1 ' *1“ WV {3.23-
fihiéttg-‘E "‘M‘ 13"“ z '
. . .
Eva->3- ~_, 5. .‘1‘:- :55:
‘

—.-91~1:;2.Z'. "My:
3'1 I:

 

‘3‘}: 18:42am?“

2
. n...;_:
5 r.‘

“;::n’_..‘:'

p5-

97‘
1.1.

253F372.“
-- .. 1

N
45.)“,
53:6 «#1
J 1' $23.1"
.35:

1.1- .1. 1..
v 1.

-.»~..‘
.1... “#95:.“
\A ‘ “ , .131“
'C' '

 

".1”
~

1‘} . ,
“it?” ~ “+34“ .3?
“QM-“s“ _
’"4 > 11:1-
--. nm...,_,
. .
1 .
“1:41.“.-- .e. .‘2‘
5.11.», a.-. -

”,4 ‘ ~ 7 I '
.;.j.:*‘..... '21-“; ~ ‘-

1
_ 5“" ft“ :10.-.‘4‘4
_ _ . . . ‘. “.13
J... $11.4..." ”.131! I 8-1:, — 1 an 1' ,l .«....;..j:
" -:Nv_x:.,-J r‘ .15 : £931.14"?
I L

- "i
«’3 u, ,_,

1....
“4":fvhh

lam ‘O'iu. “um.

.53 VII-u

 

A“
.v r. ,
\. :21 -u"'.‘ _

mp- 5".1—3.’ I at.) .l v

',xf.“.;a,
‘.#..‘

 

 

 

 

I. v1.11). m.‘ ‘1 o,
H \
‘5‘.

:r
:9me -, a,

. 1...”...MJ:
vyn- uv

v.1! vu—x U-n:

51"“-
DH 4!
. ." MT"-
' ‘rTn -

u. I.

l.--,

.413”; m;

‘ 7"... 35-1 A. E, ,
2.. 1 4.2::
f..{...;¢~;-'.Wr..ur‘§bv
I

“5 F171;,
35" ’3‘},

 

 

Ml IICH IGAN STATE

III SIIIIZI II

 

 

 

 

 

 

 

 

 

 

 

 

IIIIIIIIIIIIIIIIIIIIIIIII

m'14055

This is to certify that the

dissertation entitled

BRIDGING THE GAP BETWEEN MASS SPECTROMETRY AND SPECTROSCOPY:
AN INSTRUMENT DESIGNED FOR SPECTROSCOPIC INVESTIGATION OF

MATRIX-ISOLATED, MASS-SELECTED IONS
presented by

MARK STEVEN SABO

has been accepted towards fulfillment
of the requirements for

Ph 0 Do degree in Chemi S try

 

 

W. 4a»;

Major professor

 

Date _J_Ufl.e_2.11_l9.9_]._

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

 

LIBRARY
Michigan State
Unlversity

 

 

 

PLACE IN RETURN BOX to remove thie checkout from your record.
TO AVOID FINES return on or betore die due.

  

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IL__IC:I___I

 

 

 

 

 

 

 

 

 

 

 

 

 

I
II Ii? I

MSU le An Afflrmettve ActloNEquel Oppomnlly Institution
ammo-m

 

BRIDGING THE GAP BETWEEN MASS SPECTROMETRY AND
SPECTROSCOPY: AN INSTRUMENT DESIGNED FOR SPECTROSCOPIC
INVESTIGATION OF MATRIX-ISOLATED, MASS-SELECTED IONS

by

Mark Steven Sabo

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

ABSTRACT

BRIDGING THE GAP BETWEEN MASS SPECTROMETRY AND
SPECTROSCOPY: AN INSTRUMENT DESIGNED FOR SPECTROSCOPIC
INVESTIGATION OF MATRIX-ISOLATED, MASS-SELECTED IONS

by

Mark Steven Sabo

The study of ion/molecule reactions is important in many areas of
atmospheric, environmental, catalysis, and materials science. Their
reaction mechanisms and resulting products are often studied by mass
spectrometry, which can yield kinetic and thermochemical information,
but from which little direct structural information is available. In order to
"bridge the gap" between mass spectrometry and spectroscopy we have
constructed an instrument designed to trap mass-selected ions from mass
spectrometers in a low-temperature inert gas matrix, for subsequent
structural analysis via vibrational and electronic spectroscopy. Isolation of
mass-selected 082* in an argon matrix is confirmed using laser-induced
fluorescence (LIF) spectroscopy. The role of counter-ions in matrix
isolation of ions is discussed, as well as implications of the LIF results to
infrared spectroscopic studies of matrix-isolated, mass-selected ions. Early

infrared experiments suggest that some ion neutralization may occur.

Table of Contents
List of Tables
List of Figures

Chapter 1. Introduction
1. Matrix Isolation
A. The Ideal Matrix
II...Generation of Ions in Matrices
VUV Photoionization
Windowless Discharge
Proton Radiolysis
Electron Bombardment
Ion/Molecule Reactions
Chemical Ionization
. Mass-Selected Generation
III. Trapping Ions in Matrices
A. Deposition Rate
B. The MC Ratio
C. Counter-Ions
IV. Matrix Effects on Ionic Spectra
A. Frequency Shift
B. Spectral Distortion
C. Vibrational Relaxation

masses?

Chapter 2. Instrumentation for Matrix Isolation of Mass-Selected Ions

I. Design Goals

II. Vacuum System
Vacuum Chamber
Vacuum Pump

Pressure Gauge
Interlock

$3.05”?

iii

(OQmCEO'IOOMv—I

H
O

5235533513553

ESSEEESSB

III. Cryostat
A. Temperature Controller
B. Matrix Gas Line

IV. Ion Generation

V. Laser-Induced Fluorescence Detection of Matrix-Isolated Ions

A. Lasers
B. Chamber Windows
C. Detector and Electronics
VI. Fourier Transform Infrared (FTIR) Spectroscopy

Chapter 3. Preliminary Matrix Isolation/FTIR Studies
1. NO” Deposition Studies
A. Introduction
B. Experimental Section
C. Results and Discussion
D. NO+ Study Conclusions
II. CH2NH2+ Deposition Studies
A. Introduction
B. Experimental
C. Results and Discussion
D. CH2NI‘12+ Study Conclusions

Chapter 4. Spectroscopic Identification of Matrix-Isolated C82+
I. Laser-Induced Fluorescence Studies of C82+
A. Introduction
B. Experimental
C. Results and Discussion
D. Role of Counter-Ions
E. Vibrational Spectra Implications
II. Infrared Studies of C82+
A. Discussion of Infrared Results
B. Infrared Study Conclusions

iv

ESSEESSSS

SS3D‘38ES83838

88859883863R38

Chapter 5. Conclusions and Future Work 99
1. Conclusions 99
II. Instrumental Enhancements 99

A. Cryostat 99
B. Vacuum System 100
C. Mass—Selective Ion Source 101
D. Photolysis Lamp 102
III. Research Plan 102
A. Development and Optimization of Experimental Procedures 102
B. Early FTIR Studies 104
C. Spectroscopic Studies of Metal-Containing Ions 105

Appendix 1.1 Schematic diagram for vacuum interlock control box. 107

Appendix 1.2 Simplified diagram of Cryo-Temp controller. 110
Appendix 1.3 Schematic wire diagram of Cryo-Temp controller. 111
Appendix III.1 Experiments performed in the NO+ deposition studies 113
Appendix III.2 Infrared spectra of Ar/NO co-deposition study 115
Appendix III.3 Power spectrum of an Ar/NO co-deposition study 118
Appendix III.4 Experiments Performed in the CHZNH2+ deposition

studies 119

Appendix 111.5 A portion of the infrared spectrum of n-propylamine
deposited through the Anavac MSIS source in argon

at 15K 122
Appendix IV.1 Experiments performed in the LIF studies of CSg+

depositions 126
Appendix IV.2 Experiments performed in the infrared studies of 082+

depositions 131

List of References 133

3.1
3.2
3.3

4.1

Listof'l‘ahles

Identification of infrared bands in the NO+ experiments
Mass spectrum of n-propylamine

Infrared bands in the spectrum (Appendix 111.5) obtained
from deposition of propylamine through the Anavac
mass-selective ion source.

Spectroscopic data related to the antisymmetric stretching
vibration for C82”o

vi

ListofFigures

Figure

2.1

2.2

2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
2.11
2.12
2.13
2.14
2.15
2.16

2.17
3.1
3.2
3.3

Schematic diagram of experimental chamber constructed to
isolate mass-selected ions in a low-temperature inert gas for
spectroscopic study

Schematic diagram illustrating several sources of gases and
vapors in a vacuum system .

Conflat flanges used in UHV applications
Cross-section of a Leybold-Heraeus RPK 1500 cryopump
Block diagram of a vacuum system for experimental chamber
Block diagram of the Dycor M100 RGA

Front panel of the vacuum interlock

Flow diagram of pump-down procedure

Displex 202 cryostat modified for ultra-high vacuum
Germanium resistor calibration curve

Circuit diagram used in Ge resistor calibration
Vacuum line for matrix gas introduction

Original Anavac RGA source configuration

Modified Anavac RGA source configuration

Faraday plate mounted on linear motion feedthrough

Schematic diagram of the instrument in its laser-induced
fluorescence configuration

Optical configuration for infrared analysis of mass-selected ions

Infrared spectra of N 0+ deposition and photobleach study
Summary of Ar/NO (700/1) deposition

Infrared spectrum obtained following CH2NH2+ deposition
showing weak absorptions of neutral methylenimine

vii

Page

888%}8‘388

tfi‘wifif:

{“5

E38

@8386?

3.4
4.1
4.2
4.3

4.4

4.5

4.6

Infrared spectrum obtained following CH2NH2+ deposition
Coumarin 440 laser dye power curve
Scattered light signal observed at the pmt vs pmt voltage setting

Scattered light signal observed at the pmt vs wavelength of the
dye laser

Relationship of the excitation and emission spectra of 082+ in
argon to the cutoff frequency of the long-pass filter

Total fluorescence spectrum of C82+

Infrared spectra following CSz+/Ar coodeposition

viii

836:]

88

Chapter 1: Introduction

In order to better understand the structure/function relationships
that govern gas phase ion chemistry, we have initiated a long-term project
to spectroscopically determine the geometries of ions, especially those that
are products, of ion/molecule reactions. The desire to establish the
structures of ions has lead to the development of a number of methods for
obtaining structural information via mass spectrometry (MS), such as
unimolecular and collision-induced fragmentation, isotopic labeling,
ion/molecule reactionsl:2:3’4, neutralization-reionization5, and other
MS/MS techniques? Because they are all indirect, these techniques cannot,
alone, establish gas phase ionic structures. Nonetheless, they have been
used in collaboration with high level ab initio molecular orbital theory
calculations to provide structural information in several cases. For
example, Holmes and co-workers7 have identified eleven different stable
isomeric ion structures for the [C2H302+] cation, an ion frequently observed
in the mass spectra of oxygen-containing organic molecules. In order to
differentiate between the two isomers shown below, it was necessary for the
authors to perform a series of MS experiments combined with ab initio

calculations.

11300-020 HOCHZ-CEO

1 2

A more direct method of structural determination for these two
isomers would utilize vibrational spectroscopy, which could differentiate

between the two isomeric forms of [C2H302+] by the detection of an 0-H

stretch in cation 2. Unfortunately, such conventional spectroscopic tools
are not yet available to the mass spectrometrist. This is due principally to
the limit of the number of ions that can occupy a given region of space in the
gas phase. This "space-charge limit" is ~105 ions per cubic centimeter,8
which corresponds to femtomolar concentrations, and requires the use of
highly sensitive and sophisticated spectroscopic techniques. Structural
information from high-resolution infrared spectra of some gas phase ions
have been reported; to date such studies have been limited to diatomic and

small polyatomic ions.9
1. Matrix Isolation

Matrix isolation (MI), first introduced in 1954,10 is frequently used to
spectroscopically characterize chemical intermediates and free radicals.11
In the last two decades, MI has been adapted to the study of ions.12 Matrix
isolation involves the trapping of reactive species in a rigid cage consisting
of a chemically inert medium (matrix) at low temperature. Typically the
analyte is co-deposited with an inert gas onto a substrate that is held at a
temperature of 4 - 20 Kelvins. The rigidity of the matrix cage prevents
diffusion of reactive species. Furthermore, the matrix (host) can extend
the lifetime of the intermediates almost indefinitely while they are
characterized by a number of spectroscopic techniques(infrared, Raman,
visible/ultraviolet, and ESR).

Cryogenic and vacuum technologies are important in performing
matrix isolation experiments. Temperatures as low as 1.8 K are obtainable
with liquid helium cryostats and 10 K with closed-cycle refiigerators. The
temperature needed depends on the matrix gas to be utilized. High vacuum

conditions must be attained in order to reach these temperatures. The

matrix gas and precursor are deposited on an optical substrate maintained
at low temperature, where condensation occurs almost immediately. The
precursor must be in a vaporized form so that it may be mixed with the host
before deposition, or deposited directly. Photolysis, electron bombardment,
and proton bombardment are just a few of the techniques that have been
used to generate species of interest within the matrix. Matrix isolation is
recognized as one of the two primary methods of investigating short-lived
molecular systems. The alternative approach involves the use of rapid
spectroscopic techniques such as picosecond spectroscopy, flash photolysis,
and other related methods13 to monitor gas and liquid phase species

ATheIdealMatrix

The choice of the matrix host would seem to be fairly important since
the matrix provides the necessary isolating environment so that guest
species may be properly characterized. Several factors and characteristics
affect the choice of the best matrix for a given experiment. One of the most
important requirements is that the matrix material be free from
impurities. Non-reproducible spectra result from even the smallest
contaminant. This makes a clean vacuum system a necessity for matrix
isolation work. The host material should have a sufficiently high vapor
pressure at room temperature so that it can be handled easily and be
deposited on the cold surface. The vapor pressure at the cryostat's
temperature should be sufficiently low to prevent loss of matrix material.
The matrix should also be optically transparent in the spectral region of
interest. Bands arising from absorption, emission, or scattering can
interfere with the spectroscopic study of the matrix-isolated sample. The

matrix utilized in a particular experiment should, in most cases, be

chemically inert. This is a requirement only for those experiments where
the major interest lies in studying the intrinsic properties of the guest.
Some experiments are designed to monitor the interactions between the
guest and host, so that an inert matrix might not be the best choice;
however, for most matrix isolation studies, the best matrix would be one
that does not react with the guest.

The ideal matrix should be rigid at the temperature of the cryostat.
This is an important characteristic since lack of rigidity leads to diffusion
which results in recombination or complete loss of guest species. The
temperature at which diffusion occurs is approximately 0.5Tm, where Tm
is the melting point of the matrix material. Rigid matrices result when the
temperature is below 0.3Tm. Temperatures ranging between 0.3Tm and
0.5Tm result in a process called annealing where the matrix rearranges at
the atomic level to adopt the most stable crystal structure.

The matrix must possess excellent thermal properties in order to
trap reactive species on the cryostat substrate. When the matrix gas is
deposited on the cold surface, there is heat released upon condensation
called the latent heat of fusion (Lf). This energy, along with the total energy
of the gas, must be conducted away from the substrate through the cryostat.
Therefore, it seems reasonable that the ideal matrix should have a high
thermal conductivity so that the excess energy may be removed quickly.
Hosts characterized by low thermal conductivities experience local heating
at the deposition surface, which produce opaque and highly scattering
matrices and possible loss of reactive species through diffusion. 14

In reality the ideal matrix cannot be achieved, but some materials do
come quite close to approximating the ideal case. Most noble gases have a

spectroscopic "window" (free from transitions) from the UV to about 100

cm;l where lattice vibrations of the host are prevalent. The rare gases
possess sufliciently high thermal conductivities, rigidity at temperatures
attainable by the present technology, and are chemically inert. These
properties make the rare gases (with the exception of He) the obvious choice
for matrix isolation work and they are the host materials most often

employed.
1]. Generation of Ions in Matrices

As discussed above, matrix isolation is used to study chemical
intermediates and free radicals and has recently been adapted to the study
of ions. The trapping of ionic species was not reported until 1970, when
Kasai wrote the pioneering paper describing the entrappment of Na+ in an
argon matrix.15 Ions are difficult to characterize spectroscopically in the
gas phase because of the difficulty of preparing appreciable concentrations
due to the long-range coulombic forces that prevail. Discharges, flames,
and plasmas are other methods by which ions can be studied; however,
because of the high temperatures needed, a large number of states can be
thermally populated which increases the spectroscopic complexity and
difficulty in species characterization. Rotational and vibrational structure
is the leading cause of complicated electronic spectra. An advantage to
matrix experiments is that the species generally absorb radiation from
their ground vibronic state, and emit from the lowest vibrational level in the
upper electronic state which greatly simplifies the resulting spectrum.
Vibrational spectra are also less complicated since rotational motion is
usually quenched in a matrix. 15

Ionic species have been generated from stable precursors already

trapped in the matrix environment, by vacuum-ultraviolet

photoionization,17 multiphoton ionization,18 windowless discharge,19
electron bombardment,20 and proton radiolysis.21 More recently, methods
such as fast atom bombardment, laser vaporization and chemical
ionization,22'23' 24 where the ions are generated external to the matrix
environment and are co-deposited with the matrix gas, have been used to
trap ionic species at low-temperatures. The first successful isolation of a
mass-selected cation was reported by Maier and coworkers in 1989.25 Ions
in matrices have been characterized by a number of spectroscopic
techniques including laser-induced fluorescence (LIF),26 electronic
absorption spectroscopy,27 ESR,28 and vibrational spectroscopy.29
Experiments where the ions have been generated in situ are often difficult
to interpret due to the fact that a complex mixture of ionic and neutral

species can be formed in the matrix.30

Charged species in matrices can be classified into two general
categories: 1) chemically bound ion pairs, and 2) isolated or coulombic ion
pairs. The first category refers to a species of the type M+X', in which the
wave functions of both atoms overlap considerably. The isolated ion is of the
type M+, where the cation is separated from its counter-ion by an unknown
number of matrix atoms.31 Described below are several ways in which ions

of the isolated type have been generated in a matrix.
A. VUV Photoionization

The first technique used to produce isolated ions involved the co-
deposition of Na and H1 in an argon matrix and subsequent photoionization
of Na with a mercury arc lamp to produce electrons which migrated

through the matrix. These electrons were captured by HI, which reacts

with the electrons by dissociative electron capture.15 The resulting ions
(Na+and I') are well separated fi'om one another in the matrix so that there
is no overlap between their orbitals. This process can be summarized by the

following reaction.

Na+HI -l‘-’—>Na++H+I'

An interesting aspect of photoionization in matrices is that the
ionization energy can be altered relative to that for the gas phase because of
solvent effects and ion pair formation. Kasai discovered that electron
transfer to weaker electron acceptors could be stimulated by irradiation
well below the metal's ionization energy. The electrostatic interactions as
well as the electron affinity of an acceptor play a major role in the ionization
processes if the electron tunnels from a highly excited state of one guest to
another. The energy of this process is given by:

 

e2
E=IE - EA - 4
neoer

where IE is the ionization energy of the donor species, EA is the electron
affinity of the acceptor separated from the cation by a distance r, 60 is the
permittivity of vacuum, and 8 is the relative permittivity (dielectric
constant) of the matrix. Since the separation of two nearest sites in an
argon matrix is 7.5 Angstroms, the coulombic stabilization is on the order
of 1.3 eV. This discussion indicates that even when two species are too far
apart to affect each other's spectra, they may be close enough to help
stabilize ion pairs, so that photon energies less than the ionization potential

of a gas phase species may suffice to produce ions in a matrix.32

Photoionization can also be performed during the deposition process,
which results in direct trapping of ions. Photolysis in this manner cannot
take advantage of the solvent interactions; hence, high-powered sources
with sufficient energy must be used to form ions.

A major limitation to VUV photoionization is that it is a non-selective
technique. Several ionic species may be formed in the matrix in addition to
the species of interest, particularly when photolysis is carried out during
deposition. The high energy photons can cause bonds to break, which leads
to the formation of fragment ions as well as of dimer and trimer ions
generated from recombination of fragment ions. This may make

unambiguous assignment of infrared or electronic spectra quite difficult.
B. Windowless Discharge

Another popular technique for ion formation in matrices utilizes a
windowless discharge. This method involves the co-condensation of a
sample with argon that has passed through a discharge and has been
applied to the formation of BX3* (X = Cl, Br) from BX3,33 002+ and 002'
from 002,29 04* and 04‘ from 02,34 N4+ from N2,35 and CF4.an+ from CF4-
an (X = Cl, Br; 11 = 1, 2, 3, 4).36 In these experiments, both photoionization
from argon resonance radiation, and chemiionization by collision with
metastable argon atoms in the region between the end of the discharge tube
and the cryostat substrate are responsible for ion formation from precursor
molecules.37 Secondary photoprocesses are less important than in typical
photolysis studies, thus giving the windowless discharge technique
somewhat higher selectivity and yields than the conventional photolysis
methods; however, secondary ion formation does indeed occur. This can

result in a variety of ionic species in the matrix.

GPmtonRadiobrsis

This technique for producing ions was first reported by Andrews and
coworkers in 1975.38 Protons produced in a RF discharge were accelerated
to 2 keV and focused onto a matrix during deposition of carbon tetrachloride
and argon.21 Ions such as Cle and 013+ were identified using infrared
spectroscopy. The mechanism of ion formation seems to involve energetic
argon ions, formed from the 2 keV protons, which collide with CC14
causing fragmentation and ionization. Incidentally, some of the protons
are thermalized and react with other species; products identified in the

matrix include ArH", HC12', CHC13’ and CH2C12.
D. Electron Bombardment

In 1983 Knight reported a new experimental procedure utilizing
electron bombardment for generating ions in matrices.20 The bombarding
electrons (40 uA, 60 eV) are focused onto the matrix during deposition.
Knight found it necessary to float the deposition target at a positive potential
(25 - 35V) in order to overcome the negative surface charge that developed.
He studied species such as H217O*, 1300+, 15NH34’, and 14N2+ in a neon
matrix. Comparison of the esr spectra of these small cations produced by
electron bombardment with the esr spectra of the same ions produced by
more conventional means (such as the windowless discharge method)
demonstrated that electron bombardment produced a greater number of
ionic species. This method allows for a great deal of flexibility since the
electron flux and energy can be varied over a wide range; however, it may

not be selective for only one specific cation, and suffers the disadvantage of

10

being a complicated experimental procedure. Knight went to great lengths
to electrically isolate the deposition target.

More recently, Suzer and Andrews developed a thermionic electron
source capable of delivering 10-200 mA of electrons at 30-200 eV to the
cryostat window.39 Continuous bombardment of an argon/H20 matrix
(M/G = 1000/1) with electrons during condensation at 12K produced
sufficient amounts of OH' and ArnH+ ions as well as of OH radicals to

record their infrared spectra.
E. Ion/Molecule Reactions

One of the most interesting methods of generating ions in matrices,
promulgated by Knight, involves the use of ion/neutral reactions.
Photoionization and electron bombardment were employed as a means of
cation formation, and the products have been studied by ESR.“0 Reactions
that have been investigated are listed below.

00+ + CO —> (3202+ (1)
N2+ + CO ——'> N200+ (2)
N2+ + N2 —" N4+ (3)

In reaction ( 1), CO+ reacts with a neutral CO molecule to form C202+
which becomes trapped in a neon matrix. The mechanism of C202+
formation is most likely due to condensation of CO** with a neutral CO
trapped in the same cage. The frequency of collision is increased due to the
cage effect and 0202* is formed.

11

Products trapped from reactions such as these would provide a
wealth of information if infrared (IR) and laser induced fluorescence (LIF)
studies were performed. This technique can also be used to study the
dynamics of ion/molecule reactions in matrices; however, the need for high
concentrations of reactants with subsequent photolysis can yield an
abundance of secondary ions in the matrix.

Two other methods of ion formation have been applied to ion
generation (albeit only to a limited extent to date): Fast Atom Bombardment
(FAB),22 and laser sputtering.23 Both prove to be valuable techniques that
can aid in the understanding of processes involved in trapping ions in

matrices.
F. Chemical Ionization

A chemical ionization source from a Finnigan 3200 mass
spectrometer was modified for matrix isolation studies 1989 by Hacaloglu
and Andrews.“ Ions generated in the source were directed to the cryostat
substrate (12 K) and deposited with excess argon. Electronic spectra were
recorded for the naphthalene cation when naphthalene vapor and argon
flowed into the source block. Investigation of negative ions from a 1%
Cly/Ar mixture resulted in ions such as Clz‘ and 013'. The HClz' anion was
observed when a Cly/HCI/Ar mixture was studied. It was not determined
whether CgH10+ was produced in the source and trapped in the matrix or if
argon cations accelerated to the matrix produced CgH 10+ by colliding with
the neutral molecule during condensation. Electron attachment to 012
produced 012', and at higher pressures resulted in 013' formation. These

species were produced through the following mechanism: (Note: eth'

represents thermalized electrons)

12
Ar+e'——->Ar++e'+eth'
C12+em-——->C12.
C12+e'———>Cl'+Cl

C]. + 012 -——>' 013.
G.Mass-selectedgeneration

One of the most troublesome aspects of the generation methods
described above is that a number of ionic species could be obtained from the
same precursor, which will inevitably complicate the matrix spectrum. In
1989, J. P. Maier and coworkers developed an ion generation method that
employs a mass spectrometer for separation and selection of the ions before
they are trapped in a low-temperature neon matrix.25 Ions are produced in
an electron impact source and mass-selected with a quadrupole mass
analyzer before being deposited in the matrix. Ion currents of 1-10
nanoamps were produced with energies of 50 - 350 eV. The electronic
absorption of ions such as N2+, C02+, C2N2+, ClCCH+, H(CC)2H+, C2+, and
C4H2+ were measured in the 220 - 1200 nm region. This technique has the
potential to be applicable to spectroscopic characterization of ionic species of
interest to mass spectrometrists and will be discussed in the following

chapters.
11]. Trapping Ions in Matrices

The entrapment of ions in matrices seems to be somewhat of an
"art". A great many factors influence the trapping process, such as
deposition rate, the matrix/guest (M/G) ratio, and the presence of counter

ions.

13

A. Deposition rate

The matrix deposition rate is an important parameter involved in the
trapping process. In order to prevent diffusion of trapped species, the
matrix must attain a temperature of 0.3Tm which means condensation
must occur as rapidly as possible. The heat released on cooling must be
conducted to the cryostat through previously deposited layers of the matrix
fast enough to keep diffusion from occuring. This can only be accomplished
if the matrix is deposited slowly. The power liberated in this process is
given by

1...: {W}

where Lf (cal mole'1)is the latent heat of fusion, 11 is the deposition rate in
moles sec‘l, A is the deposition area, 1 is the thermal conductivity in cal
cm'1 sec'1 K'l, T and T0 are the surface and interior temperatures

respectively, and w is the thickness of the matrix; that is:

where t is the length of deposition time and p is the molar density of the
matrix material. Thus the surface temperature is given by the following

equation:

 

T = To + LPZI
2W

Note that the surface temperature rises as the square of the deposition rate

and linearly with time. Substitution of typical parameter values into this

equation shows that under normal deposition rates (1-18 millimoles hrl)

14

the temperature only rises one degree or so even after several hours of
continuous deposition.“1

The deposition time will depend on the method used for guest
characterization. For instance, because infrared spectroscopy is less
sensitive than LIF, one would have to deposit more sample in order to
record an infrared spectrum, with consequently longer deposition times. It
is imperative that the matrix gas be deposited slowly so that adequate
cooling of the gas will result in the trapping of ions, but not their

subsequent diffusion.42
B. The MIC ratio

In order to avoid excessive interactions between guest species, it is
important that the ratio of the matrix gas to guest species (MIG) be
monitored. In a rare gas lattice, each guest molecule would have twelve
nearest neighbors, so that for a M/G of 100, statistically one out of eight
guests would have another guest as a nearest neighbor. Typically M/G
ratios of 1000 to 10,000 are used so that resulting spectra are free from
guest-guest interactions.

C. Counter-Ions

It may be expected that an accumulation of positive charges would
lead to destruction of the matrix due to repulsive coulombic interactions.
An important aspect of ion entrapment in matrices is the fact that the
matrix must remain electrically neutral.32 How is this neutrality attained?
The answer depends on the mechanism of ion formation and trapping,
described previously. When cations are generated in situ (such as in

photoionization), it is generally accepted that the counter-ions result from

15

electron attachment to impurities such as 02, C02, or OH (from H20
dissociation) in the matrix environment;28 nevertheless, these species have
not yet been identified with certainty.31 (It has been shown that electrons
are mobile in matrices and travel through the matrix until they encounter
some sort of electrophile.32 Attempts to form counter-ions by intentional
introduction of an electrophile into the matrix have also been reported.43
Despite the fact that this resulted in an increase in the cation signal, a
positive identification of the anion (counter-ion) was not obtained. Suzer and
Andrews observed the coexistence of ArH+ and OH' in a matrix after
electron impact on an Ar/HzO matrix during condensation;39 however, no
direct relationship between the relative concentrations of these two species

was found.
IV. Matrix Efl'ects on Ionic Spectra

There are many effects that matrices have on the electronic and
vibrational spectra of ions. The vibrational spectra that result are
characterized by sharp, well-resolved bands; however the matrix usually
has the opposite effect on electronic spectra due to the various sites in which
the guest may be trapped. The spectrum of a matrix-isolated species may
differ from the gas phase counterpart with respect to the positions of the
peaks, the shape of the peaks, and the number of peaks. A discussion of
some of the various effects of the matrix on vibrational and electronic

spectra follows.

A.Frequencyshift

Quite generally, the electronic or vibrational transition frequency of a
matrix-isolated guest is shifted relative to the gas phase value. Molecules

16

in the gas phase are free from complicating solvent phenomena. Physical
properties of gas phase species are often the standards to which properties
measured in solution are referenced. Intermolecular interactions are quite
prevalent in solution. These interactions arise from dipole/dipole,
dipole/induced-dipole, and London dispersion forces. These interactions
depend directly on the polarization of the molecule and usually result in
attractive forces which lower the electronic transition energies of the
system (red shift). A matrix can, in principle, be treated as a solution and
attempts have been made to calculate the frequency shift in matrices;44
however, in the matrix, solvent molecules surrounding the trapped species
are fixed, forming a rigid cage. The rigidity of the cage can introduce
another important interaction - repulsive forces - which must be taken into
account. The neglect of repulsive forces (increasing energy, blue shift) is a
serious shortcoming because these forces may be strong enough to cancel
the attractive forces.
The frequency shift may be expressed by?“5

An pg“ = A0 + Aoind + A'odi813 + Aurep

tot = 1)matr'ix ' elec
The electrostatic term arises from the interaction between permanent
charge distributions of two molecules. The inductive term is described by
the dipole/dipole and dipole/induced-dipole forces, whereas the dispersion
term is characterized by the induced-dipole/induced-dipole interactions
between neutral molecules. As the polarizability of the matrix host
increases there is a trend to lower frequencies.46 The general trend is that
the shift becomes less negative as the vibrational frequency decreases.47

This is due to the fact that the matrix cage tends to distort the guest to fit the

17

available site. Normal coordinates characterized by small force constants
are more easily distorted by matrix forces, which will generally result in
decreased equilibrium bond distances. This causes the potential energy to
increase, hence increasing the vibrational frequency (blue shift). Loose
cages have a tendency to produce red shifts.

Jacox reports that most vibrational matrix shifts [100' (1)3” -
DmameUgu] for diatomic molecules are less than 1% for neon and less
than 2% for argon with red shifts being more common than blue shifts.48
The heavier rare gases induce somewhat larger shifts from gas phase
values. Larger deviations than these norms are due to 1) hydrogen bonding
with the matrix, 2) cage stabilization of weakly fonded species, 3) incorrect
assignments in either matrix values or gas phase values, and 4) species
with a large dipole moment. Neutral molecules are less affected than ions

by this effect since isolated ions more strongly polarize the host atoms.
B. Spectral distortion

Guest-host interactions for neutral molecules in a matrix are
negligible compared to those for molecular ions; ionic spectra can be
significantly perturbed by the strong interactions between the guest and the
matrix. The spectrum of the diacetylene cation (C4H2‘t) was studied by
Bondybey in both neon and argon matrices.17 The laser-induced
fluorescence excitation spectrum and the dispersed fluorescence spectrum
were not strongly afl‘ected in solid neon, being similar to the gas phase
counterparts; however, the spectra in solid argon showed large
perturbations due to the fact that the ionization energy for argon is much
lower than that of neon (15.76 eV and 24.6 eV respectively). Since the
electron aflinity of C4H2+ is 10.17 eV and 12.62 eV in the ground and excited

18

states, one would expect a greater charge transfer interaction to take place
in the more polarizable argon matrix. This interaction will also be stronger
in the excited state.

In order to get sharp spectra with argon as a matrix, one must work
with cations that have smaller electron affinities ( < 10 ev). The guest-host
interactions are not strongly sensitive to the vibrational state of the guest,
and since the interactions are not as great for the ground state of a
molecular cation as they are for excited electronic states, argon may be
quite suitable for infrared studies that probe the ground state directly. This
fact is documented by the fact that the IR spectra of a number of molecular
cations have been reported and shifts in the vibrational spectra from the gas

phase values are indeed quite small, especially for a neon matrix.42
C. Vibrational Relaxation

Following the absorption of radiation into various vibrational levels in
an electronic excited state, a number of processes can occur. Consider the
fluorescence process, which can follow one of two paths. The system can
fluoresce from the vibrational level initially populated, or it can relax to the
v=0 level in the excited state before fluorescing. The first pathway is
observed for many gaseous systems, whereas matrix isolated species tend
to follow the second pathway.

Vibrational relaxation is considered to be a process that occurs only
for large molecules with a high density of vibrational states; however,
vibrational relaxation can occur even for small matrix-isolated molecules
because of the energy sink the matrix provides.42 When a molecule is in an
upper vibrational state in the electronic excited state, the vibrational energy
is usually lost to the lattice vibrations. This phonon energy transfer process

19

is faster than the radiative "process" and results in emission from the v=0
level of the excited state. Although vibrational relaxation greatly simplifies
the resulting emission spectrum, it limits the amount of spectroscopic

information obtainable.

2)

Chapter 2: Instrumentation for Matrix Isolation of Mass-Selected Ions

In order to better define the nature of the ionic species in the matrix
environment, an instrument designed to trap mass-selected cations and
anions emanating from quadrupole mass spectrometers, in a low-
temperature inert gas matrix has recently been constructed and tested.
The ions are co-deposited with excess argon until sufficient numbers are
present to be characterized by laser-induced fluorescence (LIF), to obtain
their electronic spectra, and/or Fourier transform infrared (FTIR)
spectroscopy, which can aid in structural determination of the ion of
interest. An obvious advantage of this instrumentation over other methods
employed to isolate charged species is that W ions can be
deposited, including isotopomers.

The instrument was designed for the purpose of determining the
structure of positive ions of interest to mass spectrometrists, as well as for
investigating the role that counter-ions play in the effective trapping of ionic
species in inert cryogenic hosts. The necessity for, or the role of, oppositely-
charged ions in approximately equal number to that of the cationic species
in matrix isolation experiments has not been established to date. The
instrument described here uses two sources of mass-selected ions, one for
cations, the other for anions. Ions in these two ion beams can be deposited
simultaneously to reduce charging of the matrix. This instrumentation is
extremely versatile, and should permit the spectroscopic investigation of a
wide variety of ionic species, and establish the role of counter-ions in the
spectroscopic analysis of cations by matrix isolation.

The new instrumentation has been successfully constructed and

tested; its efficacy is demonstrated here by the observation of LIF spectra of

21

CSz” isolated, from a mass-selected ion beam, in an Ar matrix. Described
below are the design requirements of the instrument, their
implementation, and the demonstration of its operation. Initial results

from vibrational spectroscopic studies are reported in subsequent chapters.

LDesignGoals

To ensure successful completion of this project, certain design
criteria of the instrument must be met: 1) The vacuum system must be
designed to house two mass spectrometers and a cryostat, as well as to
provide optical ports for examination of the ionic species via LIF and
infrared spectroscopy; 2) Since infrared spectroscopy is inherently less
sensitive than LIF, rather long deposition times may be required to obtain a
vibrational spectrum; therefore, a clean vacuum system with low base
pressure is imperative. The pump on the experimental chamber must be
able to compete with the cryostat (which is itself a vacuum pump) in
removing the neutral gas molecules (which are required to form ionic
species) that are introduced into the ion source of each mass spectrometer;
3) The instrument should be equipped with a way to monitor the residual
gases in the chamber (rather than simply measuring the chamber
pressure); 4) There should be some means to measure the ion flux
impinging on the substrate; and 5) The temperature of the matrix
substrate should be measurable and controllable. A schematic illustration
of the experimental chamber designed with these goals in mind, in its
FTIR configuration, is shown in Figure 2.1. The important subsystems of

the instrument are described in the following sections.

 

 

 

 

 

 

 

 

 

 

 

a
E

 

 

 

 

Figure 2.1: Schematic diagram of the experimental chamber constructed to
isolate mass-selected ions in a low-temperature inert gas matrix for subsequent
structural analysis via infrared or laser-induced fluorescence spectroscopy.

[(A) - CaF2 windows on conflat flanges, (B) - cryostat, (C) - CsI substrate at

15 K, (D) - residual gas analyzer (RGA), (E) - mass-selective ion sources,

(F) - neutral gas inlet, (G) - electrical feedthroughs, (H) - linear motion
feedthrough for tilting mass-selective ion source away from window,

(1)- cryopump, (J) - sapphire window for fluorescence observation, and (K) -
matrix gas inlet.]

II. Vacuum System

Design of a vacuum system to operate in the high or ultra-high
vacuum region is not a trivial task. In order to achieve pressures <10'6 torr
in the system, great care must be taken when considering the construction
materials for the chamber and components within the chamber, the seals
to be used, the dimensions and shape of the chamber, the type and size of
pump to use, and the kind of device to use to measure pressure. Described
in the following sections are the components utilized in the design and

construction of the experimental chamber.

A. Vacuum Chamber

Knowledge of the basic properties from which vacuum systems are
fabricated is essential for maintaining low base pressures. Because a large
pressure differential exists between the vacuum and the outside world, the
chamber must be constructed from a material of high mechanical
strength. This material should be relatively easy to machine and join by
welding, brazing, soldering, or demountable seals. Gas release
(outgassing) from solids at low pressure is one of the most troublesome
properties of materials used in vacuum applications. The significance of
reducing the outgassing load from the vacuum system cannot be over-
emphasized. There are several potential sources of gases and vapors in a

vacuum system.49
Vaporization
Backstreaming
Leaks (real and virtual)
Permeation
Difl‘usion
Desorption

24

These are illustrated in Figure 2.2. W usually occurs from
components inside the vacuum chamber; therefore, materials with low
vapor pressures should be utilized. W is the transfer of
vacuum pump fluids from the pump to the chamber. This can be
minimized by use of oils with low vapor pressures, traps, baffles, and
proper operating procedures. Wk: are small holes that are frequently
found around the seals. WM: arise from small pockets
of trapped air inside the vacuum. Virtual leaks are practically impossible
to detect. Care must be maintained when assembling the components that
are housed within the vacuum chamber to avoid trapping small amounts of
air between two parts. W is a three-step process. Gas first
adsorbs onto the outside wall, diffuses through the wall material, and
desorbs off the interior wall. Diffusion can be defined as the transport of
one material through another. Gas dissolved in the wall of the vacuum
chamber diffuses to the interior surface and can desorb into the vacuum,
contributing to the total outgassing of the system. This phenomenon arises
from the pressure gradient between the chamber wall and the internal
vacuum. Desolation is the release of gas adsorbed on the interior wall of
the vacuum chamber. These gases may arise from permeation, diffusion,
or from gases that adsorbed on the chamber surface while it was exposed to
the atmospheric environment. (Any solid surface has a weak attractive
force for at least a monolayer of gas or vapor.) The gas load from
vaporization, backstreaming, and leaks can be eliminated by proper design
and construction of a high vacuum system. It is much more difficult to
eliminate the load from permeation, diffusion, and desorption. The
adsorbed gases consist typically of a large amount of water vapor and some

oxygen and nitrogen.

Vacuum Chamber

 

Internal leak

 

Realleak

 

 

 

Figure 2.2: Schematic diagram illustrating several sources of gases and vapors

in a vacuum system.

 

Desorption

Vaporization

 

__\_+,_,_

 

 

Backstreaming

Pump

 

Difl‘usion

Permeation
(—

 

 

 

 

 

 

as

Polished vacuum-baked stainless steel is the material of choice for
high vacuum systems. Its outgassing rate is relatively low compared with
other materials, and it provides the structural integrity needed for vacuum
system design. Further reduction of outgassing by baking is very important
in reaching the lowest possible pressure. Warming of the chamber with
heating tape or a heat gun speeds the desorption rate and helps to remove
the gases from the chamber walls. Care must be taken to keep from
warming the chamber to a temperature so high that melting or
vaporization of the seals or joints occur. This would result in loss of
vacuum and could damage the components inside the chamber. Those
components should possess low outgassing rates as well. Solder and solder
flux should be avoided in high vacuum applications, due to their high vapor
pressure and outgassing rate. Bare wire should be used inside the
vacuum; if insulation is necessary, teflon coated wire is recommended.

The seals and joints employed in vacuum systems are an important
consideration. Elastomers, such as Viton and Buna-N, are used to form 0-
ring seals for static, sliding, or rotary seals. O-ring seals are made by
compressing the elastomer between two flat surfaces or in a confined
groove of rectangular or triangular cross section. More common to ultra-
high vacuum applications is the copper gasket for use with conflat flanges.
The "knife edge" bites into the copper gasket, forming the seal, when the
mating plates are tightened. Figure 2.3 shows how conflat flanges are used
to join vacuum components. More permanent seals are made by welding or
brazing metal parts together, and should be utilized inside the vacuum
chamber where possible. (Welds on the outside can cause virtual leaks

within the vacuum system.) The correct configuration for welded joints is

 

 

 

 

 

   

Weld Copper Gasket

 

 

 

l Vacuum I

Figure 2.3: Illustration of: A) conflat flange, B) cross section of conflat flange,
and C) conflat flange with copper gasket and weld joint.

%

also shown in Figure 2.3. Vacuum accessories are welded to conflats and
joined together with bolts and nuts.

The experiments described in this dissertation are performed in an
8" diameter stainless steel vacuum chamber, equipped entirely with conflat
flanges, and designed for ultra-high vacuum operation. Ports (configured
with conflat flanges) are provided for the entrance and exit windows for the
light source, two ion sources, purified and pre-cooled matrix gas, a
Faraday plate on a translational mount, electrical feedthroughs, a
roughing valve, a high vaccum gauge, and a window for observation of the

matrix.

B. Vacuum Pump

Because long deposition times are anticipated, a "clean" vacuum
system is a necessity to minimize interference from trapped neutral species
in the vibrational spectrum. Thus the chamber is evacuated using a
Leybold-Heraeus (L—H) RPK 1500 Series 8" cryopump, with a L-H RW3
water-cooled compressor unit, which can be isolated from the experimental
region with a pneumatically-operated gate valve (L-H, #DN 200 CF) to
facilitate access to the instrument without venting the entire system.
Unlike diffusion-pumped systems, cryopumps provide, essentially, an oil-
free vacuum environment. The fast pumping speed (2000 Us) also assists
in the removal of neutral species from the chamber.

The fast pumping speed achieved by cryopumps is due to the
pumping mechanism. Cryogenic pumping is defined as the entrainment
(condensation) of molecules on a cold surface. In principle, any gas can be
pumped, provided the temperature is sufficiently low. Most modern

cryopumps are closed-cycle systems and use the Gifford-McMahon

Q

refrigeration cycle (with helium gas) to provide continuous cooling of the
cold surfaces. The Gifford-McMahon cycle operates on the thermodynamic
principle that gas expanding against constant pressure does work; hence,
the surfaces in contact with the gas are cooled. The main working
components in a cryopump are a compressor (which compresses the
helium for subsequent expansion), and a cold head that contains a piston
(displacer) enclosed in a cylinder which moves the helium through a
regenerative heat exchanger. The compressor is connected to the cold head
with two gas lines. The cooling process takes place in two stages. The first
stage reaches temperatures of ~50 Kelvins (K), while the second stage can
attain temperatures as low as 10 K.

Helium closed-cycle refrigerators can effectively remove most gases
from the vacuum chamber. The vapor pressure of all gases, except helium,
hydrogen, and neon, is less than 7.5 x 10'12 torr at 20 K.50 In order to
reduce the vapor pressure of He, H2, and Ne, activated charcoal in thermal
contact with the second stage (10 K) for W pumping (the process
of having the cold charcoal adsorb the gases instead of condensing them) is
often employed. A cross-section of our cryopump is shown in Figure 2.4.
The radiation shield and baffle are cooled to ~50 K by the first stage and are
used to condense water vapor and hydrocarbons. The surfaces of the
radiation shield and battle are blackened to prevent radiant heat from being
reflected onto the cryopanels. The second stage produces temperatures
between 10 and 20 K and is used to cool the cryopanels which remove the
gases that are not condensed by the first stage. The charcoal on the inner
surface of the cryopanels adsorbs He, H2, and Ne. The pump is equipped
with a pressure relief valve in case the gas load reaches a pressure > 1

atmosphere. Also shown in the figure is a hydrogen vapor pressure

 

 

 

 

 

 

 

 

 

 

 

 

10" Conflat Flange
Radiation an-Baflle
Shield
en __ Cryopanels
Fore-Vacuum
Port X t E
\ Relief Valve
let S
”1' Hydrogen Vapor
_-/Pressure Thermometer

 

 

Helium Gas /

Connections

 

 

 

 

Figure 2.4: Cross section of a Leybold-Heraeus RPK 1500 cryopump.

31

thermometer which reports the temperature when the cold head is between
14 and 27 Kelvins.

The block diagram of the vacuum chamber and pumping system is
shown in Figure 2.5. The mechanical pump evacuates the cryopump to a
pressure of ---10'2 torr (measured with a thermocouple gauge) when
roughing valve 1 is opened. This valve is a L-H #298-22-B1
electropneumatic valve. Care should be exercised in this roughing
procedure to avoid backstreaming of oil vapor into the cryopump. This can
occur when the pressure gets too low (<10‘3 torr) or when no trap is used.
The mechanical pump on our chamber is a L-H Tri-Vac #D16A. A L-H
#85415 molecular sieve trap is attached to the roughing line to prevent oil
backstreaming. W to «:10'2 torr, W
W and the cryopump is turned on with a switch on front of the
compressor. Under no circumstances should the cryopump be turned on
when the roughing valve is open! Oil backstreaming can occur if this is
done! Once the cryopump reaches its operating temperature (1 hour), the
chamber can be evacuated by opening roughing valve 2 (Hughes Aircraft
#HVV-150-2) and allowing the mechanical pump to evacuate the chamber
down to 10 qu . This roughing valve is then shut, and the gate valve is
then opened. High vacuum (10'6 torr) is attained almost instantly once the
gate valve is opened.

The cryopump retains all the gases it removes from the vacuum
chamber; hence, the cold surfaces will eventually become saturated and
lose their efl‘ectiveness. To offset this phenomenon, the pump must be shut
off so the gases can be released. System shut-down and regeneration
commences by closing the high vacuum valve and removing power to the

cryopump. If a great deal of water vapor was trapped, this could liquify and

 

 

 

 

 

 

 

Residual
Gas
Analyzer
¢ Chamber
Gate valve
Cryopump
Temp

 

 

 

 

Roughing valve

/ Thermocouple gauge

a)!

 

 

J Mechanical

 

'2)—

LM”

Roughing valve

Figure 2.5: Block diagram of vacuum sytem for experimental chamber.

3'3

drip to the bottom of the pump (or onto the charcoal adsorber). Therefore
gas flushing, preferably with a hot (140°F) gas, combined with rough
pumping is recommended when regeneration is performed; however, if
this flushing is done too soon after power is shut off, the cryopump may still
be sufficiently cold to trap the flushing gas. This is also true of mechanical
pump oil. Precautionsshouldbetakentoinsurethattheroughingvalveis
not opened too quickly after power-down! Thus, gas flushing with helium,
combined with a heater attached to the cold stages is a better method. Since
our cryopump lacks a heater and flushing line, the best method for
regeneration is to simply allow the pump to warm-up overnight. At times
the pump was backfilled with dry nitrogen or argon. This procedure
shortens the regeneration time by approximately 10 to 20 percent; however,
due to the absence of a proper backfill valve we used the leak valve to inlet
the gas. This eventually led to clogging of the leak valve, and therefore this
procedure was terminated. Future modifications for the cryopump would
include the addition of a regenerator kit (which heats the radiation shield
and cryopanels) and/or a flushing line for helium purging/roughing of the
pump.

The vacuum chamber rests on a movable table that facilitates
transportation within and between the laboratories which house the
infrared and LIF spectroscopic equipment. This table frame is constructed
of 1.5" square steel tubing, whereas the top is made from 1" aluminum.
The gate valve rests on the table top and supports the chamber above, while
the pump hangs beneath the gate valve through a hole cut in the table top.
The table was designed such that the legs wrap around the Bomem FTIR
spectrometer; this provides easy access to the back port, which is utilized in

the infrared experiments.

C. Premise gauge

Low base pressures (10‘9 torr) are routinely achieved with this
vacuum system, as measured with a Dycor model M100 residual gas
analyzer (RGA). In addition to measuring total pressure, the RGA can
measure the partial pressures of the individual gases in the vacuum
system; this enables the operator to continuously or periodically monitor the
relative amounts of matrix and sample gases present in the chamber
during an experiment.

RGAs are best described as miniature quadrupole mass
spectrometers. Their total length is approximately five to eight inches. A
block diagram of the Dycor RGA can be found in Figure 2.6. It consists of a
power supply, electrometer, control unit, and analyzer head. The mg:
ml! produces the voltages necessary for operation of the analyzer head
and converts the signals from the electrometer into digitized signals for
transmission to the control unit. The W converts the ion current
at the detector to a voltage that is sent to the power supply. The W
is used to monitor and adjust the RCA parameters. The data can be
manipulated and displayed on the monitor. The W51 is housed in
a 160 mm long, 38 mm diameter tube that contains a high-efficiency
electron ionization (E1) source, injection optics, a small quadrupole mass
filter, and a detector. (The Dycor M100 contains a Faraday plate detector;
however, units are available which utilize electron multipliers). The
housing is terminated at both ends with a 2.75" conflat flange; one end is
connected to the electrical feedthrough which joins to the electrometer, and
the other end can be mounted to the vacuum chamber. Electrons emitted
from a hot filament in the source ionize the residual gases (also producing

fragment ions) in the vacuum chamber, and the ion optics directs them into

Analyzer Head

 

V

    

Optics

PF 1

E1 Source Quadrupoles

Electrometer

 

 

 

 

 

 

 

Control Unit

 

 

 

 

 

 

 

 

 

Power Supply

Figm‘e 2.6: Block diagram of the Dycor M100 residual gas analyzer (RGA).

$

the quadrupole mass filter where they are separated according to mass-to-
charge ratio; the selected ions impinge on the detector. By scanning the
quadrupole, a mass spectrum of the residual gases in the vacuum can be
obtained. This is important in identifying any contaminants (such as air)
in the system. Typical operating pressures for residual gas analyzers
range from 10" to 10'13 torr. Pressures above 10" torr cause the filament to
burn out; however there is a protection circuit in the M100 that shuts the
filament current off if the total pressure exceeds a pre-set value. (This
value is set at 10‘5 torr for the experiments reported here). The RGA is
interfaced to an AT-compatible computer, which can control the RGA or be

used for data storage or manipulation.
D. Interlock

The instrument is protected with an interlock system that prevents
damage to the ion sources, RGA, and cryopump should a power failure or a
vacuum leak occur. An interface box plugs into the 15-pin connector on the
back of the RGA controller and is connected to the interlock control box.
The interface contains the necessary circuitry for RGA control of the
interlock. Should the total pressure of the vacuum system (as shown on the
RGA "total pressure" reading), or the partial pressure of an individual gas
exceed that of the operator's pre-set value, the gate valve will be shut,
protecting the cryopump from overload. The front panel of the interlock
control box is shown in Figure 2.7. It contains a lighted power switch, a
system failure light, an override switch, and nine circuit breaker power
switches. The control box is powered by 220 volt, 50 amp service. The
vacuum and instrument components plug into the back of this box for

protection by the interlock system. In order to understand how the

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Switch # Description

Mechanical pump
Rough valve
Cryopump

Gate valve

RGA

Cryostat
Temperature
Auxiliary 1
Auxiliary 2

 

chQO’UthNH—i

Figure 2.7 : The front panel of the vacuum interlock.

 

 

 

38

interlock works, it is first necessary to understand the detailed procedure
required for pump-down (evacuation) of the instrument. Figure 2.8
outlines this process. Operator decisions are indicated by the dark boxes.

The features of the interlock are described below.
1. Power to the control box is supplied by a power switch.

2. The cryopump cannot be turned on until power is removed from
the foreline valve.

3. The interlock contains an "override" switch that must be engaged
in order to turn on the cryopump, gate valve, and residual gas
analyzer.

4. Once the RGA is on, and the pressure falls below the pre-set value,
the overide switch is disengaged and the vacuum system is protected.

5. If the pressure becomes too high, power is removed from switches
4 - 9.

6. If a power loss occurs in the building, power to switches 3 - 9 will
not return when building power returns.

7. If the cryopump fails, power to switches 3 - 9 is removed.
8. A red indicator light signifies a "system failure".
9. A fuse protects a short in the control box from reaching the RGA.

10. A bread-board and 15 V power supply is included in the control
box for further expansion of the interlock.

The schematic diagrams for the interlock control box, RGA interface,
and the changes made to the RGA control board are shown in Appendix
11.1

 

MILLERTIME

 

 

 

 

 

 

 

Rough to
0.01 torr

 

 

 

 

 

 

 

 

 

 

 

 

 

Turn on
Cryopump
Temperature Rough chamber Close
<14K to 0.01 torr roughing valve
Turn on Open gate
ion gauge valve

 

 

  

Pressure No power to
< 10-5 torr? switches 3 - 9!
YES YES

Building

 

     

Figure 2.8: Flow diagram of pump-down procedure. Dark boxes indicate
a decision or a task to be implemented by the interlock control box.

 

III. Cryostat

The sample window (substrate) is maintained at 15 K with an Air
Products Displex 202 cryostat which has been adapted for use in ultra-high
vacuum by the addition of a 4.5" conflat flange which interfaces the cryostat
to the experimental chamber. An illustration of the cryostat is shown in
Figure 2.9. The base plate of the cryostat motor was welded onto a 3.5" o.d.
stainless steel pipe upon which the conflat was attached. A ten-pin
electrical feedthrough is attached to the pipe via a mini (1.33") conflat
flange that provides the connections necessary for temperature control of
the matrix. Not shown in the figure is a radiation shield that screws into to
the first stage (50 K) and protects the second stage (15 K) from thermal
radiation from the walls of the chamber. The second stage is connected to a
copper block where a germanium resistor which is used to monitor
temperature and a heater for controlled warming of the matrices are
embedded. These two components are part of the temperature controller
and will be discussed in the following section. The copper support, upon
which the substrate window is attached, is secured to the copper block.
Indium gaskets are used between the copper components to maintain
thermal contact with the second stage. This insures that the matrix
window is sustained at 15 Kelvins. The matrix is deposited on a 2.5 cm
diameter sapphire substrate in the LIF studies, whereas CsI or an
appropriate alternative window is employed for the infrared experiments.
The substrate is held tightly against the copper with a face plate and
screws. Indium gaskets are also used between the substrate and the
copper.

It is worthy of note that the copper components are fastened together

with screws; therefore, care must be taken to safeguard against virtual

 

Cryostat Motor

Electrical Feedthrough

 

 

 

Conflat

illlll \

 

 

 

 

 

 

 

 

First Stage (50 K)

 

 

 

Temperature %

controller

' es
wrr Vb— Second Stage (15 K)

 

 

Zener diode
(heater) N

Matrix window

/

Capper support

 

Figure 2.9: Displex 202 cryostat modified for ultra-high vacuum.

leaks that may arise due to trapped air in the screw holes. All screws in
the vacuum chamber have holes drilled through the screw axis, or drilled
perpendicular to the screw hole through the block, to remove any air

trapped within the holes.

A. Temperature controller

The Cryo-Temp controller was designed by M. Raab and is used for
controlled warming of the matrix window. It operates on a feedback
mechanism which senses the temperature at the cryo-head with a
germanium resistor. The voltage obtained from these resistors is amplified
and converted to a current which drives the zener diode heater mounted in
the copper block. (Wood's metal, a low melting-point alloy, was used to
attach the Ge resistor to the cryostat.) The temperature setting switch on
the front panel is calibrated in Kelvins to facilitate direct setting of
temperature.

The germanium resistor has a negative resistance characteristic
which makes it quite useful, since resistance increases as the temperature
drops to absolute zero. This resistor can be used to measure temperatures
below 60 K, although it is non-linear. The resistor chosen for this
application is model #N2 supplied by Scientific Instruments, Inc. (West
Palm Beach, FL). When resistance versus temperature is plotted on a log-
log scale, the graph is almost a straight line between 10 and 30 K. (See
Figure 2.10.)

A simplified diagram of the circuit is shown in Appendix 11.2. A
constant current source of 100 uA is fed into the germanium resistor. The
resulting voltage is amplified by an instrumentation amplifier (gain = 10)

and is further amplified with a logarithmic amplifier. This voltage is

43

compared to that supplied by the temperature setting switch, which has
been calibrated to give a voltage corresponding to the desired temperature
via the calibration curve shown in Figure 2.10. When the difference in
voltage indicates the cryostat is too cold, the comparing amplifier develops a
voltage which causes current to flow in the zener diode heater via an opto-
isolator and a current amplifier; when the cryo-head is too warm no
current flows, which allows the refrigerator to cool the head, thereby
maintaining the temperature at the desired value. A front panel meter
indicates the magnitude of the heater current (2A max). A light emitting
diode (LED) indicates when the temperature of the cryostat is at least 10%
higher than the dial setting. A BNC connector at the back of the controller
provides a way to monitor the voltage (x 10) at the Ge resistor. A detailed
schematic of the Cryo-Temp controller as well as the cable wiring diagram
can be found in Appendix 11.3.

The calibration curve shown in Figure 2.10 was obtained by using a
cryostat from Professor D. G. Nocera's laboratory. That cryostat contains a
gold—chrome] thermocouple that measures temperature. The germanium
resistor was mounted on the end of the cryostat and the circuit shown in
Figure 2.11 was used to record the voltage at various cryostat temperature
settings. The Ge resistor has three colored wires that are attached to the
casing. The black wire supplies the positive excitation current (100 uA), the
yellow carries the positive voltage output, and the blue carries the negative
voltage output. The voltage measured at the meter is converted to a
resistance for the calibration curve. The temperature of the cryostat can be
determined by referring to the calibration curve. This circuit was used
more often than the Cryo-Temp controller for recording the temperature of

the substrate. The temperature controller has not been utilized for

 

E
o

A. AA‘ All All AAIAAAALJAAAALl‘A‘AA

Resistance (0)
s a 8 §

[0
0

All.

 

 

 

O

 

Temperature (Kelvins)

 

Log<R>
’5.

 

 

 

 

Figure 2.10: Germanium resistor calibration curve.

 

 

__i_) Atmosphere Side ' Vacuum Side
I
I
L— I
15 V
' F Current |
meter I
: B180]!

 

 

1

Ge
Blue I

Yellow

6

Volt meter
fv\
V

 

————-b-

Figure 2.11: Circuit diagram used in Ge resistor calibration.

 

46

warming of the matrix because it was not yet completed at the time the
experiments were performed. Any warming experiments were done by
removing power to the cryostat, then turning it back on. It is be
recommended that any annealing of the matrix be done using the Cryo-

Temp, since a controlled, slow warming is then possible.

RMauixgasline

The vacuum line used to introduce the argon into the instrument
chamber is shown in Figure 2.12. The line is pumped by a 3" air-cooled
diffusion pump backed by a Welch Duo-Seal #1376 rough pump. A liquid
nitrogen trap helps prevent backstreaming. Pressure measurements are
made with a G100P ion gauge tube combined with a Veeco model RG-830
controller to measure low pressures (-~~10'6 torr), as well as a Celesco DM2
digital capacitance manometer to measure pressures in the torr range (1-
1000 torr, with DP-30 transducer) and to monitor the argon flow rate. (The
volume of the vacuum line with the 5 liter argon reservoir is 6.024 liters; the
line itself is 0.801 liters). The manometer measures pressure by the change
in capacitance that occurs between a diaphragm, which is referenced to
vacuum, and a fixed electrode. Only occasional calibration is necessary
since a difference in pressure is recorded when monitoring the argon flow
rate. The transducer has a low internal volume (4 x 10'3 in3) and is
constructed from stainless steel, which makes it ideal for working with
corrosive gases.

Matheson high-purity argon (99.9995%) was used as the matrix gas
in all experiments described in the following chapters. A Vacuum
Generators MD7 high-precision leak valve (right angle valve with 1.33"
conflat flanges) is used to regulate the argon flow rate through 1/8"

47

dean—~33 as» fine you on: 8353’ ”3d earn

 

 

 

 

 

 

 

 

. «24
v“. \
mo )
5008.334
3:38.250
on.
a.

 

 

 

 

 

 

 

 

ufi a. _
as> on..xHH”.H.H.H.M.H.mm.mwnHm.“.Hmmmfimfl .... mm
9 a. a .

03:82.52;

 

 

 

 

 

48

stainless steel tubing. The leak valve is mounted onto a liquid nitrogen
(LN2) Dewar for pre-cooling of the matrix gas, should it be necessary.
Several sprayer designs were tried in attempts to deposit the argon onto the
substrate in a uniform fashion; however, the best design, within the space
limitations of our experiment, seemed to be the stainless steel tubing
pointing down onto the substrate. The argon flows through the sprayer and
condenses on the cryostat window, thus forming the host matrix for species

emanating from the mass-selective ion source.

N. Ion generation

The desired experiments can employ two sources of mass-selected
ions, one for negative ions and the other for positive ions, which are to be
trapped in the argon matrix. These sources of mass-selected ions are
modified residual gas analyzers, which have been described in detail
elsewhere,51:52 and will be covered only briefly in this section. The
modifications made to the VG Anavac-2 RGA (which was selected to
produce a beam of negative ions) are illustrated in Figures 2.13 and 2.14.
Figure 2.13 shows the RGA in its original configuration. It contains an E1
source, quadrupole, and detector mounted on a 2.75" conflat flange much
like the Dycor RGA described in Figure 2.5. The changes made to the
Anavac are illustrated in Figure 2.14. Similiar modifications were made to
a Dycor M200 RGA, which serves as the source of positive mass-selected
ions. In our application the RGA detectors have been removed so the ions
may be directed towards the cryostat window. Other major modifications
include rotation of the original RGA configuration by 180°, the addition of a
sealed source region (with a conductance-limiting aperture and a gas

inlet), and focusing and deflection optics to direct the mass-selected ion

 

 

 

    

 

   

/

Extraction
Lens

 

 

 

 

 

 

 

<~ xx
0 \I §
5 Quadrupole
= 0 ——
Conflat I
Faraday
Flange T Plate

 

Figure 2.13: Original Anavac RGA Source configuration.

Filament Gas inlet
Feedthrough

  

Sealed high pressure
source enclosure

 

Conductance
Limiting
Aperture

  
  

Quadrupole

Ion optics for beam
"""'"" focusing and deflection

Figure 2.14: Modified Anavac RGA source configuration.

51

beam onto the spectroscopic window, while discriminating against neutral
species which emanate from the source. Several factors make these RGAs
ideally suited for this application: 1) they are small so they can be readily
adapted to our chamber; 2) they are designed for high-efficiency ionization;
3) they yield high current ion beams (>10'9 A); and 4) they exhibit unit mass
resolution for a mass range of 1 to 200 daltons for the modified Dycor M200
source utilized to select cations, and 1 to 60 daltons for the VG Anavac-2-
based negative ion source. The positive and negative ion sources are
mounted separately on 6" conflat flanges and are directed off-axis from the
substrate to reduce the number of neutral species following a "line-of-sight"
path from the ionization region to the cold window. The ions exiting the
quadrupole are then focused with an electrostatic lens, deflected onto the
substrate, and are trapped in an argon matrix at 15 K. The source flanges
contain electrical feedthroughs for the lenses, source, and quadrupole rods,
as well as a gas inlet for introduction of neutral species.

A 2.5 cm diameter Faraday plate, mounted on an MDC LM-133-2
linear motion feedthrough, can be moved in front of the cryostat window to
record the ion current impinging on the substrate, and then withdrawn
during sample deposition and spectroscopic examination. This feature
allows the operator to periodically adjust and re-optimize the source optics
and deflection lenses to maximize the current at the substrate. The
Faraday plate is positioned 6 mm from the cryostat window so the current
reading at the plate is fairly indicative of the number of ions hitting the
window. The 2.5 cm plate is spot-welded to a ceramic insulator which is

connected to the linear motion feedthrough as shown in Figure 2.15.

 

"I

Ceramic insulator Mini conflat flange

1:- —‘\' T

Faraday plate Linear motion feedthrough

 

 

 

 

 

 

 

 

 

 

Figure 2.15: Faraday plate mounted on linear motion feedthrough.

53

V. Laser~induced fluorescence detection of matrix-isolated ions

A schematic diagram of the laser-induced fluorescence experiment
is shown in Figure 2.16. Tunable radiation from a NszAG-pumped dye
laser (Quanta-Ray DCR2A/PDL-2/WEX) enters the instrument and is used
to excite the sample. The laser beam entrance and exit ports are fitted with
0.8m long stainless steel arms containing light baffles and terminated by
Brewster angle quartz windows to eliminate much of the stray laser
radiation. The dye laser is scanned with a computer-controlled stepping
motor through the region of interest and the total fluorescence from the
sample is monitored as a function of excitation wavelength at an angle of
22° with an EMI model 9558B photomultiplier tube (PMT). Cutoff filters can
be placed in front of the PMT to effectively discriminate against scattered
laser light, which might otherwise lead to saturation. The PMT signal is
integrated (SRS 250 Boxcar Integrator), digitized, and transferred to an
IBM-XT compatible microcomputer for storage and further manipulation.

The LIF signal is then averaged and displayed on the monitor.

A. Lasers

The LASER laboratory at MSU has two NszAG-pumped pulsed dye
lasers (PDLs): one for red dyes and the other for use with blue dyes. The
NszAG lasers (DCR2A) are identical; however, the blue dye laser
amplifier is in a side-pumped configuration and has a delay line for
suppression of amplified spontaneous emission (ase), while the red dye
laser contains an end-pumped amplifier and is without a delay line. The
pulse widths of the DCR2As are 8-9 nanoseconds (ns) and the repetition rate
can be varied from 1 to 30 Hertz. NszAG lasers have a fundamental

output at 1064 nanometers (nm); however, with a harmonic generator the

 

 

 

 

 

Trigger

 

 

 

 

 

 

 

 

 

 

 

 

PMT S'E
[ l J l ' BOXCAR I Power Suppfl

 

 

 

 

 

 

 

 

 

_l-l_ Gm mrscaxroa
Oscilloscope
._ vi
a 2
§ Filter
5 \
LIF "\
HM _
—_ U —_

 

Figure 2.16: Schematic diagram of the instrument in its laser-
induced fluorescence configuration.

second, third, and fourth harmonics (532 nm, 355 nm, and 266 nm) are
available.

The blue dye laser was utilized for most of the work described in this
dissertation. It is pumped with 355 nm vertically-polarized radiation (420
mW) that results from harmonic generation from the fundamental output
of the DCR2A. The radiation is used to pump the dye in the oscillator, pre-
amplifier (pre-amp), and side-pumped amplifier (amp). The PDL output
wavelength varies with the dye used, but usually spans 10 - 20 nanometers
for each dye. Coumarin 440 and 460 (Exciton, Cincinnati, Ohio) were the
dyes used with most of the reported experiments. Two 1 liter dye/methanol
solutions are mixed to concentrations specified in the Exciton catalogue.
One of these solutions is poured into the oscillator dye circulator and the
other into the pre-amp/amp circulator. Each dye is circulated to prevent
burning of the solution by the 355 nm radiation. The output from the PDL is
directed to the experimental station (10 meters away) with a series of
prisms, mirrors, apertures, and a long focal-length telescope. The spot size
from the dye laser is usually 2-3 millimeters in diameter by the time it

reaches the cryostat window.
B. Chamber windows

To minimize PMT saturation by background scattering, special input
windows were designed to eliminate most of the scattered light. The laser
radiation is directed into the vacuum chamber through a quartz input
window sealed with Torr-Seal?" to a 1" diameter pyrex tube cut to
Brewster's angle. A 1" Cajon adapter joins this window to a 2.75" conflat
flange that has a 1" diameter stainless steel tube welded to it. Inside this
stainless steel tube are two 45 mm diameter apertures through which the

 

56

laser light passes. The total length of the input arm is ~25 cm. The laser
beam exits the chamber through the same type of window, sealed at
Brewster's angle to a 0.6m long, 38 mm diameter stainless steel tube. The
long exit arm also contains two apertures, which minimize reflected laser

light at the exit window.

C. Detector&electronics.

The EMI 9558B photomultiplier is a 2" diameter tube with a flat-faced
spectrosil end window. The cathode is a 44 mm tri-alkali type with a
spectral range that extends from 8500 to 3000 Angstroms. There are 11
venetian blind dynodes that contain Cs/Sb secondary emitting surfaces.
The PMT is housed in a black enclosure that abuts to the chamber
observation window. A high voltage power supply provides the dc voltage
for the PMT.

Although great pains have been taken to reduce the scattered
radiation at the entrance and exit ports of the chamber, there still exists a
considerable amount of scattered light due to reflections off the matrix
window. Therefore it is necessary to use cut-off filters to eliminate much of
this scattered laser radiation (bandpass filters pass too much light in this
application). A Schott GG475 filter was used to filter the Coumarin 440 dye
radiation that was employed in our work, since the transmittance is < 10'5
for wavelengths below 450 nm.53 Even through this filter, laser light is still
observable with the PMT. When the PMT became saturated, the signal
began to broaden in time. The result was that no fluorescence was
observable with the integrator. In addition to a proper optical design, a
number of factors can be employed to prevent the photomultiplier from

becoming saturated; these include adjustment of the laser excitation

 

57

wavelength, PMT input voltage, laser power, and use of a second, or
different, cut-off filter.

The fluorescence signal must be extracted from the PMT current in
order to record a total fluorescence spectrum. A Stanford Research
Systems (SRS) model SR250 gated integrator and boxcar averager module
were employed to obtain the fluorescence signal; a model SR245 computer
interface module was utilized for computer control of the detection
electronics. The SR250 integrator consists of a gate generator, a fast gated
integrator, and averaging circuitry. The gate generator is triggered by the
variable output of the Nd:YAG laser operated at 10 Hz. The gate generator
provides an adjustable delay from a few nanoseconds to 100 milliseconds
before it generates a gate having widths adjustable between 2 ns and 15 ms
(Typical values for the gate delay and width in the LIF experiments
reported here were 45 ns and 50 ns respectively.) The signal is integrated
only when the electronic gate is open. The signal can then be averaged over
1 to 10,000 samples and the output sent to the computer. This technique is
useful for extracting small signals from noisy backgrounds. As more
samples are averaged, the fluorescence signal converges to a mean value
while the noise is averaged to zero. A Tektronix model #2230 digital storage
osciloscope was used to set the gate delay and width so that the open gate
overlaps (in the time domain) with the PMT signal output. (Note: a
"reference" PMT {RCA #1P28] samples a portion of the laser beam and
passes through the same detector electronics to compensate for shot-to-shot
variations in the laser output.) The dye laser is then scanned over a
selected wavelength range with the stepping motor and the total

fluorescence spectrum is aquired and stored in the computer.

"'7

 

To increase the solid angle of the fluorescence captured by the PMT

and maximize the fluorescence collection efficiency, a collection lens was
placed in the vacuum chamber in front of the photomultiplier tube. This
was a 30 mm diameter, 55 mm focal length lens epoxied to a 1" aluminum
tube and secured to the observation port with a set-screw. The lens may
have improved the fluorescence collection efficiency; however, it also
"improved" the scattered light collection efficiency, which caused PMT
saturation! Therefore this lens was removed and not used for subsequent

experiments.

V1. Fourier Transform Infrared (FTIR) Spectroscopy

Infrared spectra were recorded with a Bomem model DA3-01 FTIR
spectrometer. The DA3 laboratory model FTIR features five available
output beams; two sample compartments in the front, and three external to
the instrument which exit the beam switching compartment at the sides
and the rear. The experimental chamber described above is designed to
interface to the back port. Special Optics were designed by our laboratory in
collaboration with Bomem Inc.54 to direct the infrared radiation from the
spectrometer through our chamber to an auxiliary detector. The optical
configuration is illustrated in Figure 2.17. The 3" diameter collimated
infrared beam from the DA3 encounters mirror M1 (90° off-axis paraboloid,
6" effective focal length, 3" aperture) and is focused and deflected 90°
towards mirror M2 (90° off-axis ellipsoid, 2" and 22" effective focal lengths,
1.25" aperture), where it is directed through the chamber. The beam
diameter at the cryostat window is matched to the substrate size. Mirror
M3 (90° off-axis paraboloid, 0.75" focal length, 1.25" aperture) focuses the

radiation onto a 1 millimeter diameter Infrared Associates (W22-1)

av

 

 

Bomem DA3 FTIR

 

 

 

 

 

 

M1

 

 

 

 

 

 

 

 

 

 

 

 

 

Detector

 

 

 

Figure 2. 17: Optical configuration for infrared analysis of mass-selected ions.

fl)

mercury-cadmium-telluride (MCT) detector. The chamber side arms and
Brewster angle windows that were used for the LIF experiment are
replaced with CaF2 (or KBr) windows mounted on conflat flanges for the

FTIR work.

The Bomem FTIR is equipped with a globar source, Michelson
interferometer with a KBr beamsplitter, and a helium-neon laser which is
used for the dynamic alignment system unique to the Bomem DA3 model
FTIR spectrometers. To facilitate the removal of air within the FTIR
spectrometer, o-rings are included on all the panels that house the
spectroscopic components, to facilitate either evacuation or purging with
dry nitrogen. To remove air from the interfaces between the FTIR, detector
housing, and our instrument, air-tight plastic bags have been secured to
these components for purging with dry nitrogen obtained from the bleed-off
of the MSU Chemistry Department liquid nitrogen tank. This procedure
insures that the infrared spectra of any atmospheric gases will not

interfere with the sample spectrum.

The LNZ-cooled MCT detector, obtained from Infrared Associates,
rests in an aluminum housing (purchased from Bomem) that contains
mirror M3. The housing is supported by an aluminum plate that is
attached to a table containing roller bearings for height and side
adjustment of the detector. A Bomem variable-gain MCT pre-amplifier is
used to amplify the signal from the detector before it is processed by the
DA3.01. (An extension cable connects the pre-amp at the back to the input
at the front of the Bomem.) The pre-amp must be "matched" to the detector
to provide the correct current which operates the detector. This process is

w as

61

explained in the Bomem manual and the supplementary material provided

with it.

Alignment of the infrared beam exiting the back of the FTIR
spectrometer through our instrument is a critical factor in obtaining the
best infrared spectrum possible. The optics that direct the beam out the
back have been optimized and should not have to be adjusted. The cryostat
chamber should be maneuvered around the IR beam. The cryostat has
been designed such that radiation passing through the two outer ports will
pass through the center of the substrate. The table that supports our
instrument contains casters and adjustment legs for proper positioning of
the instrument. Since the infrared beam is not visible to the human eye,
the quartz halogen lamp in combination with the mylar beamsplitter is
used for this alignment procedure. The detector table and housing are also
adjusted with this white light source by using the "dummy detector" for
initial alignment. (Mirror M3 should be approximately 22" from mirror
M2). Once this rough alignment is completed, the globar source (and KBr
beamsplitter) can be replaced for the final alignment procedure. The
process that seems to work best for this fine alignment procedure is to run a
"test" phase spectrum with 1000 scans. This gives the operator sufficient
time to adjust the detector to maximize the "peak % gain" on the front
panel. The aperture may have to be decreased several times before the

detector is in its best position (largest signal).

An absence of signal (peak % gain = 0) during this alignment
procedure could be due to a number of factors. The detector must be filled
with LN2 and be connected to the pre-amp and the front input with the

extension cable. The BNC input cable on the front of the Bomem must be on

 

62

input #3 when using the back port (input #2 for front port). A final item to
check is that the globar source is on with the KBr beamsplitter in place.
Once the detector is in position the instrument is ready for operation.

Several experiments were performed utilizing laser-induced
fluorescence and FTIR spectroscopy as detection methods. The
experimental details and results are described in Chapters 3 and 4 of this
dissertation.

Chapter 3: Preliminary Matrix Isolation/FTIR Studies

Once construction of the instrument was completed, the choice of a
molecular cation for investigation via MIIFTIR began. It was important to
find a cation with an ionization energy below 11 eV, but also one that could
be generated with a significant ion current. The choice of cations was
limited to those with molecular weights below 60 daltons because the Dycor
mass-selective ion source (MSIS), which extends the mass range for our
experiments to 200 daltons, was not completed at the time; however, the
Anavac MSIS was available for use. These first experiments, described
below, were performed Mm co-deposition of a negative ion (counter-ion)
and helped us to understand the instrumentation and the experimental
parameters that will eventually govern the isolation of cations in a low-

temperature matrix.

I. NO+ DEPOSITION STUDIES
A. Introduction

The nitric oxide cation was chosen as the first experiment because of
its low electron affinity. The low electron affinity results because nitric
oxide, the precursor, contains an unpaired electron that is readily lost (IE =
9.264 eV).55 The low ionization energy for NO is unusual for small diatomic
molecules. Because it is a gas, nitric oxide is easily handled using
standard vacuum line techniques for facile introduction into the ion source;
however, it must be handled with extreme caution since it is highly toxic
and is rapidly oxidized by air to nitrogen dioxide. Although nitric oxide is

noncorrosive, in the presence of oxygen and moisture corrosive conditions

 

1i. AWL-W 6.31:.-
l

64

will develop due to the formation of nitric and nitrous acids. It is
recommended that stainless steel be used for materials that could come into
contact with the gas.56 The NO cylinder regulator is made of stainless
steel, as was the transfer line to the source inlet of the MSIS.

BExper-imentalSection

The NO (98.5%) used in these experiments was supplied by
Matheson. A glass bulb, filled to ~1 atm, was used as a reservoir for NO in
the ion deposition experiments. Appendix 111.1 lists the various
experiments that were performed in the N 0* study. High purity argon was
deposited onto the matrix substrate in all the experiments. In order to
determine which vibrational bands belong to the neutral species, NO was
co-deposited with argon (M/G=5000/1)through the vacuum manifold. The
ion deposition experiments were performed by introducing =5-6 mtorr of NO
into the ion source of the Anavac MSIS. Ionization with 70 eV electrons
formed ions that were extracted from the source region, injected into the
quadrupole, focused, and deflected onto the matrix substrate with an
energy of 12 eV. Currents of approximately 1 nA were observed at the
Faraday plate. The infrared spectrum was recorded at a resolution of 0.5
cm‘1 over the range of 4000 - 900 cm’l. Photobleaching experiments were
performed with a small Eveready® flashlight. The residual gas analyzer
was not turned on during the experiments to avoid saturating the cryostat

window with excess electrons.

C. Results and Discussion

Several vibrational bands were observed after deposition of NO+;

bands not observed in the Ar/NO co-deposition. These peaks were

I. IV '. 2.1K “'11.“.1

1'- ' " '_

$

positioned at 1863.3 cm'l, 1832.3 cm'l, 1776 cm'l, 1296.1 cm4, and 1282.8
cm'1.. Since these peaks occurred only in the NO+ deposition, it seemed
possible that they might be due to the isolated NO cation, or to some
molecular species formed from deposition of "high energy" ions. Figure 3.1
illustrates the effect of photobleaching the matrix for various time intervals.
The band at 1872 cm'1 is due to neutral NO. Photobleaching experiments
resulted in a decrease, and eventual disappearance, in the 1832 cm’1 and
1296 cm-1 (not shown) absorptions. The 1863 cm‘l, 1776 cm'1,and 1282 our-1
hands were unaffected, which indicates that they are most likely neutral
species.

To help ascertain whether the two vibrations that decreased upon
photobleaching were due to NO*, an experiment was executed whereby all
the experimental conditions used to produce these absorptions were held
constant, except that the M818 was turned off so that NO+ was not being
deposited. All four bands appeared disaffirming the theory that NO+ was
trapped in the frozen matrix! If these peaks are not due to an ion, then
where did they come from? To answer this question, one must investigate
the differences between direct deposition of the neutral and the deposition
experiments where NO was deposited via the MSIS. The infrared spectra
from the two experiments indicate that the water and carbon dioxide
absorptions were much greater in the M818 deposition experiments. This
was most likely due to the gas inlet system for the M818; since the inlet was
evacuated with a mechanical pump prior to the experiment, water and
carbon dioxide are not removed as effectively as in the co-deposition
experiments, where a diffusion pump is used to remove the residual gases
prior to the experiments. Another comparison between the two

experiments reveals relatively greater absorption of the NO neutral band

 

F Irish"?!

.2889 023 so 8389; .83 2 a as seen?

8 368332... 5 sense 8 868388 6 .osofia 8 s. 858232... < 3 figs a

pace as ca .02 do 838% 2 5% £3588... 83 - o8m 69a 38% 832 a.” .55
6:55:95?

8.2 8.3 8%: 8m: 88“

 

 

 

eoeqrosqv elitists};

67

(187 2 cm'l) in the ion deposition experiments. This probably results from a
large amount of NO reaching the matrix substrate from the ion source of
the MSIS. In the neutral co-deposition experiments the gaseous mixture
was directed onto the window by the sprayer; however, the gas introduced to
the MSIS diffuses into the chamber and can be frozen onto the cryostat
substrate. Experiments (source-on or source-off) where NO is deposited
through the MSIS must result in a lower Ar/NO ratio trapped on the
matrix window than those of the co-deposition experiments, where an
Ar/NO ratio of 5000/1 was used. This could also explain the greater water
and carbon dioxide content in the matrix . This observation is deliterious to
our goal of obtaining infrared spectra of ionic species in the matrix, since
absorptions due to the ubiquitous neutrals have the potential to mask the
bands of ions in certain experiments.

Are the bands at 1863, 1832, 1776, 1296, and 1282 cm'1 due to the
formation of adducts of nitric oxide with water, carbon dioxide, or another
NO molecule during deposition on the low-temperature substrate? Why are
two bands (1832 cm'1 and 1296 cm°1) affected in the photobleaching
experiments? To test whether a NO-NO adduct was formed, a second
argon/NO co-deposition was performed with the Ar/NO ratio at 700/1. The
infrared spectrum recorded after several hours revealed the same bands at
1863, 1832, 1776, 1296, and 1282 cm4. The 1863 cm'1 and 1776 cm‘1 bands
have previously been assigned as the NO dimer isolated in an argon
matrix.57v58 The assignment for the other three bands has proven to be
more difficult; however, they are probably due to some unstable neutral
species. The two bands that respond to photobleaching may be due to the
same species, while the absorption at 1282 ch is most likely a more stable

species. Figure 3.2 illustrates the effect of photobleaching experiments on

 

meadows—£3649 30:33.5 .«c met—£8 mm no»? @0280.— one?
coach :3 25. deflation gonna—:8 .«c nude; 3 has @0288 983
8E 835 on... 8388.8 38: 023 co 555m "a..." .5»:

30:85:96?
am: ON»: 83 $63

Low —
1 d

c)

 

 

 

 

\
CD
33
N
3
eouaqrosqv situate};

 

Q

the 1832 cm'1 and 1296 cm'1 absorptions in the Ar/NO co-deposition
experiment. Fately, et al. have reported the infrared spectra of a number of
frozen oxides of nitrogen, and has observed a band at 1282 cm'1 which was
assigned to the staggered form of OzN-NOg with Vd symmetry. They also
observed bands at 1829 cm'1 and 1290 cm'1 which were assigned to the
unsymmetrical dinitrogen tetroxide molecule ONO-N02 (other names for
this dimer are nitrosyl nitrate [ON-ONOz], nitryl nitrite [OgN-ONO], or
simply iso-N204). These vibrations are very close to the bands observed in
our experiment and, until further investigations prove otherwise, will be
assigned to iso-N204. Table 3.1 contains a summary of the vibrational

bands (from 2400 cm'1 to 900 cm'l) of all species observed in the NO study

 

Table 3.1 Identification of infrared bands in the NO experiments.
Dem’gnation Wavenumber Assignment

a 2345 C02

b 2339.1 002

C 2218.4 N02

d 1871.7 NO

e 1863.3 cis-(NO)2 [sym. stretch]

f 1832.3 ONO-N02

g 1776.3 cis-(NO)2 [asym.

stretch]

h 16885 H20

i 1633, 1630, 1632 H20

j 1611.2 N02

k 16062 H20

1 1296.1 ONO-N 02
m 1282.8 N 204 [Va symmetry]

 

70

(Appendix 111.2 contains the corresponding spectra of the peaks designated
in Table 3.1). The assignments have been made (where possible) with the
help of the references previously cited, as well the work of St. Louis and

Crawford59 and Loewenschuss and Givan.°0
D. NO+ Study Conclusions

The NO+ deposition experiments proved to be very helpful in
understanding and coordinating experiments with our newly-constructed
instrument; however, the experiments were plagued by many problems,
the main one arising from the corrosive conditions that developed in the
source region of the MSIS. Filament burn-out made it impossible to extend
an NO+ deposition longer than five and a half hours under the current
conditions. This severly restricted the number of ions impinging on the
cryostat substrate, hence limiting the number of ions that could be trapped
in the low-temperature matrix. It may be possible to extend the filament
life by using a thoriated-iridium filament, but these were unavailable to us
from the manufacturer of the MSIS. Attempts were made to spot-weld a
thoriated filament obtained from an old ion gauge, but the weld would not
hold on the filament post of the MSIS source.

Some very interesting neutral chemistry has taken place in these
experiments. A more thorough investigation, utilizing isotopically-labeled
precursors, might provide clues to the identity of the unassigned bands in
this experiment; however, since these are neutral species and outside the
scope of this work, namely, to isolate and identify mass-selected ions in low-
temperature inert gas matrices, this NO study was not pursued.

Incidently, Jacox and Thompson have recently detected the infrared
absorption of NO+ and NO', as well as that of (NO)2+, (NO)2', N20+, and

 

71

N N 02' in neon matrices“:62 These species were produced by the
windowless discharge technique as described in Chapter 1 of this
dissertation. A band at 2345.2 cm:1 was assigned to NO+ which is very close
to the gas-phase value derived from data given by Huber and Herzberg.63

11. CHszlg+ DEPOSITION STUDIES

A. Introduction

In an effort to increase the likelihood of trapping a mass-selected ion
in the frozen argon matrix, a search for the "ideal" ion began. The "ideal"
ion should possess a relatively low electron affinity, have a substantial
abundance in the EI mass spectrum, and have an m/z value less than 60
daltons. The methaniminium cation (or methylenimmonium ion),
CH2NH2+ was selected because this ion seemed to be well suited for this
experiment due to its low electron affinity (- 6 eV). Moreover it is easy to

make, since the E1 spectrum of many amines contain this ion at m/z = 30.

B. Experimental

Propylamine (Aldrich) was chosen as the precursor to CH2NH2+ and

was used without further purification. Table 3.2 represents the mass

spectrum of propylamine. The CH2NH2+ fragment at m/z 30 is the most

abundant peak and provides an ion current in the nanoamp range from the

Anavac MSIS.

 

 

12 0.17 27.5 0.45 45 0.04
13 0.26 $ 12.02 50 0.06
14 0.78 29 2.68 51 0.28
15 2.89 m m 52 0.60
16 0.36 31 1.66 53 0.08
17 0.45 $ 0.06 54 0.49
18 2.96 37 0.40 55 0.08
19 0.03 {E 0.73 56 1.17
21 0.01 13 2.68 57 0.18
24 0.07 40 0.92 58 2.34
% 0.34 41 5.53 59 10.60
$ 2.40 42 3.33 60 0.43
26.5 0.55 43 2.20
27 6.17 44 1.17

The experiments and conditions that were utilized in these studies
were similar to those described for the NO+ deposition. Pr0pylamine is not
as corrosive as nitric oxide in the source region of the MSIS and allowed
our depositions to continue for several days; however, the inlet system for
the MSIS had to be re-designed without O-rings due to their attack by
propylamine. Appendix 111.4 contains a list of all the experiments

performed involving CH2NH2+ deposition; the results will be discussed in

the following section.

C. Results and Discussion

Table 3.3 contains a list of many of the vibrational bands (and the
species to which they are assigned) that appear in the spectrum (Appendix
111.5) of the cryostat substrate after propylamine was deposited through the
ion source of the Anavac MSIS for approximately 43 hours to provide a

 

Table3.3 Infraredbandsinthespectrum‘Appendkm)obtainedfmm
deposifionofpmpylaminethmughtheAnavacmass-selecfiveionsoume.

 

Designation Wavenumber Species
A 3776.5 H20
B 3756.9 H20
C 3731.0 H20
D 3711.6 H20
E 29743 pr0pylamine
F 2934.0 propylamine
G 2883.6 prOpylamine
H 28562 propylamine
1 2345.4 C02
J 2339.6 C02
K 1768.5 propylamine
L 1722.0 propylamine
M 16242 H20
N 1608.3 H20
0 1593.5 H20
P 1573.2 H20
Q 1557.0 H20
R 1483.0 propylamine
S 1471.3 propylamine
T 14608 propylamine
U 1361.8 propylamine
V 1304.7 propylamine
W 1216.8 propylamine
X 10763 propylamine
Y 1034.0 propylamine
Z 974.4 propylamine

74

reference of the vibrational bands due to the precursor and the residual
gases in the vacuum chamber. Although the spectrum is very rich and
contains an abundance of vibrational bands, these bands are not of interest
to us since they are neutral species. There were, as expected, many peaks
that can be assigned to propylamine, as well as water and carbon dioxide
absorption bands. Any new vibrational bands observed in the ion deposition
studies would be due to ionic species in the matrix, or to exotic neutrals
formed following ion deposition onto the low-temperature matrix.

The first deposition of CHgNH2+ continued for 41.5 hours and
revealed no new vibrational bands other than those neutrals found in the
propylamine deposition. In an attempt to provide a counter-ion (anion) in
the matrix, the RGA was left on during the second CHzNH2+ deposition.
This procedure resulted in concurrent deposition of electrons from the RGA
and CH2NH2+ ions from the MSIS. It was hoped that the electrons might
migrate through the matrix until they found a suitable electron acceptor
species to attach themselves to, thus providing a source of counter-ions.
(Electron currents on the order of 0.1 nA were measured the Faraday plate
when it was positioned in front of the substrate.) The infrared spectrum
obtained after 24 hours uncovered four bands not present in the neutral
deposition experiment. The absorptions, appearing at 3306 cm'l, 1123 cm'l,
1064 cm'l, and 1059 cm°1, showed no response to photobleaching of the
matrix. (Afier observation of these absorptions, the spectra from the first
deposition was examine more closely and found to contain the same
absorptions, but with diminished intensity.)

Could these new vibrational bands be due to CHzNH2+ trapped in the
argon matrix? The vibrational bands at 1123, 1064, and 1059 cm'1

 

75

correspond to those found by Jacox and Milligan65 for ‘03, ‘09, and '07 of
methylenimine, CHzNH, which were very strong absorptions in their
study. They prepared eleven isotopic species of methylenimine by mercury-
arc photolysis of methyl azide isolated in argon, nitrogen, and carbon
dioxide matrices. Identification of all the vibrational fundamentals were
made except the N-H stretching mode, which they calculate to be- at 3296
cm'l. Figure 3.3 shows the D7, 03, and ‘09 vibrational bands of CHzNH.
Other bands assigned to methylenimine are either buried under different
neutrals or are too weak to observe in our experiments. Could the
absorption at 3306 cm-1 be the N-H stretch of methylenimine? An
interesting assignment of a weak band at 3308 cm'1 HCN was made by
Jacox. This band was very weak compared to the CHgNH bands in their
experiments. However, since the absorption at 3306 cm'1 in our
experiments (Figure 3.4) is considerably stronger than the identified
CH2NH bands, it more likely arises from HCN. Formation of HCN almost
certainly proceeds through a different mechanism in the ion deposition
experiments than in the methyl azide photolysis, which could explain the
stronger absorption in this work. It is also possible that the band at 3306
cm'1 in the ion deposition could be ‘01 of CH2NH2+, which has not been
previously observed; however evidence that sufficient ions have been

deposited for infrared detection is lacking.
D. CH2NH2+ Stmb' Conclusions

It is clear that electrons from the residual gas analyzer as well as
mass-selected CH2NH2+ ions both played a role in the formation of
methylenimine. Deposition of CH2NH2+ with the RGA turned off resulted

in an absence of CH2NH vibrational bands, as did deposition of propylamine

 

 

76

 

.AEZNEUV 058303508 350a ma as. find £9 Kc. 30395030 A003 05 ”530:0
893 a go a 533%.. «£335 333% 83.8% 3383 a...” charm

Eggngakw
caps cm: 34: - can.“

0 0 fi
_ -

.. cred

\cit/i
.2. 47/- -
2 1....
N 3
h a
I). 23

.. cad

 

 

eoueqiosqv

 

 

.. sad

 

a- 83

 

 

800.3 Ego .«0 850.506 083.50.“ 00.80008 3-80 Swmécci 85.30090 v0.8.9: “v.0 0803

00038880203
8.00 8.8 8.3

E

 

 

 

 

‘ 0.8mm

£5 8%

St
17

 

 

.. :6

.. “Nd

mad

eoueqiosqv

.. wad

. and

- and

78

through the MSIS with the RGA on. In order to isolate CH2NH in the
matrix, the RGA and the MSIS must be turned on; however the
mechanism of methylenimine formation is still unclear. Is it possible that
the electrons are neutralizing the CH2NH2+ ions that might have been
trapped in the matrix environment? Neutralization of CH2NH2+ would
most likely result in formation of CH2NH in the matrix. If this were the
case, then CH2NH2+ should be observable in the absence of electrons. Or
are the electrons interacting with some exotic neutral species during
condensation, with subsequent formation of methylenimine? A worthwhile
experiment might be to detune the MSIS so that no ions are impinging on
the cryostat substrate, but the source would still be on. Observation of
methylenimine in the absence of cations, but with the RGA on, would
suggest that CHgNH is formed from an unknown neutral species.

It is possible that the absorption at 3306 cm'1 is due to an N-H stretch
in CH2NH2+. A more thorough investigation of the ion deposition
conditions and utilization of isotopic precursors are needed in order to

determine to which species the 3306 cm‘1 vibration belongs.

 

79

Chapter 4: Spectroscopic Identification of Matrix-Isolated CS2+
I. Laser-induced fluorescence studies of CS2*
A. Introduction

The initial goal of this work was to obtain sufficient numbers of ions
within the matrix for the spectroscopic analysis of mass-selected ions. The
deposition of NO+ and CHgNHg+ into the argon matrix, described in the
previous chapter, provided some interesting results; however, confirmation
of an ion isolated in a low-temperature matrix was not accomplished. It is
possible that the concentration of isolated ions was below the detection limit
for infrared analysis. Laser-induced fluorescence is inherently more
sensitive than infrared spectroscopy and can afford some structural
information. However, for many ionic species the necessary fine structure
is absent when argon is employed as the inert gas host; especially for small
cations, neon matrices must be utilized to provide well-resolved spectra.66
Temperatures near 4K must be achieved to use neon as a matrix;
unfortunately, they are not accessible with our present cryostat. The best
approach within the limitations of our instrumentation was to confirm the
presence of a mass-selected ion in argon, to maximize that signal through
LIF measurements, and then attempt an infrared study of the same cation.

Several factors limit the choice of cations to study. For purposes of
comparison, a system that had been well defined both in the gas phase and
in matrices was required. Most LIF/M1 experiments have employed neon
hosts because charge transfer interactions between the guest cation and the
host in the more polarizable argon matrices can or may severely perturb

and broaden the sample absorbance and fluorescence spectra.42 The

 

80

limited mass range of the mass-selective ion sources (200 daltons for the
Dycor source) restrict our studies to small cations, which in many
instances have fairly high electron affinities. For example, the electron
affinity of CS2+ in the excited electronic state of the LIF transition (12.7 eV),
is sufficiently close to the ionization energy (IE) of argon (15.76 eV), that the
guest is rather perturbed; on the other hand, neon has an IE of 21.6 eV and
effectively traps CS2+ with minimal perturbation to the electronic
spectrum!“68 It has been suggested that cations with electron affinities
greater than about 11 eV cannot be trapped effectively in argon matrices,69
although there have been some exceptions reported.18 Species with lower
IEs have been found to give fairly well-resolved LIF spectra.70 Also, the
mass-selective ion sources are currently limited to ion formation by electron
impact ionization of gas phase neutral precursors. Thus, our selection of
CS? as the test cation involved some compromise, particularly in light of
the limitations imposed by the argon matrix. It was clearly a practical
choice since the LIF spectra of 082+ in the gas phase and in inert matrices
are well established;71’67:58 082* is the most abundant ion in the 70 eV mass
spectrum of 082; the molecular ion is in the mass range of the our mass-
selective cation source; sufficiently high currents of 082+ can be generated;
and the excitation wavelength range required to produce LIF spectra is
readily accessible with tunable lasers available in the MSU LASER

laboratory.
B. Experimental

In the feasibility experiments, a beam of 082+ ions was co-deposited
with excess argon on the cryostat substrate held at 15 K. Matheson high-

purity argon (99.9995%) was used as the matrix gas in all experiments

 

81

described below. A Vacuum Generators MD7 high-precision leak valve
was used to regulate the argon flow rate. The precursor, CS2, was
purchased from Aldrich and subjected to a number of freeze/pump/thaw
cycles before use. Carbon disulfide was introduced into the ionization
region of the Dycor MSIS through a Granville-Phillips Series 203 leak valve,
where it was ionized by electron bombardment (70 eV), injected _'into the
quadrupole, mass-filtered, focused, and deflected onto the window. Typical
CS2+ currents (measured at the substrate window) in the experiments
reported ranged from 0.75 - 2.0 x 109 amperes, with ion kinetic energies of
68 eV. Deposition continued until sufficient numbers of ions were obtained
for spectroscopic characterization via infrared or laser-induced
fluorescence spectroscopy. Experiments were performed with and without
simultaneous deposition of negative charge carriers; this will be discussed
in more detail later.

A Quanta Ray DCR2A NszAG-pumped pulsed blue dye laser (10Hz)
was utilized for the excitation radiation in the LIF experiments. It was
found that operating the dye laser in the oscillator-only configuration
provided ample power to produce CSg+ fluorescence. Typical pump powers
of 400-420 mW produced dye laser outputs of 0.1 to 20 mW depending on the
excitation wavelength. Coumarin 460, Coumarin 440, and Stilbene 420
were the laser dyes used to provide tunable excitation of the CSg+
fluorescence, although Coumarin 440 was used routinely. Figure 4.1 is an
illustration of the power curve for Coumarin 440 measured at the output of
the dye laser. The laser was scanned with a computer-controlled stepping
motor that utilized the same SR265 software that is used to control the
boxcar integrator. The undispersed fluorescence signal from the sample

on the cryostat substrate was collected and monitored as a function of the

 

€32

 

.asaaxzacuaHra2nQHUA—
h0m792835ucanyoaaneaconuua.xzaad0nu0Pnnrmzziuchv.800_cvvaaufiusso suvofinwm—

00:85:95?

cccvu cocmm cccua OOOHN
- p P p p -

r..¥¢~

(MW) Jamod

 

[uAWON

83

excitation wavelength. Appendix l'V.1 contains a list of all the experiments
performed in the LIF studies of C824:

Several experiments were performed to determine the optimum data
collection conditions under which the CSz+ fluorescence signal could
integrated. Attempts were also made to minimize the scattered light inside
the matrix chamber. These studies will be discussed further in the

following section.
C. Results and Discussion

Elimination of the scattered laser radiation proved to be a difficult
task due to the construction of the stainless steel vacuum chamber. Many
experimental parameters such as 1) laser alignment, 2) detection
electronics, 3) photomultiplier voltage, 4) laser power, 5) cut-ofl‘ filters, and
6) an appropriate reference signal had to be Optimized in order to detect the
fluorescence from species trapped on the cryostat window. With the series
of apertures and Brewster-angle windows described in Chapter 2, we were
able to reduce the pmt scattered light signal to <100mV (at a pmt voltage of -
IOOOV). Alignment of the laser was critical when the substrate was in
place for an experiment. It was imperative that the laser light which
reflected off the front surface of the cryostat window came back out the
same path it entered, while the light passing through the window exits the
center of the long baffle arm which was equipped with apertures and a
quartz Brewster angle window.

The boxcar integrator can be used to discriminate against scattered
laser light by setting the delay to collect data after the laser has fired. In

most cases this delay was set to ~50 nanoseconds; however, if too much

 

84

scattered light strikes the pmt, broadening of the pmt signal can result;
thus it remains important to minimize scattered laser light.
Photomultiplier afterpulsingn»73 proved to be a major obstacle to
extracting the fluorescence emanating from species trapped on the cryostat
substrate. Afterpulses of the pmt were influenced by a number of factors
such as laser misalignment, pmt voltage, laser power, and the use of
longpass (cut-off) filters. Afterpulsing effects included a broad laser pulse
observed on the oscilloscope which at times lasted several hundred
nanoseconds, thus hindering detection of the fluorescence signal. By
reducing the amount of scattered light in the chamber, one can more
effectively monitor the fluorescence signal. Two studies which characterize
the dependence of the pmt signal broadening on instrument parameters
are depicted in Figures 4.2 and 4.3. Figure 4.2 shows the relationship
between the voltage on the pmt and the scattered light signal (peak to peak,
in mV) observed on the oscilloscope. The laser was set to 22,727 cm°1 and a
power of 13 mW. The cryostat window was in position and a GG475 cut-off
filter was place before the pmt to help filter the scattered laser radiation. If
the pmt voltage is kept below -1200V, signal broadening can be avoided.
Figure 4.3 illustrates the relationship between scattered light signal at the
pmt and wavenumber of the Coumarin 440 dye laser radiation. Afterpulses
occur when the cut-off filter no longer discriminates against laser light.
Figure 4.4 shows the relationship of the excitation and emission
spectra of CSZ“ in argon to the cut-ofi' frequency of the Schott GG475
longpass filters that were employed in the CS2 experiment. The spectra
have been adapted (by permission) from work reported by Bondybey,
English, and Miller.67 Their spectra were obtained following in situ

vacuum-ultraviolet photolysis of 082 cocdeposited with argon. Since most of

 

‘I

85

5002- 08.000 0.880 00300300 E
.803 a 8:0 .038 08 38.8 2: a? was 5.00 0a 383 0a .838 83 a an 3282 as:
0000800 0009—. 00.300 0000—3 080 .0> 080 05 00 0030000 0000 I»: 000033 Harv 0.803—

01 «03:5 0.20

08m 003 83 2x: 000 80 8*

(Am) [Bufirs ma

 

 

86

 

.033 000.500— 05 .3 b0 03 0&5.— 20 00 an»: 000$ 05 00:3 00:80 mam—000.000:
03000 .3 0a 3 Ba 2 0. aka 08.: a Ba >82- 3 Ea 20 a? 38:8 2o: 3% use.
.0003 0%.. 05 .00 0090500203 0b 080 05 «a $000003: E0 «A»: 0000— 30303 ”ma. 0%

 

00905000203
8ch 8000 ccoum 080a
. . F p . . c
I 80
acacia; 1&0\\V ..

I
8
N
(All!) [units um

I
.3.

 

l8?

87

000500 05 .00 0000000000 .3 NM 000000.000 0000.0 0000000 0003 0000000 +Nm0 03000 001 00 m4.

0003b 00 80.50000 00.00.3000 05 03:0 3 0000000 09000 0.3 0000— c: 000000000 00S. A0300 00000000 05 .00
00000 0000 90030000000 80m 0m 0003 05 00 “0000000.. 00 000 00000000030 05 .00 :0 000001 “305 .0008 0000.900—
05 .00 50005 E00000 003 3 00w00 E +Nm0 .00 0.00000 00000000 “.00 003000000 05 .00 030003303 #0 005mm

83 N v 00090002003

23 van ma NON SN 3 N 3 N 3 N NNN @NN cmN 3N

- q u q d d a a q q u 0 q .1!

.
\   ‘ /

000000003

 

 

 

 

088008050. 0.80 .0300 03:0

88

the ng+ fluorescence is to the red while the excitation radiation is to the
blue of the cutoff wavelength, efficient discrimination against scattered
laser light results. Also shown in Figure 4.4 is the range of the Coumarin
440 laser dye scanned to obtain the LIF spectrum reproduced in the
following illustration.

A portion of the excitation spectrum of CS2+ in argon, obtained over
the 22300 - 23000 cm'1 range, is shown in Figure 4.5. The spectrum is
representative of the first successful isolation of a pre-formed ion in a
matrix with this instrument. The LIF spectrum can be compared to the
excitation spectrum in Figure 4.4, obtained from CSZ“ ions formed by
photoionization of the neutral precursor co-deposited into the matrix
environment. Unfortunately, the spectral features in the LIF spectrum of
C82+ in argon are not nearly so well-resolved as those of the ion in neon;67.58
nonetheless all of the fine structure and the spacing shown in the earlier in
situ Ar spectrum are perfectly matched in the LIF spectrum of matrix-
isolated mass-selected CSZ“ ions. The excitation bands centered around
21500 cm'1 and 22000 cm'1 in Figure 4.4 could not be effectively scanned
with our current instrumentation because of the inefficiency of the cutoff
filters at those energies. Both bands were readily observable under our
experimental conditions; however their shapes were rather perturbed due
to the scattered laser radiation.

Several control experiments were performed to ensure that the
observed CS2+ LIF signal originated from deposited ions. The observed
signal is not due to trapped C82, which is also present in the matrix,
because the neutral does not absorb at wavelengths longer than 380 nm to
which our excitation was restricted. Furthermore, fluorescence was not

observed when the deflection lens was flagged so that the ion beam no

 

89

hp

409E920000afi§£00409833.
3Y0oaaflhaaak0£0203nflfifiu00£§P500quv00m0009325”0056
80008 080 90.. 03. 053:8 503 one: 0.8 880 .. 8000 2:
026;500340+§Wu2250592008fi038820090880uaflfia"mégzaflfl

83 N v 0000000000003
HmN mNN bNN mNN MNN

 

manem
eouaosaxonm

 

90

longer impinged on the substrate during a deposition. In another
experiment a 2 nA beam of 6 eV Ar+ ions was co-deposited with a 5000:1
Ar:CS2 mixture. No fluorescence was observed even after prolonged
deposition times (18 hours); this demonstrates that CS2+ is not formed from
ion bombardment of neutral CS2 on the cryostat substrate. The well-
reproduced LIF signal was observable only after 5 or more hours of
continuous ion deposition with at least 0.75 nA of CSZ” current (as
measured in front of the substrate window). This corresponds to a flux of at
least 5 x 109 ions per second (1.8 x 1013 ions per hour).

It is worthy of note that the instrument described in this dissertation
was not optimally designed for laser-induced fluorescence experiments.
Scattered light is a serious problem due to the matrix (argon matrices
scatter more than do neon hosts) and reflections within the stainless steel
chamber. A better-designed LIF chamber would substantially increase the
sensitivity of the experiments as well as decrease sample deposition times.
Nevertheless, because of the high sensitivity of the technique, it seemed
reasonable to begin with LIF to optimize the many parameters (e.g. ion

kinetic energy) that influence effective trapping of mass-selected ions.
D. Role of counter-ions

One would assume that approximately equal numbers of positive and
negative species must be present within or proximate to a given matrix
environment in order that sufficient numbers of ions are trapped for
spectroscopic examination. This insures the overall neutrality of the
matrix. Yet the nature of the counter-ions has not been firmly established
in any cation matrix spectroscopy experiment reported to date. We have

carried out several experiments to address these questions. One involved

 

91

the co-deposition of CSz+ with Cl' from the mass-selective anion source. It
was anticipated that isolation of CI' in the dilute matrix might improve our
cation signal because a greater number of cations could be isolated upon
reduction of charging effects. Although more extensive studies are
merited, these experiments have to date been unsuccessful; no CS2+ LIF
signal could be detected after six hours of co-deposition of Ar, CSz“, and 01'.
Future experiments will involve alternate (layered) deposition of cations
and anions. As noted earlier, an RCA on the vacuum chamber allows us to
monitor pressure in the chamber during the course of an experiment.
When this RGA is on, an electron current of ~ 0.1 nA is measured at the
window. Since electrons can readily migrate through the matrix, it was
our hope that they would form counter-ions by electron attachment to
impurities in the matrix. However, no LIF signal was observed when this
gauge was turned on during an experiment, i.e. when Ar, CSZ‘“, and e' are
co-deposited. This suggests that the cations were neutralized by the
electrons during deposition, and therefore the RGA was switched off
during subsequent CSf depositions. Thus, while approximate overall
charge-neutrality must be maintained, by some mechanism which is not
yet understood, our best results with LIF are achieved when no attempt is

made to deposit negative charge carriers into the matrix.
E. Vibrational spectra implications

Does there exist a sufficient number of cations in the matrix to
acquire an infrared spectrum in an 8-hour experiment? In the LIF
spectrum, fluorescence with good signal-to-noise was observed following
five to eight hours of 082+ deposition. At an argon flow rate of 0.5 mmole/hr

and an average ion current of 1 x10‘9 amperes the rare gas-to-cation guest

 

92

ratio is about 10 million to one near the substrate window. Over an eight
hour deposition, assuming 50% trapping efficiency, there would be
approximately 2 x 1014 ions isolated in the matrix. Of this number, only
"-1013 ions are irradiated in the LIF experiment, since the area of the laser
beam (at the substrate) is only four square millimeters, while the matrix
window is one inch in diameter. Moreover, much of the fluorescence will
not be captured due to the disparity in the boxcar gate width (50 nsec) and
the long 082* excited state fluorescent lifetime (4 nsec).67 The infrared
beam, on the other hand, will interact with all of the ions trapped on the
CsI substrate, hence increasing the sensitivity by approximately two orders
of magnitude. The lower cross-sections for absorption of infrared photons
will also be partially offset by the longer integration times of the vibrational
spectroscopy experiments. Nonetheless, tests of the FTIR spectrometer
sensitivity (for analytes in solution) suggest that as many as 1016 absorbers
may be required before reliable spectra can be acquired, which indicates

that rather long ion deposition times may be required.
11. Infrared studies of 082*

After experimental conditions that lead to the successful capture of
082+ ions for LIF analysis were established, the instrument was re-
configured for vibrational analysis and interfaced to the FTIR
spectrometer. While we have not yet been successful in obtaining the
vibrational spectrum of mass-selected 082*, the initial results reported here
provide important information, when coupled with the LIF data, on the fate
of ions impinging on an inert gas matrix while it is being formed.

The LIF studies suggested that counter-ions need not be co-deposited,

so all FTIR experiments were performed without using the negative ion

 

93

source. A beam of 082* (> 10'9 A) was directed towards the substrate and
co-deposited with argon for 48 hours. Infrared spectra were recorded
periodically during the deposition, and many bands grew with time,
including absorptions due to matrix-isolated CS2. Several of the bands were
n91, observed when the same experimental conditions were maintained,
except that the quadrupole was not mass-selecting CSz“. This indicates
that these absorptions do not arise from neutrals formed in the ion source
and deposited onto the substrate. Could these features be due to ions in the
matrix? There have been a number of approaches reported to test for the
presence of ions in a matrix. The most common is direct comparison of the
spectrum to gas phase results when they are available. Photobleaching is
another method used by some scientists, whereby low-intensity visible
photons are used to remove electrons from the anions (counter-ions) in the
matrix.74 Since electrons can migrate throughout the matrix, they can
neutralize cationic species in a non-selective manner. Loss of signal after
photobleaching can thus indicate the presence of an ion.69 However,
electron photodetachment may require energies higher than that of visible
radiation; as a result, photobleaching is not always reported in ionic matrix
isolation experiments. A third method used to test for the presence of an
ion is a visual inspection of the matrix during warm-up. Luminescence
may result as positive and negative species combine on the substrate,75
although this is not definitive since recombination of neutrals/radicals can
also produce emission. Thus, the CSf/Ar matrix (48 hour deposition) was
allowed to warm to 40 K and then cooled down for further spectral analysis.
Weak luminescence was observed during this warm-up period.

Figure 4.6 shows a portion of the FTIR spectra obtained a) after the 48

hour ion deposition, and b) after the subsequent warm-up/cool down

 

94

 

0:888 35088 08 00 3 a 05883

03.00 00.008 0800 0.00 .00 00 3 00003 .40 0.0 .00 0000000 000000 0m000>0 00 00
000000000000 0000.— Q0. 0300 0000000000 003 0 80000005 00000000000000 040+Nm0
900300000 00000000000 0.80 0 .0.80 8: .. 80.00 0.50000 E 6.0. 00020.0

 

 

 

 

00x00000>02w
cmHH OVNfi ONmfi OOVH
furl/UPI o T .r w x
a i
8.032 a m
m.
m
m
0..
._ 0
mad "Md

 

9S

process. None of the peaks shown were observed in an equally long control
experiment with the detuned cation source. Table 4.1 lists information
from which the wavenumber of the infrared active antisymmetric
stretching (1)3) fundamental of 082+ and of 082 can be obtained. This mode
is well separated for the neutral and cationic species; the bending
vibrations of these linear triatomics lie below the infrared cut-off

wavenumber of the CaF2 chamber windows.

 

Tableuspecmscopicdatsmlstedtomesnfisymmeuicsuetshing
vibration for 082W.
Species Assignment 0(cm'1) Medium serum
C52+ 21)3 2418 Ne 67,68
203 2406 Gas 71
21)3 2401 Gas 76
082 03 1528.6 Ar 77

The tabulated data suggest that if C82“ is present, an infrared absorption
should be observed at =1210 cm‘l, although it would be expected to be
slightly red-shifted by the argon matrix.‘57»68 (A band at 1211 cm'1 has been
tentatively so assigned for C82+ trapped in neon.)42 No absorption can be
detected in this region; however, peaks at 1300, 1275.2, and 1270.9 cm'1 are
observed. Vibrational data obtained from gas phase measurements give
1)(CS+) = 1376 crn'1,78 and D(CS) = 1284 cm‘1.79 On this basis we infer that
the peaks in the 1270 cm'1 range are due to neutral CS trapped in the argon
matrix. The doublet attributed to CS in Figure 4.6a becomes a singlet when

 

96

the matrix is briefly warmed (Figure 4.6b); this may reflect the presence of
multiple matrix sites initially, which upon annealing change to provide a

more uniform matrix environment for the trapped species.

A. Discussion of infrared results

The combination of the infrared observations with the LIF data
provides a very important assessment of this system. The results suggest
that a fraction of the 082* impinging on the growing matrix remains intact
(since the LIF spectrum of C82+ was obtained), but that most of it is
converted into CS and S+/°. How is C82+ being converted into CS? We have
considered two mechanisms for this conversion.

One reasonable explanation might be that CS is the product of
collision-induced dissociation (CID). The C82+ ion, with a kinetic energy of
as much as 8 eV, may undergo CID as it approaches the matrix. The
lowest energy CID pathway, requiring 4.75 eV, is to CS and 8*. However,
the center-of-mass collision energy for even the first CSZ'WAr collision is <3
eV. (Other possible collisional processes, those leading to CS+ and S or to
CS, S, and Ar“, are energetically less feasible.) We note that Maier and co-
workers have successfully deposited C2+ and C2C12+ at ion energies higher
by more than an order of magnitude (150 eV) than those used here and
obtained the spectra of the intact ions.25 Much higher energies (340 eV)
were required before products of CID were detected. Thus, we believe that
the CID route from C82+ to CS would be very inefficient, and could not
explain the observed conversion of most of the ions to neutral CS.

The second proposed mechanism of CS formation is that the ion may
capture an electron and be neutralized. This would yield CS2*, which

contains ample energy to induce fragmentation to CS and S.

 

97

(Unfortunately, sulfur cannot be detected with infrared spectroscopy; nor
can the products of the other exothermic dissociative pathway, C and 82.)
What is the origin of the electron? The copper support for the substrate is
held at electrical ground; local fields due to ions in the matrix might result
in extraction of electrons from the metal. We note that as yet we have been
unsuccessful in obtaining a LIF spectrum of CScf in experiments when an
"insulating layer" of argon is deposited on the cryostat window prior to
initiating deposition of ions, which lends plausibility to this explanation.
The neutralization pathway is further supported by a preliminary
experiment in which mass-selected CS+ (rather than CSf) was co-
deposited with argon; a weak doublet at the frequencies assigned to CS was
observed.

Ionic candidates other than C82+ were considered for spectroscopic
analysis, in the context of the limited mass range of the RGAs and
restricted IE range available with argon matrices. The CH2NH2"’ ion (IE: 6
eV), was a good candidate for IR analysis, but not for LIF. As was
described in Chapter 3, we have co-deposited CHZNHZ+ (~1 nA, ion energy =
7-8 eV) with argon and observed the growth of bands in the infrared
spectrum which could not be attributed to the neutral precursor of the
cation, as the deposition proceeded over two days. Again, our preliminary
analysis suggests that these new spectral features are due not to the mass-
selected ion being deposited, but to a product of neutralization, CHzNH.65 If
CHzNHz+ captures an electron, energy equal to the IE (6 eV) is available.
Only 2 eV are required to form CHZNH and H from CHZNHz; thus this
neutral product of dissociative electron capture is certainly reasonable. We
are now optimizing the experiment for infrared analysis of ions with these

initial experiences and insights in mind.

 

 

98

B. CSz+ Stub Conclusions

Several conclusions can be drawn from the above discussion. We can
generate fluxes of mass-selected ions that 31111311911 vibrational spectra to
be obtained. The infrared studies are crucial for understanding the fate of
these ions, at least the final products, and will be important indicators in
the success of alternative deposition approaches. The results indicate that
neutralization and fragmentation processes complicate the simple
entrapment of mass-selected ions. Thus, the experimental conditions
which provide sufficient numbers of matrix isolated cations for LIF
experiments may not lead to the continued accumulation required for the

less-sensitive FTIR measurements.

99

Chapter 5: Conclusions and Future Work
1. Conclusions

The ultimate purpose of the research described in this dissertation is
to obtain structural information for ions exiting quadrupole mass-
spectrometers. Although that goal has not yet been reached, we have
demonstrated the first objective: that mass-selected beams of cations can be
matrix-isolated and observed by laser-induced fluorescence. Preliminary
experiments suggest that neutralization does occur, although the
mechanism has yet to be verified. Neutralization may enable us to obtain
spectra of new transient species as well as of mass-selected ions trapped in
matrices. Several enhancements must be incorporated into the current
instrumentation before spectrosc0pic examination of ions will become
routine. These instrumental improvements are described below, as are
improvements to previously-described experiments that can aid in a more
complete understanding of trapping processes. Finally, specific chemical

systems to be studied with the new instrumentation are suggested.
11. Instrumental Enhancements
A. Cryostat

It is imperative that the Displex 202 be replaced by a a 4 K cryostat so
that neon may be used as an inert gas host. This will make selection of the
cation far less critical. Then the instrument can be optimized first with
small cations, and subsequently employed to study ionic species of interest
to mass spectrometrists. There are two types of cryostats that will attain
temperatures below 4 K. Liquid helium cryostats such as the Supertran-

 

 

100

B®30 can be purchased for under $10K and can reach temperatures as low
as 1.8 K for brief periods of time. Such liquid helium cryostats are relatively
easy to use and require very little maintenance since there are no moving
parts. A disadvantage is the cost of liquid helium required to run the
cryostat continuously for extended periods of time. The HS-4 Heliplex®81 is
a three-stage closed-cycle refrigerator that can attain temperatures as low
as 3.6 K. For the purchase price of $45 - $50K, low-temperature operation
can be achieved without the added cost of liquid helium. However, routine
maintenance on this type of cryostat can become very expensive. The pros
and cons of each of these cryostats should be carefully weighed before the
deciding on which type to purchase. Ancillary equipment such as
temperature controllers, temperature sensors, and transfer lines can be
purchased for the cryostat. Both types of cryostats can be built to
specifications required to adapt to the matrix chamber and simultaneously
place the cryostat substrate along the optical axis.

B. Vacuum System

The regeneration time of the cryopump can be shortened by adding a
regeneration kit to the current cryopump. If funds for this accessory are
unavailable, time can be saved by purging the pump with warm helium, or
by backfilling the chamber and pump with helium. In either case, it would
be advantageous to have a vent valve on the chamber that would vent the
chamber to a nitrogen or argon atmosphere before admitting room air.
This modification could save regeneration time as well as pump-down
time.

Another improvement to the vacuum chamber would be the addition
of a removable baffle that could be placed in front of the RGA. A baffle

 

101

would shield the cryostat substrate from the abundance of electrons
emanating from the RGA filament. This would allow continuous
monitoring of the vacuum environment without electrons impinging on the
cryostat window. For experiments requiring electrons, the baffle could be
removed. (Note: it may be advantageous to incorporate an electron gun
that can provide a higher flux of electrons to study the effects of electron
bombardment.)

The calcium fluoride chamber windows currently provide sufficient
transmittance of light in the UV region for LIF studies; however, the
windows cut off light below ~1100 an1 in the infrared region. In order to
study species with absorptions below 1100 cm'l, it is necessary to replace the
CaF2 windows with some other material. Flanges that will retain a
window of 1.75" diameter on the 2.75" conflats have already been
constructed. They utilize a 2.75” Viton O-ring to seal the window to the
chamber. The windows can be purchased, or cut from an old KBr
beamsplitter removed from the Bomem FTIR; this will extend the IR cut-off
to 450 cm4.

C. Mass-Selective IonSounce

Modifications to the M8188 have been discussed elsewhere,82 and
will only be highlighted here. The implementation of better focusing optics
following the quadrupoles is extremely important. Ions exiting the mass
filters are divergent and this divergence increases with increasing m/z. In
several experiments an ion current of up to 1 nA was observed at the
Faraday plate when it was fully retracted. Focusing these stray ions onto
the cryostat window would increase the efficiency of the experiment and

shorten the time it takes to trap sufficient numbers of ions to be observed

 

‘2—

102

spectroscopically by LIF or FTIR. There are two approaches that can be
used to achieve this goal. One is to design more emcient lenses at the end of
the quadrupole to focus the ions onto the window. The other method
involves the utilization of a ring electrode at the front of the cryostat
substrate to "pull" the ions into the matrix, such as the one described by
Andrews.39 Sensitivity of the infrared experiment could be greatly
enhanced if the ions were tightly focused on the cryostat window. The IR
beam should also be focused to the same diameter. The smaller the
deposition area, the better the sensitivity.

D. Photolysislamp

If sufficient funds become available, it would be worthwhile to invest
in a UV photolysis lamp that could be adapted to our chamber. This
traditional method of ion generation could be employed to compare (or
verify) the formation of ionic species from the M8188 with those formed
from the lamp.

111. Research Plan

In the research plan presented below a series of increasingly difficult
projects is outlined to test the feasibility of our ultimate goal: to trap mass-

selected products of organometallic gas phase ion/molecule reactions.

ADevelopmentandOptimimtionofExperimentalProcedures

Upon acquisition of a 4 K cryostat, a series of experiments is
envisaged to optimize the conditions for ion trapping. A complete
understanding of the CS2+ results is still lacking. It would be beneficial to
reexamine the LIF experiment using neon to trap 082+. The sharper bands

 

103

that would be observed may permit a greater level of control over
experimental parameters. Investigation of C4H2+, C5F6+, and smaller
cations could also be undertaken with neon matrices. For example, 00*
could be employed in the optimization experiments. This is a simple
molecular cation that can be generated in high yield, has been studied
extensively,83 and trapped in matrices and characterized by ESR.84 The
spectroscopy of CO+ in the gas phase is well known, and wavenumber shifts
due to the possible formation of (CO)2+ aggregates should be readily
distinguishable.”86 Since laser-induced fluorescence (LIF) is much more
sensitive than infrared spectroscopy,87 LIF should be used to monitor the
presence of CO+ in the matrices prior to the FTIR measurements. The
shorter deposition times required for LIF measurements will more
efficiently facilitate optimization of the many variables (e.g. ion kinetic
energy) that must be adjusted for effective trapping. LIF will also be
utilized in the evaluation of the necessity of the counter-ion in the cation
trapping experiments; CO+ can be co-deposited with Cl' (generated by
electron attachment to CCl4) or Fz'. Although the vibrational frequencies of
diatomic molecules are rather insensitive to the matrix environment ,
somewhat larger shifts may be expected for ions.88 The utility of several
matrix gases should be evaluated to determine the most appropriate
matrix. Thus early studies on ion trapping in the matrix using LIF
detection will allow the variables that will lead to the minimum time for
building a matrix containing sufficient ions for infrared analysisto be
determined, while avoiding sputtering of the matrix due to ionic

"bombardment”.

 

104

EEaflyFI‘IRStudies

Once the vibrational spectra of 082+ and CO"’ have been obtained,the
experimental methodology can be evaluated by choosing one of the many
small organic systems of interest to mass spectrometrists. For example, a
number of experiments have been reported which attempt to distinguish
among the various C2H30+,39 C2H302+,90 Cv7H74',91 CvzHr-IO‘F,92 or
CNOH+,93 isomers that are known to exist, and can be formed by electron
impact ionization from different precursors. These ions may be easier to
trap than CO+, although their spectra will be more complex. The planned
matrix isolation experiments parallel the mass spectrometric studies. The
products of E1 of various precursors at a selected m/z will be trapped and
their vibrational spectra recorded. Ab initio calculations and normal
coordinate analysis of the experimental spectra will be used to differentiate
the isomeric ions and to determine their geometries. This work will be
performed in the mid-ir region where sensitivity is highest. The results
will be compared to other theoretical and experimental treatments, both
from the literature and being undertaken simultaneously within the
department. This investigation of organic ions is only indirectly related to
our goal of characterizing metal-containing ions; yet these experiments
should be relatively simple, and will demonstrate the power of the
technique.

In the absence of a 4K cryostat, further investigation of NO+ is
warranted. Longer NO+ deposition times may be accomplished with a
thoriated-iridium filament. The NO+ band at 2345 cm'l, assigned by Jacox
and Thompson,94 may be observed. It would also be interesting to trap

ionic products of electron bombardment ionization of CHC13. J acox and

. m

 

105

Milligan have observed the infrared bands of CC13+, HCClz+, and HCClz' in
argon after UV photolysis.95

c. specunseopie Studies ofMetal-Containing Ions

With the methodology established, relatively simple organometallic
cations that can be produced by direct electron ionization of readily available
neutral precursors can be studied. The series MCO+ (where M = Fe, Co,
Ni, V, Cr), obtained from corresponding metal carbonyls,95 is an obvious
initial choice. There is also interest in ions such as CoNO+, which may be
linear or bent, depending on the nature of the metal-ligand bonding.97
Other candidates, which have already been treated theoretically or are
amenable to electronic structure calculations, include MnH+,93u99
MCH",100 CrX+ (X = F, O, N, and C),101 MCO+,102 AlCH1,2,3+, and
A1C2H5+. The vibrational modes which involve significant metal atom
participation will lie at lower frequencies, and many require the use of a
liquid helium-cooled germanium bolometer detector rather than the
somewhat more sensitive mercury-cadmium telluride mid-IR detector.
The molecular geometries can be deduced from the vibrational spectra and
will be correlated with the theoretical results.

This work will lead to the ultimate goal of our experimental
structural studieszthe entrapment and subsequent spectroscopic analysis of
mass-selected products of ion/molecule reactions, such as the products
formed when transition metal ions react with methanol, ethanol, methyl
iodide, etc. Ions such as MOH+, MCH30H+, and MCH31+ will be of interest.
Moreover the structure of the MCH31+ ion formed by the gas phase reaction

MCO+ + CH31 00> MCH31+ + co

 

If) 1‘. 40

106

can be compared with that of the MCH3I+ ion which might be formed upon
direct co-deposition of M+ and CH3I into the inert matrix. One would expect
a different distribution of possible product structures from these two
experiments. With a working system of the type which is now under
construction, a number of interesting variations may be envisaged.
Consider as a variable the kinetic energy of the ions impinging on the
matrix. Can collision-induced dissociation experiments be performed? For
example, if the structure of FeCH30H+ is CH3-Fe+-OI-I, will a "soft landing"
of this ion in the matrix maintain its geometry? In contrast, will deposition
with a higher kinetic energy yield FeCH3+ and OH- in the matrix? If co-
deposition of metal ions and organic reactants yields interesting products,
the kinetic energy-dependence of the incoming metal ion on the product
structures may yield useful insights, complementing the primary
structural studies. The vibrational spectra of the trapped mass-selected
products of ion/molecule reactions will provide unambiguous information
regarding ionic geometries and force constants. Infrared spectra should
distinguish absorptions due to the molecular ion from those of fragment
ions resulting from CID. Optical spectra obtained from the same matrix
will provide electronic transition energies, and possibly excited state
vibrational frequencies. Many variables in these ion/molecule reactions
can be investigated: the role of the metal, the organic neutral and the
ligand/degree of coordinative unsaturation. In conjunction with theoretical
calculations performed on these systems, these results will lead to a greater
understanding of the factors that govern the chemistry and bonding in
these organometallic systems, and ultimately those for neutral systems as

well.

 

 

 

Appendix 1.1: Schematic diagram for vacuum interlock control box.

108

 

 

>- 0.1}
«4—4

 

 

—0

b - 4H»
“—1
.4—4

4
”(H—4
4*- J—i:

-I ‘1
1
‘FP .4—

 

 

~ ) 4h—1 -4r-o1>—

t
.4
#1»
U—«r—
L.
L1
. ...L

 

$

 

 

 

 

’ILP
l
H
L
F s-

 

-4

L1)
)

 

d

J
4<~

4

~HH

{-4-4 .4
.4

 

 

- -<

 

A)
IN

I
J
0 4»— L-
- 4 r—h-
pap-4
—» +0: LA «r
l
H)
”-0- >- -0

ii
I
I
*i
’.
7o
l
l
«Mb «fiw

 

- >- .

 

 

 

 

 

 

 

-1)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cpl ..... '. - " - _;_' ’ . -_ H2
F : . . . : ;:; 77:1: ". ::.:.;.::.i::::'-“ "2 "H":
'''' '7" :§?f:;..:\21~;;» 0F.;:?}E':-:-—‘—r+‘r—~

 

 

I l . V . . . . . . . , . . _--,._. _-.. a . 0&5: .
......._,,.....-:0 ..s. em”
- . . .. .... L.--.. ...._.-_..._..

. .. .. '..Cs..‘r.—..€- “.2225; (3"7‘»

 

 

 

 

----- 0 .- . C..._.'*_—.—

m: . . "i L . I'.'TII:'__.:::E:TZI
.,1\ ISA. 1" 599mg. . -..
.""V l 1.. ”'7‘
cs slag: €°3.T'.".‘.‘. .(Wfi;
'W" QlSESA): ..

 

 

 

 

 

 

 

 

sows-i. . "::.' ";’.:;;;'::.::.:
G‘- @450 s-o0_RcE- -——. - --
—/\l 1 \ “1.7..- ____-.°

0——-O—- ‘° CQL-SLREL5)-..
' ' - ::.|IM- TI:

 

 

‘ -.0..-

Appendix I.1 continued: Schematic diagram for interlock control box

3.»

 

109

 

 

 

 

 

 

 

 

 

 

 

 

 

 

||
III Ilall III.
a I. Il‘lll .l| -l‘.‘ |.'|".|.|" 11"- II I

. II I"- -l‘. | I "I. |"‘l-|' {I I ll'l'
--n! i II. I: I I , i ..III at!
l.lxl..l|. .. nl all!!! ..I.

In.“' l..- I: 'l ' ‘Il ‘I'lll'Iil I-
-0 -.‘ -I .0 I'd. 'l'll'li Ii 0.. 1 -il ..IIIIIIII
-.om..-' ‘ Ull'll'-’

0“! dill]. -ll'llv‘llll .llsll

+-- i.." Ill|l|l'|"|
I! ffipElalllllllll-..-sf - II
A - .n e--t""..l '-.‘| "lllll-'l'llllu|"l-.- ll
..IIAU -. g

 

 

 

 

I
u -00 u I 0.-
O . I. .0 l
. v a I
O
. u I s

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ing to

gee to RGA control board for interfac

Appendix 1.1 continued: Chan
the vacuum interlock control box.

110

 

.uozobaoo 9809-930 mo Scum—flu poem—mamm ”NA “550%?

  

médooé 5.;

...7 A...

H”.

:83; W \

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.:

ac

 

.uo=o.5:oo mEoB-o.¢O .«o 8393c out» omuafiofim ”mu “3333.3.

 

 
  

 

         

 

 

 

 

r a w .332!“ an”... u do
I d .0
gr _H\fiu.é r . o , o 1 o... 5': 595.: .... own a... .30... {DUO

nil .. II Ole-{l p v 0 av.
pipe-lulu \O pl-lpl1Ano Elu>ul3 =<= 83.5.! . _ . O a a

. 1's. 1“ C
(343313 9.5. -03.. I E . 9.3 . ...:.
I... 1.3. Fall ...:. «do a... "a... w“ 00—
o 5! Tart; . ‘ .035... (a a a

. . u . s o
In)... < 301:. .5 < 91.3.3.4: . . f ..l (...: Q

03...... e
0“ 0.639.131; 01016 3 OJ...) O

111

.....

(:31 . 2

 

l3.- OMIM; .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:"— 11:- __‘, ”“_':———'"-—C.7i‘“‘- —‘ ' ' “ "" ‘--“'—
—_- -_—_—_- - . -——;\—D< were -- --- 31:”
:2": 1172?; - _ _______ *‘—“_‘l“—
'_::_:‘::: :“ “:7 man} -- - ---—-- 10 3
_______ “77%;. W V .______ ._-_._.--._T '2‘
__ ;- __‘_:_‘:._;:.‘- _ _:._:— ::._ _.:__ ____._+_2,_..
:‘t 21;; 'T__".".‘:'I.—."’.”____.-':—'T;;;‘_'.’.I.’-‘.‘i “‘—‘—"‘"”'"'-?I—‘"
“:lkiéLmfic lpsfi' ‘ ‘lét;_1 341-;(Bwe imam-6056'“ WW

 

 

 

----———-— (nee-- ——en<-— flaws-I nab-eunu-
--ru=.5- -- -~ - - T“

.—-————-—_..- .-_._J--_- __ _

:_——— ._.-.wr :4__ -- :9

 

 

 

 

 

 

..- ‘— ‘ I r 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.x
‘W Q 981‘, \fi'7

 

 

 

 

 

 

- m- +'—- 175:3— 49)

- .--______——— — —-—~——LQ——m._EDl 2.

--- --9——‘ mfike 3:55.—

T______.__.t.ary~’ _Eub Q"- 5_ ' -- i

._.._...._._...-__. .__._._--_._. -___,_ [I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

"“"2‘2— “77;.-.“ -.JZLLEID! ""22"
"" -...-- .. __ WEED I Eat:
-_.... . - - _QLu L you?
"‘"""—“_2"_'.’._‘-- " '5' "299 o‘ _
._.--." .. _--.-.__ 153.53%- $-55»
U_._1>u..,

 

 

 

 

 

 

 

—'—“""‘33‘E§—6mwm ; . ‘

P

7 --
{1" Gem - ‘7?"

E
W

 

W

Appendix 1.3 continued: Cable wiring diagram of Cryo-Temp controller.

 

 

 

93:58.8 gong

 

 

 

 

 

88.8.. ohms—anew.
28 ea .23 8a ._. 13m ommfiamoz uses... 8 3833...:
cameo—3oz em
came—3oz 3. ace—ca... «an 33 Eu .2.
.65... .2338...» 0.5.8382 .2 b3... 3 u gag 8.8m
82 ...:. .«82. .oz .«oo .93 Bad—3oz ..... o... 858 £25: 3.. oz 8:3
Badge: new: be 3.93
.33.... 9.232 239232 8336 2. 8 <3
88.8% amen—282 8.2:... 8 mm
28 $3 3... «a: 3 Ban omm.52«oz 83%.. 2. mm
.693 1838...» 0.53232 8.8.... 8 308232..“
«02 ...: «88 .oz .«oo .8: 8.0.3232 3 3 8.88 2.3.2 d3 +02 95:...
833.. 25 8a. «8. 2. 3.5 8m. 8802 .355... 2 398.4%...—
ommgomoz 2. a u .888 8—
.6.3.. .3338...» EQSSOZ =8... c .. 8:38.. 8.58
fl «oz ...:. 08% .oz .80 .8: 308802 3 a.» 2:22 533.33 +02 gem
1
omm.§m<oz 98 28 a... n B.
0%....832 SN 5253... «.133. so: .3
...::5 unwound: ...: .83282 v.8 3
.235: 02 v.3 .«oo .8: 83232 ..... 3 83.33.32 :83. 023 33
unnamed»? e ".2
.653 .2325... «8 .8: "twice: 3: mum). 1.3... 8.5. 25a
28: 8:30: $5 25an
$320 Eon .556 can 892.8855 85

 

 

 

883m 85803 +02 2: a. Bataan angry—am .2: £533..

114

 

 

88...... 0.5.32.0... 83...... 8 0..
.-.E 08. ...:. an... 2. 0...... 020.380.. 5.5.... on 3.08.32...
...:... .358...» 0.5.283
~02 ...:. .082. .02 .80 .0“: 0.6.35... 3. ..s.S........ .3 u 3... 3......
0.0.880... ..... o... 80.8378 38... 02.2
...:... m25:25... 3:: 2:35....
...:036 Ems—Om 23......0 don .332 2.05:0..me

 

3.5.230 ...: 5.58....

llS

 

 

5.5: 83333 383 023. .3 2.5 8383... «him "2: Nana?

mfiflfififiaisg
omm~ owcm ovum ccflw
. n u n u

p
—

H

L

1 Q36

a card

, oomd

anmnnnqv'

 

 

 

 

. coca

 

 

$3» 8583.3 385 023 no 2.3 82 58: 38mm fiasv «.5 £52“?

 

 

 

 

 

 

 

 

 

mugs—unngag
coma com: 8: coma coma
Will- 4. fl . 3i"; “ “ fl “
:2. J? -A s...

m A m # 1. «.3

A H '
1, a;

m +
1 , S..."

Q

4. and

4,
.. «3

a m a .

 

2V6

aouaqxosqv

 

117

 

53% 8332.33 585 023. .6 2.3 83%: Exam fiaé «.5 £333

309550.33
cam: ow: av»: 83

IIIP .
d W 1

 

 

 

s a;

 

 

 

 

118

.noEmfioeé 388v 02E 5 we 358% 3.8m ”3: 5:33

 

 

 

Ewan—5:253
8w . 32 r 8% 8mm 83.
L1cad
L. and
oz:

.. and
.. and
t
.. cad

 

 

cfiuwfimx<z
NH

119

 

 

 

.253 .2385? «00 .8: 0.5.8520 3 q: 20: 632.5% +3290 $03
”.235 «5502 02.255 .28.. a 5 20m
0.5.2238 835... an 5 <00
0.5.2250 .332. n 8 20:
0.8.3530 “3:5... 8 258532:
0.8.2520 .3558 3. 258333..
0.5.2250 «3:58 om 258232:
”3:3 32 ..z 0.8.255 ca 92
0.6.2520 «a 3.»
0.5.855 3 can
0.3.855 ed 9mm
0.58on 3 .03
0.5.8218 3 98 >0 S u .920 :3
0.5.35.5 «a S: 250...... S u 333.: .95
25 a: .22 .32 an 35.0 02.8520 3 a3 :35 on 3 238.5
28 89... a 230 0.3.8250 3 ...: 85: 2.22329:
60.2: cacao—.329: 0.5.8238 «A 9m” .0953 32:3 :98 to
855 1.322.? «00 .8: 0.5.8250 3 3. 20m 5.58%.. +3220 gag
0.2.2.22 ..... Q8
02.8.22 ..... ad
02.8.22 ..... 2:
02.2552 ..... 3
02.8.22 ..... 3.
0.6.8052 ..... 3 25°55 3 u 3.: .80
.85.. 152.5: «.00 .8: 0.84352 ..... 3 53.2.2. as... 5&2 gas
2.2. 3.35:... :2: £505;
.a?_o2_O Emzcm 23.2.5 .39 33: 2.25.339 3

 

.8630 5560900 RZZNEU 05 E 00530qu 3.5810an "VA: 50:00.3.

120

 

 

 

 

 

 

5:25 can .853 «58232.. ommao<mm< ..... ma managesacmouéo .39... $58
3:3 80 E... 8: ommd<zv=o ..... Sm 55.28.. ._....z .59... $23
.053 m 5 626330 59.59 02 3:3 535m no > cma-
.3255. u...:...z Rafizvmo .32.... 8 858332...
0.565550 8.555 on cocoaoZBofi—
.253 378:0 oz 0.5.5.235 m: oéa
0.5.3.2300 m: odm >0 «H u .835.”— :3
0mwd~2v=0 fig 9: ._.—35 ma}. 5 95395
0.5.2.7330 a; 5.3 3.58 0552.395-..
0.5.5.va0 EH Qua .353 0965.. :98 :o
.353 5.52.5 So .8: 95.2250 3 9... <2. 5.28%.. +5250 $38
Bmfimfléfl 8555 cm 3533305
95.3935 .95 385:5..8oaa 2.5m
Badmfléd ..... 99. to 52.55 .0958
...:—«5 «552.395-: 0mmdovm<m ..... aim £9.25 new 0552595-: 33¢”
.1525: misaoz 2m6m2v=0 3555 2. cozuamfiouofi
0.5532300 35:5. cv 358—385
0.5.3235 3555 cm vmfiuaozsofi
95.37250 Ea ndm .2525: 3. u Ans 39. 39
95653.30 vd aim >mmmugoco 57:35 v... .5
4-59 MN: .33 .83 an 555 0.355550 ad can 95395 3.58 1895-:
759 8mm .3 9:5 9.95m 0.5.8235 m4 9: .358 2.2.2 :98 co
2...: 2.3.3:... 3:: 252......
.5320 2020: 59.50 .3: 332 595...»an 85m

 

3.5.2.8 ...:. £2.23...

121

 

 

 

 

953:5 ..... 2:
95.825 ..... 2:
95.825 ..... 3. ..co 2255 .858
. 95.825 ..... 2. £9.25 3a «£52329: 533
95.885 ..... 3:
05.885 ..... 2:
95.885 ..... 2. 53:82. ass 89¢ 3%
.1325 9.322 95.5535 5 u .2
05:35 .3356 5 363232:
2% £25 333% 25 a: 95.253. v u .2 05 :85
as. .82 .88 .88 3 355 95.584 5 :2
25.283. 3 ...:
85.883. 3 .5
85.853. 5 3:
05.853 3 3
95.883. 3 3.
95.883. 3 3 t3 2: u a “a 23.
253 1852.? So .9": 25.38:... 3 3. .329: 5.3 2:225
153 no 035900 Umwdoomb< od EASE—o .353 ~83 ..3
.322 to 3:900 o5g83< ..... 2. 43 53:33 3:238 833
2.2. «Enzuzm 3;: $5259
.atomnc SHEO: 39:50 .35 332 anon—_uonum 8.5

 

32:50 5: 5:33

122

.0958 mum: 0952 498% Sign“. 0583.389: «6 3.80 smug. «500mm "menu £35.34

Boga—53a?
cmwm . 2: m con 85m 83.

. v _

-.

..mmd

 

mad

Q . mvd

 

O
o
aatquosqv

 

 

 

.. and

m .. bmd

 

 

%. cwd

123

 

552 8.65 flag: 83°38 vagina: .3 :18 88358 .535 ”383 3: £533

 

 

 

 

name—5:353
coca ccmm Svm 8mm 8mm
“ . n . fl .1 u .
4 mad
.. 36
m
. mm c m.
m
. 3.0
H 1r
.. Sac
2. c2

124

 

.55: 822$ $53 832.8% «Eugen... “a 2.5 82 33. 38.5 5382 3: 58...?

 

 

 

 

 

 

 

waging—w?
83 82 8: 8m: 83
.- 1 7:; T -...--- fl 1.
-83
21.1 <1:<\ .-
- x .535
A
m
-52. m.
m
-532
.55.:
2
%

 

- cmmd

125

 

.552 0&3 3:85 8.263% 833383 .8 2-3 3.5.85: 38.5 finsv 3: 5595‘

5.8955963
cNS cm: 3.2 83

p p bl r
d . u 1 - q

:2;

.- mwad

 

 

 

 

 

 

 

j

N 3.9. << .. 33
> w
.. 1.. m.
.55... m
m m

x 3 D
.. 53¢
a.
5
.. 52

 

P. 83

 

126

 

 

 

 

 

oz 882: «a 3 as. 508:8 825
3 3 20 80338 +50 an 853
oz 835)" A: md :80“ on 005500»—
: 3
8.8 3
62 82¢: 52 3
oz 882: 8." 3 80 as: 8552
84 3 3.. «8532... 85
..z 8822 0 A 3 20 55300 +50 53 88$
50» 88,2 3 :55 03 :83
3.82 8834 8.8 3
382 8852 3 3
2 3
oz 88¢: 3 3 80 2.50 8:202
3 3 .8 83832... 825
S 3 20 58830 +50 05 8:3
Oz 88.? 8.8 3 20 8:. 8 225:8 m=55 8.55.
Oz :83 3
«a 3
5.8 3 023+ 3: .85
8.8 5 >+ a . m u 920 :3
oz .58.? 3 3 20 853: +50 EE 8205.
_auu5 2.25:...— 25 2505:. 2.85
+ mu m: 2.9—.50 do: <6: 5302 .5520an 80:

 

5.835300 +«m0 00 50:55 m: 05 E 00808.89 3:0Et0axm ASH xmvC0QQ<

127

 

 

 

 

 

88.8 8.8
8.0 0.0 20 88388809 +50 £0 005%
oz 888.852 8.8 8.8
8.8 8.8
oz 888.852 8.8 8.8
82 885.2 8.8 8.8.
8.8 8.8
”6 GA mnco— ComaOOSOo “826:5“
8.8 8.8 20 88888888 +50 .88 888.3
82 885.2 8.8 8.8 22»: :3
~ A 0.: :0 d0. 00380009
oz 854: 8.8 8.88
8.8 8.8
82 88502 3 8.8
8.8 8.8
oz 8558 8.8 8.8
8.8 8.8
06 GA 25— Comte—~00 @003
8a 8.8 20 8888888 +50 58 8853
82 88822 88.8 8.8
88.8 8.8
82 88822 88.8 8.8 883233 .820
«A 9v 40051098 35 E
.28.»; 255:8 3:: 3:28; 2:85
+ m0 ...: 2.0....an .009 <6: 80.02 8:00.20an 88

 

”30:58:03 82 50:009.‘

128

Hi

 

 

 

 

82 888808 8.8 8.8

82 888808 8.8 8.8 382882 .828
82 88808 8.8 8.8 8.3888. :3 80
0 Z «8830 8.0 0.8 038000003 003 cows.

882888: 8880.. 88.8 8.8 8.80 88888888 +50 5: 88888
82 8858 88.8 8.88 888 828 885888.0 88$
82 888.. 8.8 8.8 288 8.5 83
8z 8885.. 8.8 8.8
82 88858 8.8 8.8 .82 88 8.2.. <0: 895

.820 8858 8.8 8.8 8.80 88888888 +50 588 88:8
:08; 88.5.. 8.8 8.88 288 .96 83 8838
88> 8885.. 88.8 8.8
88% 8858 8.8 8.8
88882 8858 8.8 8.8
82 8858 88.8 8.8
8.8 8.8
...8 8.8 8.80 8888888 +50 58 88888
:80» 88852 8.8 8.88 8.8. .280 82.88 88
8.8 8.88 882883328 885. 888.88
oz 888858 8.8 8.8 .880
8.8 8...
8.8 8.8
oz 8888.2 88.8 8.8
.8885 8285:..— 25 3:258. 2:85
+8mU 8:.— .notau due <0: 80.02 30.580088— 3

 

”80:58:08 fit 50:09:.

129

 

 

 

 

 

 

 

:02 8885.. 88.8 8.8
- -- 8.8 8.8 20 88008888 +50 588 88: S.
:02 88838 8.8 8.88
3 8.8 20 88888888 +50 588 8883.
88> 8838 8.8 8.8
:80» 8838 3 8.8
88.32 88838 8.8 8.8.
8.8 8.8
8820 8838 3 8.8 8.80 8888888 +50 .8: 8888.
00> 28:. ad ad 0000.80.80.— 935 0083
:8 8883.. I 8.8 88588.8 .823: 89:38
0 Z 88:. a; 0d 00:08.80» 02
. -. 8.8 8.8 8.80 8888888 +50 .888 888:.
8z 8882. 8.8 8.8
82 888808 8.8 8.8
oz 888808 8.8 8.8
”.6 DA 60.3880.— 928—
8.8 8.8 8.80 88888888 +50 588 88888
82 88808 8.8 8.:
8.8.8.8 28888:... 25 0.5.5:. 2:38
+ m0 8:..— .cotao do: <0: 80.02 «nun—_uonum SB.

 

”€035.88 ".3 500095

130

 

 

 

 

:oz 88:. 5 9% +2. a 2589 $03.. SSS
3. 2. ...Eo 5.333% «mo 2:. :4 83g
--- -.- anon—Immune oz :32 o3 acute ._O
ed 3383.8 -6 £8 can 8&5

5332.2. ...«mo ._ofid
2mm; 82.... .52 9 5 .3 RES <3
2mm» 83:. .3 3" £52. 8 $8 :3 25.3.
3:: sad. 3 3 .52 3 a ma .6
3?: Sad 23 no 38.... u 29 _
3 a... <53- u 20:
oz 83:. 3 ca <53. u 20:
we a... €26- u 20:
ES £3.58 .6 a a8 3 SEE

:Eua «5.2.2:. 2:: 3:25... 2.35

+~mU m: 2.9—.50 don <6: 332 2.25.3an 88

 

”Atacamaaouv H .2 5959.5

131

 

 

22mm «A 3.398 85:32. 8-3.. +qu
83mm <9 3 v 2: 238 can an“?
833 <9 adv ...:
82mm <9 3 vv ma manna .238 Sn sun?
82mm 3 3 v 3. took—Eu“): “on :2: $23
82mm 3 2. v a.» 22.59 2.32. 32.95
to ”.3395 inc—m 89mm :6 ad econ—tag vaccum—u +Nmo gata
5am: van—«396
82mm 5.: h3.52% @353
82mm and .3. «to 2335 23%
.83 83mm :3. 3 33:5.
3.9mm gum: .283 2: gun?
Oufloouoflmadd nammm Sam 3 @3603
$8 2.: £8. 2:: «5432 83mm 2 93.
mo 3. ~a2. 25 £2 a. 48m 23mm 3 93. as": .233 8a Jum<
82mm 3 3.4 can: .238 Sn in?
33mm n3 9:
83mm 5 3m 83%
83mm 3 2m gum—x .23: Sa dug
82mm 93 93
82mm 3... 3a 9ch .233 8d .32
83mm 3 q: 833
82mm 3 9m
38%.. 25 22 a. ._an 32mm 3 ca
,8an x: a; no"? .238 2: .52
.3 .38an £55 82% a... 9c .5 Egtéeémo 83$
2.2. 2:23:5— 25 252:;
.2.—93¢ EHEOG 23.2.5 .3: 382 305...“..an é-

 

.msofimmaov +30 .3 863m wwuméfi 05 E cantata chfitmaxm ”NE 538%?

132

139 mo :25 b3 32%

 

 

«58% .28. 9 .3: «3.82 83mm 3A.:— ..=8. 2: .32
:38 S .3: 3:22 “saw 3 A 5.3
825:: agroun— anoa «am A 6.3
Saw «a A 5.3
85m 2. A 5.3. as. :8 3 E56
Saga 3 3a gum: .238 8a an“? 23$
88mm 2 9:
855 a a.» gun:— ..:§ emu sun?
83mm 9 ad 233 man—Log? 322.2
.co 38an £55 88mm 3. ed 5332:... :5 ...E $8 25%
88mm 2 6.25.8 21mm .233 8a .32 583
mafia 3 8.393
23: «Ego—r..— A<5 7.52:;
.a>.32.¢ Susan 33.30 .3: 332 23.5.3an 8.5

 

”8253.83 «.3 Amiga?

10.

11.

13.

IJSTOFREFERENCFB

J. L. Holmes, Org. Mass Spectrom. 20, 169 (1985).
K. Levsen and H. Schwarz, Mass. Spectrom. Rev. 2, 77 (1983).

L. Howe, Mass Spectrometry: A Specialist Periodical Report, Vol. 7,
R. A. W. Johnstone, Ed. (Burlington House, London 1984), Chap. 4. l”

M. A. Baldwin, Mass Spectrometry: A Specialist Periodical Report,
Vol. 8, R. A. W. Johnstone, Ed. (Burlington House, London 1985),
Chap. 2.

P. O. Danis, c. Wesdemiotis, and F. W. McLafl‘erty, J. Am. Chem.
Soc. 105, 7454 (1983).

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

M. C. Blanchette, J. L. Holmes, C. E. C. A. Hop, F. P. Lossing, R.
Postma, P. J. A. Rutfink, and J. K. Terlouw, J. Am. Chem. Soc.
108, 7589 (1986).

R. Stepnowski and J. Allison, Anal. Chem. 59, 1072A (1987).

J. V. Coe and R. J. Saykally, "Infrared Laser Spectroscopy of
Molecular Ions," in Ion and Cluster Ion Spectroscopy and Structure,
J. P. Maier, Ed. (Elsevier Science Publishers B. V., Amsterdam,
1989), Chap. 5.

E. Whittle, D. A. Dows, and G. C. Pimentel, J. Chem. Phys. 22,
1943 (1954).

R. N. Perutz, Ann. Rep. Prog. Chem., Section C 82, 157 (1985).

L. Andrews, ”Spectroscopy of Molecular Ions in Noble Gas
Matrices,” in Molecular Ions: Spectroscopy, Structure, and
Chemistry, T. A. Miller and V. E. Bondybey, Eds. (N orth-Holland
Publishing Company, New York, New York, 1983), Chap. 5.

A. J. Downs and S. C. Peake, Mol. Spectrosc. l, 523 (1973).

 

14.

15.
16.
17.
18.
19.

21.

31.
32.

134

H. E. Hallam and G. F. Scrimshaw, Vibrational Spectroscopy of
Trapped Smcies, H. E. Hallam, ed., John Wiley and Sons, New York
(1973).

P. H. Kasai, Acc. Chem. Res. 4, 329 (1971).

M. E. Jacox, Rev. Chem. Int. 2, 1 (1978).

V. E. Bondybey and J. H. English, J. Chem. Phys. 71, 777 (1979).
N. P. Machara and B. S. Ault, J. Phys. Chem. 93, 1908 (1989).
M. E. Jacox, Chem. Phys. 12, 51 (1976).

L. B. Knight, Jr., J. Chem. Phys. 78, 6415 (1983).

L. Andrews, J. M. Gryzbowski, and R. 0. Allen, J. Phys. Chem. 79,
904(1975).

L. B. Knight, Jr., P. K. Miller, and J. J. Steadman, J. Chem. Phys.
80, 4587 (1984).

L. B. Knight, Jr., S. T. Cobranchi, J. Herlong, and T. Kirk, J. Chem.
Phys. 92, 2721 (1990).

J. Hacaloglu and L. Andrews, Chem. Phys. Letters 160, 274 (1989).
D. Forney, M. Jakobi, and J. P. Maier, J. Chem. Phys. 90, 600 (1989).
V. E. Bondybey and T. A. Miller, J. Chem. Phys. 73, 3053 (1980).

L. Andrews and B. W. Keelan, J. Am. Chem. Soc. 103, 99 (1981).

L. B. Knight, Jr. and J. Steadman, J. Chem. Phys. 77, 1750 (1982).
M. E. Jacox and W. E. Thompson, J. Chem. Phys. 91, 1410 (1989).

V. E. Bondybey, ”Time-Resolved Laser-Induced Fluorescence Studies
of the Spectroscopy and Dynamics of Matrix-Isolated Molecules," in
Chemistry and Physics of Matrix-Isolated Species, L. Andrews and
M. Moskovits, Eds. (Elsevier Science Publishers B. V., Amsterdam,
1989), Chap. 5, p. 129.

L. Andrews, Ann. Rev. Phys. Chem. 30, 79 (1979).

R. N. Perutz, Chem. Rev. 85, 97 (1985).

M. E. Jacox, Chem. Phys. 51, 69 (1980).

W. E. Thompson and M. E. Jacox, J. Chem. Phys. 91, 3826 (1989).

37.

39.

41.

42.

{E

47.

49.

51.

52.

135
W. E. Thompson and M. E. Jacox, J. Chem. Phys. 91, 3856 (1989).
L. Andrews and F. T. Prochaska, J. Phys. Chem. 83, 368 (1979).
M. E. Jacox and W. E. Thompson, Res. Chem. Int. 12, 33 (1989).

R. 0. Allen, J. M. Gryzbowski, and L. Andrews, J. Phys. Chem. 79,
898 (1975).

S. Suzer and L. Andrews, J. Chem. Phys. 88, 916 (1988).

L. B. Knight, Jr., J. Steadman, P. K. Miller, and D. E. Bowman, J.
Chem. Phys. 80, 4593 (1984).

M. Moskovits and A. Ozin, Cryochemistry, M. Moskovits and A.
Ozin, eds. (John Wiley and Sons, New York, New York, 1976).

V. E. Bondybey and T. A. Miller, Molecular Ions: Spectroscopy,
Structure and Chemistry, V. E. Bondybey and T. A. Miller, eds.
(North-Holland, New York, New York, 1983) Chapter 4.

L. B. Knight, Jr., and J. Steadman, J. Chem. Phys. 78, 5940 (1983).
D. E. Milligan and M. E. Jacox, Adv. High Temp. Chem. 4, 1 (1971).
A. J. Downs and S. C. Peake, Molecular Spectroscopy: A Specialist
Periodical Report, Vol. 1, (Burlington House, London, England
1973), Chapter 9.

S. Cradock and A. J. Hinchclifl‘e, Matrix Isolation, (Cambridge
University Press, London, England, 1975), p. 90.

G. C. Pimentel and S. W. Charles, Pure Appl. Chem. 7, 111 (1963).
M. E. Jacox, J. Mol. Spectr. 113. 286 (1985).

J. F. O'Hanlon, A User's Guide to Vacuum Technology, (John Wiley
8: Sons 1980), Chapter 6.

Leybold Heraeus cryopump manual, p. 62.

J. R. Gilbert, G. E. Leroi, and J. Allison, submitted to Int. J. Mass
Spectrom Ion Processes.

J. R. Gilbert, Ph.D. Dissertation, Michigan State University 1990.
Schott catalogue, 1990 edition.
Critical insights were provided by H. Buij s, President of Bomem Inc.

 

57.

59.

61.
62.

.33

67.

70.

71.
72.

73.

136

S. G. Lias, J. E. Bartmess, J. F. Liebman, J. L. Holmes, R. D. Levin,
and W. G. Mallard, J. Phys. Chem. Ref. Data 17, Supplement No. 1
(1988).

Product information sheet, The Matheson Company, Inc.
W. A. Guillory and C. E. Hunter, J. Chem. Phys. 50, 3516 (1969).

W. G. Fately, H. A. Bent, and B. Crawford, Jr., J. Chem. Phys. 31,
204 (1959).

R. V. St. Louis and B. Crawford, Jr., J. Chem. Phys. 42, 857 (1965).
A. Loewenschuss and A. Givan, Spectrosc. Lett. 10, 551 (1977).

M. E. Jacox and W. E. Thompson, J. Chem. Phys. 93, 7609 (1990).
M. E. Jacox, J. Chem. Phys. 93, 7622 (1990).

K. P. Huber and G. Herzberg, Molecular Spectra and Molecular
Structure IV. Constants of Diatomic Molecules (Van Nostrand
Reinhold, New York, 1979)

U. S. Bureau of Mines, Serial No. 2096 (1969).

M. E. Jacox and D. E. Milligan, J. Mol. Spectrosc. 56, 333 (1975).

V. E. Bondybey and T. A. Miller, "Vibronic Spectroscopy and
Photophysics of Molecular Ions in Low Temperature Matrices," in
Molecular Ions: Spectroscopy, Structure, and Chemistry, T. A.
Miller and V. E. Bondybey, Eds. (North-Holland Publishing
Company, New York, New York, 1983), Chap. 6.

V. E. Bondybey, J. H. English, and T. E. Miller, J. Chem. Phys. 70,
1621 (1979).

V. E. Bondybey and J. H. English, J. Chem. Phys. 73, 3098 (1980).
L. B. Knight, Jr., Acc. Chem. Res. 19, 313 (1986).

V. E. Bondybey, J. H. English, and T. A. Miller, J. Am. Chem. Soc.
101, 1248 (1979).

W. J. Balfour, Can. J. Phys. 54, 1969 ( 197 6).

G. A. Morton, H. M. Smith and R. Wasserman, IEEE Trans. Nucl.
Sci. ‘NS-l4, 443 (1967).

R. W. Engstrom, Photomultiplier Handbook, (RCA Corporation,
Lancaster, Pennsylvania, 1980). P. 57.

74.
75.
76.
77.

78.
79.

81.
82.

87.

89.

91.

137
B. S. Ault and L. Andrews, J. Chem. Phys. 63, 1411 (1975).
V. E. Bondybey, private communication.
M. Endoh, M. Tsuji, and Y. Nishimura, Chem. Phys. Letters 109,

A. Givan, A. Loewenschuss, K. D. Bier, and H. J. Jodl, Chem. Phys.
106, 151 (1986).

D. Gauyaq and M. Horani, Can. J. Phys. 56, 587 (1978).

S. Bell, T. L. Ng, and C. Suggitt, J. Mol. Spectrosc. 44, 267 (1972).
Janis Research Company, Inc., Wilmington, MA.

APD Cryogenics, Allentown, PA.

Jeffery Ross Gilbert, Ph.D. Dissertation, Michigan State University,
1990.

R. D. Brown, P. D. Godfrey, J. G. Crofts, Z. Ninkov, S. Vaccani,
Chem. Phys. Lett. 62, 195 (1979).

L. B. Knight, Jr., and J. Steadman, J. Chem. Phys. 77, 1750 (1982).

S. C. Ostrander, L. Sanders, and J. C. Weisshaar, J. Chem. Phys. 84,
529 (1986).

L. B. Knight, Jr., J. Steadman, P. K. Miller, D. E. Bowman, E. R.
Davidson, and D. Feller, J. Chem. Phys. 80, 4593 (1984).

R. N. Zare and P. J. Dagdigian, Science 185, 739 (1974).
M. E. Jacox, J. Mol. Spectrosc. 113, 286 (1985).

M. C. Blanchette, J. L. Holmes, C. E. C. A. Hop, F. P. Lossing, R.
Postma, P. J. A. Ruttink, and J. K. Terlouw, J. Amer. Chem. Soc.
108, 7589 (1986).

Y. Apeloig, M. Karni, B. Ciommer, G. Depke, G Frenking, S. Meyn,
J. Schmidt, and H. Schwarz, Int. J. Mass. Spectrom. Ion Proc. 59, 21
(1984).

A. G. Harrison, L. R. Honnen, H. J. Dauber, and F. P. Lossing, J.
Am. Chem. Soc. 82, 5593 (1960).

D. H. Russell, B. S. Freiser, E. H. McBay, and D. Canada, Org. Mass.
Spectrom. 18, 474 (1983).

C. E. C. A. Hop, K. J. van den Berg, J. L. Holmes, and J. K. Terlouw,
J. Am. Chem. Soc. 111, 72 (1989).

3.8.3???

100.

101.
102.

138
M. E. Jacox and W. E. Thompson, J. Chem. Phys. 93, 7609 (1990).
M. E. Jacox and D. E. Milligan, J. Phys. Chem. 54, 3935 (1971).
M. S. Foster and J. L. Beauchamp, J. Am. Chem. Soc. 93,4924 (1971).

F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry - A
Comprehensive Text”, (John Wiley & Sons, New York, 1980).

A. E. Alvarado-Swaisgood and J. F. Harrison, J. Phys. Chem. 89,
5198 (1985).

A. E. Alvarado-Swaisgood, J. Allison, and J. F. Harrison, J. Phys.
Chem. 89, 2517 (1985).

A. Mavridis, A. E. Alvarado-Swaisgood and J. F. Harrison, J. Phys.
Chem. 90, 2584 (1986).

J. F. Harrison, J. Phys. Chem. 90, 3313 (1986).

J. Allison, A. Mavridis, and J. F. Harrison, Polyhedron 7, 1559
(1988).

"‘111133111131‘5