAU

mOOAU

owoAu

mPOA.

BONVBHOSGV

CXP
obs
1’01.
wh
15
Or
th:

fe

at

144

experiments are not clear; perhaps the concentration of ions in our matrices is too small to
observe the opposite effect. The “rotating” and “nonrotating” labels, referring to the final
rotational state in each spectral transition, can be misleading. When sites are considered in
which H20 could be either held rigid or allowed to rotate, the two observed features at
1595.8 cm‘1 and 1631.0 cm'1 could represent a single site in which rotation is possible.
Only upon annealing, when the relative peak intensities change, can it be stated that more
than one site is available and present for H20 molecules in this experiment.

Finally, it should be noted that annealing does not result in many new spectral
features. In this regard, the most substantial feature that appears upon annealing grows in
at 1189 cm“. While possible “reaction products” have been considered, the peak remains
unassigned. Its annealing profile most strongly parallels the CO4" absorption at

1852 cm".

V. A Comparison of the COz'H" and CSZH" systems

When incoming ions with initial kinetic energies of more than 100 eV encounter
the growing matrix, is the result simple trapping, or do ions undergo “matrix chemistry”?
Specifically, does collision-induced dissociation (CID) of the incoming ions occur or are
the ions decelerated by the matrix potential such that CID is insubstantial? Also, do the
deposited ions react with the variety of low concentration matrix components such as
H20, N2, etc? In the C02” depositions, there was no evidence in the spectra which
suggested that the C02” ions fragment as they encountered the matrix. Table 4.1
indicates that the CS2” ions are less stable with respect to fragmentation. Thus, incoming

C82” ions may be more likely to undergo CID reactions such as:

'——> 08‘“ + s 1
CSZ‘h + Ne(s or g) ‘7 ( )

'L-> CS + S+' (2)

 

 

They

path1

mole
mat:
add
0.01
0.01
[CS
obs

ext

gr:
rat
00
[C

of

CC

21;

145

These reactions have been extensively studied by Futrell and coworkers.22 Prinslow and
Arrnentrout23 show that, at low collision energies, pathway (2) is the dominant CID
pathway.

As shown in Figure 4.1, CS is observed in these experiments. However, it is
suggested that CID of the mass-selected C52” is not the dominant source of the CS
molecules detected. Comparison of many experiments in which CS2 and C02 are used as
matrix additives reveals that the CS signal is clearly much more intense when C82 is the
additive. The peak absorbance of CS in the 24 hour [CSf‘ —) NezCOZ] experiment was
0.00156 A. U., whereas that of the 23 hour [CSf’ —>Ne:CSz] experiment was
0.00555 A. U. The peak absorbance of CS in the previously reported1 48 hour
[CSf' —) Ar] experiment, which utilized mass-selected ion currents of ~1 nA, was
observed to be ~0.3 A. U. The ion source of the mass spectrometer used in the argon
experiment was directly inside the matrix chamber and lacked differential pumping. The
amount of neutral precursor C82 which diffused to the matrix was presumably much
greater in the argon experiment. This suggests that the CS evolves fiom neutral CS;
rather than from the mass—selected ions. It may well be another product of ion
bombardment of CS2 on the heat shield, which difiiises to the matrix. Even in the
[CSf' -) NezCOZ] experiment, there is CS2 on the heat shield - due to diffusion of the
C82 introduced in the ion source into the region of the matrix. Thus, although CS is
observed, it is probably not indicative of the CID process.

As discussed earlier, CS+' may be present in these matrices, but its low extinction
coefficient makes its detection unlikely. While CID may occur to some extent, it does not
appear to be a dominant process. In both the COf‘ and CS2” depositions, intact cations
are detected, and results suggest similar trapping efficiencies for both, even though the
CS? ions require less energy for fragmentation. Perhaps this is indicative of high average
matrix potential, which is effective in lowering the kinetic energies of the incoming ions

during the deposition process.

addi

is d'
The
abs
spe
shc
gel

res
the
ho

is
le

ir.

146

Finally, a comment should be made on the selection of C52 vs. C02 as a matrix
additive. Based on its positive electron affinity, it was thought that C82 would be a more
efi‘ective additive for the generation of counterions. The results when the neon matrix gas
is doped with CS2 during cation depositions show that its use is not preferable over C02.
The use of C82 as the additive results in relatively complex spectra, showing a number of
absorptions in addition to that attributable to C82". Also, C02 is preferable from a
spectrosc0pic perspective, since C02" can be detected at lower levels than C82". It
should be noted that relatively high Ne:COz ratios appear to adversely affect the
generation of C02". Mass-selected CS2+' depositions using a NezC02 ratio of 300:1
resulted in a notable decrease in the amount of C02" which could be detected. Perhaps at
these high levels, the increase in C02 coverage on the radiation shield prohibits ionic

bombardment of bent C02 adsorbed directly onto the copper surface.

VI. Observation of Emission during Cold Diffusion of Matrix Components

The present results also suggest that matrix chemistry in which the incoming cation
is converted into another chemical species does not hinder the experiment. In these
lengthy depositions, even at very low pressures, a variety of background gases condense
into the matrix, and are candidates for reaction. The annealing data for the (C02);
species show that matrix chemistry can occur, but it does not do so in the deposition step
to any great extent. The spectra suggest that the ions are quickly incorporated into a rigid
host upon deposition. Ions accumulated over a 24 hour period undergo very little matrix
chemistry with other components, whereas 1 minute at 8-10 Kelvins provides sufficient
mobility for such reactions to be observed. The data obtained for the C02 and C82
systems suggest that larger organic ions could be successfully trapped for subsequent
analysis using the methods employed here.

We believe that, at the temperatures sampled in these cold diffusion experiments,

increased mobility is achieved without substantial losses due to sublimation. As mobility

inc
d8
wl
en

Cf

147

increases, recombination and other types of reactions occur. We have collected additional
data related to these processes. Controlled annealing experiments have been performed
where a photomultiplier detector is attached to a vacuum chamber window such that
emission from the matrix can be detected and recorded. Figure 4.8 displays the visible
emission observed during the warming of a matrix afier a 24 hour [CSf‘ —-)Ne:C02]
experiment. Unlike the earlier annealing experiments, the temperature of the matrix
during this experiment was set to oscillate over an approximate range of 6-10 K. Each
increase in matrix temperature is accompanied by an increase in emission. Results from
experiments similar to this one and those in which the dispersed emission was collected
upon warming of the matrix can be found elsewhere.24 A combination of IR data and the
visible emission spectra should provide a more complete description of . the matrix
components and their chemical evolution as a 4 K neon matrix containing ions warms and

is vaporized.

Intensity

148

 

 

 

c;
L

41101111111111

1

 

o 50 100 150 200 250 300
Time (seconds)

Figure 4 8 Visible emission detected, following a 24 hour
[CS,' —) Ne: C01] deposition, upon repetitive cycling of the temperature
between 6 and 10 K, at a rate of ~0 25 sec".

149

References

l.

10.

Sabo, M. S.; Allison, J.; Gilbert, J. R.; Leroi, G. E. Appl. Spectrosc. 1991, 45,
535.

Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard,
W. G. J. Phys. Chem. Ref. Data, 1988, 17, Suppl. 1.

Frisch, M. J.; Trucks, G. W.; Schlegel; H. B.; Gill, P. M. W.; Johnson, B. G.;
Wong, M. W.; Foresman, J. B.; Robb, M. A.; Head-Gordon, M.; Replogle, E. S.;
Gomperts, R.; Andres, J. L.; Raghavachari, K.; Binldey, J. S.; Gonzalez, C.;
Martin, R. L.; Fox, D. J.; Defrees, D. J .; Baker, J.; Stewart, J. J. P. and Pople, J.
A. Gaussian 92/DFT, revision F.3,° Gaussian Inc., Pittsburgh, PA, 1993.

(a) Chin, 8.; Person, W. B. J. Phys. Chem. 1984, 88, 553. (b) Wong,,M. W.;
Wentrup, C.; Flammang, R. J. Phys. Chem. 1995, 99, 16849. (c) Sohlberg, K.;
Chen, Y. J. Chem. Phys. 1994, 101, 3831. (d) Tseng, D. C.; Poshusta, R. D. J.
Chem. Phys. 1994, 100, 7481. (e) Hiraoka, K.; Fujimaki, S.; Aruga, K.; Yamabe,
S. J. Phys. Chem. 1994, 98, 1802. (f) Brommer, M.; Rosmus, P. Chem. Phys.
Letters 1993, 206, 540. (g) Froese, R. D. J.; Goddard, J. D. J. Chem. Phys. 1992,
96,7449.

Smith, D. F.; Overend, J. J. Chem. Phys. 1971, 54, 3632.

Bondybey, V. B; English, J. H. J. Chem. Phys. 1980, 73, 3098.

Miller, T. A.; Bondybey, V. E. In Molecular Ions: Spectroscopy, Structure and
Chemistry; Miller, T. A., Bondybey, V. E., Eds.; North-Holland Publishing
Company: Amsterdam, 1983; p 132.

Fischer, 1.; Lochschmidt, A.; Strobel, A.; Niedner-Schatteburg, G.;
Muller-Dethlefs, K.; Bondybey, V. E. Chem. Phys. Letters 1993, 202, 542.
Jacox, M. E.; Thompson, W. E. J. Chem. Phys. 1989, 91, 1410.

Jacox, M. E.; Thompson, W. E. J. Phys. Chem. 1991, 95, 2781.

ll.

12.

13.

15.

16

17

18

19
2f

11.

12.

l3.
14.
15.

16.
17.
18.
19.
20.
21.
22.

23.
24.

150

Schallmoser, G.; Wurfel, B. E.; Thoma, A.; Caspary, N.; Bondybey, V. E. Chem.
Phys. Letters. 1993, 201, 528.

Burkholder, J. B.; Lovejoy, E. R.; Hammer, P. D.; Howard, C. J. J. M01.
Spectrosc. 1987, 124, 450.

Harrison, J. F., private communication.

Kawaguchi, K.; Yamada, C.; Hirota, E. J. Chem. Phys. 1985, 82, 1174.
Brommer, M.; Chambaud, G.; Reinsch, E.-A.; Rosmus, P.; Spielfiedel, A.;
Feautrier, N.; Werner, H.-J. J. Chem. Phys. 1991, 94, 8070.

Gauyacq, D.; Horani, M. Can. J. Phys. 1978, 56, 587.

Colman, J. J.; Trogler, W. C. J. Am. Chem. Soc. 1995, 117, 11270.

Andrews, L.; Bondybey, V. E.; English, J. H. J. Chem. Phys. 1984, 81, 3452.
Cataldo, F. Inorg. Chim. Acta 1995, 232, 27.

Forney, D.; Jacox, M. E.; Thompson, W. E. J. M01. Spectrosc. 1993, 15 7, 479.
Jacox, M. E.; Chem. Phys. 1976, 12, 51.

(a) Tosh, R. E.; Shukla, A. K.; Futrell, J. H. J. Phys. Chem. 1995, 99, 15488. (b)
Shukla, A. K.; Tosh, R. B; Chen, Y. B.; Futrell, J. H. Int. J. Mass Spectrom. Ion
Processes, 1995, 146/147, 323.

Prinslow, D. A.; Armentrout, P. B. J. Chem. Phys. 1991, 94, 3563.

Godbout, J. T. Ph.D. Dissertation, Michigan State University, 1996.

Chapter 5

Investigations of Mass-Selected C7H7+ Depositions with Neon

I. Application of the MS/MI Technique to Organic Cations

The previous chapter discussed the successful detection of mass-selected CSf‘
cations in neon matrices by FTIR spectroscopy. Utilization of the MS/MI method to
investigate the structures of larger mass cations, such as those found in organic mass
spectrometry, is one of the ultimate goals of this research. While the spectroscopic study
of CS2” is interesting in its own right, the ubiquitous nature of organic ions in chemistry
makes the study of them very appealing. Several molecular ions have been chosen to
examine the utility of MS/MI spectroscopy in the structural elucidation of organic cations.
These ionic systems include CH3+, C2H3O+, C7H7+, and C6F6+2 The remainder of this
chapter will focus on the results from the mass-selected C7H7+ ion depositions.

The choice of C7H7+ for investigation by MS/MI spectroscopy has several
interesting merits. As will be explained, the isomers of the C-,H-,+ ions have played a
special role in the development of organic mass spectrometry. The successful detection of
this cation by MS/MI spectroscopy would generate considerably greater interest in this
method from mass spectrometrists than the detection of other species such as C82)".

The study of C-,H-,+ depositions may also provide additional thermochemical
comparisons to the COz/CSZ systems. The ionization energy of the cycloheptratrienyl
radical (C7H-f) has been reported1 to be 6.24:0.01 eV. This is considerably lower than
the values for C02 (13.8 eV)1 and C52 (10.1 eV).1 Results from the study of C-,H7’r
depositions can be used to verify the conclusion, from the CS2” study, that mass-selected
cations do not undergo extensive “matrix chemistry” during their deposition into the

matrix.

151

152

The study of the COZICSZ systems also included a comparison of energetics of their
individual fragmentation mechanisms. As discussed previously, the lower bond
dissociation energies (BDEs) for C82)", as compared to COf', did not result in any
observable increase in the occurrence of collision-induced dissociation (CID) of the cation
upon entrapment in the matrix. The following bond dissociation pathway for a C7H7+
cation is slightly less endothermicl than the lowest bond dissociation pathways for the

CO? (5.4 eV) and CS? (4.7 eV) cations (see Table 4.1):

C7H7+ ——’ (:sz + AH(1'X11) = 4.5 CV (5.1)

Other dissociation pathways, such as the formation of 2 acetylene molecules and C3H3+,
are more endothermicl than any of the values listed for the COz/CSZ systems in Table I of
Chapter 4. The results from depositions of C7H7+ ions may prove beneficial to the

investigations of the extent of CID processes that may occur in MS/MI experiments.

11. The Tropylium Cation and Mass Spectrometry

The formation of C7H7+ cations from alkylbenzenes has been one of the most
widely studied reactions in mass spectrometry.2 Much of the research has been involved
in trying to understand the relative abundances of the possible C7H7+ isomers which may
be formed during the ionization of alkylbenzene molecules. The three most studied forms

are illustrated below:

101

ma

the

of

106

1112

W]

pc

is

11.

153

(W

(1) tropylium cation 2)ben2yl cation (3) 3 tolyl cation

Although hydrogen scrambling and skeletal rearrangement reactions were known
to occur at the time, the identity of the cation represented by the m/z = 91 peak in the
mass spectra of alkylbenzenes was assumed to be that of the benzyl cation. As stated in
the 1957 report3 that first suggested otherwise: “formation of (the benzyl) ion by cleavage
of a B-carbon-hydrogen or B-carbon-carbon bond had seemed so straightforward that its
identity was not questioned.” This report, which has since become a landmark in organic
mass spectrometry, hypothesized that tropylium (also known as the cycloheptatrienylium
cation) was the identity of the cation. This was based on mass spectrometric studies
which showed singly deuterated C7H7+ ions exhibiting a 2:5 loss of C2HD:C2H2. This is a
possibility for the tropylium cation, but not the benzyl cation. Further labeling studies
with carbon-l3 gave additional support for the tropylium structure.4 The details of the
isomerization mechanisms involved in the formation of the tropylium cation from
alkylbenzenes such as toluene have been investigated both experimentally5 and
theoretically.6

Clearly, the m“: values and the relative intensities provided by a mass spectrum of
an alkylbenzene molecule are not sufficient to determine the structure of the C-,H7+ 1on
unambiguously. As discussed in Chapter 2, there are several mass spectrometric
techniques that have been developed to discern ionic isomers. The numerous
investigations of the C7H7+ isomeric cations since 1957 have been reviewed elsewherez‘ 7
8

MS/MS experiments have been particularly useful in investigating these structures.

However, these spectra offer only minor differences in the relative abundances of the

have
C711
muc
MS
suc

an 1

int

Fa

154

fragment ions in the m/z = 74-77 region to distinguish among the isomers. These small
differences may be difficult to reproduce.9
Had spectroscopic techniques been available to the early organic mass

”2' 7b of the formation of C7H7+ ions from toluene could

spectrometrists, the “old problem
have been solved quickly and with little ambiguity. The NMR or IR spectra of each
C7117+ isomer, for instance, could have been used as its unequivocal signature. Although
much is known about the formation of the tropylium cation, it was chosen for study by IR
MS/MI primarily because of its historical significance in organic mass spectrometry. The

successfirl detection of mass-selected, matrix-isolated tropylium cations could be used as

an example of the utility of this method for other organic cations.

III. Generation of Mass-Selected Tropylium Cations

Mass-selected C7H-,+ cations formed by electron-impact of toluene were deposited
into growing neon matrices. An example of mass spectral currents observed at the matrix
Faraday plate during such depositions is given in Appendix A.5. Based on previous

3' 8‘ it is believed that tropylium is the most abundant

studies of the ionization of toluene,
ion in the mass-selected ion beam with an m/z value of 91.

As the mass spectrum in Appendix A.5 shows, the resolution of the quadrupole
was degraded to much less than unity in these experiments to achieve the 30-40 nA C7H7+
beam currents. In the previous CF3+ and C024" experiments, the degraded resolution was
not of much concern due to the lack of abundant ions formed with m/z values close to the
selected ion. However, in the case of the mass spectrum of toluene, as shown in

Figure 5.1, there is an intense peak at m/z = 92, which corresponds to the C7H8'“

molecular cation.

155

 

 

 

 

 

 

a?
m
.53.
.E
d)
3.
r_u
2
I- 411.11111' 1.1
l'l'r‘l'l'l'l'lflf
10 20 30 40 50 60 70 80 90
m/z

Figure 5.1 The mass spectrum of toluene

The relative amounts of the two C7117+ and C7H8+' cations that are deposited cannot be
ascertained from the mass spectrum in Appendix A.5. In an effort to limit the amount of
C7H8+' which is transmitted through the mass filter, the lowest m/z value of the maximum
current was selected during all experiments.

As discussed in Chapter 2, although it would be possible to increase the resolution
of the mass filter to exclude the C7H8+' ions from being deposited into the matrix, the
transmission of all ions through the quadrupole would significantly decrease. This would
preclude detection of the ions in the matrix within reasonable deposition times. Since
some C7H8+' ions may be deposited as well, possible assignments to the matrix IR spectral
features must also include isomers of the C7H8+° ions, such as the toluene and
cycloheptatriene cations. In the remainder of this dissertation, all C7H7+ depositions will
refer to depositions in which both 0,11; and C7H8+' may be present in the mass—selected

ion beam.

IV. Previous Spectrosc0pic Investigations of C7H7+lo and C7118”,
The processes occurring during the deposition of mass-selected cations in neon
matrices are only partially understood at the present time. In addition to the trapping of

the mass-selected species, neutralization and/or fragmentation of the cations may also be

me
me
181
101

181

131

CI

51

156

expected. As mentioned in the introduction of this dissertation, there exists a number of
methods that can be used to assign cation vibrational frequencies, and the data from such
methods are extremely usefiil in interpreting the results of MS/MI experiments. However,
few spectroscopic studies have been done on the large number of possible neutral and
ionic fi'agmentation products of C-,H-,+ and C7H3+1 This is certainly expected for such
large cations.

There does exist, however, much information on vibrational frequencies of both
C-,H-,+l0 and C7H8+/0. The data from these studies are extensive, and no single report
provides complete assignments to all vibrational modes in either the gas phase or other
environments. Due to its volume, a complete listing of all vibrational information for these
species in this dissertation is not possible. Pertinent vibrational data will be given when
needed, using the Mulliken convention10 for the numbering of normal mode frequencies.
Additional vibrational frequency information of these species can be found elsewhere:
tropylium cation,ll tropylium radical,12 benzyl cation, ‘3 benzyl radical,l4 toluene cation,15
and neutral toluene.16

Of particular importance to this C7H7+ study are the IR absorption
investigationsuc‘“! of tropylium bromide, which show that the tropylium cation exhibits a

1

very strong absorption at ~1480 cm' . This absorption has been assigned to v13, a C—C

Ila at

ring stretching motion. This vibrational mode has also been detected in the gas phase
1424 cm‘1 with PE spectroscopy. On the basis of these observations, the successful
trapping and detection of the tropylium cation in a neon matrix should result in the

observation of a band in the 1400-1500 cm'1 region.

V. Results of Mass-Selected, Matrix-Isolated Tropylium Cation Studies
Mass-selected C7H7+ ion currents of ~40 nA, formed from toluene, with 130 eV
initial kinetic energies, were deposited with neon matrix gas doped with carbon dioxide

(Ne:COz 150021). To investigate the possibilities of neutral fragments difiiising from the

ion

16!

DE

11

157

ion source to the matrix, a control experiment was also performed in which all
experimental parameters were held constant, except that the mass filter was set to
W: = 100. Several absorptions due to toluene, C02, CO, and H20 were observed in these
experiments. Absorptions due to C02" were observed in all experiments involving
mass-selected ion depositions. Unless otherwise noted, all FTIR spectra of the resulting
matricies were obtained by averaging 256 scan files and were recorded with 1.0 cm‘1
resolution.

Figure 5.2a displays the 1590-1250 cm'1 region of the absorption spectrum of a
neon matrix taken afier 19 hours of continuous deposition of C7H7+ ions. The
1590-1250 cm‘1 region of the spectrum taken from a 19 hour control experiment is
displayed in Figure 5.2b. Amongst a neutral toluene absorption at 1500.8 cm‘1 (a band at

d16c to v13 of toluene isolated in solid krypton), no new absorptions

1483 cm‘1 was assigne
in this region were observed. The band at 1472.8 cm‘1 is also observed in Neztoluene
300021 depositions and can therefore be attributed to neutral toluene, however, no
vibrational mode is presently assigned to this absorption. The relatively large 28 cm’1
difference between this band and the V” band makes it unlikely to be a site splitting.

It is therefore concluded that the tropylium cation has not been detected, due to its
low abundances and/or small extinction coefficients relative to other detectable species in
the matrix. Low abundances of C7H7+ in the matrix may be the result of extensive
fragmentation in the matrix region due to the relatively high 130 eV initial ion kinetic
energies used in these studies.

Some new absorptions were observed during C7H7+ depositions, but were absent
during the control experiment. The most reproducible of these absorptions were observed
at 731.1, 2080.5, 2895.5, and 3283.7 cm". These absorptions are shown in Figures 5.3-6.
The spectra displayed in Figure 5.3 were obtained with 0.5 cm“1 resolution and are the

result of averaging 1024 scan files. The 2080.5 and 2895.5 cm’1 absorptions are presently

not assigned. The 2895.5 cm'1 band has been observed in some ”C8; and ”cs2“

158

 

 

1500.8 cm“

1472.8 cm"
01

8
i

 

WWW“ 1111/(1

1422.2 cm'l %

 

 

 

I 0.001 A. U.
C
11110111111110
1
l ' I ' j ' I 7 ”
1600 1500 1400 1300

Figure 5.2 Infrared absorption spectra in the 1590-1250 cm'| region of
neon matrices after 19 hours of deposition of (a) m/z = 91 (CHI, 40 nA)
generated from electron impact of toluene, (b) m/z = 100 (no ion current)
while subjecting toluene to electron impact, and (c) m/z = 91 (C,H,+, 30
nA) generated from electron impact of toluene, while the copper grid on
the Csl window was floated at a potential of +80 V during deposition (18

hours).

 

159

 

 

736 9 cm '
732.3 cm"
i731.1 cm"

1

730.1 cm"

 

 

I 0.002 A. U.

41.11.111.110th

 

 

1 ' I 1 1 r I f 1 ' 1
800 780 760 740 720 700

Figure 5. 3 Infrared absorption spectra in the 800- 700 cm" region of neon
matrices after 19 hours of deposition of (a) W: = 91 (C ,H, , 40 nA)
generated from electron impact of toluene, (b) m/z = 100 (no 1011 current)
while subjectin toluene to electron impact, and (c) m/z= 91 (C, H, , 30
nA) generated from electron impact of toluene, while the copper grid on
the C81 window was floated at a potential of +80 V during deposition (18
hours). These spectra were taken with 0. 5 cm 1resolution and are the result
of averaging 1024 scans.

' '(l'lOI-A

 

160

 

 

2141.7 cm'l
2080.5 cm‘I

 

 

i

[0.001 A. U.

W111 WNW/“WWW

 

 

 

 

 

 

F3 H C
.11 l
W I I I ' I ' l
2200 21 50 21 00 2050 2000

Figure 5.4 Infrared absorption spectra in the 2200-2000 cm'1 region of
neon matrices after 19 hours of deposition of (a) m/z = 91 (C,H,+, 40 nA)
generated from electron impact of toluene, (b) m/z = 100 (no ion current)
while subjecting toluene to electron impact, and (c) m/z = 91 (C,H,+, 30
nA) generated from electron impact of toluene, while the copper grid on
the CsI window was floated at a potential of +80 V during deposition (18

hours).

 

161

 

2895.5 cm'I

[.0 0004 A. U.

 

 

 

 

l l l l l I I
3000 2950 2900 2850 2800 2750 2700

Figure 5. 5 Infrared absorption spectra in the 3000- 2700 cm'1 region of
neon matrices after 19 hours of deposition of (a) m/z— - 91 (C,H , , 40 nA)
generated from electron impact of toluene, (b) m/z = 100 (no 1on current)
while subjecting toluene to electron impact, and (c) m/z = 91 (C ,H, , 30
nA) generated from electron impact of toluene, while the copper grid on
the CsI window was floated at a potential of +80 V during deposition (18
hours).

162

 

 

 

 

'8 .5
If _ 0
a 'E .‘2
a ,3 2
'8 9' ..
V. 2 a
8.
'E
U
V‘.
1 0.0004 A. u. g C

 

 

1 ' r ' F ' 1 T 1 ' 1
3350 3300 3250 3200 3150 3100

Figure 5. 6 Infrared absorption spectra in the 3350-3110 cm" region of
neon matrices afler 19 hours of deposition of (a) m/z- - 91 (C,H, , 40 nA)
generated from electron impact of toluene, (b) m/z = 100 (no 1011 current)
while subjecting toluene to electron impact, and (c) m/z = 91 (C ,H, , 30
nA) generated from electron impact of toluene, while the copper grid on
the CsI window was floated at a potential of +80 V during deposition (18

hours).

163

 

WWW

E
g

 

 

 

_ -. -. 1855.8 -'
g 5 5 F1533 :iii'
2 8' 8'
0‘ cc 00
0..001A u -°-‘i T "
' 1 3 1 ' 1 ' 1
2100 2000 1900 1800 1700

Figure 5. 7 Infrared absorption spectra in the 2070-1720 cm' region of
neon matrices after 19 hours of deposition of (a) m/z — 91 (C ,H, , 40 nA)
generated from electron impact of toluene, (b) m/z = 100 (no 1011 current)
while subjecting toluene to electron impact, and (c) m/z = 91 (C, H, , 30
nA) generated from electron impact of toluene, while the copper grid on
the CsI window was floated at a potential of +80 V during deposition (18

hours).

 

 

2.111

164

depositions and is therefore not due to the tropylium cation or its fragmentation products.

d11c to the tropylium cation, which is

A weak absorption at 2060 cm'1 has been assigne
similar to frequency of the 2080.5 cm‘1 band observed in this study. However, the
absence of a band near 1480 cm‘1 suggests that this band is also not due to the tropylium
cation. The 2080.5 cm'1 band may be due to a linear carbon chain as several linear carbon
chain species show strong absorptions in the 2000-2200 cm’1 region. Strong absorptions
have been reported at 2010.0, 2036.4, 2134.6, and 2166.4 cm'1 in neon for (v6) C9,
(v3) C3, (v4) C7, and (v3) C5, respectively.” A strong band at 2060.6 cm'1 in argon has
been assigned to a CEC stretching mode (v3) of C4H.l8 The formation of a linear carbon
chain species in this experiment would presumably be the product of fragmentation of the
mass-selected C-,H—,+ cations or a fragmentation product of neutral toluene which has
diffused into the matrix region.

The 731.1 cm'1 and 3283.7 cm'1 bands are assigned to v5 and v3 of acetylene,
respectively. Infrared spectra obtained from a NeIC2H2 1000:] matrix deposition confirm
this assignment. Table 5.1 provides a summary of the pertinent vibrational frequencies
and peak absorbance values obtained in this experiment, along with those observed in the
C7H7+ deposition for comparison. Both the frequencies and the relative peak absorbance
values observed in the Ne2C2H2 study agree well with those observed during the C7H7+

deposition. Presumably, the v,, + v5 vibrational mode was not observed during the C7H7+

deposition due to its low intensity.

VI. The Origin of Acetylene in Tropylium Cation Depositions

There are several possible sources for the formation of C2H2 during C7H-7+
depositions. Diffusion of (3sz from the ion source is one possibility. The most intense
peaks in the mass spectrum of toluene, besides the molecular cation, are at m/z values of
39, 65, and 91. The difference of 26 between these m/z values can be interpreted as due

to the loss of 1 or 2 C2H2 fragments from the C-,H-,+ ions. However, the absence of the

165

Table 5.1 The observed vibrational frequencies and peak absorbance intensities of
matrix-isolated Csz. The middle column lists results from a direct deposition of
Ne:C2H2 100021. The column on the right lists the frequencies observed in a C7H7+
deposition which are assigned to C2112. (A. U. = absorbance units)

 

 

 

 

 

*1
vibrational modes NezC2H2 matrix C7H-,+ deposition ‘
v5 732.1 cm'1 731.1 cm“
0.31063 A. U. 0.00553 A. U. _ 1
v3 3284.2 cm'1 3283.7 cm"
0.04613 A. U. 0.00107 A. U.
(v, A. U.)/(v5 A. U.) 0.15 0.19
v,, + v5 1330.1 cm‘1 not observed
0.02137 A. U.

 

 

 

166

absorptions due to CZHZ in the control spectrum (Figure 5.3b and 5.6b) indicate that
difl‘usion of this neutral fragment from the ion source is not a dominant source of the
matrix-isolated C2H2.

As illustrated in equation 5.1, (:sz is an expected product of dissociation of the
mass-selected C7H7+ ions. Investigations19 of collisions of C7H7+, formed from toluene,
at solid surfaces show that the peak at m/z = 65 (C5H5+) in the fragment ion spectrum of
C7H-,+ is the most intense. The ion kinetic energies used in these studies were in the range
of 25-44 eV. The initial ion kinetic energy used in the matrix depositions was 130 eV.
However, as discussed in Chapter 3, the matrix may hold a positive potential, which
results in lower ion kinetic energies as they approach the matrix. The actual ion kinetic
energy as the ions enter the matrix region may very well be similar to the surface-induced
dissociation (SID) experiments. The lack of vibrational frequencies and intensity
information for C7H7+, Csz» and CSH;r makes it difficult to draw fiirther conclusions on
the extent of CID occurring during ion depositions.

It is also possible that neutral toluene molecules, adsorbed onto the cold surfaces
surrounding the matrix, yield CZHZ fragments upon bombardment by the mass-selected
ions. The evidence presented in the previous chapter suggest that ionic bombardment of
CS2 adsorbed on the cold surfaces may be the dominant source of CS molecules detected
in the matrix. A similar process may very well be occurring here. If this is the case,
extensive fragmentation may not play a large role in the C7H-7+ depositions, and the
absence of detectable 0,117+ ions in the matrix may be due to low extinction coefficients
relative to other detectable cations. Further experiments, such as doping the neon with
toluene while depositing CFf, are needed to ascertain which source of C2H2 in the matrix

dominates during ion depositions.

167

VI]. Reduction of the Incident Mass-Selected Ion Kinetic Energy

To further investigate the extent of CID processes which may occur, mass-selected
C—,H7+ depositions were also performed while floating a copper grid on the matrix
substrate (see Figure 3.5) at a positive potential to slow the 130 eV ion beam as it
approaches the matrix. It was hoped that the potential would reduce fragmentation of the
ions, and increase the likelihood of detection of the matrix-isolated C7H7+ ions. This
method of decreasing the kinetic energy of the ions was chosen because it was the most
convenient. As illustrated in Figure 2.18, the current of the mass-selected ions is
negligible below 80 eV. Modifications to the ion source or the construction of a
deceleration lens would be required to reduce the ion energy before the matrix.

A potential of +80 V was chosen on the basis of SIMION studies which show that
this potential on the matrix substrate reduces the mass-selected ion kinetic energy but does
not drastically reduce the level of ion current by deflecting the ion beam from the matrix
region. With this potential, it is expected that the 130 eV ions will be deposited into the
matrix with no more than 50 eV energy.

Several regions of the spectrum, collected afier an 18 hour C7H7+ deposition with
the copper grid floating at +80 V, are reproduced in Figures 5.2c-7c. This experiment will
be referred to as the +80 V-grid experiment. The spectra from the previous deposition,
where the grid was at ground potential, and for a control experiment in the same regions
are also given in Figures 5.2a, b-5.7a, b. The experiment in which the grid on the matrix
substrate was at ground potential will be referred to as the 0 V-grid experiment. As can
be seen fiom Figure 5.2c, no absorption is detected in the 1590-1250 cm’1 region which
can be attributed to the C7H7+ cation. The absorption at 1422.2 cm"1 is assignable to v3
of C02“ The observation of this cation has been discussed in Chapter 3 as being due to
ionization of neutral C02 molecules in the matrix by electrons that are accelerated towards
the matrix by the potential applied to the grid. The electrons are formed by cation

bombardment of surfaces and/or from the dissociation of metastable C02“ anions during

 

 

 

168

their flight to the matrix region. Although the toluene partial pressure in the ion source
was kept at a constant 30 mTorr in all of the experiments which produced in the three
spectra shown in Figure 5.2, the VB toluene absorbance at 1500.8 cm'1 is not as intense in
the +80 V-grid experiment as it is in the 0 V-grid and control experiments. The only
source of matrix isolated toluene observed in the m/z = 100 control experiment
(Figure 5.2b) is diffusion of the neutral precursor from the ion source. It is not clear why
the diffusion of toluene from the ion source has varied to such an extent in these
experiments.

Figures 5.3 and 5.6 show that the v3 and v5 absorptions due to C2H2 have been
reduced in intensity in the grid experiment. This may be due to a reduction in the
fi'agmentation of the mass-selected C7H7+ ions due to the applied potential, a reduction in
the amount of adsorbed neutral toluene molecules on the cold surfaces surrounding the
matrix, or a result of the lower mass-selected ion current used in the +80 V-grid
experiment (30 nA) relative to the 0 V-grid experiment (40 nA). ' The inability to control
all experimental parameters makes quantitative assessments of processes occurring during
ion deposition difficult.

The absorption observed in the 0 V-grid experiment at 2080.5 cm'1 is not present
in the +80 V-grid experiment. Since no assignment is given to this band, speculations on
the reason for its disappearance are difficult to make. Perhaps it is due to a positive ion,
such as C5H5+, which is repelled from the matrix region when a positive potential is
applied to the grid. The absorption at 2147.0 cm'1 also remains unassigned. As discussed
in Chapter 4, it may be due to a complex of CO with some as yet identified species since it
is slightly blue shifted from the CO absorption at 2141.7 cm']. The absorption observed at
2895.5 cm'1 is present in both the 0 V-grid and the +80 V-grid experiments, as shown in
Figure 5.5. No assignment can be given to this band at the present time.

Five new bands (1897.2 cm'], and two doublets at 1864.9, 1862.5, and 1855.8,

1853.4 cm“) appear in the 2070-1720 cm'1 region, as displayed in Figure 5.7. These

‘v '1.“ mama—p,-

 

169

bands are assigned to the (C02 . . Oz") complex, which may be represented as C04". The
appearance of these bands has been discussed in Chapter 4. It is interesting to note that
these bands are only observed in the C7H7+ experiment when the matrix grid is floated at
+80 V, while these same bands are observed in the CSf‘ investigations in the absence of
the grid. This may suggest that the average potential of the matrix during CSf’
depositions is more positive than the average potential during C7H7+ depositions.
Negative species, such as CO4", would be expected to be attracted and trapped in the
matrix in greater abundances under such conditions. It would also follow that the C7H7+
cations would be more likely to undergo fragmentation at these low matrix potentials.
The relatively high matrix potentials during CSf‘ depositions reduce the incident ion
kinetic energy, reducing CID fragmentation of these ions, allowing for intact CS; ions to

be detected.

VIII. Vibrational Intensity Information of the Matrix Components

The study of vibrational intensities from ab initio calculations for the C82 system
suggested that some matrix components may be present in abundance but are not detected
because of low absorptivity coefficients. It was concluded that CS+' was not detected for
this reason. If the origin of Csz in the tropylium study is from fragmentation of the
mass-selected C7H7+ ions, would the vibrational intensities of any of the C2H2 normal
modes allow for its detection? The calculated vibrational intensities of v3 and v5 of C2H2,
performed at the same level (HF/6-311+G*) which was used to obtain the values listed in
Table 4.3, are given in Table 5.2. Several species from the CSZ/COZ systems are listed for
comparison. Available experimental values are also given. The reported experimental
values were converted to the km/mole intensity unit with use of the conversion factors
provided in reference 20.

The v5 mode is the most intense of all infrared active modes in C2H2. The

calculated result for this vibrational intensity, 130 kin/mole, agrees well with the

 

 

170

Table 5.2 A comparison of the calculated vibrational intensities of several infrared-active
modes of species pertinent to MS/MI investigations. Available experimental values are
also listed. The reference which provided each experimental value is given as a
superscript next to the value.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Species Vibrational HF/6-31 1+G“ Experimental
Mode (km/mole) (km/mole)
csf v3 361 — 1i
co,+- v3 231 _ .
CS+’ v 2 —
CS v . 155 —
C,H, v3 79 7221
v5 130 18021
v4 +v5 —— 2121
C0, v3 1160 65222

 

 

 

 

 

 

 

 

(:8, v, 1567 55023

 

171

experimental value of 180 km/mole. If each C-,H-7+ cation fragments to form one Csz
molecule as it impinges on the growing matrix, would it be possible to detect this
fragment? To investigate this possibility, consider that the calculated intensity of v3 for
C52“ is 361 km/mole. A 21 hour CS2+' deposition yielded a v3 intensity for this cation of
0.00304 A U. A 19 hour C-,H7+ deposition resulted in an observed v5 intensity of
0.00553 A U. for C2H2~ The relatively small 130 km/mole calculated value for v5 of
C2H2 would suggest that the fragmentation process:

C,H,* —) C,H, + C511; (5.1)
could not yield the abundance required of the relatively large observed peak absorbance.
Two possibilities could provide an additional source of Csz- The fragmentation process:

C,H,“ -9 2 C,H2 + C311,+ (5.2)
is possible, considering the previously noted SID studies. However, this process does not
occur to the same extent as (5.1). The second possibility, neutral fragments formed by
ionic bombardment of adsorbed toluene on the cold surfaces surrounding the matrix,
seems more plausible. It has been previously suggested that the CS observed during CSf‘
depositions originates from adsorbed CS2 on the radiation shield. The calculated intensity
for this fragment is 155 km/mole, a very similar value to that for v5 of (5H2. On the basis
of the data provided in Table 5.2, it is suggested that the dominant source of the fi'agment
C2H2 in C7H7+ depositions is through ionic bombardment of neutral precursor molecules
adsorbed on surfaces in the matrix region. Additional data are needed from other possible

systems before this conclusion can be confirmed.

IX. Conclusions from the Mass-Selected C7117+ Deposition Studies

Depositions of mass-selected C7H7+ cations into growing neon matrices did not
result in the observation of infrared absorption(s) assignable to the tropylium cation.
Previous assignments”16 to the C7H7+m and C7H8‘"0 species do not match any of the

bands observed in the ion depositions and are absent in the control experiment. However,

 

172

two bands were attributed to CZHZ, a possible fragment of the mass-selected C7H7+ ion or
the neutral toluene adsorbed on the cold surfaces near the matrix.

The experiments performed with a positive potential on the metal grid placed on
the matrix substrate to reduce the mass-selected ion kinetic energy were also unsuccessful
in the detection of matrix-isolated C-,H-,+ ions. It is difficult to predict the effect this
reduction in ion energy will have on the percent of the ions which fragment. SID studies24
of several different ions have shown that 11-15% of the laboratory collision energy is
converted into internal energy. 7.5 eV of internal energy is still sufficient to induce bond
dissociation in the C7117+ ion. It remains to be determined whether extensive
fragmentation is the dominant reason for not detecting the tropylium cation or if it is
related more to a low extinction coefficient of this ion relative to the cations that have
been detected in the same deposition time period. It is concluded that successful detection
of tropylium cations will require either much larger mass-selected ion currents or the
lowering of the initial ion kinetic energy before the ion beam enters the matrix region.
Both of these adjustments require considerable modifications to the present apparatus.

Neutral cycloheptatriene has the lowest ionization energy of the corresponding
neutral counterparts to all systems studied to date. However, products due to “matrix
chemistry” of the mass-selected cations were not observed in these experiments. This
fiirther supports the conclusion from the CSZJ" study. The results also suggest that neutral
precursors, diffusing from the ion source, may adsorb on the cold surfaces near the matrix,
and yield fragments which then find their way to the matrix. Although the occurrence of
this process would presumably not interfere with the deposition of the mass-selected
cations, it could provide ambiguity to spectral assignments. Because of the possibility for
this process, it is difficult to ascertain the extent of fragmentation of the mass-selected

cations.

 

173

References

1.

10.
ll.

Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard,
W. G. J. Phys. Chem. Ref Data, 1988, 17, Suppl. 1.

Lifshitz, C. Acc. Chem. Res. 1994, 27, 138.

Rylander, P. N.; Meyerson, S.; Grubb, H. M. J. Am. Chem. Soc. 1957, 79, 842.
Bursey, M. M.; Hoffman, M. K. In Mass Spectrometry: Techniques and
Applications; Milne, G. W. A. Ed; Wiley: New York, 1971; p 379. -
(a) Bartmess, J. E. J. Am. Chem. Soc. 1982, 104, 335. (b) Kelsall, B. J.; F
Andrews, L. J. Am. Chem. Soc. 1983, 105, 1413. f
(a) Cone, C.; Dewar, M. J. S.; Landman, D. J. Am. Chem. Soc. 1977, 99, 372.

 

(b) Dewar, M. J. S.; Landman, D. J. Am. Chem. Soc. 1977, 99, 2446. E!
(a) Kuck, D. Mass Spectrom. Rev. 1990, 9, 187. (b) Grotemeyer, J.;
Gn'itzmacher, H.-F. In Current Topics in Mass Spectrometry and Chemical
Kinetics; Beynon, J. H., McGlashan, M. L., Eds; Heyden: London, 1981; p 29.
(a) Olesik, S.; Baer, T.; Morrow, J. C.; Ridal, J.; Buschek, J.; Holmes, J. L. Org.
Mass Spectrom. 1989, 24, 1008. (b) McLafferty, F. W.; Bockhoff, F. M Org.
Mass Spectrom. 1979, 14, 181. (c) McLafferty, F. W.; Winkler, J. J. Am. Chem.
Soc. 1974, 96, 5182.

Heath, T. G.; Allison, J .; Watson, J. T. J. Amer. Soc. Mass Spectrom. 1991, 2,
270.

Mulliken, R. S. J. Chem. Phys. 1955, 23, 1997.

(a) Koenig, T.; Chang, J. C. J. Am. Chem. Soc. 1978, 100, 2240. (b) Fritz, H. P.
In Advances in Organometallic Chemistry; Stone, F. G. A., West, R. Eds;
Academic Press: New York, 1964; Vol. 1, p 239. (c) Fateley, W. G.; Cumutte,
B.; Lippincott, E. R. J. Chem. Phys. 1957, 26, 1471. ((1) Nelson, R. D.; Fateley,
W. G.; Lippincott, E. R J. Am. Chem. Soc. 1956, 78, 4870. (e) Fateley, W. G.;

 

13

11

12.
13.

14.

15.

16.

174

Lippincott, E. R. J. Am. Chem. Soc. 1955, 77, 249. (f) Doering, W. von E.;
Knox, L. H. J. Am. Chem. Soc. 1954, 76, 3203.

Johnson, R. D. 111. J. Chem. Phys. 1991, 95, 7108.

(a) Weisshaar, J. C. SPIE: Optical Methods for T ime- and State-Resolved
Chemistry 1992, 1638, 453. (b) Eiden, G. C.; Weinhold, F.; Weisshaar, J. C. J.
Chem. Phys. 1991, 95, 8665. (c) Houle, F. A.; Beauchamp, J. L. J. Am. Chem.
Soc. 1978, 100, 3290.

(a) Langkilde, F. W.; Bajdor, K.; Wilbrandt, R.; Negri, F.; Zerbetto, F .; Orlandi,
G. J. Chem. Phys. 1994, 100, 3503. (b) Langkilde, F. W.; Bajdor, K.; Wilbrandt,
R. Chem. Phys. Letters 1992, 193, 169. (c) Gunion, R. F.; Gilles, M. K.; Polak,
M. L.; Lineberger, W. C. Int. J. Mass Spectrom. Ion Proc. 1992, 117, 601. (d)
Fukushima, M.; Obi, K. J. Chem. Phys. 1990, 93, 8488. (e) Negri, F.; Orlandi,
G.; Zerbetto, F .; Zgierski, M. Z. J. Chem. Phys. 1990, 93, 600. (f) Selco, J. J.;
Carrick, P. G. J. Mol. Spectrosc. 1989, 13 7, 13. (g) Orlandi, G.; Poggi, G.;

Zerbetto, F. Chem. Phys. Letters 1985, 115, 253. (h) Miller, J. H.; Andrews, L. J.

Mol. Spectrosc. 1981, 90, 20. (i) Cossart-Magos, C.; Leach, S. J. Chem. Phys.
1976, 64, 4006. (j) Lloyd, R. V.; Wood, D. E J. Chem. Phys. 1974, 60, 2684.
(k) Cossart-Magos, C.; Leach, S. J. Chem. Phys. 1972, 56, 1534. (l)
Watmann-Grajcar, L. J. Chim. Phys. 1969, 66, 1023. (m) Ward, B. Spectrochim.
Acta 1968, 24A, 813. (n) Grajcar, L.; Leach, S. J. Chim. Phys. 1964, 61, 1523.
(a) Lu, K.-T.; Eiden, G. C.; Weisshaar, J. C. J. Phys. Chem. 1992, 96, 9742. (b)
Takahashi, M.; Okuyama, K.; Kimura, K. J. Mol. Struct. 1991, 249, 47. (c)

Meek, J. T.;Long, s. 11.; Reilly, J. P.J. Phys. Chem. 1982,86, 2809. (d) Debies,

T. P.; Rabalais, J. W. J. Electron Spectrosc. Relat. Phenom. 1972/73, 1, 355.
(a) Hameka, H.; Jenson, J. O. J. Mol. Struct. flheochem.) 1995, 331, 203. (b)
Lau, C. L.; Snyder, R. G. Spectrochim. Acta 1971, 27A, 2073. (c) Katz, B.;
Brith, M.; Sharf, B.; Jortner, J. J. Chem. Phys. 1971, 54, 3924. (d) Fuson, N.;

I‘ ‘ ‘ij

 

 

Gar
K3
800
11. 81111
19‘)
1%. She

19. ta)

20. p

21.

22.

23.
24.

17.

18.
19.

20.

21.

22.

23.
24.

175

Garrigou-Lagrange, C.; Josien, M. L. Spectrochim. Acta 1960, 16, 106. (e)
Kahane-Paillous, J.; Leach, S. J. Chim. Phys. 1958, 55, 439. (f) Pitzer, K. S.;
Scott, D. W. J. Am. Chem. Soc. 1943, 65, 803.

Smith, A. M.; Agreiter, J.; Hartle, M.; Engel, C.; Bondybey, V. E. Chem. Phys.
1994, 189, 315.

Shen, L. N.; Doyle, T. J .; Graham, W. R. M. J. Chem. Phys. 1990, 93, 1597.

(a) Morris, M. R.; Riederer, D. E. Jr.; Winger, B. E.; Cooks, R. G.; Ast, T.;
Chidsey, C. E. D. Int. J. Mass Spectrom. Ion Proc. 1992, 122, 181. (b) Ast, T.;
Mabud, Md. A.; Cooks, R. G. Int. J. Mass Spectrom. Ion Proc. 1988, 82, 131. (c)
Mabud, Md. A.; Dekrey, M. J .; Cooks, R. G.; Ast, T. Int. J. Mass Spectrom. Ion

 

Proc. 1986, 69, 277.

Pugh, L. A.; Rao, K. N. In Molecular Spectroscopy: Modem Research; Rao, K.
N. Ed; Academic: New York, 1976; Vol. 2, Chapter 4.

Varanasi, P.; Bangaru, B. R. P. J. Quant. Spectrosc. Radiat. T ransfer 1974, 14,
839.

Burch, D. E.; Gryvnak, D. A.; Patty, R. R.; Bartky, C. E. J. Opt. Soc. Amer. 1969,
59, 267.

Person, W. 8.; Hall, L. C. Spectrochim. Acta 1964, 20, 771.

(a) Ast, T.; Riederer, D. E. Jr.; Miller, S. A.; Morris, M.; Cooks, R. G. Org. Mass
Spectrom. 1993, 28, 1021. (b) Despeyroux, D.; Wright, A. D.; Jennings, K. K;
Evans, 8.; Riddoch, A. Int. J. Mass Spectrom. Ion Proc. 1992, 122, 133. (c)
Wysocki, V. H.; Ding, J .-M.; Jones, J. L.; Callahan, J. H.; King, F. L. J. Am. Soc.
Mass Spectrom. 1992, 3, 27. (d) Wysocki, V. H.; Kenttamaa, H. 1.; Cooks, R. G.
Int. J. Mass Spectrom. [on Proc. 1987, 75, 181. (e) DeKrey, M. J.; Kenttamaa,

H. 1.; Wysocki, V. H.; Cooks, R. G. Org. Mass Spectrom. 1986, 21, 193.

 

Chapter 6

Conclusions and Future Directions

I. Conclusions

The FTIR MS/MI cation spectroscopy experiments described here have been
successful for three cations to date: CF3+, COf', and C82“. Mass-selected depositions of
CH3+, C2H3O+, C7H7+, and C6F6+' have not resulted in detectable spectral features
assignable to the cation under investigation. To extend this method to these and other
interesting species, it must be first resolved which process(s) are involved in determining
the success or failure of an experiment. There exist three possible reasons for the lack of
mass-selected cation detection in the unsuccessful experiments: these ions react more
readily with other components of the matrix as they are entrapped into the growing
matrix; these ions undergo extensive fragmentation as they impinge on the growing
matrix; and/or these species have inherently low absorptivity coefficients.

It could be rationalized that cations whose neutral counterparts have relatively high
ionization energies would tend to react more readily with other matrix components, such
as C02, H20, N2, and 02, which are present in high abundances. Comparison of the C02
and C82 systems show that this probably does not occur to any great extent. While the
ionization energy of C02 is more than 3 eV higher than C82, only intact cations are
observed in the matrix. Spectral features observed in these experiments have yet to be
assigned to possible reaction products of these cations, such as a (COf' + H20) complex.
Furthermore, if reactions are responsible for the decrease in abundance of the
mass-selected cations in the matrix, one might expect C7H7+ to be readily detected, due to

the low ionization energy of its neutral counterpart. Depositions of C7H7+ have not

176

 

 

177

resulted in spectral features assignable to this cation to date. These results show that
“matrix chemistry” does not occur to any great extent.

The second possibility of fragmentation processes seems very plausible,
considering the 130 eV initial ion kinetic energies which are utilized. There are two
difficulties in discerning the extent to which this process occurs. As discussed previously,
while the initial ion kinetic energy is ofien 130 eV, the matrix may hold a net positive
charge, reducing an ion’s kinetic energy as it approaches the matrix. Experimental
measurement of the average potential on the matrix during or afier mass-selected ion
depositions is not readily carried out. The actual ion kinetic energy as it reaches the
matrix remains unknown. The second difficulty in determining the extent of fragmentation
is the lack of spectroscopic detection of fragmentation products of the mass-selected
cation. The possible infrared-observable fragments of CSf‘, which are CS“ and CS, have
not been observed or have been observed but cannot be attributed as originating directly
fiom the cation. Calculations show that CS“ has a relatively low absorptivity coefficient;
it therefore may be present in abundance in the CS2+' study but is not detectable. CS was
detected, but it was concluded that it originated from dissociation of neutral precursors,
adsorbed onto the radiation shield, by ionic bombardment. The absence of spectroscopic
information about the possible fragmentation products of C7H7+, such as C5H5+, has also
limited any usefiil insights to the extent of fragmentation.

The third possibility for the lack of mass-selected cation detection is that these ions
have low absorptivity coefficients as compared to those which have been detected. Ifthis
is the case, the mass-selected ion current must be greatly increased to allow observation of
these species. Ab initio calculations should provide usefiJl information about the
magnitude of the extinction coefiicients of the mass-selected ions under study. The
absorptivity information on CSP provided by these calculations has clearly shown why this
cation should not be expected to be detected after typical depositions into a neon matrix.

Not only can the results from such calculations provide insight on the vibrational

‘.
'1: MI.

 

 

178

frequencies and absorptivities of the mass-selected cation, they can also provide the same
information on the possible fragmentation products which may also be detected in the

matrix.

I]. Possible Modifications to the Experimental Apparatus

As discussed above, unsuccessful detection of MS/MI cations may be due mostly
to fragmentation of the mass-selected ions, or too weak ion vibrational intensities.
Extensive modifications to the present apparatus are needed to provide smaller ion kinetic
energies or to substantially increase the mass-selected ion current. Efforts to date to slow
the incoming cation beam by placing a positive potential on a grid positioned on the matrix
substrate have not proven to be effective in increasing the likelihood of spectroscopic
detection of mass-selected ions. Reduction of the ion kinetic energy before the matrix
region remains the only viable option. In this regard, utilization of the magnetic
sector/deceleration lens instrument constructed in this laboratory,l may prove to provide
mass-selected ion beams with lower kinetic energies. However, its implementation to the
present matrix vacuum chamber would require several modifications.

The detection of ions with low absorptivity coefficients would require
modifications to the ion source or lower detection limits of the FTIR spectrometer. Long
periods of ion depositions have not been a useful option owing to filament burnout and/or
quadrupole performance degradation due to contamination of the electrodes by ion
bombardment. It should prove possible to replace the original Finnigan 3200 ion source
used in these experiments with one capable of generating much larger amounts of ions.
Not only would this increase the mass-selected ion current, it may also allow a reduction
in the initial ion kinetic energy and/or the ability to increase the resolution of the
quadrupole mass filter.

The results from this dissertation have also shown that it is possible to sputter

neutral species off the radiation shield and subsequently detect their presence in the matrix.

 

\‘ {+1 ylw'h‘ r. ‘1 -"1..-.
h!-

 

 

 

180
de

re;

179

The presence of such species may hinder the investigation of future ionic systems. It
would also be advantageous to include an electrostatic quadrupole2 in the ion source
design to decrease the amount of neutrals diffusing from the ion source to the matrix
region. Maier and coworkers3 have successfully implemented this device in their MS/MI
instrument and it should prove 1.15le in the one constructed in this laboratory.

The results from the studies discussed here verify that counterions are generated
during cation MS/MI investigations, and that this allows the mass-selected cations to be
continuously deposited. However, the utilization of two mass spectrometers to select
both cations and anions, as discussed in the original proposal of this project, may prove to
be the most important “modification” to allow MS/MI spectroscopic investigations of a
wider variety of species. The results obtained from the cation MS/MI experiments show
that the mass-selected cations are not neutralized by negatively-charged species as they
enter the matrix region, and these oppositely-charged species do not undergo extensive
“matrix chemistry” as they impinge upon the matrix. They suggest that cations and anions

can be deposited at the same time and be separately isolated in the growing neon matrix.

III. Future Studies

If fragmentation of the mass-selected ions is the dominant reason for the failure to
observe some cations spectroscopically, in addition to physical modifications to the
present instrument, it may prove possible to “chemically” reduce the mass-selected ion
kinetic energy by selecting an optimal matrix additive to create counterions. As mentioned
previously, the potential of the matrix may be more positive in the successful experiments.
This is likely due to the interaction of the mass-selected cations and the adsorbed C02
matrix additive on the radiation shield. Perhaps other matrix additives, such as 02 or
H20, can be used to generate counterions as well, and at the same time provide the
optimal matrix potentials for reduction in mass-selected ion fragmentation. Likewise, by

allowing an additional mass-selected cation to impinge on the adsorbed C02, the matrix

 

 

180

potential might be affected. Consider the possibility that COZ+° produces high matrix
potentials, and the failure to observe spectroscopic signals assigned to C2H3O+ during its
deposition is due to low matrix potentials. This may be due to a specific reaction of the
CD; with the metallic radiation shield and/or the adsorbates on it, which as a result limits
the rate at which C02" anions can be sputtered off. In this scenario, the simultaneous
introduction of both C02 and acetone into the ion source, along with the necessary
degradation of the resolution of the mass filter, might allow for the generation of an ion
beam of both CO; (m/z = 44) and C2H3O+ (m/z = 43) and an optimal potential of the
matrix might be attained, allowing for the spectroscopic observation of intact C2H3O+
cations.

Once the MSM spectroscopic method for investigating molecular ions becomes
routine, it can be used to assist in the study of several interesting areas of molecular ion
science. As discussed previously, this method can be used to aid mass spectrometrists in
the differentiation of ionic isomers, such as C2H3O+, C2H3Oz+, and C7H7+. A discussion
of the utilization of this method in the spectrosc0pic study of metal-containing ions has
also been given.4 The study of distonic ions might also be undertaken. Distonic ions are

5 It is commonly

species which contain spatially separated charge and radical sites.
proposed in the literature that molecular ions isomerize to distonic ions prior to
dissociation. However, the structures of the resulting isomerized ions have only rarely
been verifieds The following section will discuss additional areas in which the MS/MI

method can contribute.

A. Interstellar Molecular Ions

Although the interstellar medium was once supposed to exist as a vacuum, several
neutral and ionic molecules have since been detected in these regions.‘ There exist both
diffuse and dense interstellar clouds in which many of these species have been discovered.

Diffuse interstellar clouds are very low in density and consist mainly of H atoms, H2

 

 

181

molecules, and micron-sized “dust” particles. It is in these regions that the first interstellar
molecules (CH', CW, and CN‘) were detected7 in the late 1930s. The intense stellar
visible and ultraviolet radiation ofien limits the existing species in these regions to simple
diatomic species. The dense interstellar clouds, on the other hand, contain greater
concentrations of “dust” grains. The interior regions of the dense clouds are protected
from the stellar radiation, allowing the formation of complex polyatomic species.8

The assignment of the discrete absorption features originating from these clouds
has been termed “the longest standing unsolved problem in all of spectrosc0py.”9 In
particular, it is the diffuse interstellar bands (DIBs) which have captivated the astronomy
community since 1921 when the first few bands were detected.10 Following these initial
observations, well over a hundred bands have been detected.11 The “unsolved problem”
refers to the lack of assignment of these bands. They likely originate from ionic species,
formed in the harsh environment in the diffuse interstellar clouds. However, as discussed
in this dissertation, little laboratory spectroscopic information about molecular ions exists.

As Maier and coworkers have already shown, MS/MI spectroscopy is well suited to

investigate ionic species which may be responsible for these interstellar bands.12

B. Matrix Chemistry

From a spectroscopic viewpoint, the absence of “matrix chemistry” during the
deposition of mass-selected cations into a growing neon matrix is the ideal situation.
However, one of the long-term goals of this project is to spectroscopically study the
products of ion/molecule reactions.‘3 The promising results of the C02" + C02 cold
diffusion experiments show that this technique may be used to obtain spectroscopic
information of other ion/molecule reactions. Several investigations of near thermal,
gas-phase reactions between CF3+ and neutral molecules such as unsaturated aliphatic
hydrocarbons,“ fluoroalkenes,15 and benzene derivatives“ have already been undertaken

by mass spectrometric methods. By d0ping the matrix with both C02 and one of these

 

 

182

neutral species, it may be possible to obtain spectral information about reaction products
formed during the controlled annealing process. A particularly interesting study" has
shown that one of the products of the reaction between CF3+ and toluene is C7H7+.
Perhaps it is possible to spectroscopically detect matrix-isolated tropylium cations by IR
spectroscopy through use of this particular reaction. In this work, we’ll ultimately shoot

for the stars, and perhaps settle for the dust between them.

 

 

183

References

l.

10.
11.
12.
13.
14.
15.

O’Connor, P. J .; Leroi, G. E.; Allison, J. J. Am. Soc. Mass Spectrom. 1991, 2,
322.

(a) Zeman, H. D.; Rev. Sci. Instrum. 1977, 48, 1079. (b) Huber, B. A.; Miller, T.
M.; Cosby, P. C.; Zeman, H. D.; Leon, R. L.; Moseley, J. T.; Peterson, J. R. Rev.
Sci. Instrum. 1977, 48, 1306. (c) Farley, J. W. Rev. Sci. Instrum. 1985, 56, 1834.
(d) Bower, J. E.; Jarrold, M. F. J. Chem. Phys. 1992, 97, 8312. (e) Laaksonen,
R. T.; Goetsch, D. A.; Owens, D. W.; Poirier, D. M.; Stepniak, R; Weaver, J. H.
Rev. Sci. Instrum. 1994, 65, 2267.

(a) Freivogel, P.; Fulara, J .; Lessen, D.; Forney, D.; Maier, J. P. Chem. Phys.
1994, 189, 335. (b) Maier, J. P. Mass Spectrom. Rev. 1992, 11, 119.

Sabo, M. S. Ph.D. Dissertation, Michigan State University, 1991.

Kenttamaa, H. l. Org. hiass Spectrom. 1994, 29, 1.

Smith, D. Chem. Rev. 1992, 92, 1473.

(a) Swings, P.; Rosenfeld, L. Astrophys. J. 1937, 86, 483. (b) Dunham, T., Jr.;
Adams, W. S. Am. Astr. Soc. Pub/s. 1937, 9, 5.

Rydbeck, O. E. H.; Hjalmarson, A. In Molecular Astrophysics: State of the Art
and Future Dire/ions; Diercksen, G. H. F., Huebner, W. F., Langhoff, P. W., Eds;
Dordrecht: D. Reidel, 1985; p 45.

Krelowski, J. Astron. Nachr. 1989, 310, 255.

Heger, M. L. Lick Obs. Bull. 1921, 10, 146.

Snow, T. P. Nature 1993, 366, 407.

Fulara, J.; Lessen, D.; Freivogel, P.; Maier, J. P. Nature 1993, 366, 439.

Sabo, M. 8.; Allison, J.; Gilbert, J. R.; Leroi, G. E. Appl. Spectrosc. 1991, 45, 535.
Tsuji, M.; Aizawa, M.; Nishimura, Y. J. Phys. Chem. 1995, 99, 3195.

Stanney, K.; Tedder, J. M.; Mitchell, A. L. J. Chem. Soc. Perkin Trans. 11 1986,
1383.

 

 

184

16. Tsuji, M.; Aizawa, M.; Ujita, H; Nishimura, Y. Bull. Chem. Soc. Jpn. 1995, 68,
2385.
17. Tsuji, M.; Aizawa, M.; Nishimura, Y. Chem. Letters 1995, 211.

 

 

 

APPENDIX A

 

185

Conventional Mass Spectrum

 

100
80 1
60 "
40 ‘
20 ‘

010203040 50 60 70 80 90100
m/z

Relative Intensity

 

 

 

 

Ion Current at Matrix Faraday Plate

 

35

30- ,
i .

25"

 

20*

15"

Ion Current (nA)

10- l

 

 

 

 

mlz —-—>

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV octopole (quad axis) = -28.4 V
total filament emission = 7.50 mA octopole = -14.3 V

ion source pressure = 50 mTorr post-quad einzel lens = -800 V
chamber pressure = 5.9x l 0'9 Torr post-aperture einzel lens = -61.7 V
ion source collector = +200 V left plate = +819 V

ion source extractor = +393 V right plate = +200 V

ion source lens = 0.0 V top plate = +101.2 V

bottom plate = +94.7 V

Appendix A] The mass spectrum of acetone.

186

Conventional Mass Spectrum

 

 

 

 

 

 

 

100
2‘ .
'17: 802
8 .
g 60‘
3 40-
'fi; 1
22 2°?
0 '1 if 1 1 .1
' ‘ I ‘ I I I ' fiI—r I ' I 7
0 10 20 3O 40 50 60 70 80 90 100
m/z
Ion Current at Matrix Faraday Plate
40
30-
if
5 ,.
E
g 20~
3
0 i
c r i
2 \1
10’- KM
0_ _J

 

 

 

m/z —->

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV octopole (quad axis) = -97.5 V
total filament emission- — 7. 41 mA octopole = -24.5 V

ion source pressure" — 30 mTorr post-quad einzel lens = -900 V
chamber pressure — 2. 3x10'7 Torr post-aperture einzel lens = -200 V
ion source collector- - +101. 7 V left plate = +47.2 V

ion source extractor = +1038 V right plate = +200 V

ion source lens = +78.6 V top plate = +854 V

bottom plate = +893 V

Appendix A2 The mass spectrum of carbonyl sulfide.

 

187

Conventional Mass Spectrum

 

100
sol
60 i
40 i
201

1

Relative Intensity

 

.11

111

 

 

l

 

OTIII'I'

I

' r

'l'ri'r'r'
0 10 20 30 40 50 60 70 80 90100

mlz

Ion Current at Matrix Faraday Plate

 

35

30’

25’

20F

15"

Ion Current (nA)

10~ fl

 

 

 

 

mlz -—>

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV

total filament emission = 7.80 mA
ion source pressure = 50 mTorr
chamber pressure = 4.9x10'9 Torr
ion source collector = +200 V

ion source extractor = +39.3 V
ion source lens = 0.0 V

octopole (quad axis) = -54.8 V
octopole = -18.2 V

post-quad einzel lens = -700 V
post-aperture einzel lens = -181.9 V
left plate = -68.2 V

right plate = +52] V

t0p plate = +5.7 V

bottom plate = +2.1 V

Appendix A3 The mass spectrum of methyl. ethyl ketone.

 

188

Conventional Mass Spectrum
100

 

A

on
0
L1

0)
o
s

A

N #
O o
s . 1

Relative Intensity

A

 

 

.l

‘ I I I ' I ' I ' I ' [4' I ‘ I ' T f
0 10 20 30 40 50 60 70 80 90 100
m/z

 

o

 

Ion Current at Matrix Faraday Plate

 

80

 

O)
O
I

Ion Current (nA)
A.
O

 

N
O
I
f)
7“?
,_._,

 

 

 

mlz —>

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV octopole (quad axis) = -80.6 V
total filament emission = 7.50 mA octopole = -75.9 V

ion source pressure = 25 mTorr post-quad einzel lens = -600 V
chamber pressure = 2.3x l 0'7 Torr post-aperture einzel lens = -200 V
ion source collector = +1200 V left plate = -6.7 V

ion source extractor = +1008 V right plate = +1405 V

ion source lens = +268 V top plate = +45.2 V

bottom plate = +35? V

Appendix A4 The mass spectrum of carbon disulfide.

189

Conventional Mass Spectrum

 

 

 

 

 

 

 

 

 

 

100
b 1
‘6 801
g l
g 60-
.2 40«
fl 1
g 20:
0 *r'rrfif i‘filer‘Lirrr'lilr
0 10 20 3O 40 50 60 70 80 90 100
m/z
Ion Current at Matrix Faraday Plate
40
30-
2‘
5
E
g 201-
3
o . N
C
2
10—
ii
i jLK/ //\1\/\/\./\/\
0- j m] L

 

 

 

mlz ——->

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV

octopole (quad axis) = -100.7 V

total filament emission = 7.50 mA octopole = -85.5 V
ion source pressure = 30 mTorr post-quad einzel lens = -700 V
chamber pressure = 5.3x 10'9 Torr post-aperture einzel lens = -200 V

ion source collector = +200 V
ion source extractor = +3 9.3 V
1011 source lens = 0.0 V

left plate = -51.9 V
right plate = +729 V
top plate = +5.7 V
bottom plate = -8.7 V

Appendix A5 The mass spectrum of toluene.

 

190

Conventional Mass Spectrum

 

100
80 1

4

601

4

40'

20 "
1
0 .11

Relative Intensity

 

 

 

 

 

1,1. 1.11

 

Ion Current at

0 1'0 20 30 4'0 50 80 70 80 90 100

f

mlz

Matrix Faraday Plate

 

30'

20"

Ion Current (nA)

 

 

 

 

 

 

m/z -—>

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV

total filament emission = 7.20 mA
ion source pressure = 30 mTorr
chamber pressure = 1.6x10’8 Torr
ion source collector = +121.6 V
ion source extractor = +95.0 V
ion source lens = +87.3 V

Appendix A6 The mass

octopole (quad axis) = -63.0 V
oct0pole = -39.0 V

post-quad einzel lens = -500 V
post-aperture einzel lens = -200 V
left plate = +15.3 V

right plate = +1493 V

t0p plate = +53.l V

bottom plate = +47 .6 V

spectrum of methyl bromide.

 

191

Conventional Mass Spectrum

 

100
80 -
60 -‘
40 4

20-
0‘ v.1..l 1

Relative Intensity

 

 

 

 

' I T ' I ' l ' I
O 10 20 30 40 50 60 70 80 90 100
mlz
Ion Current at Matrix Faraday Plate

 

35

 

25’
20"

15’ 11

‘l
l

1
l
1
_1

Ion Current (nA)

10»

 

 

.3 .1

mlz -—>

 

 

 

Mass-Selected Ion Source Operating Parameters

ion energy = 130 eV octopole (quad axis) = -56.5 V
total filament emission = 7.50 mA octopole = -74.8 V

ion source pressure = 40 mTorr post-quad einzel lens = ~500 V
chamber pressure = 5.8x 10‘9 Torr post-aperture einzel lens = -200 V
ion source collector = +200 V left plate = -56.1 V

ion source extractor = +394 V right plate = +792 V

ion source lens = 0.0 V top plate = -4.7 V

bottom plate = -14.0 V

Appendix A7 The mass spectrum of chlorotrifluoromethane.

 

192

Conventional Mass Spectrum

 

100
80 2

1

601

403

1

20‘

1

0 r"! ‘T ‘ I ' T ' 1 '
0 10 20 30 40 50 60 70 80 90 100
mlz

Relative Intensity

 

 

 

 

 

Ion Current at Matrix Faraday Plate

 

35

30"

25“

Ion Current (nA)

10"

 

 

 

Appendix A.8 The mass spectrum of methane. The mass-selected ion
source operating parameters were not recorded for this spectrum.

 

APPENDIX B

 

 

APPENDIX B

Calculation of the Electric Potential at a Distance from a Charged Disk

1. The electric field due to a uniformly-charged ring
The first step in calculating the electric potential due to a charged disk is to

determine the electric field of a uniformly-charged thin ring, as shown in Figure A2. 1.

 

 

Figure A2.1

The electric field (E) at point P can be calculated by placing a test charge (go) at point P.
The field is then directly proportional to the force (1") applied to the test charge by the

total charge on the disk (q), as the following relationship shows:

F
E = —
0. (1)

Coulomb’s law states that the force (F) between these two charges is inversely

proportional to the distance between them (r):

F= I (”2"; (2)

Jae, r

 

 

thus, the electric field at point P due to the charge on the disk is:

._ I 9
5-41137. (3)

 

193

 

194

Now consider a differential element of the ring of length (d5) located at an
arbitrary position on the ring in Figure A2.1. It contains an element of charge given by
dq = Ads, where 21 is the linear charge density (units of C/m). The differential field (dE) at

point P is then:

1 Ads
dE = 4718 7 (4)

 

The formula r2 = x2 + R2 can be incorporated into this equation to produce:

1 Ms
dE = .
47mg (1:2 +R2)2 (5)

 

The resultant field at point P is the sum of all the field contributions dE made by the
differential elements of the ring. The a’El components will add up to 0, due to the
symmetry of the ring. The only components of the field that will contribute to (III are the

ones parallel to the axis of the ring (dEn = cos 6). The following relations:

3: X
6 = — =
COS r (6)

(x’ + 111")’/2

 

can be used to derive the electric field at point P in the following steps:

 

 

] Ads x
a’E 6 = 7
COS 4775,,(61'2 +R2)2) (x2 +R2)'/2 ( )
x71
E = IdEcosG = [615. (3)

4113,02 + R2 )3/2
The integral ldS is simply 27rR, the circumference of the ring. Since q, the total charge on

the ring is A(Zrdi), we can then write the previous equation as:

qx
E = . . 9
47:8,,(x2 +R’)3/’ ( )
This is the equation that relates the electric field at point P to the total charge on the ring.
It will be used in the next section to derive the electric field at point P due to a charged

disk rather than a ring.

 

195

II. The electric field due to a uniformly-charged disk
Figure A2.2 illustrates a circular disk of radius R, carrying a uniform surface
charge of density don its surface (a has units of C/mz). The ring of radius s and of width

ds is drawn on the disk.

 

ds
s P dE ,
.................... \ x 3 p x
R
Figure A2.2
The total charge on the disk (dq) is:
dq = 061.4 = o(21rs)ds, (10)

where (M is the differential area of the ring. Upon substitution of dq for q in equation 9,

and replacing R with s, the following equation is obtained:

dE = 7:027:st 3 . (11)
41tt»:,,(.ir2 +s2)A

 

The electric field (E) at point P can then be derived by integrating over the surface of the

disk, froms = 0, tos = R:

 

ox i; 25ds (12)

E = 675 = .
I 458 o (x2 -1-s2 )%

Integrating and substituting the limits, the final equation obtained is:

E = 3—(1 —- ——3‘-——] 13
a s/x’+R2 ( )

196

Appropriate substitution of the constants 0’ (C/mz) and £0 (NmZ/Cz) results in the electric

field having units of N/C.

III. The electric potential due to a uniformly-charged disk
The electric potential at a point in space is always defined as the difference in
potential between that point and a reference point at some other position. In this case, the

reference point will be taken as x' = co. If a test charge go is moved from this reference
point to point P in Figure A2.2, the electric field from the charged disk will exert an

element of work on the test charge given by

(q.E)dx' (15)
where an is the force exerted by the field on the test charge. The total work will then be

the integral
f
WV = qu‘Eahr'. (15)

The difference in potential between points i and f is defined as

V _ Wy (16)
f ' qo'

Substitution of WV from equation 15 into equation 16 results in the following relationship

I
V, — L; = —[de'. (17)

The initial point has been previously defined to lie at x’ = co and we may also define the

potential at this point to be 0, relative to all other points. Therefore

V = —ide'. (13)

 

197

The electric field E has been previously derived in equation 13. Substitution into
equation 18 results in
f

c x'
V = — 1 - —————d '. (19)
280 IL \/x'2+R2j x

This equation allows us to calculate the potential between any two points in space. The

 

electric potential at point P (VP) due to the charge on the disk can then be calculated by
integrating equation 19 from 00 to point P which lies at x,

0' I0 - ——£'———jdx'
260w [xr2+R2

Integrating and substituting the limits results in

V = a [W—x] (21)

P _
280

 

VP = (20)

The electric potential at any point P along the axis of a charged disk is a fiinction of the

charge density on the surface of the disk (0'). the radius of the disk (R), and the distance

fi'om the center of the disk (x).

 

"‘llllllllllllll“