)V1ESI_J RETURNING MATERIALS:
P1ace in book drop to
”saunas remove this checkout from
._:_. your record. FINES will

 

 

be charged if book is
returned after the date
stamped below.

 

 

 

THE EVALUATION OF OPTICAL MASS SPECTROSCOPY
AS A BASIS FOR A UNIVERSAL/SELECTIVE
DETECTOR FOR GAS CHROMATOGRAPHY

BY

Christine A. Gierczak

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Chemistry

1984

ABSTRACT

THE EVALUATION OF OPTICAL MASS SPECTROSCOPY
AS A BASIS FOR A UNIVERSAL/SELECTIVE
DETECTOR FOR GAS CHROMATOGRAPHY

BY

Christine A. Gierczak

The need for universal/selective detectors for gas
chromatography has increased significantly in the past
few decades. This thesis investigates a new basis for
such a detector which is known as Electron Impact Induced
Emission Spectroscopy (Optical Mass Spectroscopy). Optical
Mass Spectroscopy utilizes electron impact to produce
excited state fragments of gaseous samples. The radiation
emitted by such species is collected and dispersed into
its various wavelengths and is then used to identify
the excited fragments. It has been found through this
work, that a number of functional group fragments (CO, OH,
NH, CH) can be identified using this technique.

ACKNOWLEDGEMENTS

The author gratefully acknowledges Dr. John Allison
for his guidance throughout this work. Appreciation is
also expressed for the expert craftsmanship of machinists,
Mr. Deak Watters, and Mr. Russel Geyer, in the Michigan
State University Instrument Shop. Finally, great thanks
is expressed for the kind help and patience extended by
my husband, G. Mark Allen in the production of the final
copy of this thesis. '

ii

TABLE OF CONTENTS

I. IntrOduction ............OOOOOOOOOOOOOOOOO

I.A. The Electron Impact Process........................3

I.B. Techniques Currently Used to Investigate Electron

Impact Processes...................................4
I.C. Optical Mass Spectroscopy: Fundamentals and
Current Interests..................................

I.C.l. Fundamentals...... ..... ... ...... .................
I.C.2. Current Interests ...... . ......................... 8
II. Experimental ..... .... ................. .... ....... ..19
II.A. Instrumentation..... ..... ........................19
II.A.1. The Emission Source............................19
II.A.1.a. The Dual Chamber Vacuum System. ..... .........21
II.A.1.b. The Gun Chamber..............................22
II.A.1.c. The Collison Chamber ..... . ..... ..............24
II.A.2. The Optical System.............................27
II.A.3. The Spectrometer...............................27
II.A.4. The Detection System .......... .................29
11.3. Experimental.............. ..... . ..... ............29
III. Results and Discussion............................33
III.A. The Determination of the Relationship

Between Electron Current and b-H Emission

Resulting from Electron Impact on n-Propanol....33
111.8. The Determination of the Relationship

Between Total Pressure and g—H Emission

Resulting from Electron Impact on n-Propanol....37

III.C. The Evaluation of Optical Mass Spectroscopy
as a Basis for a Universal/Selective Detector

for Gas Chromatography..........................42

II.C.1. Oxygen Containing Compounds ..... ...............48

iii

III.C.Z.
III.C.3.

III.C.3.

III.C.3.

III.D.
III.E.
LIST OF
LIST OF

LIST;0F,

Nitrogen Containing Compounds.... ...... . ....... 55

Emission from Compounds with Miscellaneous

Functional Groups ....... ..... ............ ......62
a. n-Propanol ....................... ............62
b. n-Bromopropane.. .......................... ...65
General Observations.... .................. . ...... 68
Conclusions...................... ..... ...........73
TABLES ..................... . ............ ..........v
FIGURES.... ................................... ..Vii
-REFERENCES... .................. ..... ............. 74

iv

LI ST OF TABLES

I. Products Resulting from Electron Impact on
Carbon Monoxide........................... ............ 2
II.A. Emission Resulting from Electron Impact
on Hydrocarbons.................... ................ 9
II.E. Emission Resulting from Electron Impact
on Oxygen Containing Compounds............. .. ...9
II.C. Emission Resulting from Electron Impact
on Nitrogen Containing Compounds.......,... .. ..10
II.D. Emission Resulting from Electron Impact
on Sulfur Containing Compounds.... ........ . .. ..lO
II.E. Emission Resulting from Electron Impact
on Halogen Containing Compounds............ .. ..ll
II.E. Emission Resulting from Electron Impact
on Compounds Containing More Than One
Functional Group........................... .. ..12
III. Emission Resulting from Electron Impact
on Inorganic Compounds...................... . ..... 12
IV. Compounds of Interest and Their Sources...... .. ..31
V. Experimental Conditions for the Determination
of the Relationship between Electron Current
and -H Emission................................. ..34
VI. Listing of Filament Currents, Electron Currents,
and the Intensity of 5-H Emission............... ..35
VII. Experimental Conditions for the Determination
of the Relationship between Total Pressure and
-H Emission................................... ..38
VIII. Listing of Total Pressures and p-H Emission
Intensities ...... ................. ......... .. ..39

IX. Comman Experimental Parameters......................43
X. Experimental Parameters that are Specific for
Each Compound........................................44
XI. Emission Wavelengths for the Fragments Observed
in this Work........................................45
XII. The Heats of Reaction for the Formation of CO
and OH from Oxygen Containing Compounds.... ....... .53
XIII. The Heats of Reaction for the Formation of
CN from Nitrogen Containing Compounds....... ..... .59
XIV. The Heats of Reaction fro the Formation of
SH from Thiols.... ..... ................... ......... 64
XV. Comparision of the Compositions and the Emission
Intensity Ratios fro All of the Compounds Studied...71

vi

LIST OF FIGURES

I. Instrumentation ........... . ......... ..... ............ 20
II. Transmission Curve for Optics................ ..... ..28
III. Quantum Efficiency Curve for PMT......... ..... .....30

IV. Plot of Emission Intensity versus Electron Current..36

V. Plot of Emission Intensity versus Total Pressure.....40

VI. Spectrum of n-Propanol ...... .. ......... .............49
VII. Spectrum of n-Butyraldehyde ........... ..... ..... ...50
VIII. Spectrum of 2-Butanone.... ....... ... ...... ........51

IX. Spectrum of Diethyl Ether...........................52
X. McLafferty-type Rearrangement for the Formation
of OH from n-Butyraldehyde...........................54
XI. Spectrum of 2-Nitropropane.... ............. . ...... ..56
XII. Spectrum of n-Butyronitrile........................57
XIII. Spectrum of n-Propylamine ........ . ................ 58
XIV. McLafferty-type Rearrangement for the
Formation of NH from n-Butyronitrile...............60

XV. Spectrum of n-Propanethiol...... .................... 63
XVI. Spectrum of n—Bromopropane..... ...... ..............66
XVII. Spectrum of Air....; ............ ... ............... 67
XVIII. Diagram of the Specific Wavelength Regions

Used for the Detection of Functional Groups ...... 69

XIX. Plot of the Ratio of H Emission Intensities

versus Ionization Potential.... ..... .... ........... 72

vii

CHAPTER I

INTRODUCTION

The use iof column chromatography in the separation of
complex sample mixtures has increased significantly in the
past few decades. Its growth has prompted the development
of substance-selective detectors that simplify the problem
of the positive identification of a wide range of compounds
(1). At the present time, the only detector that provides
both universal and wide range specific detection
capabilities is the mass spectrometer. Mass Spectrometry
employs electron impact processes to ionize analyte
species. The monitoring of the total ion currents produced
by electron impact provides universal detection
capabilities. The monitoring of selected ion currents
provides specific detection capabilities.

This thesis concerns the investigation of a technique
that might serve as a basis for a gas chromatographic
detector. The technique, Electron Impact Induced Emission
Spectroscopy, is similar to Mass Spectrometry in that it
utilizes electron impact to form ionized and .neutral
fragments in their ground states and in excited states. The
excited state species can emit radiation upon relaxation.

This radiation can be optically detected and the emitting

2

Table 1. Products Resulting from EIectron Impact on Carbon Monoxide

PROCESS

‘VIBRATIONAL EXCITATION

ELECTRONIC EXCITATION

IONIZATION
IONIZATION AND
EXCITATION
FRAGMENTATION

DISSOCIATIVE
EXCITATION

DISSOCIATIVE
IONIZATION

IONIZATION ,
FRAGMENTATION
AND nxcxrnlou

CO

CO

e- o C0(v>0) + e-

e- e 00* + e-

e- . 00+ + 2c-

e- * (00+)*'+ 2e-

e- 9 C +’O + e-

C* + O + e-
C + 0* + e-

{'0

e- 9 0+ +'O + 2e-
C + 0+'+ 2e-

{v

e- 9 0* + 0+‘+ Ze-
a C + (O+)* +
2e-
» 0+ + 0*‘+ 2:-
+ (O+)* + 0 +
2e-

PRODUCTS

VIBRATIONALLY
EXCITED NEUTRAL

EXCITED STATE
NEUTRAL

GROUND STATE
ION

EXCITED STATE
ION

GROUND STATE
NEUTRAL FRAGMENTS

EXCITED STATE
AND GROUND STATE
NEUTRAL FRAGMENTS

GROUND STATE
NEUTRAL AND IONIC
FRAGMENTS

EXCITED sun:
NEUTRAL AND IONIC
WIS

 

(*- ELECTRONIC EXCITATION)

3
species can be identified. Such information can be used to
determine the structure of the analyte, as is done in Mass
Spectrometry. Since the fragments are detected optically,
one might refer to this technique as Optical Mass
Spectroscopy.

This introductory chapter is intended to serve four
purposes. Its first purpose is to provide a fundamental
understanding of the electron impact process. Secondly, it
serves to review the techniques that are currently used to
investigate electron impact processes. Thirdly, the merits
of the specific technique of Electron Impact Induced
Emission Spectroscopy (Optical Mass Spectroscopy) are
explored in greater detail. Finally, the objectives of this
thesis, which concerns studies that employ Optical Mass

Spectroscopy, are briefly stated.

I.A. The Electron Impact Process

The electron impact process can be most simply described
as the interaction between a gaseous atom or molecule and
an energetic electron. A consequence of these interactions
is the deposition of a wide range of energies into the
sample species. This can result in the formation of
products that vary greatly in composition, structure and
energy content. Table 1 lists the entire range of products
resulting from electron impact on a fairly simple molecule,

carbon monoxide. Each of the products reveals diverse

4
information about the original sample species and the
unimolecular reactions leading to its fragmentation.
Unfortunately, there exists no single method that can fully
characterize all of these reaction pathways in order to
provide a complete understanding of the electron impact

process.

I.B. Techniques Currently Used to Investigate Electron

Impact Processes

Mass Spectrometery widely employs electron impact as a
method for. the ionization and fragmentation of analyte
molecules. The technique determines the mass—to-charge
ratios and the relative abundances of the ionized fragments
resulting from the electron impact process. This
information is used to determine the composition, the
structure, the molecular weight, and the functional groups
contained within the original molecule. Unfortunately, a
large number of neutral fragments that retain important
information cannot be detected by mass spectral means. In
addition, the internal (electronic, vibrational, and
rotational) energies of these ions cannot be determined
using Mass Spectrometry.

Emission spectroscopy can be used to obtain information
from both neutral and ionic fragments, if they are formed
in electronically excited states that emit radiation upon

relaxation. Ground state and metastable excited state

species can also supply useful information if lasers are
employed to induce electronic transitions. The optical
techniques used to study such electron impact fragments are
Electron Impact Induced Emission Spectroscopy (Optical Mass
Spectroscopy) and Laser Induced Emission Spectroscopy.

The information available from spectroscopic experiments
parallels as well as diverges from that provided by Mass
Spectrometry. A closer look at the types of information
that are furnished by each method will help emphasize the
important role each plays in the characterization of the
electron impact process.

As mentioned above, Mass Spectrometry is employed in the
determination of ionic products and their relative
abundances. This sort of information is most frequently
used to ascertain the composition and structure of an
analyte. Mass Spectrometry can also be useful in
characterizing the decomposition pathways that lead to a
particular ion fragment.

Theshold energies and appearance potentials for a
particualar ion fragment can also be acquired using mass
spectral techniques. The determination of ion abundances as
a function of electron energy is useful in determining
ionization potentials and bond strengths.

Unlike Mass Spectrometry, emission spectroscopy provides
information about neutral, as well as, ionic species. For
example, the structural information furnished by Optical

Mass Spectroscopy can compliment the information available

B
from mass spectra because of an ability to detect neutral,
as well as, ionic fragment emission. Unfortunately, this
information is limited due to the energy range over which
the emission of fragments can be detected, and by the
number of molecules that produce fragments that have well
defined, accessible spectroscopic parameters.

The study of the relationship between emission intensity
of a particular fragment and the electron energy supplies
useful information. Such studies can be used to postulate
fragmentation and excitation pathways, as well as their
relative probabilities. These studies can also be used to
determine whether fragment emission occurs via optically
allowed or forbidden transitions. In addition, the
threshold energies and appearance potentials for a given
excited state of a neutral or ionic fragment can be
determined.

Laser induced emission studies can be used to complete
the picture of the electron impact process. Laser induced
studies on electron impact products serve to identify and
study the amount of ground state species formed, as well as
their vibrational and rotational temperature parameters,
which cannot be determined using Mass Spectrometry or

Optical Mass Spectroscopy.

I.C. Optical Mass Spectroscopy: Fundamentals and Current

Interests
I.C.l. Fundamentals

Optical Mass Spectroscopy is the study of the radiation

produced by electron impact. This radiation can be
collected and directed into a spectrometer and its
wavelength dependent intensity measured by a

photomultiplier tube. Two basic types of spectra can be
obtained using this optical technique; the excited species'
excitation spectrum, and its emission spectrum.

The plot of emission intensity as a function of electron
energy at a given wavelength, is known as the excitation
spectrum. This spectrum provides information about reaction
pathways and ‘appearance potentials associated with a
particular fragment.

The emission spectrum is a plot of emission intensity as
a function of wavelength at constant electron energy. The
electronic structure of an emission spectrum can be used to
identify the type of fragment emitting radiation. This, of
course, provides insight into the composition and the
structure of the sample. Electronic transitions also reveal
information about the various excited states occupied by a
particular fragment. The intensity of the emission
resulting from these electronic transitions and the
energies associated with them can be used to assess a

molecule's energy deposition function, which is the

7

8
probability of depositing a given amount of energy into a
molecule by energetic electron impact. The population
distribution of the vibrational and rotational levels of an
electronic state can be used to determine a temperature

parameter that describes the emitting species.

I.C.2. Current Interests

The majority of the molecules which have been studied
using Optical Mass Spectroscopy are listed, along with
their observed emission, in Tables 2A—2F and in Table 3
(2). The organic species studied seem to fall into one of
two catagories; small straight chain aliphatics, such as
substituted methanes and ethanes, or highly unsaturated
compounds, such as, benzenoid compounds and conjugated
olefins. Approximately twenty or so inorganic compounds
have been studied; the majority of these are simple di- or
triatomic species.

The extent to which information on analyte structure
from emitting electron impact product fragments is
available is not completely evident from these listings.
This is a result of the fact that only specific features of
the emission spectra of many of the molecules were of
interest to the investigators. (Examples of such
experiments will be discussed in greater detail in the
following paragraphs.) Consequently, only those. transitions

that were of interest could be included in these tables.

Table 2A. Emissions Resulting from Electron Impact on Hydrocarbons

SATURATED HYDROCARBONS

Ethane
Methane

UNSATURATED HYDROCARBONS

Ethylene

1,3-Butadiene
1.3.5-8exatriene
Trans-1,3,5,7-0ctatetrsene
1,3 and 1,5-Cyclooctadiene
Benzene

dg-Benzene

Toluene

O,M,P-Xylene

Cumene

P—Cumene

Tetraline

Naphthalene

l-Methyl Naphthalene
Azulene

Acetylene

Discetylene

1,2 and 2,4 Hexadiyne
Tetraacetylene
Triacetylene

Table 28. Emissions Resulting from Electron Impact on Oxygen

In:

\’ ><><>< ><>4

NNNNNNNNNNNNNUNN

Containing Compounds

ALCOHOLS

Methanol

Phenol

l-Nsphthol

ETHERS

Anisole

0,}! ,P-Dimethoxybenzene
Phenetole

KETONES

AcetOphenone
Benzephenone

.2

X

In

NNN

NNN

SE;

MN

NNNNNNNNNQNN

NNN

CH+ C

‘2
X
X X
X X
X
X
X
X
X X

NNN

§4§2t

£21
x
x
x

NNN

PARENT*

.913.

NM

HH

2222222222

PARENT

222

 

10

Table 2C. Emissions Resulting from Electron Impact on Nitrogen
Containing Compounds

AHINES AND IMINES E

In:
[8
>4 x IE

C“ 92§4t 2'2 2&3!!!

 

Ethyleneimine
Analine

N ,N-Dimethyl Anal ine
léNaphthylamine
Methylsmine

NNNN
NNNN
NH
NNNN
N
222

CN . MNTAINING MOLECULES

Hethylcyanide
Benzonitrile
Dicyanoacetylene
Dicyanodiacetylene

MN

HHZ

Table 2D. Emissions Resulting from Electron Impact on Sulfur
Containing Compounds

 

THIOLS _C_ 3 E _C_§2-.+ _(§_ 811+ PARENT
Hethanethiol X X X X X
Ethanethiol X X , X X X

11

Table 2E. Emissions Resulting from Electron Impact on Halogen
Containing Compounds

BRDMINE CONTAINING c g C2§4+ g; Br+ man CBr 'PARENT

>< 53
L?

Bromoacetylene X
Dibromoacetylene X
Dibromodiacetylene
Bromobenzene

l and 2 Bromonaphthalene

NM
H

NM

CHLORINE CONTAINING g

 

Dichloromethane

Chloroform X
Chlorobenzene
Dichlorobenzene (1 ,381 ,4)
1,3,5 Trichlorobenzene
2-Chloronaphthalene
Chloroacetylene
Dichloroacetylene X X I
Dichlorodiacetylene

gm
NNN '9 NM
NM +
L‘?
no
E-
H- ‘
I53
l3
...
A. 8
In
" £3
.53

NNNNNNN
H

MN
H

X
FLUORINE CONTAINING C H CH CH+ C2 gggzi F P+ HE+ CF PARENT

 

Trifluoromethane X X X
Tetrafluoromethane X X
Cis-l,2—Difluoroethy1ene X
Difluorodiacetylene
Perfluoropentadiyne 1,3
Perfluorohexadiyne 2,6
Fluorobenzene X
1,3—Difluorobenzene X X X
1,3,5 and 1,2,4- X
Trifluorobenzene
Tetrafluorobenzene X
(1,2,3,4-1,2,3,5 and
1929695)
Pentafluorobenzene I
Hexafluorobenzene I

HHZHHHHHH

H

IODINE CONTAINING g

_C_H_ + 92 94th:; IiEI-O-EIPARENT
.Iodoacetylene X X
Diiodoacetylene ' X

I
I
Diiododiacetylene I

NNN

12

Table 2F. Mission Resulting from Electron Impact on Compounds
Containing More than One Functional Group

 

 

E g; §__1_ _If_ CCl+ CCl PARENT
Chlorotrifluoromethane X X X X X I
Dichlorodifluoromathane X X X X X X I
Trichlorofluoromethane X X X X X X I

C £N_ 25 gl_ _1_ PARENT
Cyanogen Bromide X I
Cyanogen Chloride X X I
Cyanogen Iodide X I

Table 3. mission Resulting from Electron Impact on Inorganic
Compounds

NITROGEN (DNTAINING

 

 

 

hmnia N N+ H

Cyanogen C CN CN+ CZNZ
Hydrixine NH

Hydrogen Aside NR

Nitrogen N2 N2+

Nitric Oxide N 0 N04-

Nitrous Oxide N200-

O_X}CEN mNTAINING

Carbon Dioxide 002+ 00 00+ 0 C+ C
Carbon Monoxide CO 00+ 0 C+ C
Deuterium Oxide D20+ 01)!-

Oxygen 02+

Hater R O

HALOGEN mNTAINING

Deuterium Chloride D DC1+ Cl+

Hydrogen Chloride H HCl+ 61+

SULFUR CONTAINING

Carbon Disul‘fide CS CSz-O- CSz

Hydrogen Sulfide ‘ B 38+ 328+

13‘

None the less, certain trends that imply that Optical Mass
Spectroscopy could serve as a basis for both a universal
and a group specific detector for gas chromatography can be
identified.

The universal detection capabilities are exhibited by
hydrogen and CH emission that is seen for all organic
compounds. More specific capabilities are indicated in the
spectra of oxygen containing compounds. Alcohols, esters
and ketones reveal emission from OH, CO, or CO+. Many
nitrogen containing compounds, such as amines and and
nitriles, show NH and CN emission, respectively. Thiol
emission indicate the presence of CSZ+’ CS, and ISH+.
Halogen containing compounds have been extensively studied,
and it appears from their listings that X, HX, or CX (X=Cl,
Br, F, or I) emission is not as consistent as would be
expected. In addition, it is interesting to note that only
the larger, highly unsaturated compounds show emission from
either the neutral or ionic parent, and that emission from
smaller aliphatic compounds is limited to atomic or
diatomic emission.

Although, the aliphatic emission is limited to smaller
fragments (H, CH, CN), such emission is simple in structure
and is more often used to obtain specific information about
the fragment species. For example, the analysis of the
intensity distribution in the rotational structure of the
electronic spectrum of the CH fragment produced by electron

impact of simple aliphatics has been investigated

14

extensively by Beenakker , and colleagues (2,3). The
objective of their studies was to determine whether the
distribution over the rotational levels of the zeroth,
first, and second vibrational levels of the Afil state of CH
could be 'described by a Boltzmann distribution. A plot of
the intensity of the rotational lines versus the energy
associated with the transitions revealed that the AUX state
of CH could be described! using a rotational temperature
parameter of 3500i 500K.

Kuchitsu and associates (5) have vibrationally and
rotationally analyzed the CN (BZZ+—> X2Z+) emission
resulting from electron impact on a number of simple
molecules that contain the CN functional group. The spectra
of HCN, CHzCN and CzN2 revealed that the vibrational
populations produced by electron impact are comparable for
these three compounds and correspond to a vibrational
temperature parameter of 7000K. They concluded from these
results that the differences in the bond dissociation
energies for these cyanides do not affect the B‘Z-l- state
vibrational populations. The fact that no irregularities
were found in these measurements also implies that the BZZ+
states of these molecules are formed by the direct
dissociative excitation of each molecule.

The intensities of vibrational and rotational' lines can
also be used to estimate the contributions of cascading

processes from higher excited states to a lower state or to

the ground state, as is implied in the discussion above.

15

Ogawa and associates (6) studied such contributions in the

spectra of nitrogen and carbon monoxide.

The analysis of the vibrational lines of the N2+ (B?L,+€>
X1Z9+) and the N2 (CLHu->>B%T3) band systems produced by
electron impact on nitrogen revealed that no cascading
occurs from higher excited states to either the BZXO+ or
the CHE, states. They concluded that both states were most
likely formed by direct electron impact excitation from the

ground state.

Cascading from higher energy short-lived states to the
Bli+ level of the CO+ fragment is expected from the
vibrational anaylsis of the emission sspectrum of carbon
monoxide. Ogawa's results, in this instance, agreed with
results obtained by Ajello (7) who performed similar
investigations under more rigorous experimental conditions.

Atomic emission, such as the Balmer lines of hydrogen,
can be used to determine fragmentation schemes in molecules
such as methanol or methylamine. The quantitative analysis
of Balmer lines emitted from isotopically labeled methanol
and methylamine revealed that the ratios of the scission
probabilities of CH and OH or NH, q,/% and e /e,, were
0.76 and 0.80, respectively. This indicates that the
formation of the H fragment by electron impact occurs more
readily by OH or NH scission than by CH scission in these
two molecules.

The above examples describe only a few of the specific

16

studies that have been performed on straight chain
aliphatics using Optical Mass Spectroscopy. They are
included in this discussion to emphasize the fact that the
majority of the studies performed on these molecules have
been essencially incomplete and provide very limited
information about the analyte.

The emission spectra of highly conjugated polyatomic
species are far more complicated than the spectra of
atomics or diatomics. The population of many vibrational
and rotational levels is quite significant at room
temperature. This leads to a highly congested spectrum that
may be difficult t0' resolve, and hence, provides little
useful information about the excited species. As a result,
the information obtained from such spectra is usually
qualitative in nature and detailed studies of such systems
are limited to the determination of the lifetimes of the
polyatomic excited states. There are two research groups
which have performed the majority of the studies on
polyatomics using Optical Mass Spectroscopy. Ogawa and
co—workers have done much of the work in the area of
aromatic polyatomics (6). Maier and colleagues have
explored numerous polyatomic open shell cations using
Optical Mass Spectroscopy (8).

Ogawa has investigated over twenty aromatic compounds.
Most of these were derivatives of either beneze, toluene or
xylene. Unlike straight chain aliphatics subject to

electron impact, these aromatic compounds showed intense

l7

emission from excited polyatomic species. In all cases
except for chloro-, bromo-, and nitrobenzenes, emission
from the parent species was. observed. Such emission was
described as being quite intense for benzene and its
derivatives, and was observable for toluene, xylene and
their derivatives.-0gawa also noted that the band emission
from such excited species became broader and more congested
as the sample species became larger and less symmetric.

Maier has investigated polyatomic open shell cations
that are either olefinic, acetylinic or benenoid in nature.
The ground states of these cations are formed by removal of
an electron from the highest filled U' orbital of the
neutral parent. The excited states are formed when
electrons from inner 'U orbitals are excited to the hole
which exists in the highest occupied TT orbital. Maier has
obtained the electron rimpact induced emission spectra of
over one hundred of these cations. The spectra have been
vibrationally analyzed and the lifetimes of the
electronically excited states have been determined for the
majority of these cations.

The vibrational analysis of the emission spectra of
these polyatomic cations is difficult due to an increase in
congestion. Attempts to relieve the congestion include the
supersonic cooling of the sample species prior to impact by
expansion through a supersonic jet. The neutral parents are
cooled to ~2K and are then ionized and excited by electron

impact. Under such condition, only a small number of

18

vibrational levels are occupied. Maier (9,10) and Miller
(11) have employed this techniques and have recorded the
spectra of many of the polyatomic species previously
studied by conventional ambient means.

It is apparent from the above exemplary review that the
studies performed to date using Optical Mass Spectroscopy
have been confined in most cases to the investigation of
specific properties of single species. It is the opinion of
this investigator that a wider range of molecules should be
explored using this technique in order to determine its
capabilities as a specific functional group detector for
‘gas chromatography.

The work described in the following chapters attempts to
evaluate Optical Mass Spectroscopy as a basis for a
universal/specific detector for gas chromatography. The
purpose of this work is actually two fold; to design and
fabricate a simple electron impact induced emission
spectrometer, and to evaluate its utility as a functional
group detector for a series of slightly larger (three to

four carbon chain) aliphatic compounds.

CHAPTER II

EXPERIMENTAL

II.A. Instrumentation

The electron impact emission spectrometer (Figure 1)
consists of the following components: 1) The source of
emission-the intersection of the molecular and electron
beams. 2) The optical system that interfaces the emission
source to the wavelength dispersive device. 3) A wavelength
dipersive device-the spectrometer. 4)The detection system
Each of these components will be discussed in detail in the

following sections of this chapter.

II.A.l. The Emission Source

The induction of emission from electron impact requires
a source of energetic electrons which can be directed
towards a collimated beam of sample gas. The electron
source, or the electron gun, is most frequently comprised
of a tungsten, a rhenium, or a combination tungsten-rhenium
filament. Simultaneous heating and biasing of the filament
are required to produce the electron beam. The. evacuation
of the area surrounding the electron gun is neccesary in

order to avoid contamination and subsequent failure of such

19

20

coaumucmaduumcH .H muomflm

 

 

 

 

 

        

 

 

 

 

 

532810022
N 8s_-x..tm
x
was.” “an.“
zombuufi
14¢... J<U.Z<IUg
.38
g
< 29mins
<
14m:
0400

 

 

 

 

 

 

I 230 Zgbwqw

mdmidm

 

 

 

mun—Souk
#1410 Exhw

21

filaments. Most frequently, differentially pumped vacuum
systems are employed. This provides for separate gun and
collision chambers, which further reduce the chance of

filament contamination and oxidation leading to burnout.

II.A.l.a. The Dual Chamber Vacuum System

The walls of the dual chamber vacuum system, that
contain the emission source, were machined from 0.75" thick
aluminum stock. Aluminum stock was chosen instead of
stainless steel stock because of its ease of machining and
its reduced weight. All walls, exclusive of the top wall,
were arc welded into place. The top wall acts as a large
flange which allows for easy access into either the gun or
the collision chamber. The exterior dimensions of the dual
chamber system are 12"x12"x23". The collision chamber
interior dimensions are 10.5"x10.5"x14.25". The gun
Chamber's interior dimensions are 10.5"x10.5"x6.5". All
flanges that required soldered connections were machined
from 0.75" thick brass stock. All other flanges were
machined from 0.75" thick aluminum stock. All flange seals
are of the o—ring type.

Each chamber can be individually rough pumped or
diffusion pumped. The gun chamber has a separate rough
pumping flange. This flange connects the gun' chamber to a
vacuum pump (Dou Seal Model R—1376) via a Veeco o-ring seal

valve and 1" diameter copper tubing. In addition, the gun

22

chamber also accomidates a diffusion pump flange. This
flange connects the chamber to a liquid nitrogen trap that
is attached to a 2" diffusion pump (Consolidated Vacuum
Corp.). The collision chamber may be rough pumped or
diffusion pumped through the appropriate valve with a 4"
diffusion pump with a liquid nitrogen trap (Varian Model
M—4).

Each chamber also accomodates an ionization gauge tube
flange that allows one to determine the pressure of either
chamber. Ionization. gauge tubes (Grandville Phillips Model
274002) are inserted into the Cajon fittings of these
flanges. The chamber pressures are monitored using a
ionization gauge controller (Grandville Phillips Series
260). The system’s fore pressure can also be monitored
using the thermocouple connections of this same gauge
controller. Thermocouple gauges (Grandville Phillips Model
L2) are placed between the base of each of the diffusion

pumps and their vacuum pump connections.

II.A.l.b. The Gun Chamber

The gun chamber serves two purposes; it houses the
electron gun, and it provides for electrical connections
between the_ gun and the exterior gun controls. An interior
flange, situated on the wall that is comman to both
chambers, supports the gun components. An exterior flange

provides for four ceramically isolated electrical

23

connections to the interior of the gun chamber.

The electron gun is comprised of three basic components;
the filament holder, the repeller, and the accelerating
grid. The filament holder block was machined from 1" thick
stainless steel. This block is 3.75" long and 1.25" .wide.
The filament supports consist of two copper rods, 0.15" in
diameter and 2.25" long. The ends of the supports are
threaded into 0.5"x0.25"x0.25" copper blocks. Each block
accomidates a 0.08" diameter screw and washer that is used
to center the filament and to hold the filament tightly and
tautly in place. The filament supports pass through two
ceramic isulators that have been inserted into each end of
the filament block. The positions of bOth the filament
supports and the ceramic insulators are secured by set
screws. The repeller is a piece of coarse steel mesh that
can be easily inserted into the back of the filament block.
The grid was machined from 0.032" thick stainless steel. A
strip of 0.25" wide, 1.75" long strip of fine steel mesh
was spot welded over over a hole in its center. The
filament block and the grid have 0.20" diameter holes
drilled through each corner. Steel threaded rods that are
covered by ceramic insulators are passed through these
holes and are then threaded directly into the flange.
Vespel spaters are used on each rod to electrically isolate
the filament holder and the grid. The gun components are
secured into place by four nuts that are tightened onto the

end of each rod. The flange contains a 0.25" wide, 1.70"

24

long slit to facilitate the passage of the electron beam
into the collision chamber.

Two independent power supplies are required to operate
the electron gun. A constant current supply (Harrison
Laboratories Model 808A) heats the filament, and a constant
voltage supply (Hewlett-Packard Model 6207B) biases the
filament with respect to the grid. The repeller and
filament block can be independently biased if a third power

supply is employed.

II.A.l.c. The Collision Chamber

The collision chamber houSes the sample inlet system,
the electron collector, the light shields, the interior
lens holder and the lens mount, as well as the windows
through which emission is observed. The chamber is designed
so that the electron beam, the molecular beam, and the
optical axis are mutually perpendicular. This is achieved
by placing the sample inlet system on the top of the
chamber (the molecular beam is along the z-axis), the
electron gun on the interior comman wall (the electron beam
moves along the y-axis), and the two observation windows on
each of the elongated sides of the collision chamber (the
optical-axis comprises the x—axis).

The inlet system was constructed from 0.25" diameter
stainless steel tubing. All connections between .the tubing

pieces were made using 0.25" diameter Swagelok fittings.

25

Sample gas flow is controlled by a needle valve. The sample
tube is connected to the inlet system using a 0.25" Cajon
fitting. The sample inlet extends 6" down into the
collision chamber. The first 2" of this interior extension
is 0.25" diameter stainless steel tubing. A 0.25" diameter
Cajon fitting has been soldered to the end of this
extension. A 4" long, 21 gauge stainless steel needle has
been placed in this interior Cajon fitting and serves to
roughly collimate the sample beam.

The electron collector is comprised of a 2" diameter
aluminum cone which is attached by a wing nut to a 0.5"
diameter aluminum post. The pbst is seated in a 3" diameter
Plexiglas base (to prevent electrical contact of the
collector with the collision chamber floor). The collector
is attached by a bare nichrome wire to an electrical
connector. This ceramically coated electrical connector is
located in the flange on the collision chamber wall that is
oppisite the wall on which the gun is mounted. An
electrometer (Keithley Model 610C) is used to evaluate the
electron current that passes from the filament to the
collector.

Two L—shaped aluminum light shields have been attached
to either side of the 0.25" wide slit that has been
machined through the gun flange. These shields extend the
entire length of the slit and serve to block the light
emitted from the hot filament from being collected by the

optical system.

26

The interior lens mount was machined from 0.25" thick
aluminum stock. The three mount components allow for
focusing adjustments along all three axes. The lens was
placed into a flat black plastic holder that is then
attached to the lens mount with a thumb screw. The lens,
when properly focused, will transfer the emission image
through the observation window to the exterior optical
system.

The observation windows are used to tramsmit radiation
from the interior, low pressure region of the collision
chamber to its surrounding atmospheric pressure region. Two
observation windows are employed. One window is used to
view the emission source, and the second window is a
component of the actual optical system. This window
transfers the light collimated by the interior lens to the
exterior optical system.

Each window is attached to the exterior walls of the
collsion chamber by a dual flange system. The outer most
flange provides an o—ring seal between itself and the outer
surface of the window. The inner flange provides an o-ring
seal between the flange and the inner surface of the
window, as well as between the inner flange and the

exterior surface of the collision chamber wall.

II.A.2. The Optical System

The transfer of the image of the emission source to the
spectrometer is achieved using a combination of two lenses
and a single window. The lenses have a focal length of
101.6 mm and a 38.1 mm diameter (Ealing Corp. Cat. No.
34—6791). The fused silica windows are 9.5 mm thick and are
50.8 mm in diameter (Ealing Corp. Cat. No. 35-9471). The
transmission curve for these fused silica optics is shown
in figure 2.

The first of the two plano convex lenses is placed in
the lens holder inside the collision chamber. The lens is
positioned so that the emission source is located at its
focal point. The light emitted from the source is then
collimated by the lens and passed through the observation
window to the second exterior lens. The exterior lens is
positioned so that its focal point is located at the center
of the entrance slit of the spectrometer. This lens serves
to refocus the collimated image onto the entrance slit.

This lens configuration results in a 1:1 image transfer.

II.A.3. The Spectrometer

A Spex (Model 1700—11) spectrometer with curved,
adjustable entrance and exit slits is used to disperse the
emission into its various wavelengths. The Spex 1700-11 is
a 0.75 meter. focal length spectrometer that employs a

symmetric Czerny-Turner configuration. This spectrometer is

27

28

 

100

80

 

 

 

 

 

 

70'

60

 

 

 

50
40

 

% TRANSMISSION

30

 

 

 

20

 

 

 

 

10

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1

 

Figure 2.

.2 .3 .4 .6 .8110 2.0 3.04.0 6.0 100

Wavelength pm .
FUSED SILICA (10 mm thick)

Transmission Curve for Optics

 

 

 

 

 

 

29

most frequently used in the scanning mode. The scan rate
can be controlled using a Servo Tek (Type ST—584-1) DC

velocity servo motor.

II.A.4. The Detection System

The dispersed radiation is detected using photoelectric
techniques. The light impinges on a photomultiplier tube
(EMI Type 9558QA- quantum efficeincy curve shown in figure
3) is powered by a high voltage power supply (Fluke Model
4128). The currents produced by the tube are converted to
voltages using a picoammeter (Keithley Model 480). The
voltage changes resulting from the scanning of'a wavelength
region are monitored using a strip chart recorder (Soltec

Model 1242).

II.B. Experimental

Nine separate samples were investigated using Optical
Mass Spectroscopy. Table 4 lists the samples chosen and
their respective sources. All samples were in their liquid
phase at room temperature and were placed in a Pyrex sample
tube (0.5" in diameter and 3.0" long) with a teflon
stopcock. Approximately two millimeters of the sample were
required for each run. AlI samples were prepared for use by
sucessive freezing (using liquid nitrogen) and degassing
steps.

The following steps must be followed in order to

h) 00 CO 45
CD CD (Ii CD

Quantum Efficiency %

-a NO
. 01 O

..s
O

5
0

30

l T

 

1
..-- '0' variants
' i

 

 

 

 

 

 

 

i
I‘--.0'\l‘~,/\ 4820

,s20(96583)
MI[\ 1 i I

1
595588)

 

 

T‘ l 1
I.V\%*\Extended 820 (9659)

 

\

 

 

 

 

 

 

 

 

 

 

 

 

 

100 200 300 400 500 600 700 800 900

Figure

Wavelength nm

3. Quantum Efficiency Curve for PMT

31

Table 4. Compounds of Interest and Their Sources

COMPOUND
n-propanol
2-butanone
diethyl ether
n-butyral dehyde
2-nitropropane

n-butyronitrile

n-propylamine
n—propanethiol

n-bromOprOpane

m
Mallinckrodt-Paris, Kentucky
Hallinckrodt

Mallinckrodt

Chem Services-Nest Chester, PA.
Chem Services

Natheson, Colman, and Bell-
Cincinatti, Ohio

Aldrich-Milwaukee, Wisconsin
Fluka-Eauppage, N.Y.

Fisher Scientific-Fairlawn, N.J.

32

initiate and record the sample emissions from elctron
impact: 1) Set PMT voltage to 700V and turn on Fluke power
supply. 2) Pass 6.5A at ~10V through the filament (3%
Rhenium, 97% Tungsten 0.006" diameter filament, Scientific
Instrument Services Cat. No. W45) using Harrison power
supply. 3) Introduce a sample pressure within the range of
l-4x10E—4 torr through needle valve of inlet system. 4)
Raise the bias on the filament slowly to 150V using
Hewlett— Packard power supply. (At this point localized
emission from the excited species should be observed at the
tip of the inlet needle.) 5) Adjust the picoammeter range
and the full scale voltage on the recorder according to the
intensity of H* emission at 4861A. 6) Scan the 4900—2000A
wavelength region at 106A/min. and record the resulting

spectrum using a chart speed of 30mm/min.

CHAPTER III

RESULTS AND DISCUSSION

Three distinct types of experiments have been performed
using the instrumental set up described in Chapter 2. The
first two sets of experiments evaluate the general
performance and efficiency of the system. A third set of
experiments assesses the system's ability to serve as a
functional group detector for a series of three to four

carbon monofunctional organic compounds.

III.A. The Determination of the Relationship between
Electron Current and b-H Emission Resulting from Electron

Impact on n-Propanol.

The electron beam current that results in optimum
emission intensity and optimum sensitivity can be
determined by varing the current used to heat the filament
and measuring the effect of this variation on the emission
intensity of a given transition. The E-H emission (n,=4+»
n =2, at 468 nm) resulting from electron impact on
n—propanol was chosen for this study. Table 5 lists the
experimental conditions under which this study was
performed. Table 6 lists the resulting data. Figure 4 is a

plot of emission intensity versus electron beam current

33

34

Table 5. Experimental Conditions for the Determination of the
Relationship between Electron Current and p—B mission

PARAMETER SETTING

CURRENT THROUGH FILAMENT VARIABLE

BIAS 0N EILANENT 150 VOLTS

BASE PRESSURE 2.0 x 10-5 TORR
TOTAL PRESSURE 1.0 x 10" TORR
PEOTONULTIPLIER BIAS 700 VOLTS

SLIT wIm'E 250 pm
PIOOANMETER RANGE 1-20 x 10-9 AMPS

ELECTRONETER RANGE 1-10 x 10"2 AMPS

35

Listings of Filament Oxrrants, Electron Currents, and

Intensity of p-H mission

Table 6.

EMISSION (AMPS x 10-9)

(AMPS x 10-2)

FILAMENT CURRENT ELECTRON CURRENTS INTENSITY 0F F-H

(AMPS)

TRIAL N0 .

9‘0505
o a e a a s
222110

50.50.50.
6.6.554.‘

111111

067320
so a o o s
321111

050505
000000
322110

505nw50.
6.65.56u4.

222222

957210
0 O. O O 0
221.11....

0.30.30.5
322110.

30.30.50.
0633‘.“

333333

444044

nU.6~I.211
321.111..

J50.30.3
weaning/.110.

50.50.50.
6..m554.4.

533535

PEAK INTENSITY / A"

C) l L

36

 

l

o I 2 3 4
ELECTRON CURRENT (IN A")

 

Figure 4. Plot of Emission Intensity versus

Electron Current

37

produced from these results.

Figure 4 indicates that the intensity of the B —H
emission resulting from electron impact on n-propanol
increases linearly with electron currents above 1.5x10E-2
A. This implies that maximum sensitivity is achieved at the
maximum electron currents attainable using this

instrumental set—up.

III.E. The Determination of the Relationship between Total
Pressure and b—H Emission Resulting from Electron Impact on

n—Propanol.

The determination of the optimum total pressure
(pressure resulting from base pressure and sample pressure
in the collision chamber) was performed for two reasons.
First, it is important to maintain the lowest possible
total pressure in order to reduce the chances of
contamination and oxidation leading to filament burnout.
Secondly, one must assure that the product emission is
representative of single collision events which accurately
characterize the electron impact process.

The fl—H emission resulting from electron impact on
n-propanol was again chosen for this study. Table 7 lists
the experimental conditions and Table 8 lists the
experimental results. Figure 5 is a plot of emission
intensity versus total pressure produced using this data.

Figure 5 shows that the relationship between the b—H

38

Table 7. Experimental Conditions for the Determination of the
Relationship between Total Pressure and p41 mission

PARAMETER SE'I'I‘ING

CURRENT THROUGH FILAHENT 6.5 AMPS

BIAS ON FILAMENT 150 VOLTS

BASE PRESSURE 2.0 x 10-5 TORR
IOIAL PRESSURE VARAIBLE
PBOTOMULTIPLIER BIAS 700 VOLTS

SLIT wrnra 250 pm

PICOAHMETER RANGE 1-100 x 10-9 AMPS

39

Listing of Total Pressures and p-H Emission Intensities

Table 8.

INTENSITY 0F p-H EMISSION
-9)

(AMPS x 10

9‘3

TOTAL PRESS
(TORR x 10'

TRIAL N0.

8
J5937°~J29

6.6.4.26.2110.
1111

6.6.0.242
0508-].7539m
21‘.

111111111

5
970.8,“:“270”

9.0n83200210.
11111

222222222

1.12.3060

1162511
2211

3333333

20.6378973Qn

QHQMLLo.6.1.110.
11111

“0.4.6.J.b.9.6
059.987.6531
21....

4444444444

J4750.0.63°n

663182110.
1111

804445

.0 O. O 0
940886431
111

555555555

40

h) h)
C) 69
T 1

 

PEAK INTENSITY/A"
5 fi '5 6 '6

 

 

J l J l 1

.0 3.0 7.0 I l.0 ISO 9.0

CDC) '0 35 (h 00
1

TOTAL PRESSURE / no" TORR

Figure 5. Plot of Emission Intensity versus
Total Pressure

41

emission intensity and the total pressure is linear within
the pressure range of 7.0-12.0xlOE—S torr. The leveling off
of this curve at presssures above 12.0x10E—5 torr may be a
result of the instrumental design. The ionization gauge
tube that is used to monitor the total pressure is situated
on the far side of the front wall of the collision chamber.
Thus, the pressure measured by this tube might not
accurately reflect the sample pressure at the tip of the
needle (the point from which emission is collected).

The plateauing of the curve might reflect the vacuum
pump's ability to pump away sample gas at lower total
pressure and its decreased ability to do so at higher total
pressures. In other words, the molecular beam remains more
highly collimated at lower sample pressures, resulting in
localized, intense emission in the observation region. ’At
higher total pressures the sample might be dispersed over
the entire chamber resulting in diffuse, less intense
emission in the observation region. In addition, the
dimensions of the electron beam are large relative to those
used in other experiments. The presence of gaseous
molecules in an electron beam reduces its divergence. Thus,
when the total pressure is increased, there is increased
emission due to both an increase in the number of electron-
molecule collisions and a pressure-induced increase in the
electron flux. Thus, it is difficult to make any
conclusions regarding .the shape of the curve or the sole

existence of single collision events.

III.C. The Evaluation of the Optical Mass Spectroscopy as a
Basis for a Universal/Specific Detector for Gas

Chromatography.

Nine different functional groups were chosen to assess
the utility of the Optical Mass Spectrometer as a group
specific detector. Sample selection (listed in Table 4) was
based on such properties as; vapor pressure, toxicity, and
availability. Many of the compounds selected were oxygen
containing. Such samples frequently show emission from
excited CO and CO+ fragments. Such emission is also
prevailent in the specta of carbon monoxide and carbon
dioxide (6) and in order to accurately interpret the
spectra of oxygen containing compounds, a spectrum of air
was obtained under similar experimental conditions. In the
event that a sample's spectrum is contaminated by air
peaks, resulting from undetectable leaks in the vacuum
system, one can eliminate those peaks which might not
result from sample emission.

Many of the experimental parameters were held constant
for all samples. A list of ‘these parameters appears in
Table 9. The remaining parameters are presented in Table
10. Table 11 lists the transitions observed for all of the

compounds studied and their wavelength values.

42

43

Table 9. Conan Experimental Parameters

PARAMETER szrrmc
wAVELENc'm REGION EXAMINED .90-200 mi
CURRENT manual FILAMENT . 6.5 AMPS
BIAS on FILAHENT 150 vows
mnocnRouATon SCAN RATE ' lOGUIin.
PHOTOMULTIPLIER BIAS 700 vous

cam RECORDER SPEED 30 minin.

44

o.o~
o.N
o.—
o.~
o.~
o.~
o.~
c.N
o.N
o.~

uu<HAO> NA‘Um
Junk “undoomu

alo~
alo~
«lea
mica
one“
alo~
wacu
alo—
one"

«nod

u o~u~
N cul—
M on!“
I out—
N o—an
K oun—
N Out—
K cal"
N outs
u O~I~

Ammz<v

muz<& mmhwzx<OOHm

can
can
can
com
com
ocn
can
can
nNN

DON

Anne
mannn.augm

«Ian M o.e
clan u o.n
«nod u o.«
elem u c.n
«nod I Q."
clan H o.~
elem x n.n
clam u N.N
«lag u n.n
«log N c.~

fiancee
snowman. a<aos

eno— u o.~
ones x o.~
etc“ a N.~
clo— x o.n
one" x o.~
cued u c.n
one“ x o.~
onc— K n.~
muo— u o.“
eno— N o.N

Ammoav
nmammmmm umcm

has
oaunounoaouAI:
.Hoafiuocuoounsc
oualuameouenc
oauuuaoouhuanlc
onenoueouuaolw
non». Hug»...
ouoauusonn
ovhsoeauuhuanna

doaoaounln

azaomzou

noose-nu noun you uwwwoonm ou< yank nuouoauuom Huucoiuuooxu .o~ canoe

45

 

 

 

Table 11. Emission wavelengths for the Fragments Observed
in this work
PRAGMENT TRANSITION OBSERVED (v' av) PEAK HAVELENGI'HS (A)

H n2-4- nl-Z 4861
H nz-S- nl-Z 4341
B 112-6" 111-2 ‘10].
H n2-7- nl-Z 3970
11 32-8- 01-2 3889
H 02’9“ “1.2 3835
ca 126- x211 (0-0) 4315
co 312+- Alrr (0-0) 4513
(0-1) 4836

00+ Azfl- x2z+ (1-0) 4557
(2-0) 4264

(3-0) 4011

(4-0) 3790

(5-0) 3595

(6-0) 3423

(2-1) 4702

(3-1) 4396

(4-1) 4132

.(5-1) 3902

(6-1) 3700

(4-2) 4535

(5-2) 4260

(6-2) 4020

(5-3) 4683

on A2£+- x211, (0-0) 3086
03+ A3Tr; - x3z- (0-0) 3578
(1-0) 3342

(1-0) 3708

an azu- x22+ (0-0) 3876
(1-0) 3582

(0-1) 4214

(1-1) 3864

(1-2) 4191

(2-3) 4174

(3-4) ' 4161

(4-5) 4149

NH A3“: - x3z- (0-0) 3359

46

Table 11. Emission Wavelengths for the Fragments Observed

in this Work.(cont.)

ERAGNENT TRANSITION OBSERVED (v'ev) PEAK WAVELENGTHS (A)

HBr+’ A?z+- x27; (0-0) 3584
(1-0) 3421
(2-0) 3280
(3-0) 3158
Br 5. ‘r3/2- 6p ‘21/2 4785 (2)*
- 6p ‘85/2 4780 (3)
- 6p ‘83/2 4752 (5)
5." 2P - 6p" 2P1/2 4775 (4)
53' 2P - 5p.2D5/2 4694 (10)
- 6p' 2D5/2 4615 (12)
- 6p' ‘ 3,2 4576 (14)
5. ‘85 - 6p ‘ 5’2 4526 (16)
- 615-4113,2 4513 (17)
- 6p“D7/2 4478 (19)
- 6p ‘05,: 4442 (20)
53' ‘Rllz- 6p' ‘81,2 4490 (18)
- 7p ‘8 3,2 4365 (21)
Br+ 4f' 363- 5d 302 4816 (1)
- 5d 3D3 4622 (11)
63' 102- sp' 1F3 4728 (8)
- 5p' 1P1 4543 (15)
5s" 392- 5p" 303 4742 (6)

 

(* - transition that correSponds to numbered peak on spectrum
of n-bromOprOpane) '

Table 11.

FRAGMENT

Br+

47

Emission Wavelengths for the Fragments Observed

in this Work (cont.)

TRANSITION OBSERVED (v'ov)

68' 3D&- Sp' 3?

5 3R - 4 3R
p 1 p 2

5d' 30 - Sp' 10
3 2

2

5p" 302- 58" 3P1

63' 3D - 5p' 30
3 3

5d' 381- Sp' 3?

5d' 10 - sp' 3R
4 2

5d' 3R - sp' 1R
2 3

l

4786
4735
4705
4693
4601
4366
4224

4180

PEAR WAVELENGTHS (A)

(2)
(7)
(9)
(10)
(13)
(21)
(22)

(23)

III.C.l. Oxygen Containing Compounds

In addition to the Balmer series hydrogen emission and
emission from CH(A?A- XLU), which is present in the spectra
of all CH containing molecules, the presence of CO(BVZ+)
and CO+(ALU) is common to all of the spectra of the oxygen
containing .compounds. The intensity of the emission
resulting from these excited states is significantly less
for n-propanol (Figure 6) and n-butyraldehyde (Figure 7)
than it is for 2-butanone (Figure 8) or diethyl ether
(Figure 9). This difference might be explained by the fact
that n—propanol and n—butyraldehyde can undergo
decompoSitions to form OH via more energetically favorable
pathways than the pathways resulting in the formation of CO
or CO+. Such lower energy pathways are not ayailable in the
case of 2-butanone or diehtyl ether (see Table 12). The
presence of OH (A3Z+- Xzfli) emission in the spectra of
n—propanol and n—butyraldehyde reinforces this hypothesis.
It is important to note, at this point, that the presence
of this OH emission in the spectra of n—butyraldehyde
implies the possibility of the OH bond forming via a
McLafferty— type rearrangement, which is one of the basic
mechanisms of ion fragmnetation that is seen to occur
frequently in Mass Spectrometery. This fi—H rearrangement to
a radical site on an unsaturated group (the carbonyl group
in the case of n-butyraldehyde) occurs by a sterically

favorable six—membered ring transition state (Figure 10).

.48

49

        
  

 

..Euxl...N~$..6

[
.-rxia .548

9h can
. 4

....IL
ruff-E «$98

.. Wear? .3

>000. . >§ gquw

[
..w.x..=.s.8 _
P b
a. .5. .5: .extahiu

gmmwcn 44:8

.5: c: IPOZNJN>43

con 00v 0..
u . ..

rwuxaéuSBU

Hocmmoumlc mo Esuuommm

ON. One o: 00' 8' 0.3
_ la . a . .

 

8. 6.0.3.:

“— 824? u:

 

_

a. ....w. .5:

 

.m musmflm

AMSNBINI

 

 

LI

3. ......s:

50

momsmoamumusmlc mo Esupommm

E! c: Ends}

.5 musmflm

 

  

 

 

 

5.? I 0” a gmmUE J‘EH ggic>Fa 1‘
>08. . >03!” gouaw

 

 

on own 03 ofl OR 2.. 08 can 8. 9. 8. On. 0.. on. 8. 8. o... o...
d a J 4 u . l q 4 J a . u . u . . .
13.— xl’h . 3‘3 \
u ...:o
[
E a 9.923
o - .
a. ... s. .5: ...-15......
a. (...-...:
a. 6.9.5:
[ [ [ rIIIL [ r11 1L
2......0u 33.3.0”. .....Sbu 2.0.3.00 393.0”. ..Hunoc-Sbu $92.0...
82...»?!

 

A1ISN31NI

51

 

W.

 

 

 

 

 

 

maocmvnmlm mo Esuuommm .m musmflm
.6: c: Ikozm4w><3
ocm Own own Ohm can own 00v 0:. ON.» one 03. One 8.. 05¢ 00¢ 09.
MIN . IIVLJH! . d 1 q A q q u q d d
.L
E .4 Tow .900
$15.5: #
r 1 .
I E Ala ....o .|
$6.98.: ux #4 m
3
N
8
w I.
I ..A
a. ....n. «51
E FL rlL .IIIL rIL ...IIIL .IIIL
.xTSbu .xAudrou 32:43.00 €215.00 $213.00 ..Huxfcusoou .xfsoou

mo» .b_x~.~ . mgmmmma 44.5.? -
>89 . Gmwzm 28.53“; mzozflam ~

 

a. ......«5:

52

Hmnum ngpmflo mo Esnuommm ..m musmflm

“E: c: Ik02w4w><3

own Ohm on» can 00». 0:. ON». onv 03. one 80 one 00¢ 09.
{q .llr . . . T . . q . . d _ _

...

AN. .6 Nu NCvI

 

   

 

.IL
E .4 AJN .900

     

 

 

 

 

 

a. 56.8.: m
3
N
B
F L I.
EuxélduSIU A
[ ..IIL IIHIII .IIL [IL
3.38 3.5.8 3. 5.9:... ..wuxéusbu 23:8

 

Ll

«carbinnécammmca .58 5:5 ._ E 55 8......3...
>89 . .652”. 22.5me

53

Table 12. The Beats of Reaction for the Formation of CO and OH
from Oxygen Containing Compounds

 

COMPOUND PRODUCTS HEAL 0F REACTION
G3C32m208 . CH3CHZ + (I) + HZ + B 113.0 heels
-0 (330.33 + 00 + 32 . 15.6 kcels
+ (B3CHZCE2 + OH » 93.7 kcals
CB3GZCB2CDH 9 m3CHZCHZ + CD + B 101.5 kcals
9 (33012033 + CO 2.5 heels
4 04117 + OH 89.0 locals
ca3cnzocazm3 . CH3 + 00 + 112 + @3an 92.0 heels
9 0114 + (X) + 0130113 4.13 kcals
. (330120113 4» Hz 4» CO 9.0 heels
(33(DCEZCH3 4 (33 + (I) + 03263 90.2 kcnls

54

:1:
~—O
(\m

-::—-...Q+

/.-

J 1*. /
\CHHZ/ \ \CH2_.{ H

 

* H
/
+O
ll
éH2\ C
CHZ—CHZ \H
FURTHER
DECOMPOSITION

 

1

OH

Figure-10. McLafferty-type Rearrangement for the
Formation Of OH from n-Butyraldehyde

55

Unfortunately, no other aldehydes have been studied using
Optical Mass Spectroscopy, and therefore, no positive
conclusions can be made about the effects of chain length
on the probability of this OH formation.

The presence Of the OH (A1Z+-v XZTT; ) emission in the case
of n-propanol and of other alcohols investigated and the
presence of CO+(A‘TT—XZZ+) emission in the spectra of a
variety of oxygen containing compounds studied here and by
other groups (12,13,14) under similar experimental
conditions implies that Optical Mass Spectroscopy could be
used to selectively detect OH and CO bonds in oxygen
containing compounds by monitoring the 3200—30503
wavelength range (in order to detect OH emission) and the
4075—3975R wavelength range (in order to detect CO+

emission).

III.C.Z. Nitrogen Containing Compounds

The presence of CN(B2Z+) is evident in the spectra of
all three of the nitrogen containing compounds studied
here. The emission resulting from CN(BZZ++ XZZ+) is very
weak in the case of 2-nitropropane (Figure 11). Again, this
might be explained using a thermodynamic approach. The heat
Of reaction for the formation of CN from 2—nitropropane is
significantly greater (Table 13) than that for the
corresponding formations from n—butyronitrile (Figure 12)

or n—propylamine (Figure 13). As a result, lower energy

56

OCOQOHQOHUflZIN mo Esuoommm .HH musmflm
.E: c: Ihozwagds)

OnN 8N OKN 8N 8N 8n 9n Own an O'n Dan 08 Ohm 8n 0mm 00v 0; 03. one 0: 09. 8c 05' 09. owe

- d d d H d d u d I! d q q u d 1 d d 1 d u d J n J-

ll i 1"»? 1 .1441 I? 11 i i 1 {Ill H \Ii 4" '14?
r...
..w... 1.595 ~
. rIIIIIIIL
AN. .CN. “Cu... ACNX-<~‘=&U

.N. 6.? ~51

 

 

 

AlISNBlNl

 

 

“No 6.“. “Cu:

lit. #

 

 

.. . on.
«no». .95.? .wSmmuE .35; uzcaoEomtz-~ a .. ....

v>ion.. .>o¢wzw zomkuwaw

57

 

oaflnuflcoumusmlc mo Esuuoomm .Na ousmflm
:5. e: Ikozw4w><3
on» can own Gen 03 8m 00.. 0:. ON: On. 03 On. om. on. 05. 03
T 1 I1 . . . _ _ . \I—F . _ [4‘ . . _

        
   

[
£11552:

[

..u...

T

I 995

[III.L
..w.x1.w~92u

to» 70:0. . wSmmwca ..dk
>000. . KEEN zozhow

5.7. .995 .5753:

 

.N . ... Rune... E. 66.3.:

.N. ....n. «51
0....

4L waitéta -c

 

.Ll:

.w. 6.7.5...

 

 

 

ALISNBLNI

58

OGHEOHmmOHmIc mo Esuuoomm

.6: c: Ihozw4u><3

 

 

 

 

 

9% can own 95 can own 00. 0.. cm. on. o... On. ow. o... om. om.
q u d . u _ q d 4 q d a 4 Fill T I .
[ rL

51.1.54 .12 ..wux... 3.95 4. .IIL. .

_..m

Pull-IL
“OWNX’Qwflmvzo

5.7.5.95 Euxstxu

 

 

.N. .c K. «5...

mo» ...9 .6.» . mgmmumn. ..38
>89 . .655 5585

 

a. 6.98.:

 

a. ....n. «......

H— wz.2<..>n.omn. .:

 

 

5.6.83.1

 

.2 833m

AllSNBlNI

59

Table 13. The Beats of Reaction for the Formation of CN from
Nitrogen Containing Compounds

 

COMPOUND PRODUCTS HEATS OF REACTION
CH3<112632CN . m3m2cuz + on 111.2 koala
013620121“! 9 (113(112 + CH + 282 142.3 koala

., (3363 + on + 32 + B 169.15 koala
(33011102633 9 2013 + m + a + 02 2am. heals

. CH4 + (33 + 02 + CH 145.3 koala

o m3m3 + 02 + m + B 161.9 kCllfl

CH2

Figure 14.

6O

 

 

 

—b-CHZ

/////C /////
\\\\\CH2——-CH2 \\\\\CH2-—-CH2

 

O___Z\I

CH2

\\\\\CH2———CH2

FURTHER
DECOMPOSITION

 

V

NH

McLafferty-type Rearrangement for the
Formation of NH from n-Butyronitrile

61

decomposition pathways might be favored in the case of
2—nitropropane.

The presence of NH(AHE ) in the spectrum Of
n-butyronitrile again stresses the possibility of a
six-membered ring transition state in the formation of the
NH bond (Figure 14). This result and the OH bond formation
in the case of n-butyraldehyde, reinforce the fact that
unimolecular fragmentation rules that have been developed
in Mass Spectrometry can be used in Optical Mass
Spectroscopy as well.

Emission resulting from the NO(BLH-+ XfiT) and the
N0(KZ:+—.XHT) transitions within the 2100-2300A wavelength
region have been Observed for other N0. containing
compounds (6). Unfortunately, this emission is not observed
in the spectrum of 2-nitropropane. This might be a result
Of the fact that the grating used in the Spex—1700 II is
blazed at 50005 and therefore, its efficiency in the
2100-23003 range might be too low to allow for the
detection Of this emission. (NO emission has been Observed
below 30005 in the spectrum of any sample studied using
this system.)

Emission from the NH(ALW;) state in the spectra of
n-propylamine and in other amines studied (6) under similar
conditions implies that Optical Mass Spectroscopy could be
used to detect NH containing compounds by monitoring the
3350-34003 wavelength region.' The presence of CN(B§Z+) in

the spectra of all compounds containing 3 CN bond studied

62

in this work and in a wide variety of other investigations
(6,5) verifies that Optical Mass Spectroscopy can also be
used to specifically detect analytes containing a CN bond
by monitoring the 3900-3800A region.

Unfortunately, this particular instrumental set up
cannot be used to detect excited' N0 fragments however,
other studies (6) indicate that it is pOssible to do so

with other Optical Mass Spectroscopy instruments.

III.C.3. Emissions from Compounds with Miscellaneous

Functional Groups
III.C.3.a. n-Propanethiol

The only emission Observed from electron impact on
n-propanethiol (Figure 15)is the Balmer series H emission
and the emission resulting from the CH(K1 —+ XHT)
transition. The spectra of both ethanethiol and
methanethiol taken under comparable conditions (15) show
emission from the SH+(§TE -» iFZL-) transition at
approximately 3340A. In addition, emission from the CS(NW')
state is also observed for these two compounds within the
2820—2850A region. Again, it is thought that the emission
resulting from the CS(AWF—» Y§:+) transition might not
appear in the spectrum Of n-propanethiol because of the
limited grating efficiency in this region.

A thermodynamic analysis shows that the heats Of

reaction for the formation of SH+ from methanethiol,

Howzuocmmoumlc mo Esuuoomm .mH musmflm

.6: c: Ihuzu4u><3

 

CONDONOBNOQNOONOOMORONnanO'nOonownosnoaomnOOv030Nv0nvovvonv8vohv§
4.44 q

d d 1 q d

 

63

 

 

 

 

 

 

 

‘4 d d a 11" u d u J 4 d d d
A 4 1‘11 ‘ fl 1% II I]! Aw
. L
IHI . a .5
.N..c.~..~e.z :«xal M
5.56.3.1 I
N
I.
3
N
S
P u
.A
a. 6.9.5:
«10.....0. ..QN. mSmme 44.5.? .. .c
fl >89 . .655 5....53L 0.7.5230... _

.N. 6.33::

64

Table 14. The Beats of Reaction for the Formation of 811+ from Thiols

 

COMPOUND PRODUCTS HEAT OF REACTION
@333 9 (H3 '0' 33+ 314.5 koala
m3m2$H '0‘ (33512 4* 53+ 311.9 kcals

(1130112012811 9 (1130112012 + 33+ 2 312.‘ koala

65

ethanethiol and propanethiol are all comparable (Table 14).
This would lead one to believe that the only reason for the
absence of SH+ emission from n-propanethiol might be the
difference in the electron energies used to produce the
spectra (150eV was used in the case of n-propanethiol and
300eV was used for_ methanethiol and ethanethiol).
Unfortunately, to date, no other thiols have been studied,
nor has the excitation spectrum of SH+ from electron impact
on thiols been recorded. Therefore, one can only conclude
that there is insufficient evidence that SH+ emission
results from single collision electron impact on all SH or
CS containing compounds. (The possibility that SH+
formation occurs via a two electron collision should be

considered.)
III.C.3.b. n-Bromopropane

The spectrum of n-bromopropane (Figure 16) is heavily
populated with Br and Br+ emission within the 4800—4200A
wavelength region. In addition, emission from the
HBr+(#ZL+—o Xbm ) transition in the 3600-3200A region is
prevailent. Such emission has been seen for bromobenzene
and bromonaphthalene (6). However, this emission was not
seen in the spectrum of the bromoacetylene cation (16). The
spectrum Of this cation was studied by Maier using electron
energies slightly above the ionization potential Of
bromoacetylene (~10.5eV). At this low energy, emission from

only a few fragments and emission from the ionic parent was

66

OGOQOHQOEOHmIC MO Eduaommm .wH onsmfim
.E: c: zeozm4u><3
9» 8m 03 o... 8m 8n 0% 8n 8» oo. 9. 8. on. o... 8. 8. 2. 8. 8.
d 1 d 4 d 1 J H d H J Id‘ 1 q d H d d a

 

    

15.15.35: ..=.xt.u.<...oz

 
 

  

    

»
..nthOIoH a4~051

.

 

..=.xt.w.....cz

We? .

a. ...... .5:

..o. ..on . mSmmuE ..38
>89 . >655 558:

      

 

P
..w... 1.6.55

 

 

3...: 19.5...

 

32.3193...

ALISNBLNI

 

 

 

 

9.48% .c

2... col v.3...

67

HH< mo Esuuoomm .hH onsmflm

.6: .... 7:99.393)

 

08 Os.» 8 an 8' 0; ON. 3' 0: 8' 8c Os. 00¢
a q

E21533 _
..ou.x1.au.c...z * ..w~x+:~$.0o

..u.xoc.s.8
.2... l.=.u..z .. 4

 

      

 

L

.=$'.H$.Ou —

3.... [5.2.2

  

..u- x at = .350

 

 

 

AllSNHiNI

gO H.“ - nth-"0°vu

 

 

 

go: On I' 3:00.02

 

 

.5... I 3:2...

 

.

«copra. 3.. .55me 445» £4 .....a 13......
>69 .555 558.5

 

 

 

.3301 a.u.o..z

68

observed. Unfortunately, studies at higher energies
resulted in an extremely conjested spectrum which was
difficult to analyze.

The spectrum presented here and the emission observed
°for the aromatic brominated compounds would suggest that
Optical Mass Spectroscopy could be used to identify bromine
containing compounds by monitoring the 3600-32003
wavelength region in order to detect HBr+ emission.

The discussion above implies that Optical Mass
Spectroscopy can be used as a basis for a universal
detector (recall CH and H emission was seen for all
compounds) as well as a group-specific detector for gas
chromatography. Figure 18, which is a diagram of the
wavelength regions within which one might selectively
detect a particular fragment emission, serves as a concise
review of the detection capabilities of this technique.
This diagram indicates that universal detection
capibilities can be obtained within the 4325-éZOOA
wavelength region, where CH(NL - xfiT) emission can be
observed, and more specific detection capabilities can be
acquired within various other wavelength regions

(3400—3500K for amines, 3200-3000A for alcohols, etc.).

III.D. General Observations

The spectra presented here might also be used to reveal

structural information and physical properties associated

mmsouw Hmcowponsm
mo cofiuomumo may now womb maoflmmm numcwam>o3 UHMHommm mgu mo Emummao .mH ousmflm

“E: c: Iszw4w><3
COM 9m ONm on» ovm Own own 0%» 0mm 0mm 00¢ 01v Om¢ on...» 0W6

_ c i C [it . . _

IO ..m: 20 0 Io

-

 

 

 

70

with the analyte species in addition to compositional
information. For instance, one might think that the ratio
of the number of hydrogen atoms to the number of CH bonds
in the molecule might be proportional to the ratio of the
sum of the intensities of the Balmer hydrogen emission to
the intensity of the CH(AQ:- XLW) emission. Unfortunately,
.this is not found to hold true for the compounds studied
here (Table 15 lists these ratios for each compound).

The presence of the emission resulting from excited
state NH and OH fragments in the spectra of n-butyronitrile
and n—butyraldehyde is most certainly indicative of such
species undergoing McLafferty—type rearrangements via the
six-membered ring intermediates to form the NH and the OH
bonds. This offers some insight into the structure of such
compounds.

Physical characteristics, such as an analyte's
ionization potential, might be reflected by these spectra.
There appears to be some correlation between the analyte's
ionization potential and the amount of energy deposited
into the molecule by electron impact. This correlation is
shown in Figure 19 which is a plot of the ratio of the
intensity of the H(n2=5—~ n,=2) emission to the H(n,=&+
n,=2) emission versus the compound's ionization potential.
If the ratio of the intensities of this H emission is
indicative of a temperature parameter that is associated
with the excited H atoms, then this plot reveals that an

increase in the ionization potential results in an increase

71

Table 15. comparison of the Compositions and the Emission Intensity
Ratios for All of the Cempounds Studied

 

 

COMPOUND NO. OF NO. OF CH NO. OF H AIOMS H EMISSION
H AIOMS BONDS NO. OF CH BONDS CH EMISSION
n-propsnol 8 7 1.14 0.376
n-butysldehyde 8 8 1.00 ‘ 0.205
Z-butanone 8 7 1.16 - 0.220
diethyl ether 10 10 1.00 0.115
2~nitropr0psne 7 7 1.00 0.362
n-butyronitrile 7 7 1.00 0.332
n-propylsmine 9 7 1.29 0.323
n-propanethiol 8 7 1.14 0.007

n-broIOprOpane 7 7 1.00 0.168

.75
.70
.65
.60

.55
.50

1,, (5+2) /1,, (4+2)

.40
.35

l

 

72

l l

 

8 9

I0 II l2

IONIZATION POTENTIAL (eV)

Figure 19.

Plot of the Ratio of H Emission Intensities

versus Ionization Potential

73

in the amount of energy deposited into the molecule. This
agrees with the observation (made using mass spectral data)
that the relative abundance of the molecular ions formed by
electron impact decreases as. the ionization potential of
the molecular Species increases. This can be explained by
the hypothesis (17) that a greater number of low energy

("cool") molecular ions are formed for compounds with low

ionization potentials.

III.E. CONCLUSIONS

Optical Mass Spectroscopy appears to be a likely basis
for a universal/specific detector for gas chromatography.
Further studies on a wider variety of functional groups and
longer chain molecules with diverse structures might prove
to develop more specific guidelines for its use in this
area.

In addition, information, such as the evidence of
six-membered ring transition states in the unimolecular
dissociations of n-butyraldehyde and n—butyronitrile, as
well as the correlation between energy deposition and
ionization potential, shows that these studies help to

further characterize the electron impact process.

LI ST OF REFERENCES

l)
2)

3)
4)

S)
6)

7)

8)

9)

10)

11)

12)

13)

14)

15)

16)

17)

LIST OF REFERENCES

Ettre,L.S.;J.Chromatogr.Sci.,1978,l6,396.
Robin,M.L.;Schweitzer,G.K.;Wehry,E.L.;Appl.Spec.Rev.,
1981,17,165.
Baas,R.Ch.:Beenakker,C.I.M.;Comp.Phys.Comm.,1974,8,236.
Beenakker,C.I.M.;Quant.Spectrosc.Radiat.Transfer,1975,
15,333.
Tokue,I.;Urisu,T.:Kuchitsu,K.;J.Photochem.,1974,3,273.
Hirota,K.;Hadata,M.;Ogawa,T.;Int.J.Radiat.Phys.Chem.,
1976,8,205.

Ajello,J.M.;Chem.Phys.,1971,55,3158.
Maier,J.P.;Chemia,34,1980,219.
K1apstein,D.;Kuhn,R.;Leutwyler,S.;Maier,J.P.;Chem.Phys.,
1983,167. '
Leutwyler,S.,Klapstein,D.;Maier,J.P.;Chem.Phys..1983,
151. ‘
Miller,T.A.:J.Phys.Chem.,1980,84,3154.
Tsuji,M.;Bu11.of Chem.Soc.of Jpn.;1977,56,2432.
Ogawa,T.;Miyamoto,N.:Ishibashi,N.;Bull.of Chem.Soc. of
Jpn.,1978,51,394.
Donohue,D.J.;Schiavone,J.A.;Freund,R.S.;J.Chem.Phys.,
1977.67.769.

Toyoda,M.;Ogawa,T.;Ishibashi,N.;Bu11. of Chem.Soc.of
Jpn.,1974,47,95.
Allan,M.;Kloster-Jensen,E.;Maier,J.P.;J.Chem.Soc.
Faraday Trans.II,1977,73,1406. ‘
MaLafferty,F.W.;"Interpretation of Mass Spectra",
3rd.ed.,University Science Books,Mill Valley,CA.,
1980.

74

18)

19)

20)

75
Franklin,J.L.;Dillard,J.G.:Rosenstock,H.M.;et.al.;
Nat.Stand.Ref.Data.Ser,Nat.Bur.Stand.,NSRSD—NBS,26,

1969.
Huber,K.P.;Herzberg,G.;"Molecular Spectra and

‘ Molecular Structure",Van Nostrand Reinhold Co.,N.Y.,

1979.
Tech,J.L.;J.Res.Nat.Bur.Stand.,67A,6,1963.