Responses and mechanisms of positive electron affinity molecules in the N2 mode of the thermionic
ionization detector and the electron capture detector
by Christopher Stephen Jones
A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in
Chemistry
Montana State University
© Copyright by Christopher Stephen Jones (1989)
Abstract:
Very little knowledge has been acquired in the past on the mechanistic pathway by which molecules
respond in the N2 mode of the thermionic ionization detector. An attempt is made here to elucidate the
response mechanism of the detector. The basic response mechanisms are known for the electron
capture detector, and an attempt is made to identify the certain mechanism by which selected molecules
respond. The resonance electron capture rate constant has been believed to be temperature independent,
and investigations of the temperature dependence of electron capture responses are presented.
Mechanisms for the N2 mode of the thermionic ionization detector have been proposed by examining
the detector response to positive electron affinity molecules and by measurement of the ions produced
by the detector. Electron capture mechanisms for selected molecules have been proposed by examining
their temperature dependent responses in the electron capture detector and negative ion mass spectra of
the samples. In studies of the resonance electron capture rate constant, the relative responses of selected
positive electron affinity molecules and their temperature dependent responses were investigated.
Positive electron affinity did not guarantee large responses in the N2 mode thermionic ionization
detector. High mass ions were measured following ionization of samples in the detector. Responses in
the electron capture detector varied with temperature and electron affinity.
Results support a mechanism for the N2 mode of the thermionic ionization detector where a high mass
ion is formed during decomposition of the sample and reaction with inorganic species. Resonance
electron capture rate constants are not temperature independent, and electron mechanisms were
elucidated. Electron capture responses do not bear a strict dependence on electron affinity. RESPONSES AND MECHANISMS OF POSITIVE ELECTRON AFFINITY
MOLECULES IN THE N^ MODE OF THE THERMIONIC
IONIZATION DETECTOR AND THE ELECTRON
CAPTURE DETECTOR
by
Christopher Stephen Jones
A thesis submitted in partial fulfillment
of the requirements for the degree
of
Doctor of Philosophy
in
Chemistry
MONTANA STATE UNIVERSITY
Bozeman, Montana
November 1989
D31#
ii
APPROVAL
of a thesis submitted by
Christopher Stephen Jones
This thesis has been read by each member of the
thesis committee and has been found to be satisfactory
regarding content, English usage, format, citations,
bibliographic style, and consistency, and is ready for
submission to the College of Graduate Studies.
Ge.+ 2 7, '?<??
Chairperson, Graduate Committee
Date
Approved for the Major Department
Head, Major Department
Date
Approved for the College of Graduate Studies
92^. 9, /fff
Date
Graduate Dean
r
I
Il Il V J
iii
S T A T E M E N T OF PERMISSION TO USE
In
the
presenting
requirements
University,
available
further
I
to
this
thesis
for a doctoral
agree
that
borrowers
agree
that
in
the
the
under
copying
only for scholarly purposes,
prescribed
in partial
U.S.
degree at
Library
rules
of
fulfillment of
this
of
Montana
shall
the
thesis
State
make
it
Library.
I
is allowable
consistent with "fair use" as
Copyright
Law.
Requests
for
extensive copying or reproduction of this thesis should be
referred
Zeeb
to University
Road,
granted
the
Ann
to
format."
Signature
Date
Arbor,
Michigan
"exclusive right
copies of the
right
Microfilms
dissertation
r eproduce
and
International,
48106,
to whom
to reproduce
in and
from
distribute
by
300 North
and
I have
distribute
microfilm
abstract
and
the
in any
VITA
Christopher Stephen Jones was born January 10, 1961
in Monm o u t h , Illinois, the first child of Ronald L. and
Maxine E. Jones.
He lived in Ankeny, Iowa from 1966 to
1979, where he graduated from Ankeny High School.
From
1979 to 1983 he r e sid e d in I ndianola, Iowa, whe r e he
graduated from Simpson College.
Since 1983 he has been
enrolled in doctoral studies at Montana State University
in Bozeman, Montana.
ACKNOWLEDGMENT
I
would
devotion
advanced
to
like
science
degree.
Grimsrud1s research
last four years.
to
thank
gave
Thank
group
Dr.
Eric
Grimsrudz whose
me the opportunity to pursue
you
to all
for putting
the
up
members
with
an
of Dr.
me for the
vii
TABLE OF CONTENTS
Page
LIST OF TABLES ..................
LIST OF FIGURES ......................................
ix
X
ABSTRACT ..............................................
xiv
INTRODUCTION ..........................................
I
Gas Chromatographic Detection ..................
Thermionic Ionization Detection ................
Electron Capture Detection: Design
and Operation .............................
Electron Capture Response Mechanisms ...........
I
2
8
14
RESEARCH OBJECTIVES ..................................
30
EXPERIMENTAL ..........................................
32
Chromatographic Equipment ......................
GC Detectors . ....................................
ECMS Equipment and Conditions ..................
APIMS Equipment and Conditions..................
Chromatographic Conditions .....................
Sample Preparation ..............................
Data Collection and Processing .................
Computer Modeling ......................
32
33
41
43
47
48
49
51
RESULTS AND DISCUSSION .......
TID-N2 Responses of Substituted Nitrobenzenes ..
TID-N2 - APIMS Measurements ....................
TID-N2 Candidate Response Mechanisms ...........
Direct Electron Transfer to Analyte
Molecule ....................... '............
Thermal Decomposition of Analyte Followed
by Electron Attachment .....................
Reaction of Analyte with Gas Phase
Radicals Followed by Electron Attachment
to Product (s ) ...............................
Thermal Decomposition Followed by
Polymerization and Electron Attachment ....
54
54
66
71
71
78
82
86
viii
TABLE OF CONTENTS - Continued
Page
Thermal Decomposition Followed by
Combination with Inorganic Species and
Electron Attachment at the Gas-Solid
Interface ...................................
88
Thermal Decomposition Followed by Electron
Attachment and Association with Gas
Phase Species Removed from Gas-Solid
Interface ......
90
ECD Responses of the Substituted Nitrobenzenes . 95
Computer Simulations of ECD Response .....
101
ECD Response Mechanisms and Temperature
Dependence of Substituted Nitrobenzenes
and Azulene ...................................... 107
ECD Responses to Quinones and Electron
Affinity Dependence of Responses ............... 125
Comparison of Detection Techniques .............
134
CONCLUSION .............................................
137
LITERATURE CITED ......................................
141
ix
. LIST OF TABLES
Table
1.
2.
3.
4.
5.
Page
Definitions for the Computer Model of
Resonance Electron Capture.......................
53
TID-N responses and effects of experimental
parameters on responses to various positive
electron affinity molecules......................
56
Electron capture detector responses at two
detector temperatures.............................
99
ECD responses (response per mole in KHz-min x
10~y ) of 20 ng I ,4-benzoquinone (BQ), 32 ng
2-methyI-I,4-benzoquinone (MBQ) and I ng 1,4naphthoquinone (NQ)...............................
127
Maximum responses of resonance electron
capture molecules and their calculated
electron attachment rate constants...............
128
2
I JL
X
I
X
LIST OF FIGURES
Figure
1.
Page
Potential energy curves of the formation of a
molecular negative ion for electron capture
mechanism 1 ........................................
17
3/2
Ln KT
versus 1/T plot showing temperature
dependence and kinetic regions for electron
capture mechanism I .....
19
Potential energy curves for electron capture
mechanism II......................................
22
Ln KT^/^ versus 1/T plot showing temperature
dependence and kinetic regions for electron
capture mechanism II..............................
24
Potential energy curves for electron capture
mechanism III.....................................
26
Ln KT^/^ versus 1/T plot showing temperature
dependence and kinetic regions for electron
capture mechanism III.............................
27
Potential energy curves for electron capture
mechanism IV................
29
8.
Schematic of a TID ...............................
34
9.
Configuration surrounding the thermionic
ionization source.................................
35
10.
Concentric Coaxial ECD .................... '......
39
11.
Block diagram of electrical components of
CCP-ECD............................................
40
12.
Pneumatics for GC-ECD system.....................
42
13.
Diagram of APIMS .................................
44
14.
Ion source for APIMS..............................
45
2.
3.
4.
5.
6.
7.
xi
LIST OF FIGURES - Continued
Figure
Page
15.
Resonance ECD computer modeling flow chart......
52
16.
TID-N- peak height responses to varie.d
amounts of substituted nitrobenzenes.............
55
Absolute TID-N responses of five compounds
and detector baseline current as a function
of the detector block temperature................
59
Absolute TID-N- responses of five
substituted nitrobenzenes and baseline
current as a function of emitter heating
current.............................
61
Absolute TID-N^ responses of five
substituted nitrobenzenes and baseline current
as a function of emitter bias potential.........
63
17.
18.
19.
2
20.
Negative ion responses to 350 ng o-nitrotoluene (GC retention time 4.4 minutes)......... ■ 68
21.
Total APIMS negative ion responses (m/e > 400)
to 350 ng o-nitrotoluene, 383 ng m-nitrotoluene, 310 ng nitrobenzene, 230 ng o-fluoronitrobenzene, and 270 ngm-fluoronitrobenzene....
70
Comparison of TID-N2 molar responses with
electron affinity values for substituted
nitrobenzenes.............. ......................
73
Comparison of absolute TID-N2 responses of
substituted nitrobenzenes and dinitrotoluenes with their relative molar ECD
responses..........................................
77
High temperature, negative ion electron
capture mass spectra of nitrobenzene and
three isomers ofnitrotoluene.....................
81
ECD peak area responses to varied amounts of
substituted nitrobenzenes at 200° C and 30
mL/min nitrogen...................................
96
22.
23.
24.
25.
xii
LIST OF FIGURES - Continued
Figure
26.
27.
28.
29.
30.
31.
32.
33.
ECD peak area responses to varied amounts of
substituted nitrobenzenes at 300° C and 30
mL/min nitrogen...................................
Page
97
Computer simulated data for the resonance
electron capture process..........................
103
Computer simulated data for the resonance
electron capture process..........................
104
Experimentally determined ECD temperature
dependent response curve and computer
generated curve to 50 ng azulene.................
106
ECD temperature dependence response curve of
3.13 ng nitrobenzene..............................
109
Ln K T ^ ^ versus 1/T plot for ECD response
of nitrobenzene obtained experimentally by
Wentworth and Chen (37)...........................
HO
Ln R T ^ ^ versus 1/T plot for ECD response
of nitrobenzene, data obtained byauthor..........
112
ECD temperature dependence curves for 100 pg
o-nitrotoluene, I ng m-nitrotoluene, and I ng
jD-nitrotoluene....................................
113
34.
ECD temperature dependence curves for 4.53 ng
o-fluoronitrobenzene, 2.67 ng m-fluoronitrobenzene, and 3.76 ng £-fluoronitrobenzene... 115
35.
Electron capture mass spectra of 4.53 ng of
o-f luoronitrobenzene at 200° and300°.............
117
Electron capture mass spectra of 2.67 ng of
m-f luoronitrobenzene at 200° and300°.............
119
36.
xiii
LIST OF FIGURES - Continued
Figure
37.
38.
39.
40.
41.
42.
Page
Electron capture mass spectra of 3.76 ng of
jo-fluoronitrobenzene at 200°and 300°.............
121
ECD temperature dependence response curve of
128 pg of £-bromonitrobenzene....................
123
Electron capture, mass spectra of 128 pg
£-bromonitrobenzene at 200° C and 300° C ........
124
Plot of calculated resonance electron capture
rate constants versus molecular electron
affinity...........................................
130
ECD temperature dependence response curves
for 20 ng B Q , 32 ng MBQ, and I ng N Q ..... .......
131
Electron capture mass spectra of 10 ng each
BQ, MBQ, and NQ at 200° C ........................
133
Jl
xiv
ABSTRACT
Very little knowledge has been acquired in the past
on the mechanistic pathway by which molecules respond in
the
mode of the thermionic ionization detector. An
attempt is made here to elucidate the response mechanism
of the detector.
The basic response mechanisms are known
for the electron capture detector, and an attempt is made
to identify the certain mechanism by which selected
molecules respond.
The resonance electron capture rate
constant has been believed to be temperature independent,
and investigations of the temperature dependence of
electron capture responses are presented.
Mechanisms for the Ng mode of the thermionic
ionization detector have been proposed by examining the
detector response to positive electron affinity molecules
and by measurement of the ions produced by the detector.
Electron capture mechanisms for selected molecules have
been proposed by examining their temperature dependent
responses in the electron capture detector and negative
ion mass spectra of the samples.
In studies of the
resonance electron capture rate constant, the relative
responses of selected positive electron affinity
molecules and their temperature dependent responses were
investigated.
Positive electron affinity did not guarantee large
responses in the Ng mode thermionic ionization
detector. High mass ions were measured following
ionization of samples in t h e ,detector. Responses in the
electron capture detector varied with temperature and
electron affinity.
Results support a Tnor-Hsn -i c m -For- I-h o M m n ^ o of the
ion is
thermionic ionization
formed during decomposition of the sample and reaction
with inorganic species.
Resonance electron capture rate
constants are not temperature independent, and electron
mechanisms were elucidated. Electron capture responses
do not bear a strict dependence on electron affinity.
I
INTRODUCTION
Gas Chromatographic Detection
The
technique
demonstrated
of gas
chromatography
(GC)
was first
experimentally in the early 1950s (I).
Since
that time, the growth in applications of the procedure has
been so great that the technique is now the most important
and
widely
methods.
used
An
of
all
th e
column
active area of research
chromatographic
since
the
discovery
of this technique has been the study of detection devices
which
respond
to chemical
compounds
gas chromatographic column.
is
to
produce
interaction
signaling
a
with
the moment
detector provides
from
the
The function of a GC detector
measurable
these
that elute
e lectrical
compounds.
In
of elution of a given
the means
by
sig n a l
addition
compound,
by
to
the
which the quantity of the
chemical can be electronically measured and recorded.
Detection
by
directly
ions ■have
devices
converting
seen
devices
that
produce
a
that
incoming
widespread
employ
usable
produce
the
success
compounds
as
formation
analytical
an electrical signal
signal
GC
of
are
to gas phase
detectors.
negative
ions
Two
to
the Ng mode of
2
the
thermionic
electron
ionization
capture detector (BCD).
GC detectors, responding
structures
The
detector
ECD
c onducive
has
detection
been
been
to the
These are both selective
formation of negative ions.
especially
used
popular
compounds,
for
the.
of
this
study
that
for the
detection
of
It was thought at the
investigation
techniques
selective
while the TID - N 2 has
selective
nitrated and halogenated compounds.
ons e t
the
only to compounds with molecular
of halogenated
generally
( T I D - N 2 ) and
would
of
these
produce
two
negative
ionization
converging
results.
Results, however, diverged somewhat, and the two
techniques are quite distinct from one another.
Thermionic Ionization Detection
All
met hods
of
ionizing
GC
effluents
involve
the
production of gas phase ions within the working volume of
the
GC
detector by the application of energy in one form
or another.
solid
In
surface
thermionic
housing
surface,
so
a TID, this energy is provided
known
emitter
that
and
any
as
is
samples
the
positioned
eluting
ionization
adjacent collector anode.
thermionic
inside
may
produced
by a hot,
source.
the
impact
This
detector
the
is measured
source
by
an
3
In
that
be
19 6 4
Karmen
a conventional
s elective
organic
salt.
to
and
Giuffrida
flame
ionization
hal o g e n -
molecules
(2, 3, 4) demonstrated
and
detector
(FID)
could
p h o s p h o r ous-containing
if the flame was
The thermionic source used
doped
with
an alkali
in this study consisted
of a wire mesh that had been treated with sodium hydroxide
or sodium
air
sulfate and
flame.
Since
alkali-impregnated
selective
GC
containing
subsequently
that
compounds
thermionic
time,
TIDs
have
been
sources
detection
source
of
which
use
widely
used
nitrogen-
(5-10).
is
heated in a hydrogen-
In
heated
and
these
either
heated,
for the
phosphorous-
applications
electrically
reactive flameless atmosphere of hydrogen
the
in
a
and air or by a
hydrogen-air flame.
Many
and
different thermionic emitter source compositions
configurations
have
been
studied.
These
use of alkali reservoirs at the bottom (11)
top
salt
(12) of the flame,
tip
ceramic
application
(10), a potassium
cylinder with
silicate bead
Currently,
thermionic
(7, 14,
all
15)
and
commercial
emitter
of
one
or above
of a rubidium
chloride pellet
a nickel
include
sub-layer
of
the
are
four
the
silicate
(5), a cesium-
(13), a rubidium
a rubidium-ceramic bead
TIDs
the
supplied
following
(8).
with
a
types:
4
(a)
a homogenous alkali-glass bead formed
on a loop
of bare platinum wire (14);
(b)
a
ceramic
salt
cylinder .core
activator and
coated
embedded
with
with
an alkali
a heating
coil
(16);
(c)
multiply-layered
conducting
covers
a
surface
cylindrical
su b - l a y e r
loop
of
layer
of
with
nickel-ceramic
nichrome
of
ceramic
that
heating wire and
alkali-embedded
a
a
ceramic
(13, 17);
Cd)
a
homogenous
alkali-ceramic bead
on
a nichrome
heating wire (8,9,18).
The
sensitivity
phosphorous
composition
source
19,
voltage
and
ceramic
resistant
the
allows
lower
surface to be
upon
models
of
depend
nitrogen.upon
the
surrounding the source, the
hydrogen
imposed
flow
material.
than
a
rate
and
the
the detector electrodes
generally
either
to melting
use of the
a
the
Modern
composed
impregnated
more
detectors
of the atmosphere
20).
sources
selectivity of these
selective
temperature,
polarization
(9,
and
employ
cesium-
or
The ceramic
thermionic
rubi d i u m sources
are
are the silicate sources,
higher molecular weight alkali metals
effective
established
work
(8).
function
The work
of
the
source
function in this
5
case
is
defined
as
the
energy
required
to transfer an
electron from the source surface to the gas phase (21).
The
selective response mechanism of the flameless TID
to nitrogen- and phosphorous-containing organic compounds
has
been
explained
effective work
the
presence
interface.
electron
highly
and
function
The
of the
from
are
gaseous
is
the
electronegative
surface.
ter m s
response
transfer
reactive
in
of
a
thought
radical
of the
surface due
to
species at the gas-solid
low work
to
occur
due
to
function .surface
to
species, possibly
believed
boundary
lowering
thermionic
of electronegative
P O 2 , which
not
(8)
to
layer
be
CN, N O 2
produced
adjacent
to
in
the
the
hot
It has been pointed out that this mechanism may
be
operative
selective
for
all
detectors.
In
th e
nitrogen-phosphorous
detectors
where
a
flame
is
employed, it is fairly well established that the mechanism
is
a
gas
phase
ionization
alkali
impregnated
alkali
atoms,
transfer
detectors
from
are
and
the
source
the
process
generates
response
alkali
sometimes
(22-24).
atoms
called
gas
The
phase
is
due
to
to
the
sample.
alkali
an
flame
heated
neutral
electron
These
ionization
detectors to distinguish them from the thermionic process,
which is defined as the emission of electrical charge from
a heated solid surface (25).
6
More
shown
recently,
that
selective
in
an
a
Patterson
flameless
detection
inert
and
TI D
of
can
certain
atmosphere
coworkers
of
also
(13,
be
26) have
used
for
the
electronegative molecules
nitrogen.
Using
a highly
cesium-enriched ceramic thermionic source and a relatively
low
surface
Patterson
temperature
and
certain
coworkers
classes
dinitrotoluenes.
for
this
very
mode
using
a
(28).
40 0 °
600°
C,
selective responses
to
molecules,
cesium
requires
and
as
the
selectivity
a thermionic
(27), and
surface of
this is accomplished
concentration
Sub-picogram
to
such
Optimization of response
function
high
about
observed
of
(TID-Ng)
low work
of
in the
detection
source
limits
by
surface
for
these
nitro-organics has been reported (27) for this mode of the
O
T ID, along with a very high specificity factor of 10
versus
alka n e
hydrocarbons.
detector
for the
aromatic
hydr o c a r b o n s
Subsequent
use
specific analysis of nitrated
in environmental
of
this
polycyclic
samples
has also
been reported (29).
In
itself
this
version
appears
necessary
electron
however,
for
to
carry
the
more
T ID,
with
occurrence
transfer.
no
of
of
Beyond
precise,
it
a
the
analyte
all of the
molecule
ingredients
surface-to-gas
this
general
descriptive
phase
deduction,
evidence
of the
7
surface
ionization
mechanisms
occur
by
have
simplest
surface
to
which
been
one
is
the
as shown by
mechanism
the
has
TID-N
envisioned
by
direct
molecule
been
mode
2
(28,
form
Possible
responses
30).
electron
to
offered.
The
transfer
a molecular
might
first
and
from
the
negative
ion
Reaction I.
e(surface) + M ------ > M -
This
possibility
is a reasonable
one
in
(I)
view
of the
fact
that the compounds to which the nitrogen mode is known to
respond
have
positive
electron
affinities
plausible mechanism, shown as Reaction
M ---- > N
is that the analyte
+
(32).
Another
2,
X --- ----- (surface)-- > x ""
molecule
(2)
M is thermally decomposed
on
the hot surface of the thermionic source to form species N
and
X,
one
of
abstracts
negative
ion
considers
radical
an
which
electron
X-.
species
has
a
has
very
This
X
in
high
a
high
from
electron
the
surface
possibility
Reaction
and
to
the
form
is attractive
2 to be
electron
affinity
NC^.
affinity
(31,
if one
The
NO
32)
and
2
specificity
of response
has
been
to nitroaromatic hydrocarbons
previously
(13,
demonstrated
26, 29).
An
additional
possibility is worthy of consideration since it is thought
to
be
T ID .
operative
That
interface
as
to
in
is,
might
again
the nitrogen-phosphorous
gas
phase
radicals
chemically alter
form
an
the
at
mode of the
the
gas -solid
analyte molecule
intermediate
species
X,
so
which
abstracts a surface electron as shown in Reaction 3.
M + R-
While
might
is
the
------ > N + X .1T-J surface^
presence
of
reactive
X-
radical
species,
seem unlikely in an atmosphere of pure
nonetheless
impurities
in
possible
the
due
nitrogen
to
or
either
due
R-,
nitrogen, it
the
to the
(3)
presence of
generation
of
reactive radical species from the analyte itself.
Electron Capture Detection:
Design and Operation
Gas
very
phase
electron
capture
reactions
central role in environmental
used
described
for
trace
the first
ECD
analysis
in
1958
played
a
analysis and analytical
chemistry for nearly three decades.
been
have
These reactions have
in. GC
(33).
since
Lovelock
Currently the
ECD
9
and
the
FID
constitute
the
most
devices for gas chromatography.
of
this
GC
device was
of
detector,
very
namely
conducive
Many of these
soon
the
compounds
detection
recognized
that
the
selective to certain groups
those
to
used
Following the introduction
was
sensitive and
compounds,
strongly
it
widely
that
contain
formation
of
structures
negative
are of environmental
ions.
importance,
and the ECD quickly became a useful tool for environmental
analyses
(34),
even
though
the
detector
was
very
temperamental and the basic processes occurring within the
cell were poorly understood.
very
important
pesticides in the
the atmosphere
and
disposal
tool
th e
Environment
(67,
selective
(69), drugs
active
detection
(66), halocarbon
68), toxic chemicals
sites
biologically
for
Today the detector remains a
in human
of
aerosols in
in the workplace
beings
compounds
in
, the
ECD
functions
energy
necessary
(70)
physiological
and
fluids
(70).
Like
negative
is
the
ions.
derived
source.
T ID - N
from
The
The
2
beta
particles
by
for this
emitted
by
producing
ionization
a radioactive
ECD is unusual among ionization detectors in
that a high level of ionization is present within the cell
even .in the absence of analyte molecules (27).
unique
in
that
it
functions
by
It is also
redistributing
electric
10
charges between different types of ionized species instead
of
the
changing
total
detector (27).
radioactive
gas
the
electrons,
positive
thermalized
molecules.
By
electrodes
in
the
ECD
in
ions
cell collide with
the
and
by
further
the
formation
radical
detector cell.
applying
in
of electric charges within
beta particles emanating from the
resulting
atmosphere of the
are
The
source
molecules,
amount
The
cell,
the
the
secondary electrons
with
difference
thermal
collected and a current measured.
of secondary
species within
collisions
a potential
carrier
carrier gas
across two
electrons
can
be
When electron capturing
analyte molecules elute into the detector cell, they react
with the thermalized electrons present to produce negative
ions and therefore reduce the population of these thermal
electrons.
This removal
of electrons from the atmosphere
of the electron capture cell is the basis of an electrical
signal indicative of the analyte molecule.
Lovelock's
chamber
with
a
first
ECD
(33)
contained
constant
responses
ionization
tritium radioactive beta emitter and
electrodes that measured the current
small
an
potential,
hence
the
two
within the cell at a
term
DC-ECD.
The
to different concentrations of sample were very
non-linear
and
unpredictable.
r e spons e s
to
In 1963 Lovelock
different
(35) used
molecules
10%
were
methane in
11
argon
to
as the carrier gas and the pulse sampling technique
eliminate
much
of the
anomalous
the original model of the BCD.
behavior
observed
in
The presence of methane in
the detector cell aided in the deactivation of metastables
by
inelastic
the
effects
electrons.
voltage
of the
due
by
The
pulse
to
Lovelock
to
increasing
application
the
thermalization
of
of
drift velocity of the
a
several ways.
momentary,
recurrent
In the
DC
behavior
mode of the
(35) observed that anomalous responses were
space charge
detector
the
collect the electrons affected
detector in
separation
were
enhanced
secondary electrons and nullified deleterious.electron
mobility
BCD,
collisions,
of
the
cell.
positive
Anomalous
attributed
contamination
potentials
to
and
contact
electron
had
arisen
negative
responses
of cell surfaces,
unpredictable
that
in
from the
charges
the
potentials
in the
DC-ECD
arising
also
from
high applied voltages, and
energies.
By
applying
voltage
pulses, one can restrict the electron capture reactions to
a field free atmosphere in the period between
the pulses.
This
and
eliminates
potentials
behavior
the effects of
that
in
are
the
pulses,
0.5
to
1.0
period,
eliminates
the
D C - E CD.
root
space
of
muc h
of the
Application
frse c width
the
charges
and
contribution
100
of
to
the
contact
anomalous
voltage
1000
to cell current
/^.sec
from
12
the
collection of negative
voltage
pulse
relatively
ECD
also
allows
is
of
immobile
produces
ions
because
insufficient
ions
to the
a much
the width of the
duration
anode.
The
realized
pulsed
Still,
with
after
the
(FFP)
all
the
introduction
detector,
this fixed
a linear
and
conditions
improvements
of
these
pulsed mode
more stable baseline
a wider range of chromatographic
employed.
to draw
respo n s e
this
to be
that were
frequency
factor
was
available for only 10% of the dynamic range of the device.
From
physical
were
a
better
able
al.
of
the
phenomena taking place in the
to
processing.
et.
understanding
Kinetic
(36)
showing
improve
and
that
competitive
detector
studies
Wentworth
events
in
models
Chen
FFP-ECD
kinetically controlled
of
methods
the
of
difference
present
dynamic
as
range
obtaining
in
the
the
an
cell
of
were
response.
The
with
new
Wentworth,
the
1960s
series
of
a linear region
signal
and
sign a l
led to a new
detector.
analytical
current
in
a
reactions
the
an d
of
(37)
method of signal processing that produced
90%
and
E CD, investigators
designs
and
and
chemical
Previous
used
without
method
the
analyte
of using
the
decrease in cell current divided by the instantaneous cell
current led to a much
of
this
mode
required
more cooperative device.
very
clean
conditions
Operation
so a high
Il I I
I
13
standing
current
with
long
pulse
periods
could
be
achieved.
The
replacement of tritium with
beta
radiation
the
ECD.
(38)
The
considerably
^Ni
allows
^Ni
as the
broadened
operation
at
source of
the
scope of
muc h
higher
temperatures than the tritium beta emitter and lengthened
the
list
of
compounds
which
could
be analyzed
with
the
E CD.
In 1971 a new
was
introduced
cell
current
is
method of signal processing for the ECD
(39).
In this
held
constant
new
mode of detector,
and
equal
to a reference
current by varying the frequency of voltage pulses.
an
electron
capturing
analyte
molecule
the
enters
When
the cell,
the population of electrons decreases and the frequency of
pulsing
increases.
This
change
in
frequency
is used
as
the response to an electron capturing molecule, the change
being
proportional
to the
amount
of the compound.
This
constant current pulsed ECD (C C P - E CD) was shown to exhibit
linearity
dynamic
with
r a nge
processing
sample
of
is most
concentration
the
detector.
commonly
used
up
This
to
99%
mode
of
of
the
signal
in the commercial
ECDs
today.
In
the
presently,
commercial
one
of
models of the
two
detector
ECDs
being
geometries' is
employed
usually
14
applied.
by
One
design
is the
Simmonds, et. al. (38).
positioned
coaxial
. The
E CD, first
anode of this
concentrically within
the
ECD
developed
detector is
cell.
Positive
voltage wave forms are applied to the anode to collect the
thermalized
electrons within
design
was
employed
capture
data in this study.
that
is
very
ECD
reported
in
the
successfully
by
the
collection
of
This cell
the electron
The other detector geometry
used
Patterson
(40).
displaced
In
A narrow pulse of 0.64 f&sec width collects
the
electrons
from
the
the
carrier
gas
flow.
and
from
detector
detector cell.
contamination
upstream
this
coaxial
the
are
is
is the
configuration,
configuration
anode
detector cell.
the
cell of 0.3 cc volume
the
its
The
advantages
tolerability
large
linear
to
dynamic
actual
counter to
of
this
detector
range
(40).
The detector provides a linear response to many compounds
until electron
density is reduced to 0.005 of its original
magnitude.
Electron Capture Response
Mechanisms
The
were
first
postulated
coworkers
(36,
electron
in
attachment
response
mechanisms
the kinetic studies of Wentworth
37, 41)
for a F F P - E CD.
The
and
first of the
15
mechanisms
p r o p o s e d , designated
as
mechanism
I>
is
depicted in Reaction 4.
AB + e -------------- > A B “
(4)
This process is commonly known today as resonance electron
capture.
The
capture
reaction
denoted
anion
by
rate
k^
constants
and
and
for
the reverse
k_ ^
the
forward
detachment
respectively.
electron
process
The
are
molecular
formed for this process, depending upon a number of
physical
detach
and
to
electron
chemical
reform
as
properties, may
the
depicted
in
original
Reaction
be stable,
analyte
4.
or may
molecule
and
If stable it will
react further by recombination with positive ions as shown
by Reaction 5,
AB
where
k+
is
the
+ P+ ------ > Neutrals'
recombination
rate constant,
(5)
or it may
dissociate according to Reaction 6
AB
> A + B
(6 )
16
where
k2
potential
is
the
energy
dissociation
curve
for
molecular ion is shown in
the
case
molecule
the
where
electron
are
usually
containing
zero.
of
The
a. stable
This figure portrays
affinity
large
electron
constant.
formation
Figure I.
AB is greater than
process
and
the
the
rate
of
the
analyte
Molecules that undergo
systems,
often
attracting
conjugated
functionalities,
allowing distribution of excess energy among many internal
degrees
of freedom.
can
as
be
high
The rate
as
10“
cc
constants
mol-
for this process
sec-
(62)
but
vary
greatly from compound to compound.
This accounts for the
high
of
sensitivity
detection.
If
the
Franck-Condon
anion which
the
on
may
and
electron
limit,
leads
to an
(42)
than
ma y
dissociate
kinetic
derived
capture
zero,
a
unstable molecular
autodetachment,
competitive
Becker
electron
affinity is less
disappear by
these
Wentworth
selectivity
transition
dissociation
Based,
and
or, if above
to
A*
+
B-.
relationships,
the relationship
shown
in Equation 7
In KT3/2 = In A +"In —
+—
ke RT
between
a
response
where
k+
and
kg
molecule's
are
A
is
electron
comp o s e d
(7).
affinity
and
of fundamental
the. recombination
rate
its
ECD
constants,
constants
for a
17
a E =E
Figure I. Potential ener g y curves
m o l e c u l a r negative ion
mechanism I.
of the formation of a
for electron capture
18
negative
ion
and
electron
capture
affinity
of
the
plot
In
KT
of
dependence
process.
Figure
an
electron,
coefficient,
analyte
3/2
'
and
respectively,
and
molecule.
versus
kinetic
1/T
regions
EA
is
From
K
the
this
indicates
is
the
electron
equation
a
temperature
of the electron
capture
A graph of this sort for mechanism I is shown in
2.
There
( C S ) region
are two regions in this graph, the alpha
and the
beta
3
( { .)
region.
The beta region cor­
responds to the kinetic relationship shown in Equation 8
k+ > k_1 > k 2
for
mechanism
I.
In
this
region
(8)
the
electron
capture
process was proposed to be nearly temperature independent
for
molecules
indeed
(37,
43)
which
observed
for
over
The
region
via mechanism
experim e n t a l l y
about
molecules,
alpha
responded
a
40-50
Figure
This was
Wentworth
compounds,
temperature
of
by
I.
range
and
principally
of
2 corresponds
Chen
aromatic
80°
to
250° C.
to
the
kinetic
relationship shown in Equation 9
k _ 1 > k+ > k 2
for mechanism I.
(9)
The detachment rate constant is fast and
Ln KT
19
Figure 2. Ln KT^/2 versus 1/T plot showing temperature
d e p e n d e n c e and kinetic regions for electron
capture mechanism I.
20
is
increasing
electron
with
temperature,
capture
response
therefore
to
decrease.
behavior was experimentally observed
compounds.
These
affinities and
hang
on
Wentworth
to
the
negative
the
and
compounds
electron
Chen's
at
(43)
have
formed
the
Again,
the
this
(44) for a number of
generally
ions
causing
low electron
from
higher
conclusions
them
cannot
temperatures.
on
the
FFP-ECD
response of molecules which respond via mechanism I were:
(a)
If the electron affinity of the analyte molecule
is
greater
in
the
will
than
beta
be
18.5 kcal/mol,
region
regulated
the
data will be
and
the response magnitude
by
th e
thermal
electron
attachment rate constant, which is thought to be
temperature independent.
(b)
If the electron affinity of the analyte molecule
is
b e tween
indicate
an
9 and
18.5
kcal/mol,
alpha
and
beta
the
region.
data will
At
high
detector temperatures the response will decrease
with
increasing
decrease
being
temperature,
dependent
the slope of the
upon
the
electron
affinity.
(c)
If the electron affinity of the analyte molecule
is
below
9 kcal/mol,
the
data will exhibit
only
21
an
alpha
region
with
a decreasing
response
as
temperature increases.
It should
be
according
( N O 2 ).
to
these
This
(31, 32)
The
noted that one molecule that did not behave
and
conclusions
molecule
has
a
was
high
nitrogen
electron
dioxide
affinity
a low electron attachment rate constant (71).
electron
affinity
is
above
18.5
kcal/mol;
nevertheless, the data has an alpha region.
The
second
coworkers
( 36,
m echani s m
37,
41),
proposed
by
designated
Wentworth
mechanism
and
II,
is
depicted in Equation 10,
AB + e ------- > A + B -
where
a
k 2 is the
single
bimolecular
immediately
energy
can
rate constant
(10)
for the process.
electron
attachment
step
, This
is
leading
to dissociation via the dissociative potential
curve
undergo
■
depicted
in Figure
recombination with
3.
The
negative ion B
positive ions
as shown by
Reaction 11,
B
.+
+ P
> Neutrals
( 11)
22
Figure 3. Potential e n e r g y
mechanism II.
curves
for
electron
capture
JJ
11 'I
23
where
k+
is
process.
recombination
rate
This electron attachment
commonly
kno w n
mechanism
3/2
KT '
the
is
versus
as
dissociative
favored
1/T
at
plot
higher
in
constant
response
electron
4
the
mechanism
capture.
temperatures,
Figure
for
This
as the
depicts.
is
In
The beta
region of Figure 4 corresponds to the kinetic relationship
shown in Equation 12.
k+ > k _ 1 > k 2
The horizontal region of Figure
of
insufficient
magnitude
attachment to occur.
electron
which
especially
can
halogens,
respond
for
commonly
contain
ion and a radical.
4 corresponds to energies
dissociative
Compounds that undergo
attachment
molecules
(12)
electron
dissociative
are aliphatic and
electronegative
the result being
aromatic
functionalities,
a halogen
negative
It should be noted that many molecules
via mechanism
I and
mechanism
II, depending,
upon the temperature (37).
The
third
coworkers,
mechanism
designated
proposed
mechanism
by
III,
W entworth
is
shown
and
in
Reaction 13.
AB + e- --------> A B - --------> A + B~
(13)
Ln
KT
3z2
24
Figure 4. Ln K T
versus 1/T plot showing temperature
d e p e n d e n c e and kinetic regions for electron
capture mechanism II.
3
/
2
25
This
is
a
formation
two
of
dissociative
potential
versus
a
step
molecular
step
energy
1/T
dissociative
which
curve
plot
negative
proceeds
shown
shown
process
in
in
ion
via
is followed
the
Figure 5.
Figure
6
corresponds
to
the
kinetic
the
by
a
dissociative
The
In K T ^ ^
indicates
temperature dependence and kinetic regions.
region
whereby
three
The gamma (y)
relationship
in
Equation 14,
K_i > k 2 > k+
(14)
the alpha region to Equation 15,
k _ 1 > k+ > k 2
(15)
and the beta region to Equation 16.
k+ > k_2 > k 2
(16)
The fourth and final mechanism proposed by Wentworth
and
coworkers
is
very
similar
difference is that mechanism
a molecular negative
to
mechanism
III.
The
IV involves the formation of
ion followed
by
dissociation
via the
26
Figure 5.
Potential I energy
curves
mechanism
II.
for
electron
capture
Ln K r
27
Figure 6. Ln KT ^
versus 1/T plot showing temperature
d e p e n d e n c e and kinetic regions for electron
capture mechanism III.
same
potential
Figure
7.
The
energy
curve.
temperature
This
dependent
is
depicted
kinetic regions
in
are
identical to those of mechanism III.
It
should
again
be
noted
that the
device
that was
employed when these mechanisms were proposed was a FFP- E C D
suitable for fundamental physical measurements.
and
was
that
Chen
(37)
did
point
out, however,
that
Wentworth
this detector
sufficiently similar to those used for analytical work
their
measurements
analytical importance.
could
be
applied
to problems
of
29
Figure 7. Potential e n er g y
mechanism IV.
curves
for
electron
capture
30
R E S E A R C H OBJECTIVES
The
primary
goal of this study was to elucidate the
response mechanisms
molecules,
TID-N
GC
most of which
and
2
of selected positive electron affinity
the
detectors
BCD.
are
are substituted nitrobenzenes, in
Both
of these
analytically
selective to molecules whose
important
structures
are conducive to the formation of negative ions.
Very
hence,
little is
response
formation
the
mechanisms
alkali
was made
measurements
ion
that
are occurring
embedded
obtained
source,
and,
during
the
species at the surface of
source
compounds
to
occurring
in
negatively
charged
of
a
TID-N2.
An
this
GC
light
affecting
detector
on
detector
species
interface.
and
mass
spec.tr ome trie
with .a T I D - N 2 emitter employed
shed
gas-solid
voltage,
the basic processes
here, using the TID - N 2 responses of the
aforementioned
the
about
of negatively charged
heated,
attempt
known
that
the
an d
source heating
current,
to
of
formed
the
including
detector
processes
identify
are being
Investigations
response,
basic
as
the
at the
parameters
source
temperature
bias
and
detector atmosphere composition were performed not only to
facilitate elucidation of response mechanisms,
optimize the usefulness of analyte signals.
but also to
31
While electron
capture response
thoroughly
investigated
concerning
the
remain
on
factors
a ffinity
of
the
past,
basic processes
a mystery.
two
in
mechanisms have been
many
of
occurring
the
within
details
the
An attempt was made here to shed light
which
the
affect
analyte
ECD
response:
molecule
and
the
the
electron
temperature
dependence of the electron capture rate constants.
the
compounds
electron
used
affinity
nitrobenzenes.
molecules
may
mechanism,
ECD
to study
these
molecules,
Studies
many
of
addressing
respond
via more
depending
upon
than
the
factors
them
the
Again,
are positive
substituted
possibility
one electron
temperature,
that
capture
are
also
described.
The
final
objective
is
to
compare
usefulness of the two detection methods.
the
analytical
32
EXPERI M E N T A L
Chromatographic Equipment
The
TID
data
reported
in this study
were
obtained
with a Varian Aerograph 3700 GC on which the detector was
mounted.
The
E CD
data were
obtained
using
a Hewlett
Packard
5890 GC.
A
Hewlett
column
(10
m
Packard
x
0.53
fused
mm
silica wide-bore
inside
diameter)
capillary
with
50%
phenylmethyl silicon stationary phase was used in both gas
chromatographs.
The
through
a transfer
base
the
of
end
of
line in
respective
this
each
column
was
instrument
detector
threaded
and
into the
so that exposure of the
sample to metallic surfaces was minimized.
For
the
experiments,
electron
the
spectrometer
capture
samples
(MS)
were
vi a
a
mass
spectrometry
introduced into
Varian
3700
(ECMS)
the mass
GC.
The
chromatographic column used in this instrument was a J & W
Scientific
inside
fused
silica capillary column
d i a m e t e r ) with
5%
phenyl
and
(30 m
x 0.25 mm
95% methyl
as the
stationary phase.
Samples were introduced into the atmospheric pressure
ionization mass
spectrometer
(APIMS)
via a Gow-Mac
Model
750
GC.
The column
used in this instrument was the same
wide-bore capillary column previously described.
GC Detectors
The
and
TID-N
2
was
Technology,
purchased
Walnut
from
Detector
C r e e k , California.
It has been
described previously by Patterson (13, 26).
schematic of this
detector.
Its heart
thermionic source, positioned
collector electrode.
Engineering
Figure 8 is a
is the
cylindrical
on the axis of a cylindrical
Figure 9 is a detailed sketch of the
detector configuration surrounding the source.
consisted
of
multiply-layered
sub-layer
of electrically-conducting
cylindrical ceramic with
surface layer of cesium-impregnated
which
can
withstand
electrically
detector
provided
attack
by
heated
very
protection
to
materials present
allowed
the
ionization
in
migration
through
tower
which
to
the body
the
source
The
obtained
nickel-ceramic
the
wire
replaced
of the
was
source.
enclosed
a
wa s
from
the
sublayer
corrosive
surface coating
lost at the
be
from
a
source,
temperatures,
heating
electrical charge
process
ceramic.
supply
The
the
nickel-ceramic and
high
a power
manufacturer.
and
in
by
The source
layer
surface during
by
The
was
a charge
detector
heated
to
34
R E M O V A B L E
S O U R C E
AS SEMB LY
Tl D / C F I D
TOWER
TO
ELECTROMETER
C E R A M I C
INSULATOR
G A S
EXIT
CERAMIC
JET
’
M O U N T IN O
F L A N G E N
ITTTfK
BASE
x 1.3
Figure 8. Schematic of a TID.
35
I ____
I
J
I
I
GAS
2
V
I
I
\
I
I
I
I
I
SA M P LE
4GAS I
I
I
------- SAMPLE
Figure 9 . C o n f i g u r a t i o n s u r r o u n d i n g
ionization source.
th e
C O N D U IT
thermionic
36
temperatures
of
270°
chromatograph.
The
source
voltage
with
electrode
respect
which
Negative
to
held
ionization
electrometer
electrode.
more
was
that
to
370°
was
th e
as
by
biased
the
at
a
surrounding
virtual
current
was
C
negative
collector
gro u n d
was
potential.
measured
attached
to
gas
the
by
an
collector
The detector was mounted on a FID base so that
than
detector.
one
type
of
gas
could
The
chromatographic
be
supplied
effluent
to
the
(10 mL/min)
was
mixed at the detector base with additional gas so that the
total
gas
flow rate
This
high
gas
well-purged
was
of
the
flow
nitrous
volume.
nitrogen,
California,
following
oxide,
nitrogen.
were purchased
Danbury,
Connecticut.
Mont vale.
oxygen
New
in
selected
i sobutane
in
additional
in
argon,
isobutane
and
carbon
dioxide,
Air Co., San.
cases where
carbon
0.31%
and
were
The
nitrogen
Jersey.
nitrogen
All
obtained
specially
wa s
one
dioxide,
oxygen
in
nitrous oxide
gases
were
fro m
except
first
Linde
prep a r e d
purchased
a
detector gas
Liquid
used:
provided
Chemetron Co., Chicago, Illinois.
and
oxygen
The
100 mL/min.
detector
except
argon,
methane
0.31%
the
from
gases was
from
detector was
purchased
methane,
The
the
rate through
detector
usually
Francisco,
through
passed
Co.,
tank
from
for
The
of
Airqo,
the 0.31%
through
M
Il
37
oxy.gen-removing
water
and
0.31%
oxygen
traps
particles
in
and
a
entering
nitrogen
molecular
the
was
gas
passed
sieve
to remove
chromatograph.
through
The
a molecular
sieve only.
There
are
thermionic
atmosphere
described
several
characteristics
ionization
which
detection
make
it
The
first
is the atmosphere
the
nitrogen-phosphorous
chemically
source.
radical
reactive
In
as
to
detection,
the
at
of
a
source.
arising
the
to
phosphorous
the
surrounding
selective
boundary
the
reactive
hot,
surrounds
the
layer,
O * , an d
HO"
from
the
hydrogen
and
the
In
the TID-Ng
of
800°
400°
C
mode.
concentration
of
to
are
The
600°
used
TID-Ng
alkali
nitrogen-phosphorous
function surface.
mode,
mode,
the
C,
source
while
be
air flows
mode
of
surrounds
is operated
temperatures
the
source
reactive
can
TID-Ng
with
In
a
H ",
In
(5, 10).
detector,
layer
inert
source.
as
detector.
an
previously
hot, inert atmosphere of nitrogen
tem p e r a t u r e s
600°
such
from
in
selective detector
chemically
species
envisioned
supplied
this
gas
this mode of
operated
distinct
nitrogen-phosphorous
of
nitrogen-
has
(cesium)
than
resulting
in
a
a higher
does
the
lower
work
Il
/
'I,
38
Figure
used
10 is a schematic of the
in the
experiments.
has
previously
and
by
Grimsrud
detector
energy
cell
the
seco n d a r y
e n ergy
of
thermal
lined
or
electrons
reduced.
10
per
be
pulses
to a reference current
pulses
is
adjusted
hence
the
name
constant
Therefore,
when
an electron
cell,
the
to
^Ni.
which
and
of
The
ECD
maintain
(38)
the
high
are emitted
produce about
is reduced
by
population
voltage
inside
collisions.
captured
electron
et. al.
initial beta particle.
collect the uncaptured electrons.
detector
mCi
electrons
further
can
Positive
compared
The
particles,
seco n d a r y
by
the
Simmonds,
(45).
with
beta
by
carrier gas molecules
these
a result
Warden
electrons
range
thermal
As
is
Packard
Its concentric coaxial geometry
discussed
and
electrons,
collide with
500
been
Hewlett
The
The
to the
resulting
analyte molecules.
in the
applied
detector is
to the anode
The current measured is
and
the
interval of the
a constant
current
pulsed
capturing
pulse rate rises.
cell current,
ECD
(CCP-ECD).
analyte enters
the
This pulse rate is
then converted to a voltage which is related to the amount
of
electron
capturing
analyte.
Figure
11
is
a block
diagram of the electrical components for this detector.
39
Vent
Glass Column
Figure 10. Concentric Coaxial BCD.
ECO Call
Varisble
Frequency
Puller
Signaj Out
(V olts « fre q |
4*.
O
Pulie Frequency Varied T e Meinteln I eefl «
Figure 11. Block diagram of electrical components of CCP-ECD
41
The
detector was heated
t e m p e r-a t u r e s
tailing
and
of
to
80°
to
optimize
by the gas chromatograph to
350°
C.
To
prevent
detector performance,
peak
20 mL/min
nitrogen was mixed with the column- effluent (10 mL/min) at
the detector base.
These gases were first passed through
oxygen-removing traps and through a molecular sieve before
entering
the
gas
chromatograph.
A
diagram
of
the
pneumatics of the system is shown in Figure 12.
ECMS Equipment And Conditions
All electron capture mass spectra reported
here were
obtained with a VG Instruments model 7070E-HF MS.
a
double
source
300°
sector,
was
C,
medium
operated
the
thermocouple
resolution
at
being
attached
the
to
capillary chromatographic column
GC-MS
interface
to the
instrument.
temperatures
temperature
This is
The
between
200° and
determined
ion
source
ion
by
a
block. The
was threaded through the
entrance port of
the
ion
source.
Methane, which was used as the reagent gas, was introduced
into
the
ion
source
through
a different
port.
The
pressure of the methane in
the ion source,
measured by a
MKS
model
manometer,
0.3 torr. The
ion
repeller
270B
capacitive
voltage
was
held
at
-I
was
V.
The
emission
external
plumbing
detector.
internal
plumbing
column
capillary
make-up
gas
flow
manifold
block
Figure 12. Pneumatics for GC-E C D system.
43
current
typically
maintained
at
150
exponentially
was
2
eV.
down
mA.
The
at
a
Electron
mass
rate
energy
spectrum
of
5
was
was .scanned
sec/decade.
Mass
resolution was typically 1000.
APIMS Equipment and Conditions
A
homebuilt
APIMS
was
formed by the T I D - ^ .
in
13.
for
detail previously
A
specialized
this study,
designed
offered
bolted
nitrogen
so
(46,
47)
ion
source,
the
the
ions being
and is diagrammed
in
constructed
Figure
a commercially
Detector
onto
to measure
The instrument has been described
is shown
that
by
used
of
carrier gas emerging
Its
specifically
housing
was
available T I D - N 2 source
Engineering
flange
14.
in Figure
and
the
Technology
existing
from the
GC
could
AP IMS.
be
The
was mixed with
additional nitrogen makeup gas and was passed through the
ion
source
with
a
total
flow
rate
of
30
mL/min.
The
chromatographic column was threaded through transfer lines
to
the
entrance
po r t
of
the
ion
source
exposure of the analyte to metallic surfaces.
of the ion
was
heated
supply.
A
-5 V were
source was
in
the
maintained at 250° C.
usual way
with
the
heater current of 2.5 A and
applied
to the
source.
The
to
minimize
The housing
The source
associated
power
a bias voltage of
grounded
walls of
44
O O O
O
Figure 13. Diagram of API MS.
O
O
em itter
a p e r tu r e
J
Figure 14. Ion source for API MS.
il
46
the
ion
source
ch a r g e d
serve
species
is
done
the
anode
will migrate.
current .associated
as
as
with
with
the
to which
Rather
these
TID-N^,
than
negative
this
a negatively
measuring
charge
instrument
the
carriers,
detects
the
negative ions which are directed toward and pass through a
20 m aperture which connects the atmospheric pressure ion
source
with
the
vacuum
envelope
of
a
q u a drupole
mass
spectrometer.
The quadrupole rods are biased to +24 V so
as
negative
at
to attract
ground
ions
potential.
from
The
upper
quadrupole filter was 500 amu
single
ion
or
scan
mode.
the ion
when
In
source which
mass
range
of
is
the
operated either in the
the
total
ion
mode
(DC
component to quadrupole rods is then set to zero) all ions
of mass greater than
to 500 amu
of
(48)
(Galileo
ions
using
Corp.,
capacitively
meter.
selected threshold value from 20
are allowed to pass to the detector.
negative
method
any
was
accomplished
a channeltron
Sturbridge,
coupled
the
ion
4309 electron
Massachusetts)
to a counting
'
by
preamplifier
Detection
counting
multiplier
which
and
is
rate
47
Chromatographic Conditions
For
the
splitless
TID-N^f
ECD
injections
were
sample onto the column.
nitrogen.
and
The
the
compounds
detector
c a rbon
were
gives
and
compounds
one
first
studied
Sufficient
peaks
using
the
for
the
convenient
fro m
75°
at
a
between
insured.
to
the
various
This
number
of
Column
elution
of the
180°
C.
Only
therefore
were
solvent
Injection
C
(49).
time,
runs
200°
FID.
the
molecule
chromatographic
always
L
times of all
an
to
at
at
retention
the
analyzed
resolution
was
isothermally
in
ranged
was
I
introduce
w a s .maintained
proportional
atoms
required
multicomponent
to
established
a response
compound
used
chromatographic
hydrogen
temperatures
experiments
port
operated
The
APIMS
The column carrier gas was always
injection
column
temperatures.
and
no
performed.
and
analyte
reproducibility
measured as peak area was routinely better than + 10%.
For the
used
The
to
experiments, I
introduce
column
injection
ECMS
carrier
port
temperature
the
was
sample
gas
programmed
onto
the
was
always
used
maintained
fro m
fj,L split injections were
at
an
capillary
250° C.
initial
column.
helium.
The
column
temperature
The
was
of
48
50°
held
C
to
at
a
final
the
temperature of
initial
temperature
220° C.
The
oven was
for four minutes, after
which the temperature was raised at a rate of 10° C/min.
Sample Preparation
Most
from
of the
commercial
obtained
from
Wisconsin.:
compounds
in this study
suppliers.
Aldrich
The
following
Chemical
2,3-dinitrotoluene;
2,6 - d i n i t r o t o l u e n e ;
were
chemicals were
Company,
& , a,
2,3,5-t r ic hi or o nitrobenzene;
2.3.4.5 - t e t r a c h l o r o n i t r o b e n z e n e ;
2,3,5,6-tetrachloro-
n i t r o b e n z e n e ; 1,2,3,4-tetrach loro benzene;
h.exachlorobenzene;
£-f l u o r o n i t rob enzene;
nitrobenzene;
m-f luoronit robe n z e n e ; £-f luorom - dinitrobenzene;
2-methyl-£-benzoquinone;
2.3.5.6- tetrafIuo r o - £ - b e n z o q uinone.
chemicals
were
Pennsylvania:
pentachloro-
£-chloronitrobenzene;
£-dinitrobenzene,
£-dinitrobenzene;
2,4-dichloro-
3,4-dichloronitrobenzene;
2,3,4-trichloronitrobenzene;
benzene;
e-m-
o-bromonitrobenzene?
m-br omonitrobenzene ? £ - b r o mo nitrobenzene;
nitrobenzene;
Milwaukee,
2,4-dinitrotoluene;
3,4-dinitrotoluene;
trifluoronitrotoluene;
purchased
obtained
from
Chem
I,4 - n a p h t h o q u i n o n e ;
m-nitrotoluene; £-nit rotolu ene;
and
The
and
following
Service,
Media,
o - nitrotoluene;
o-chloronitrobenzene.
49
J.
T . Baker
the
Chemical
supplier
of
Co., Phillipsburg, New
£-benzo q u i n o n e
and
Jersey, was
azulene. ■ m-Chloro-
nitrobenzene and nitrobenzene were purchased from Eastman
K o dak
Co., Rochester,
University
of
Ne w
Alberta,
£-cyanonitrobenzene,
York.
Paul
Edmonton,
Kebarle
Alberta,
of
Canada
m-cyanonitrobenzene,
the
donated
£-cyanonitro-
benzene, m-nitroanisole and £-nitroanisole.
A standard
by dissolving
either benzene
stock
a weighed
toluene was
Phillipsburg,
stock
New
HPLC
grade.
was made
in 10 mL of
The benzene
EM Science, Cherry Hill, New Jersey and
obtained
solutions
5 x l 0 “ 6 g//iL.
sample (10 to 50 mg)
or toluene, both
was purchased from
the
solution of each compound
from J. T .
Jersey.
was
concentration
therefore
Successive
solution
The
1:10
Baker Chemical
between
dilutions
that
standard
concentrations
10~
g//iL were available for each compound.
of
these
IXlO- ^
were
of
Co.,
and
made
so
10- ^ g//iL to
Data Collection and Processing
The
T ID -N
2
data
reported
in
this
study
collected with a Servogor 330 strip chart recorder.
responses
respective
were
calculated
from
chromatographic peaks.
by cutting and weighing the peaks.
peak
areas
Areas were
were
Molar
of
the
determined
The. data
collection
and
processing
for
the
ECD
experiments were performed in a quite different manner.
Cyborg
model
9 1A
system
linked
an
ISAAC
analog-to-digital
Apple
11+
Packard gas chromatograph.
Appligration II
Corporation,
computer
interfacing
the
Hewlett
This hardware, along with DY SC
software purchased from
Pasadena,
with
A
Dynamic Solutions
California, allowed
data
aquisition
and analysis to be performed on the chromatographic runs.
The computer program can perform operations such as peak
area
and
peak
determination,
height
baseline
analysis,
subtraction
retention
and
time
other processes
beneficial to analyzing a chromatographic run.
For the
ECMS
experiments, a Digital pdp 8/a computer
system was interfaced to the VG
MS.
necessary
parameters
mass
chromatographic
spectrometric
runs
With this system, the
monitored, .and
the
were
mass
set, the
spectra
analyzed.
The data for the APIMS
experiments were collected on
a Houston Instruments Omniscribe strip chart recorder.
51
Computer Modeling
Simulated resonance ECD responses were generated by a
computer program written by Dr. E. P. Grimsrud and run on
an
Apple
program
terms
11+
is shown
in
detector
the
constant
in
flow
for
was
for
chart
in
the
all
A
flow
are
the
Hewlett
shown
program
Packard
in
were
and
Table
I.
The
set to levels
CCP-ECD
and
held
e x p e r i m e n t s . ' The standing
I X 10"^ electrons/sec
2 msec
chart of the
15 and the definitions of the
simulation
set at
sample interval at
computer.
Figure
parameters
appropriate
current
personal
(1.60 mA ), the
the positive ion
recombination rate product at 500 sec
density -
52
£i1xls_J£
50
Initial Vaiuesl
-1 1 3
Cl
Loop
C2 Loop
Calculate Rates
120
cN„/dt. ONs-Zdt
MAKZ ARRAY
X C X T .0 > - '
142
144
Make Array
XCXT.2) - N*Prlnt to monitor
152
144
It XT. < 279
n-
Pulse
155
162
164
Print N
Molar Response, and TZAT
to monitor_____________
If N%- <- 1.00001 o< last N a
I C2— >C2 ♦ I
200
PRINT RESULTS
. | NANLC
244
250
Want to make ARRAY?
XC0.0) - 0
376
254
500
Repeat Results?
J2 » I <--PRINT ARRAY
Want Plot of N,
Ni,- vs T ?
520
Figure 15
Resonance ECD computer modeling flow chart
53
Table I.
Definitions
for the C o m p u t e r
Resonance Electron Capture.
Model
of
TERM
VARIABLE NAM E
[ETCL]a
ETHCL
R2
KETHCL
molecules/cc
-I .
-I
cc
* sec
Standing Current
I
electrons/sec
kI
Kl
cc
k-l
KREV
cc
R1
XRECOMB
sec
wrecomb
sec
Baseline
frequency
PO
sec
[Sample]3
NSAMPLE
molecules/cc
Smallest time unit
for integration
INCT
seconds
NNE G A
molecules/cc '
Na _ last Cl loop
NANLC
molecules/cc
V
NE
molecules/cc
dNe-/dt
RE
rate of change
dNA -/dt
RANEG
rate of change
R3
"A"
a
UNITS
-I
.
-I
* sec
- I j.
-I
* sec
—I
-1
-i
Concentration or population in detector volume
54
RESULTS AND DISCUSSION
TID-NU Responses of Substituted
Nitrobenzenes
The
responses
of
T I D - N 2 over
a wide
conditions.
Peak
36 compounds
were
determined
range of concentrations
height
using
two
different
2.30 A
and
2.10 A f are
calibration
levels
shown
of
in
and
curves
in
a
detector
determined
source
heating
Figure
16.
current,
Generally,
a
region of near-linear molar response was observed only for
the first decade or two of concentration change above the
lowest
most
detectable
substances
further
concentration.
increased
increases
compounds,
in
The molar
Or decreased
sample
responses
to
significantly with
concentrations.
A
few
however, exhibited quite linear behavior over a
wide range of concentrations.
The
peak
area
r es p o n s e s
observed
for
the
lowest
concentration of each compound in Figure 16 at 2.30 A are
listed in Table 2 as the absolute response (C/mol) of that
substance.
All
the
compounds
benzoquinone)
are substituted
but
nitrobenzenes.
affinities of all the mono-substituted
recently
The
been
compounds
measured
for which
(32)
one
and
The electron
nitrobenzenes
are listed
the- electron
(perfluoro-
have
in Table
affinities
2.
are not
55
SAMPLE CONCENTRATION (g/pl)
Figure 16. T I D - N 2 p e a k
h e i g h t r e s p o n s e s to var i e d
a m o u n t s of substituted nitrobenzenes.
The
compound associated with each calibration curve
is i n d i c a t e d by the n u m b e r a s s i g n e d in
Table
2.
Detector
conditions
are:
t e m p e r a t u r e , 320° C ; gas flow, 100 mL/min
nitrogen; bias voltage, -45 V? source heating
current, 2.30 A and 2.10 A.
T I D - N 9 r e s p o n s e s a n d e f f e c t s of e x p e r i m e n t a l p a r a m e t e r s
responses to various positive electron affinity molecules.
Table 2
C o m p o iim lk
E A kk
A bsolu te
D e te c tio n
E ffe c fi
E ffect®
Effect® *
( E c a ll
m o la r
Iim il
m o l)
re s p o n s e ***
(P S )
of
detecto r
of
h e a ter
of
bias
tem p.
current
voltage
( C jm o l)
D iffe re n t d etecto r gases*
O1
90%
90%
CO1
N 2O
/5%
.
Ci
u
15'%
90%
Is o -
Ar
butane
I NB
2 Hi-CH3ONB
3 /I-C H 3ONB
4 O-CH3NB
5 Hi-CH3NB
6 /I-C H 3NB
7 o-FNB
8 hit FNB
9 /i-FNB
10 o-ClNB
11 Hi-ClNB
12 p-ClNB
13 o-BrNB
14 Hi-BrNB
15 /i-BrNB
16 o-CNNB
17 hi-CNNB
18 /i-C NNB
22.1
22.7
19.6
20.0
21.4
20.5
23.5
27.5
24.3
25.0
28.1
27.5
25.5
28.8
28.3
35.1
34.1
38.1
2.2 • 10"4
5.5 • IO""3
0.08
0.29
1.1 • IO"3
0.03
0.11
0.018
0.51
0.018
0.018
0.060
0.043
0.035
0.13
0.39
0.49
1.4
12 • IO3
780
125
65
4 • IO3
250
115
165
65
470
340
105
260
140
150
25
, 14
9
4.2
4.9
4.5
2.6
4.3
4.4
3.2
4.9
3.75
3.4
5.5
5.2
3.7
3.3
5.0
4.2
4.8
4.0
8.4
7.4
14
4.0
8.9
5.0
2.2
9.3
4.0
2.1
6.3
4.9
4.6
3.9
3.8
4.2
5.6
4.1
2.3
2.8
2.5
2.7
2.6
2.8
2.7
2.4
2.7
2.6
2.2
2.7
2.5
2.8
2.9
2.9
2.7
2.7
0.37
0.59
0.57
0.53
0.36
0.50
0.72
0.40
0.80
0.41
0.52
0.67
0.82
0.75
0.67
0.51
0.34
0.27
0.80
0.73
0.88
1.06
0.73
0.75
1.54
0.73
1.50
1.1
0.77
0.90
1.00
0.92
0.88
0.88
0.84
0.78
0.77
0.43
0.20
0.36
0.43
0.41
0.54 .
0.32
0.32
0.32
0.33
0.30
0.50
0.40
0.27 ’
0.25
0.19
0.12
0.18
0.32
0.45
0.75
0.35
0.71
1.1
0.65
1.00
0.50
0.62
0.59
0.35
0.47
0.49
0.40
0.55
0.57
0.10
0.24
0.31
0.58
0.11
0.59
0.76
0.45
0.71
0.47
0.52
0.41
0.45
0.45
0.33
0.35
0.52
0.57
6.1
4.6
4.0
2.3
3.7
3.2
1.3
1.9
1.5
1.5
1.70
1.50
1.8
1.56
1.3
2.1
1.8
1.7
* B = Benzene; NB = nitrobenzene; D N T = dinitrotolucne.
* * Electron allinity valucs\akcn from ref. 12.
* * * Determined from peak areas using following conditions: detector temperature, 320°C; emitter heater current, 2.30 A; flow-rate, 100 ml/min nitrogen.
G Ratio of molar response observed at 370°C to that at 270°C; emitter healing current, 2.1 A.
GG Ratio of molar response observed with heater current 2.4 A to that with 2.1 A; detector temperature, 320°C.
GGG Ratio of molar responses observed using bias potentials of —45 and — 15 V applied to the emittor; detector temperature, 320°C; emitter healing current,
2.3 A.
t These detector gases were mixed at the base of the detector with nitrogen so that the total detector gas flow-rate was 100 ml/min. Values listed are ratio
of responses relative to responses in nitrogen gas.
tf /x//n 2.
on
Table 2.
(cont'd)
TID-N
r e s p o n s e s a n d e f f e c t s of e x p e r i m e n t a l p a r a m e t e r s
responses to various positive electron affinity molecules.
C n n ip o im i!*
1
9
E A **
A bsolu te
( k c a ij
m o la r
m o l)
re s p o n s e ***
D e le c tio n
* Iim il
(P U )
( C /m o l)
EHecJs
E ffect®
Effect®®
of
d e le c to r
of
h e a le r
of
bias
0 .3 %
90%
90%
/5%
15%
90%
tem p.
current
voltage
O2
CO2
N 1O
CHi
Is o -
Ar
on
D iffe re n t detecto r gases^
butane
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
O-NO2NIl
Ot-NO2NB
/J-NO2NB
Ot-CF3NB
2,3-DNT
2,4-DNT
2,6-DNT
3,4-DNT
2,4-Cl2NB
3,4-ClzNB
2,3,4-Cl3NB
2,3,S-Cl3N B
2,3,4,5-CUN B
2,3,5,6-CUN B
1,2,3,4-CUB
C l5B
CI6B
Perfluorobcnzoquinone
Baseline C u r r e n t t t
0.32
0.23
3.9 .
36.2
36.2
43.6
30.9
0.011
-
60.2
T
4.4
23
6.5
1.6
0.18
0.21
0.28 '
0.30
0.33
0.27
3.2 • IO '3
0.037
0.38
0.15
10
6
3
200
8
2
7
50
90
130
90
120
200
170
8 ■ IO3
720
60
3.6
4.3
1.9
3.1
3.4
4.5
3.3
4.4
3.0
2.7
3.6
3.6
2.5
2.9
4.4
3.7
3.1
2.0
6.6
3.6
.4.3.
3.3
4.2
2.9
3.2
3.1
3.2
3.2
4.3
3.5
5.0
5.1
4.5
2.2
2.8
3.0
2.5
2.5
2.7
2.7
2.6
2.7
1.9
3.1
3.0
2.6
2.7
2.5
2.6
2.86
2.5
0.60
0.51
0.85
0.42
1.27
1.3
0.42
0.83
0.72
0.96
0.93
0.73
1.08
0.75
1.08
1.12
0.88
1.3
0.61
1.2
0.62
0.55
0.64
0.75
1.00
0.86
0.77
0.60
0.77
0.79
0.84
0.87
0.84
LI
0.36
0.24
0.30
0.27
0.37
0.50
0.22
0.59
0.58
0.75
0.83
0.76
1.15
1.25
1.3
0.92
0.56
0.74"
0.51
0.87
0.44
0.60
0.69
0.44
0.67
0.65
0.56
0.37
0.36
0.44
0.30
0.48
0.40
0.71
0.67
0.43
0.65
0.33
0.49
0.59
0.35
0.50
0.45
0.38
0.37
0.34
0.44
0.28
0.55
0.44
0.75
1.7
1.9
1,6
1.7
1.56
1.7
2.0
1.5
2.1
2.1
3.2
3.0
3.3
5.6
2.0
2.02
1.26
I . IO3
2.6
3.7
2.6
0.56
0.46
0.75
1.9
0.47
1.1
0.54
0.74
0.54
0.55
4.0
8.8
* B = Benzene; NB = nitrobenzene; D N T = dinitrotolucnc.
* * electron aIVmity valuesNakcn from ref. 12.
** * Determined from peak areas using following conditions: detector temperature, 320°C; emitter heater current, 2.30 A; flow-rate, 100 ml/min nitrogen.
H Ratio of molar response observed at 370°C to that at 270°C; emitter heating current, 2.1 A.
W Ratio of molar response observed with heater current 2.4 A to that with 2.1 A; detector temperature, 320 C.
SB Ratio of molar responses observed using bias potentials of - 4 5 and - 15 V applied to the emitter; detector temperature, 320°C; emitter heating current,
2.3 A.
, '
t These detector gases were mixed at the base of the detector with nitrogen so that the total detector gas flow-rate was 100 ml/mm. Values listed arc ratio
of responses relative to responses in nitrogen gas.
ttZxZZN2.
Ul
-J
58
known
contain
functionality
affinities.
nitro
and
As
groups
supporting
compound
high
most
shown
prior
in
Table
very
studies
detector.
detection
was
degree
about
of
certainly
respond
nitro-selective
observed
a
electronegative
have positive electron
2, compounds
strongly
(13,
26,
29)
Also listed
limits at which
five times the
in
with
the
that
multiple
detector,
this
in Table
is
a
2 are the
the response
noise level.
to each
Picogram
levels of the most strongly responding compounds could be
routinely detected.
The effect of temperature of the detector housing on
the
response
range
Figure
all
of
17
to
270°
to
manner
compound
370°’ C.
for five
compounds
linear
each
increased
as
shown.
describe the temperature
compound
by
the
determined
Typical results
compounds.
is
was
by
It
The
are
the
shown
in
absolute response
to
temperature
is
over
in
possible,
an almost
therefore,
dependence of response
ratio of responses at 370°
and
to
for each
270° C.
This is done in the fourth column of data of Table 2.
The
result is that the relative responses of almost all of the
compounds
to
are
five times,
shown
in
increased
by
Figure
this
17
in
a similar manner,
100° C
is the
various detector temperatures.
temperature
about
increase.
baseline current
three
Also
observed
at
The baseline current rises
59
0.7-
m
oin
0.5-
O
5"
(6
0
C
1
I
0)
3
Q
Ul
O
B
"O
in
0 .1-
.D e t e c t o r T e m p e r a t u r e ( 0C )
Figure 17.
Absolute T I D - N 2 responses of five compounds
and detector baseline current as a function of
the detector block temperature.
Source heater
current,
2.10 A;
b i a s voltage, -45 V.
Hexachlorobenzene = 35; o-dinitrobenzene = 19;
m-dinitrobenzene = 20; p-bromonitrobenzene =
15; pentachlorobenzene = 34.
60
proportionately
with
responses
between
270°
and
320° C
and then increases more sharply than relative responses at
higher
detector
responses
to
temperatures.
all
compounds,
Maximum
therefore,
signal-to-noise
were
observed
in
the 270° to 320° C temperature range.
The
effect of heating
e m itter
was
heating
also
current
shown
in
current
determined
supplied to the source
for
each
(I) = 2.1 to 2.4 A.
Figure
18.
Again,
a
component
Typical
near
from
results are
linear
positive
dependence of response on heater current was observed for
all
compounds.
This
dependence
for
each
compound
is
expressed in Table 2 as the ratio of responses observed at
2.40
and
2.10
A
heating
current.
For
this
variable
a
somewhat wider range of response dependencies is observed
than
was
noted
for
variations
in
detector
temperature.
Response amplifications from about 2 to 8 are observed for
most
of
the
current.
compounds
As
a
general
weakly
respo n d i n g
degree
by
the
by
this
rule,
compounds
increase
0.3 A
increase
the responses
were
in heating
enhanced
current
responses of strongly responding compounds.
Figure
18 is the effect of
baseline current.
in baseline current
source
heating
in
source
of the more
to a greater
than
were the
Also shown in
current oh the
For this detector the relative increase
is roughly
equal to relative response
61
Baseline Current OO*13amps)
O
T-
E m it t e r H e a tin g C u rre n t (am p s)
Figure 18. Absolute TID-N^ responses of five substituted
n i t r o b e n z e n e s z an d b a s e l i n e c u r r e n t as a
function of emitter heating current.
Detector
t e m p e r a t u r e , 320° C ; bias voltage, -45 V.
£-Chloronitrobenzene = 12; £-fluoronitrobenzene
= 9; o-nitrotoluene = 4; o-dinitrobenzene = 19.
62
increases
current
up
and
to
about
2.25
associated
A,
after which
the baseline
noise level increase.more
sharply
than do molar responses to analyte molecules.
The effect of variation in the bias voltage, from -5
to
-45
V
applied
responses
is
to
shown
selected cases.
t he
central
in Figure
19
emitter
source,, on
for several
arbitrarily
The obvious result here is that responses
are continuously and linearly increased by an increasingly
negative
bias
voltage.
For all compounds
the
ratios of
responses observed with bias voltages of -45 and -15 V are
listed
in Table
2. ■
Most of these
range of 2.5 to 3.0, which
ratios.
It
distinctly
operated
(8),
where
source
those
in
with
is also
these
results
experiments
the nitrogen-phosphorous
only
of similar
current
that
of analogous
a saturation
level of response
-12
V
applied
dimensions.
shown
the narrow
is nearly equal to the voltage
significant
from
TID
observe d
is
fall within
In
to
with
the
selective mode
to analyte was
the
Figure
differ
emitter
in
a
19 the baseline
to vary linearly with bias voltage.
Since this plot approaches a non-zero origin, however, the
ratio
of
superior
This
response
with
clearly
insensitive
to
the
■ to
use
physical
the
baseline
of
the
effect,
chemical
current
greater
which
differences
is
somewhat
bias
voltage.
is
relatively
occurring
among
63
Baseline Current (IOljamps)
(j0.6-
Bias Potential (volts)
Figure 19. Absolute T ID - ^ responses of five substituted
n i t r o b e n z e n e s a n d b a s e l i n e c u r r e n t as a
function of emitter bias potential.
Source
heating current, 2.30 A; detector temperature,
320° C ; m - D i n i t r o b e n z e n e = 20; £-dinitrobenzene = 19; 2,3,4-trichloronitrobenzene = 29;
2,4-dichloronitrobenzene = 27; o-f luoronitrobenzene = 7.
64
the. 36
compounds,
was
not
anticipated.
Expected
was
a
saturation level of response with the application of -15 V
or even -5 V since the-field thereby induced was expected
to
be
sufficient
negative
to
ions.
collect
This
calculations
of
of the
relative
to
anode
ventilation
following
V;
through
cc/min;
and
conservatively
cross
cell.
For
0.12
t i mes
bias
total flow rate,
2
area of flow path 0.11 cm ;
the mobility of the ion
nitrogen
predicted
to be
velocity
the
than
due
the
of the
radial
of
ions
125 cm/sec
to detector
length
in
to be at least
is
greater
15
cm;
field
Since
for the
voltage,
velocity
longitudinal
several
example,
radial
the
by
direction
th e
while
15 cm/sec.
supported
(55);
applied
phase
velocity of ions due
assu m i n g
the
only
gas
linear
in atmospheric pressure
c m 2 s e c - "*"V-
was
experiments:
sectional
all
velocities in the
distance,
along
is
the
of these
100
1.0
ion
to the
cathode-to-anode
question
expectation
expected
values
essentially
gas flow
detector
dimension,
is
all
negative ions are expected to be collected, if the applied
field
serves
Since
the
response
applied
only
to
experiments
sweep
of
negative
Figure
19
is continuously proportional
electric fields, the probable
observation
ions
to the anode.
indicate
that
the
to relatively large
explanation
for this
is that the rate of emission of negative ions
65
from
the
emitter
e l ectric
field
surface
in
is directly proportional
that
vicinity
or
that
the
to the
field
is
somehow enhancing the ionization process.
Absolute
determined
oxygen,
and
responses
using
carbon
methane
to
altered
in
compounds
detector
dioxide,
mixed
all
argon,
gases.
nitrous
were
These
oxide,
various proportions
also
included
isobutane
into nitrogen.
These results and accompanying affects on standing current
are
shown
detector
in
the
gases
last six
columns of Table
studied,
relatively
2.
minor
For all
changes
in
signals were observed.
The
TID
operated
in
th e
nitrogen— phosphorous
selective mode has always been known as very temperamental
device.
heating
hydrogen
Variables
such
current
(8,
flow
irreversible
damage
rate
such
as
(52)
alter
the
basis.
phosphorous
(15),
source by
deposits
to-day
50),
detector
source
loss of alkali
to the
can
as electrode position
sensitivity
routine
sele c t i v e
variables
results.
For
emitter
on the
and
detector,
with
50),
source
(51),
is
buildup of
source
selectivity
it
(8),
(15,
(52), and
analysis with
be monitored
analyses
the
dioxide
temperature
composition
overheating
silicon
For
detector
from
(16), source
on
surface
a day-
the nitrogencru c i a l
that
carefully to insure valid
the
TID-Ng,
it
is
also
66
important to be aware of the relevant detector parameters
to
optimize
pres e n t e d
and
responses,
data.
source
signal-to-noise
use
of
consistently
and
but
begins
and
high
is obvious
Responses
temperature
signal-to-noise
noise
as
bias
to
levels
levels
the detector gas of choice.
detector
Both
are
the
molar responses
the
responses,
increase,
tower
where
optimized
Although
molar
previously
is reached
decrease.
voltages.
baseline
the
increase with
a point
ratio
increases
from
with
the
use of argon
the
leaving
associated
nitrogen
as
Because the detector operates
in an inert atmosphere of one type of gas only, responses
are less
dependent
on
the fine
injection reproducibility
Responses
of
the
do
decrease
alkali
from
the
of
the
decrease
is not
drastic,
can
performed
of flow
rates,
was routinely better than
over time,
contamination
be
tuning
indicating
surface
source
and
+ 10%.
either a loss
the
surface.
however,
without
of
and
source
This
resp o n s e
day-to-day
significant
or
analyses
losses
in
detectability.
T I D - N q - APIMS
A
was
specialized
constructed
within
it
were
for
ion
an
source,
APIMS
identical
to
Measurements
described
so
that
those
in
Experimental,
ionizing
existing
conditions
within
the
67
TID-Ng.
The
here:
following
nitrobenzene,
compounds
the
were selected for study
three
isomers of nitrotoluene,
and the three isomers of fluoronitrobenzene.
experiments
which
were performed
detection
monitoring
attempted
this mass
the
ion
range
very
for each of these molecules in
negative
by .mass
as each
specialized
in
or
of
A number of
ions
scans
either
from
10
by
to
single ion
500
amu
was
molecule of interest eluted through the
source.
In no experiments
observed.
sensitive
signals were detected.
It was
single
were
ions
in
specifically noted that
ion mode,
no
negative
ion
Examples of these experiments are
shown in the first two chromatograms of Figure 20 for the
case of o - nitrotoluene.
When the quadrupole mass spectrometer was operated in
the total ion mode,
analyte
molecules
negative
last
however,
studied
ion signals
three
negative ion signals for the
were
in
fact observed.
for £ - nitrotoluene are
chromatograms
of
Figure
shown
20 where
Total
in the
the
total
negative ions have masses greater than 20, 200 and 400 amu
respectively.
studied
masses
were
Analogous
obtained,
between
negative
10
and
results
for all eight
compounds
therefore
no individual
ions with
were
but total
500
ion signals were.
amu
The
observed
highest
cut-off level of
400 amu produced a signal of the same intensity as did the
68
m/e=4S
(No:)
c
m/e=137
3
( M O
u 0o
21
<D
in
c
m /e>20
a
0
<U
L. 2i
C
O
I)
m/e»200
Io
0)
c 2
CO
Z
qZ 1
<
role>400
L
0
3
6
t i m e (m in .)
Figure 20. Negative ion responses to 350 ng o-nitrotoluene
(GC retention time 4.4 minutes).
69
application
Figure
21
of
the
shows
molecules which
low
cut-off
threshold
were obtained in rapid
the
analysis.
relatively
tendency
ions.
uniform
of
analyte
is much
nitrobenzene,
signals
molecules
magnitude.
previously
TID-N2
Th i s
shown
constant during
concentrations
exhibit
are
the relative
to generate
negative
were
C/mol
negative ion
prod u c e
ortho
observation
and
responses
correlates
of
with
meta
of
the
o-nitrotoluene,
found
to be
respectively.
responses
responses
chemistry
the
that of the meta isomer
m - f l u o ronitrobenzene,
nitrobenzene
0.0002
than
isomers of
intermediate
TID-Ng
data
in Table 2 for the five compounds.
responses
benzene,
larger
while
fluoronitrobenzene
to
the
sample
so that
It is shown that the signal for the ortho isomer of
nitrotoluene
and
the
and
The
amu.
different
succession
sensitivity was reasonably
of
20
five total ion signals of five
the instrument
period
of
provides
observed
by
with
o - f Iuoro n i t r o -
m-nitrotoluene
and
0.29, 0.11, 0.02, 0.001 and
Therefore
observed in
recorded
The
an
evidence
the APIMS
Figure
actual
that
the
the
order
21 is identical
TID-Ng.
negative
is also responsible
the signals observed in an actual TID-Ng.
of
This
ion
for
70
I
NO2
I
F
NO 2
tim e (min.)
Figure 21. Total APIMS negative ion responses (m/e > 400)
to 3 5 0 n g o-nitrotoluene, 383 n g m-nitrotoluene,
310
n g nitrobenzene,
2 30 n g
o-fluoronitrobenzene, and 270 ng m-fluoronitrobenzene.
T h e a r r o w in each c h r o m a t o g r a m
indicates the k n o w n retention times of the
above substances.
71
TID - N q Candidate Response Mechanisms
Direct Electron Transfer
to Analyte Molecule
The
f i r s t ,c a n d i d a t e
considered
reversible
surface
for
the
electron
to
Introduction
the
as
N
response
mode
2
transfer
analyte
Reaction
of
the
from
TI D
is
as
was
to
a
the low work
molecule
I.
mechanism
be
simple,
function
shown
in
If the response mechanism
does involve a reversible electron transfer to form a gas
phase
molecular
negative
ion,
the
resp o n s e s
might
be
expected to bear a predictable dependence on the electron
affinity
of the
analyte.
More
specifically,
the
observed
current response to a molecule, M, might then be given by
Equation 17,
I = A T 2 exp [-(W-EAm )ZRT]
(17)
which is a variation of Richardson's equation
limiting
which
magnitude
assumes
that
of
current
all
by
negative
(21) for the
thermionic
charges
emission,
formed
at the
surface-gas interface are rapidly swept to the anode by an
applied
electric field.
function
heated
for . thermionic
surface
at
In this
equation,
emission
t empe r a t u r e
T,
W
is the work
of an electron
EAm
is
the
from
a
electron
72
affinity
of
M 7 and
T herefore7 if
responses
A is a constant
this
mech a n i s m
observed
molecules
such
within
as the
a
is
for a given
operative,
closely
substituted
surface.
the relative
related
group
of
might
be
nitrobenzenes
expected to increase exponentially with increased electron
affinity.
As
has
been
mentioned,
the electron
affinity
values for all of the mono-substituted nitrobenzenes shown
in Table 2 have recently been measured (32) and are listed
in the table.
absolute
plotted
this
reversible
factor
on
of
substituted
EA^.
if Equation
line
of
indicates,
substituted
do
nitrobenzene
positive
however,
bear
a
is
17 applies to
slope
is
over
the
that
nitrobenzenes,
not
2
the
strict
observed
exponential
the
electron
affinity.
electron
transfer
to the analyte molecule and
of
the
nitrobenzenes.
one
each
figure
of
22, the logarithm of the T I D - N
straight
responses
dependence
explain
a
This
g r oup
r elative
parts,
to
as a function
system,
entire
Figure
response
expected.
the
In
set
molecular electron
responses
of
If the data in
of
isomers
Therefore
affinity
all
of
Figure
at a time,
does
the
direct,
not alone
substituted
22 is considered in
select
portions of
the data might be taken to suggest that reversible, direct
electron transfer to form a molecular negative ion may be
p artially
inv o l v e d
or
that
it
may
be
one
of
several
73
E le c tr o n A ffin ity ( k c a ls /m o le )
Figure 22. C o m p a r i s o n of T I D - N 2 molar responses with
e l e c t r o n affinity v a l u e s for s u b s t i t u t e d
nitrobenzenes.
Nitrobenzene = I; m, £-nitroanisole = 2, 3, respectively; o, m, p-nitrotoluene = 4, 5, 6; o, m, £-fluoronitrobenzene =
I, 8, 9; o, m, £-ch loroni trobenzene = 10, 11,
12; o, ITii £ - b r o m o n i t r o b e n z e n e = 13, 14, 15;
o , m, £ — c y a n o n i t r o b e n z e n e = 16, 17, 18;
£, m, £-d ini trobenzene = 19, 20, 21.
74
mechanisms involved.
and
the
For
example,
dinitrobenzenes
affinities
and
they
the cyanonitrobenzenes
have
do, indeed,
the
have responses
as great or greater than the rest.
of
these
compound
types,
highest
the
electron
which
are
Moreover, within each
isomeric
variations
of
response appear to correlate reasonably well with electron
affinity;
and
is
the
dinitrobenzene
electron
It
that
can
para
have
isomers of cyanonitrobenzene
distinctly
larger
responses
affinity values than the meta and
also
be
seen
that
unsubstituted
and
ortho isomers.
nitrobenzene
has
the lowest molar response observed and also has one of the
lowest
electron
nitrotoluenes
reverse
within
the
their
affinity values.
and
nitroanisoles
response
each
TI D - ^
electron
compounds
response
dependence
isomeric group
relative
the
even
On
para
both
of halogenated
do
values.
isomers
though
exhibit
For
always
the meta
not
the
a distinct
on electron affinity.
r esponses
affinity
the other hand,
Also,
nitrobenzenes,
correlate
these
have
halogenated
the
isomers have
with
greatest
the highest
electron affinity values.
An
electron
additional
affinity
a r gu m e n t
on
benzoquinone.
Its
60.2
(32).
kcal/mol
against
the
importance
response
is provided
electron
affinity
Ye t
its
TI D - N
by
is
2
of
perfluorovery
high,
response
is
75
lower than
about
nitrobenzenes
expected
half of the responses of the substituted
shown
that
in
Figure
successive
dichloronitrobenzene
to
electron
form
response,
analyte
if
affinity
reversible,
molecule
however,
the
have
which
2 under the two
and
have
affinity.
source
t richloronitrobenzene
and
a n d , therefore,
direct electron
As
low
th e
the
are
T ID - N
transfer
shown
of
17 predicts
greater temperature
electron
on
responses
Equation
of
have
have significantly increased
nitrobenzenes
Additionally,
one might
chlorine
operative.
molar
chlorinated
compounds
is
Also,
additions
tetrachloronitrobenzene would
the
22.
in
di-
almost
to the
Table
to
2
2,
tetra-
constant.
that the responses of
electron
affinity
dependence than
values will
those of higher
Again, inspection of the data in Table
headings
heating
concerning
current
provide
detector temperature
no convincing
support
for this expectation.
Finally,
experiments
formed
at
emitter.
experiments
the
results
indicate
the
that
gas-solid
Neither
revealed
single
the
of the molecules studied.
no
of
the
molecular
interface
ion
TID-^-APIMS
of
ions
the
monitoring
existence of M
are being
thermionic
nor
scanning
species for any
JL
76
From
and
the
above
responses,
it
transfer
to form
all
the
in
mechanism
in
it
if
ion
it
a
electron
is involved
definitely
is
affinities
reversible,
negative
is
nor
If it can
to
that
a molecular
TID-N^,
o p erative
appear
appears
involved
instances.
is
considerations of electron
not
dominant
at
the
only
in
ma n y
one
be assumed that only one mechanism
the
TID-^,
be reversible,
that
mechanism
direct electron
does
transfer
not
to the
analyte molecule.
The
TID-Ng
is related
to the
ECD
by
virtue of the
fact that both provide selective responses to molecules by
the
formation
of
negative
ions.
It
is
interesting,
therefore, to compare the relative responses obtained with
each
in
order
to
see
if
mechanism are suggested.
measured
and
a
an y
additional
similarities of
Relative ECD responses were also
comparison
is
made
for the
substituted
nitrobenzenes and dinitrotoluenes in Figure 23.
comparison
exists
it
between
substituted
appears
the
that
rates
nitrobenzenes
on
absolutely
of
electron
a heated,
surface and in a gaseous medium.
ECD
and TID-Ng
Figure
23,
response
and
no
From this
correlation
capture
low work
to
the
function
Also compared were the
responses of other molecules not shown in
with
these,
magnitudes was
also,
observed.
little
correlation
It is not
of
necessarily
77
Relative E C D
Response
Figure 23. C o m p a r i s o n of absolute T I D - N 2 responses of
substituted nitrobenzenes and dinitrotoluenes
with their r e l a t i v e molar E C D r e s p o n s e s .
Nitrobenzene = I; m, p-nitroanisole = 2, 3;
respectively; o, m, p-nitrotoluene = 4, 5, 6;
o,
m , p-chloronitrobenzene =
10, 11,
12;
o , m, p - b r o m o n i t r o b e n z e n e =
13, 14,
15;
o , m, p - c y a n o n i t r o b e n z e n e =
16, 17,
18;
o , m, p - d i n i t r o b e n z e n e = 19, 20, 21; 2,3-,
2,4-, 2,6-, and 3,4-dinitrotoluene = 23, 24,
25, 26.
78
surprising
responses
TID-Ng
work
that
no
exist,
unless
involves
function
constitute
correlation
gas
the
between
expected
phase electron
surface.
This
E CD
and
mechanism
capture
near
possibility,
TID-Ng
for the
the low
which
a slight alteration of direct electron
would
transfer
to an analyte molecule, appears also not to be operative.
' Thermal Decomposition of Analyte
Followed by Electron Attachment
The
second
thermal
candidate
decomposition
interface
followed
decomposition
by
products,
This
conclusively
because
however,
steps
it
nitrogen
of 53
as
an
was
shown
is
is
attractive
is
assisted
formed
at
gas-solid
to one of the
as
Reaction
difficult
known
involves
the
transfer
possibility
little
mechanism
analyte
in this process.
dioxide
temperature
it would
is
the
an electron
Introduction.
elementary
of
response
of
to
the
by
decomposition,
suc h
a
electron
prove
possible
As mentioned
alternative
2 in
earlier,
because
if
surface
or
transfer
to
be facilitated by its very high electron affinity
kcal/mol
mechanism
explanation
is
(31, 32).
that
for
the
it
Another inviting feature of this
offers
anomalous
the
following
plausible
relative responses of the
nitrotoluene isomers and nitrobenzene, previously shown in
Figure
22.
The
o r t ho
isomer
responds
ten
times more
11 'I
79
strongly than the second-most strongly responding £-nitrotoluene.
This might be explained by Reaction 18,
in which
a benzyl rather than a phenyl radical is formed
by
hydrogen
atom
adjacent
methyl
than
phenyl
a
transfer
group.
to the reaction
The
radical
formation
decreases
site from
the
of a benzyl rather
the
energy
of
the
decomposition by 23 kcal/mol as deduced from benzyl versus
phenyl
C-H
bond strengths (53).
stabilization
of
is possible with
nitrotoluene
surface,
but
during
the
The same type of radical
the meta
their
and
para isomers
decomposition
decomposition
on
the hot
of these molecules would
not occur as rapidly due to the increased distance between
their
methyl
nitrobenzene,
possible
no
of a higher
Table
2,
such
th e
at
the
to
to nitrogen
energy
a
reaction
assistance
According
decomposition
fast
and
as the molecule comes
surface.
as
group
phenyl
a methyl
Wit h
group
is
in close contact with the
this
model,
dioxide requires
radical,
given
temperature
relative
responses
are also reported.
by
site.
its
initial
the
formation
and, therefore, is not
of
the
of four
surface.
In
dinitrotoluenes
Again, it is noted that the one isomer
80
among these which does not have a nitro group adjacent to
a methyl
group,
3,4-dinitrotoluene,
also has
the lowest
TID-Ng response.
In
considering
proposals,
the
above
ion
spectra
toluenes
and
nitrobenzene
M-,
capture
at
(M-O)- , and
shown
NOg
At
200°
significant
abundance
distinct
feature
intensity
of
300°
C,
C.
capture,
The
for
only
the
for
all
spectra
the
result
due
from
to
four
in
gas phase
the
for
M - ion
ortho
the base peak
cause
nitrotoluenes
the
24 of the nitro-
observed
for
is
species
all
four
formed
compounds.
Figure
except
decomposition
isomer,
in the
this may
to the mechanism just proposed
the
Figure
Ions
are
the
only
it becomes
nitrotoluene.
of
mechanistic
24
in
The
is
the
NOg ion at m/e=46 is greatly increased
the
temperature,
of
C
in
which
300°
compounds.
by
other
it is interesting ■to note the high temperature
negative
electron
and
so
t-hat at
spectrum of o-
be closely related
for the surface ionization
that
step
in gas
would
phase electron
be
expected
to
follow, rather than precede, the ionization step, as shown
in Reaction 19.
H
>
>
(19)
Relative Intensity
81
Figure 24. High temperature, negative ion electron capture
mass spectra of nitrobenzene and three isomers
of n i trotolu ene.
Ion source gas, 0.3 torr
methane; source temperature, 300° C ; sample
introduction by capillary GC.
82
In this
an
case the
adjacent
than
second
methyl
step may
group
which
again
permits
a phenyl radical to be formed.
phase
it may
radical
only
assistance
formed
be possible
for the
of
a
to form
phenyl
in the production
isomers
of
nitrotoluene,
nitrobenzene.
relative
This
rates
of NO^
the more
of
That
from
explain
NO^
stable benzyl
is, without
radicals
as they must
may
a benzyl rather
Moreover, in the gas
ortho isomer.
surface,
be facilitated by
probably
the meta
be
why,
in
production
the
in
and
are
para
the case of
Figure
for
24, the
m-
and
£-
by
the
nitrotoluene and nitrobenzene appear to be equal.
Evidence
against
TID-N^-APIMS
(NC> ) were
does
indeed
experiments.
o bserved
2
This
this mechanism
not
exclude
formed,
interactions
for
any
negative ions of mass
of
the compounds
the possibility that
however,
following
No
is presented
and
their
NO^
are undergoing
formation.
The
46
studied.
ions
are
ion-molecule
possibility
of this occurring is discussed later.
Reaction of Analyte with Gas Phase
Radicals Followed by Electron
Attachment to Product(s)
This mechanism/
is considered
shown
as Reaction
3 in Introduction,
here because of its acknowledged importance
in the nitrogen-phosphorous selective mode of the TID.
In
83
this
candidate
decomposed
by
response
reactions
mechanism,
with
th e
reactive
analyte
is
species, R, which
are continuously formed at the gas-solid interface by the
decomposition
detector
gas
of
either
or from the
permanent
analyte itself.
gas is a very stable molecule,
atomic
gas,
such
as
components
argon,
such
any
of
the
If the detector
as nitrogen,
reactive
or is an
species present
would have to be formed from either carrier gas impurities
or the analyte.
In
order
detector
gas
compounds
detector
included
90%
examine
in
were
the
The
detector
in
striking
variations
90%
using
0.30%
nitrous
nitrogen,
in
potential
a
variety
oxygen,
oxide,
of
these
detector
gas
in Table
data
is
cause
argon,
the
vast
majority
of
2.
that
only
all
different
carbon
isobutane
alterations of the absolute responses.
of
of
to
experiments,
90%
15%
are reported
participation of
responses
results of these
gases
feature
the
T I D - N 2 , the
mea s u r e d
gases.
argon,
methane
to
which
dioxide,
and
15%
The most
these
wide
relatively minor
With the exception
responses
fall
within
factors of about 0.4 to 1.0 of the responses observed with
nitrogen.
suggested
their
If
for
reactive
this
production
species
mechanism,
rates,
as
were
one
well
might
as
important,
have
their
as
expected
destruction
84
rates,
to
detector
have
gases.
mech a n i s m
levels
been
is
near
however,
not
reactive
conclusion,
detection
for higher
this
the
operative
may
at
be
concentration
molecule
possible,
levels of certain
operative
is
varied
is that this
It remains
species
analyte
these
therefore,
limits.
mechanism
of
in
concentration
intermediate
destruction
different
generally
the
that
compounds,
The
very
where
produced
itself.
the
by
the
This
may
provide a clue as to why, in several cases shown in Figure
16,
the
molar
concentrations
apparent,
responses
of
for
increase
sample.
example,
This
for
with
increasing
was
particularly
trend
all
isomers
of
cyano-
nitrobenzene.
The
use
consistently
times
that
mentioned,
higher
of
greater
in
the
with
90%
pure
as
detector
responses,
nitrogen.
baseline
the
argon
ranging
Also,
as
gas
from
was
caused,
1.3
to 4
previously
emission current was significantly
use of argon.
This result may
be more
easily explained in terms of physical rather than chemical
e f fects
of
the
detector
gas
if
one
assumes
that
the
detailed nature of the emitter source surface and its work
function,
W, can
surrounding
of
the
be altered
gas.
gases
Argon,
used,
slightly by the nature of its
being
the most
interacts
least
chemically
with
the
inert
cesium
Ii
85
enriched
surface
function
to
explanation
detector
be
gases
cause
may allow
been
thermal
oxide,
at
possible physical
c o n s i d e r e d ; that
conductivities
due
which,
to their
in
turn,
differences in the temperature of the emitter
surface. The thermal
oxygen,
Another
effective work
variations of responses "with different
has
in
the lowest
established.
of the
differences
might
and
68.2;
46.1;
49° C.
conductivities
carbon
dioxide
isobutane,
44.2;
are:
43.8;
and
nitrogen,
argon,
methane,
While this factor could
45.5;
nitrous
89.3 units
explain why
65.7;
(54)
responses
and baseline are greater in argon due to its lower thermal
conductivity than that of nitrogen, and an expected hotter
surface, it does not explain the lower responses observed
in
carbon
carbon
dioxide,
methane,
source
it
nitrous
nitrous
appears
surface
effective
lowered
dioxide,
work
and
that
an
function
responses
oxide,
oxide,
and
isobutane.
isobutane,
chemical
and
probably
modification
accompanying
increase
is
cause
of
these
detector
o b se r v e d
the
in
the
For
of
the
in
the
slightly
gases.
I
8 6
Thermal Decomposition Followed
by Polymerization and
Electron Attachment
As was previously mentioned, a specialized ion source
for
an
APIMS
conditions
mode,
the
constructed
a TID-E^
detector.
potentially
expected
ions
molecular
but were
10
monitored
never
to
anion
500
amu
of the mass
resulted
in
studied.
These
clear
type
during
observed.
operation
The
to
within
continuously
from
was
did
for
of
the
each
the
the single ion
type
NO
2
or of
compound
of the mass
not produce
signals
were
spectrum
signals either, but
in the
for
all
shown in
these
ionizing
chromatographic elution
Scanning
results were
indication
In
of
spectrometer
measurable
simulate
total ion mode
the
compounds
Figures
experiments
20 and
is
21.
that
no
negative ions of mass less than 500 amu are formed in the
T ID -N
2
.
With
this
in
mind,
t he
fourth
candidate
response mechanism is considered, thermal decomposition of
the analyte molecule,
M,
followed
by
initiation
by
one of
the products of a polymerization reaction with several (n)
other, analyte
molecules
to
form
some large
species
X M fi
which then attaches an electron, as shown in Reaction 20.
M ---- > N + X
■n M --> X M n "^^S—
> XMn
(20)
87
This
appears
possible
due
to the previous mass
spectral
data discussed and the probability that this large species
may
have a considerable number of degrees of freedom and
could
is
easily
accommodate
o perative,
expected
The
T ID -N
molar
2
to have very complex
molar
complex
the
an .electron.
r e sponses
might
relationship
If this mechanism
responses
concentration
be
expected
to
might
be
dependencies.
depend
on
a
such as the n + 1 power of the analyte
concentration since n+1 sample molecules are combined into
the
terminal
concentration
negative
curves
in
ion.
Figure
The
previously
16, however,
shown
indicate that
complex concentration dependencies are not necessarily the
rule.
The
were indeed
T ID - ^
the nature of
diversions from linearity are not consistent
with Reaction
very
Also,
molar
linear
and
to
be
non-linear,
to some compounds
but
20.
shown
detector responses
responses to
some
linearity near
observed for nearly all the analytes.
compounds are often
the detection limits is
88
Thermal Decomposition Followed by Combination
with Inorganic Species and Electron
Attachment at the Gas Solid Interface
Use of the APIMS for the measurement of ions produced
by
the
T ID - ^
created
a problem
heating
current
was
applied
aperture
between
the
ion
was
consistently
plugged.
of
a
within the
as
quickly
source,
the
the vacuum
coated
and
envelope
eventually
as
foreline pressures
With
response
combination
of
the
process.
possible
Experiments were
once
heating
of
the
and no signals were observable until
a
species,
this
observation
mechanism
decomposition
inorganic
and
After a few hours of use, the aperture
replaced.
candidate
in manifold
accompany
was virtually clogged
thermal
continuously
decrease
source commenced.
was
source and
emitter
the
MS and by the steady decrease in mass spectral
which
performed
it
the
When
This phenomenon was readily noted by observation
steady
signals
and
to
immediately.
of
the
is
I, and
considered,
sample,
decomposition
in mind,
M,
that being
followed
product,
by
X, with
an electron attachment
terminal species XI, as shown in Reaction 21,
a fifth
a
an
to the
89
M ---- > N + X --- ----- > XI
where
the
addition
when
M
combination
of
is
an
a
step
urface^ > xi“ '
precedes
electronegative
substituted
(21)
attachment.
species
X
nitrobenzene)
The
(possibly NC^
to an
inorganic
species may facilitate the transport of a surface electron
to
a
species
XI
at
mechanism
seems more
response
mechanism
other
analyte
have
a
ga s
plausible
which
molecules
complex
approximately
th e
solid
than
fourth
This
candidate
polymerization
with
the responses would
concentration
linear
the
involved
because
interface.
responses
dependence.
displayed
in
not
The
Figure
16
could be explained since the large majority of the mass of
the negative ion formed originates from the source surface
and
not
the
analyte
molecules.
obvious that material was
the
heated
emitter
Additionally,
continuously
source.
The
role
being
of
it
was
emitted from
decomposition
species, X, therefore, may not be to cause the emission of
a
neutral
surface,
species
but
to
from
the
cesium-embedded
facilitate the attachment
ceramic
of an electron
to the terminal species of which it is a part.
i
90
Thermal Decomposition Followed by Electron
Attachment and Association with Gas Phase
Species Removed From Gas-Solid Interface
The sixth and last candidate response mechanism is a
more
trivial
interpretation
in this study.
analyte
of the
observations reported
Possibly the thermal decomposition of the
molecule
does
response magnitudes
indeed
determine
and that the high
the
T ID - N
2
mass ions observed
here are formed in the gas phase by ion-molecule reactions
of the
the
smaller negative ion
gas-solid
interface
X - with
(G)
as
species
X-
removed from
migrates
to
the
collector electrode, as shown in Reaction 22.
M ---- > N + X e(surface)) x --- G--- > XG
If
the
species
X-
is
indeed
NO
2
this
(22)
negative
ion
would provide a particularly good center for clustering by
a nitrobenzene molecule
greater than
500 amu
(47).
If negative ions of masses
are to be formed by this mechanism,
incorporation of many analyte molecules would be required.
The
very
ion
formation
constants
for
multiple clustering
will be
small (47), however, at the temperatures at which the
source was
operated
and
the
formation
of very
large
91
mass
negative
will not
ions
occur.
by
clustering
If, however,
large inorganic character
a
single
clustering
from
attachment
the
the
of
an
association
involves
phase,
rather
step.
limiting
process
mechanisms.
rather
both
in that the
than
species
however,
of
to
follows,
the sixth mechanism
gas-solid
analyte,
mass
This mechanism is
inorganic
the
Firstly,
they
phenomenon.
extension
consistent
at
source,
the
in the gas
interface.
could
last
be
two
The
the rate
proposed
of these mechanisms are attractive for
r e asons.
an
500 amu.
precedes,
an
for
Both
aperture plugging
are
with
of the
heated
the necessary
Furthermore,
than
species is a
candidate mechanism
electron
reaction
decomposition
many
fifth
from the
provide
result in an ion greater than
distinct
analyte molecules
the clustering
emitted
could
with
with
of
th e
the
take
into
account
the
Secondly, these mechanisms
second
supporting
mechanism
evidence
and
are
thereof.
Therefore, the anomalous responses of nitrobenzene and the
nitrotoluenes
NO2
shown
ion
in
the
(Figure
24)
of
in
high
said
each,
are
bo t h
mechanisms
experiments
in
22, the
te m p e r a t u r e
molecules,
a greement
that
Figure
offer
with
an
indicate
and
abundance of the
negative
the
ion
explanations
these mechanisms.
explanation
the
spectra
formation
for
of
for
Thirdly,
the
APIMS
ions of at
92
least
mass
500
literature
of
amu.
Finally,
there is evidence
organic
species
combining
species to form large ions.
heavy
ions
can
be
the
nitrogen-containing
can
combine
ions.
of
with
Allison
the
species
and
potassium
Whether
products
cesium
species
coworkers
(57)
(M+K) + , where
K
is
a
the
emitter
phenomena
to
is
reported that
combustion
These
form
heavy
even
of
ions
heavier
a molecular
atom,
in
is used
have
inorganic
have also reported ions
M
potassium
thermionic
these
of
(56)
organic compounds.
and
type
Shteinboc
with
in the
any
an
analyte
MS
as the
where
ion
connection
a
source.
with
the
observations described here remains to be seen.
A
one,
variation
is
transfer
that
of this
species
reaction
with
theme,
X-
is
and
certainly
undergoing
species
G
to
form
a plausible
an
a
electron
high
mass
negative ion, as shown in Reaction 23.
X™ + G ----- > X + G-
The
thermal
the
rate
limiting
experiments
and
the
decomposition
data
decomposition
of the analyte would
process
performed
for
here with
obtained
of the
(23)
analyte
tend
this
still be
variation.
The
the
available
equipment
to
support
thermal
followed
by
association with
93
an inorganic species either in the gas phase or at the gas
solid
interface,
the decomposition
mechanistic
further,
with
and
with
of
an
occurring
of the analyte.
operating
additional
use
ionization
In order to explore the
principles
experiments
MS
sometime after
as
of
the
described
of
higher
mass
c o l l i s i o n a l .activation,
MS-MS
(48)
range
TID-N2
here, but
and
with
capabilities
are
required.
One
final
additi o n a l
(Figure
bias
point
on
discussion.
19)
this
That
T I D - N 2 response
voltage.
As was
subject
is
the
is
previously
dependence
previously
worthy
on
indicated,
the
of
sho w n
source
all responses
are continuously and linearly increased by an increasingly
negative
to
bias
transport
response
voltage.
negative
If the
ions
applied
field
to the anode,
serves
only
a saturation of
should be observed at field magnitudes much less
than the maximum applied -45V, based on calculations.
The
mobility
the
of
the
ions
in question
are determined
by
Langevin equation (72) shown as Equation 24,
M 0= 13.876/(7 nv )1/2
where
c m 2/V sec
a.O is the reduced mobility,
7
(24)
is the polarizability
94
of
carrier
gas
molecules
reduced
mass
the
ion
g/mol.
Since
is
the
light
of
(28
mass
a m u ) compared
of
the
carrier
proportional
Equation
24.
voltage
to be
small
the
essentially
some
as
other
Considering
possibility
100
the
these
10,000
swept
the
in
(N2 )
amu
amu
mobility
is
as shown
by
and
greater
anode
by
an
it
the
should
applied
has
been
(55) that once the mass of an ion
level,
ion
Obviously
than
mass
Furthermore,
the field
mobil i t i e s
the
sweeping
previously
is that
The
relationships,
to the
-15V.
constant.
role
is
gas molecule
molecule.
to the reduced
experimentally observed
reaches
mr
to the mass of the negative
Considering
enough
as
a carrier
gas
mobilities of ions of mass
great
and
(>500 amu), the reduced mass converges on
inversely
be
A3
the mass of the carrier ■gas molecule
charge carrier
the
and
in
applied
ions
discussed
to
remain
field has
th e
anode.
mechanisms,
is facilitating the
one
emission
of ions, in a linear manner, from the source surface.
The
fifth mechanism proposes the formation of. a heavy mass ion
at
the
surface
of the
source,
while the sixth
mechanism
involves formation of a small mass ion at the surface.
the
field
surface,
involved
does
this
assist
would
formation
in the
lend
of
an
emission of
credence
ion
to
ions
from the
m echanisms
directly
at
If
the
that
source
95
surface.
The
assisting
the
final
possibility
ionization
is
process
that
the
itself.
field
At
any
is
rate,
further elucidation of the field's role may lead to a more
precise description of the mechanism.
ECD Responses of the
Substituted Nitrobenzenes
The
resp o n s e s
ex a m i n e d
in
azulene,
in
are
an
varies
such
from
quite
the
curves.
This
(58-61)
and
rate
that were
along with
a wide range of concentrations.
25 and
26.
compound
linear
to compound,
has
reaction
and
as
because
is
of
electron
and
300°
C,
some
analytes
others
behaving
responding
and
the
curving
been
the
calibration
previously
addressed
When
cell,
destroyed
the
compounds,
polychlorinated
follows.
enters
analyte
while
upward
explained
the
constant
have
C
The degree of linearity
strongly
phenomenon
sample
200°
behavior
Very
often
concentration
capture
also determined,
b r omonitro b e n z e n e s
n i t r o b e n zenes,
of
36 compounds
temperatures,
non-linearly.
fraction
were
over
Figures
greatly
as
same
area calibration curves determined using two
in
exhibiting
quite
ECD
detector
shown
the
the TID-Ng
Selected peak
different
of
by
a
a
substantial
the
electron
large electron
density.
For
low
small
capture
analyte
96
KHz-min x10™J /mol
8
0
0
,
GOO-
400-
g sample
Figure 25.
E C D peak area responses to varied amounts of
substituted nitrobenz e n e s at 200° C and 30
mL/min nitrogen.
The compound associated with
each calibration curve is indicated by the
number assigned in Table 3.
97
1000
600-
400-
200 -
g sam ple
Figure 26.
E C D peak area responses to varied amounts of
substituted nitrobenzenes at 300° C and 30
mL/min nitrogen.
The compound associated with
each calibration curve is indicated by the
number assigned in Table 3.
98
concentrations
the electron
in
like
with
a linear response
this
small
sizes increase,
reacts with
molecules
used
analyte molecules
concentration
however,
as
entering
the
to be reacted
range.
When
sample
a smaller population of electrons
a successively
for these
smaller portion of the
cell.
experiments,
In the
CCP-ECD,
analyte
the
device
the signal is proportional to
the instantaneous
concentration
detector.
The
response
related
to analyte
linearly
as long
density remains large and constant, resulting
fractions of the
over
is observed
of the
cannot,
analyte within the
therefore,
concentration
over
also
the
be
entire
dynamic range of the detector.
The
peak
concentration
addition
area
r esp o n s e s
of each
to several
compound
other
t e m p e r a t u r e s . are
obse r v e d
in Figures
compounds,
listed
for
in
Table
mol- 1 ] X
I O - ^)
the
25 and
at the
3 as
two
the
detector
absolute
([K H z -min
The known
electron affinities are again listed in Table 3.
at
listed
which
five
for
the
times
detection
the
in Table
3 are the
response
the
noise
limits
that
absolute
observed
that
26, in
response
Also
of
lowest
detection
to each
analyte was
level.
It
the
ECD
detection
of
is
group
limits
approximately
obvious
is superior
this
substance.
from
these
to the TID - N
2
of molecules.
These compounds are classic ECD analytes, containing large
99
Table 3
ronpound*
I.
2.
3.
4«
5.
6.
7.
8«
9.
10.
11#
12.
13.
14.
IS.
16.
17.
18.
19.
20»
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
El e c t r o n c a p t u r e detector
detector temperatures.
Electron Affinity**
SZ cal mol-'*')
IB
JD-OCH3B
D-OCH3NB
C-CHgfB
JD-CHgtB
Jp-CHgtB
fl-FIB
JirFtB
D-FIB
O-ClIB
ErCltB
D-ClIB
O-BtiB
JS-BrIB
D -BtiB
O-CMB
JirCNtB
D-CMB
D-NO2IB
ErtDgtB
P-NO2IB
ID-CF3IB
2,3-DtiT
2,4-ONT
2,6-DNT
3,4-ONT
2,4-a2IB
3,4-Cl2IB
2,3,4-Cl3IB
2,3,S-Cl3IB
2,3,4,S-Cl4IB
2,3,5,S-Cl4IB
l,2,3,4-a4B
Q 5B
Q fiB
Perfluoro
benzoquincra
37. Azulene
Detection
Limit 200*
SsaL
22.1
22.7.
19.6
20.0
21.4
20.5
23.5
27.5
24.3
25.0
28.1
27.5
25.5
28.8
28.3
35.1
34.1
38.1
36.2
36.2
43.6
30.9
5
4
9
5
5
20
3
6
2.5
2
0.5
I
I
0.5
I
30
15
3
2
3.5
1.5
0.5
2.5
10
' 1.5
7
4
2
3
3
4
4
3
2.5
5
60.2
17.3
30
30
Absolute Molar
Response***
2oa*
43.0
57.5
13.8
52.5
66.4 .
29.1
149
52.6
82.6
97
317
182
100
288
158
10.4
36.0
119 .
272
82.9
456
190
225
52.9
385
106
36.7
56.5
72.6
. 43.2
230
91
74.2
ISO
228
3.73
7.92
responses
Absolute Molar
Response
321*
21.6
35.9
77.1
7.25
14.8
4.65
99.7
66.1
91.9
152
374
306
584
253
316
104
79.5
544
440
242
769
428
289
149
393
207
180
386
399
265
384
248
186
234
462
18.6
0.388
*B « benzene; JB - nitrobenzene; CNT - dinitrotoluene
“ Electron Affinity Values taken frgn ref. 32
•“ Peak Areas in (KHz-rain mol-1)x IO-9
••••Ratio of Absolute Molar Response at 300* to that at 200°
at
two
Effect**** <
Tpmmrmrir*
0.50
0.62
5.59
0.14
0.22
0.16
0.67
1.26
1.11
1.57
1.18
1.68
5.84
0.88
2.00
10.0
2.20
4.57
1.62
2.92
1.69
2.25
1.28
2.82
1.02
1.95
4.90
6.83
5.50
6.13
1.67
2.73
2.51
1.56
2.03
4.99
0.049
100
pi
s y s t e m s . and
responses,
than
do
that
the
as
however, .vary
the
TID-N
ECD
responding
o p posed
electronegative
is
less
from
responses.
2
the
less
functionalities.
compound
This
selective
The
to compound
observation
detection
shows
device,
to high electron affinity molecules in general,
to
the
TID-N^
which
is especially
sensitive
to molecules with multiple nitro groups.
The effect of temperature of the detector housing on
the responses to each
300°
of
C
the
is
listed
Molecules
temperature
with
whose
over the range of 200° to
in the sixth
compounds
significantly
compound
have
column of Table 3.
responses
temperature,
responses
while
decreased
Most
that
increase
a
decrease.
few
with
increasing
usually contained a relatively small degree of
electronegative
functionality,
The
of
dependence
response
such
on
the
will be discussed in more detail later.
as the
nitrotoluenes.
detector
temperature
1 0 1
Computer Simulations
of ECD Response
The first
in
E CD' response mechanism that was
Introduction
forming
was
a resonance electron
a
molecular
negative
designated
mechanism
I.
ion.
capture
This
Experimental
discussed
process
mechanism was
data
obtained
with
the CCP-ECD were compared with simulated data generated by
a computer program that modeled the resonance ECD response
as a function of a number of different parameters.
parameters
constant
The
included
(k ^ ) and
detachment
affinity
and
the
t he
rate
forward
d etachment
constant
temperature,
electron
rate
wa s
capture
constant
is a function
and
These
rate
(k_j).
of electron
calculated
using
Equation 25 (65)
k_! (s'1) = k 1 2.07 X i o 16t 3/2
(25)
exp[ EAniZRT]
where
EAm
molecule.
was
assumed
is
The
the
electron
entropy
to be
rate constant was
zero,
change
affinity
of
the
analyte
for the ionization process
and the forward electron capture
assumed
to be temperature
independent.
1 0 2
Shown
in
different
Figure
forward
27
rate
the
temperatures
forward
detachment
does
not
to
for
At
a
generated
for several
at a constant electron
It is obvious that the responses
are proportional
co n s t a n t .
reform
commence
reached
mol.
rate
data
constants
affinity of 20 kcal/mol.
at low
is
the
The
molecule
program
analyte
until the
to the magnitude of
predicts
molecule
and electron
200° C temperature level
of electron affinity of
temperatures
that
below
C, the
2 0 0°
is
20 kcal/
response
is
constant with temperature, therefore the imagined negative
ions
are
electron
with
not
detaching
capture
rate constant
temperature,
long
as
Figure
the
28
the electron.
simulated
detachment
depicts
Since
is assumed
the
to be constant
responses will not
rate
simulated
constant
data
for
forward
is
change
as
negligible.
several different
electron affinity molecules at a constant forward electron
capture
rate
maximum
simulated
electron
affinity molecules
where
simulated
dependent
with
the
dependent
the
constant
upon
the
that
I
respo n s e
responses
upon
fact
of
10 ^ cc molecule
for
is the
begin
each
of
same.
the
The
the
The
to decrease,
electron affinities.
detachment
the electron
temperature.
X
rate
.
The
different
temperature
however,
This
is
correlates
constants
are
affinity of the' molecule and
program
predicts
that
once
the
103
300H
200
-
100-
T(0C)
Figure 27. C o m p u t e r s i m u l a t e d data for the resonance
electron capture process.
Electron affinity =
20 kcal/mol. Forward attachment rate constants
=
I
x
10
cc
mol"
sec"
(A);
5 x I0~
(B) ; 2 . 5
X I0~
(C) ; a n d
1.25 x 10"b (D).
104
300-,
i=100-
80
160
240
T(C)
320
400
Figure 28. C o m p u t e r s i m u l a t e d data for the resonance
electron capture, process._ Forward attachment
rate = I x 10
cc mol 1 s e c - .
Electron
affinity = 30 kcal/mol (A); 25 (B); 20 (C); and
15 (D).
105
electron
affinity
kcal/mol,
of
the
detachment
molecule reaches
should
a level of
not be a significant
30
factor
throughout the working temperature range of the CCP-ECD,
and
decreases
in
response
with
increasing
temperature
should not occur.
Comparisons
computer
ECD.
program
Figure
data from
molecule
were
made between
data generated
and. data obtained
29
shows
a CCP-ECD
azulene.
both
and
One
the
the
can
with
the
actual CCP-
experimentally
simulated
by the
obtained
responses for the
see that the computer program
predicts the
actual decrease in response
quite well.
In fact, it predicts the temperature at which
responses
for most
decrease
(therefore detachment
of the electron)
molecules whose responses actually decrease with
increasing temperature.
azulene
with temperature
actually
temperatures.
One can
increases
see that the response of
with
temperature
at
low
The forward electron capture rate constant
is assumed to be temperature independent for this program,
however,
This
and
could
temperature
not predict this increase in response.
dependence
detail in the next section.
will be discussed
in greater
106
EXPERIMENTAL
SIMULATED
IOOOO
800060004000-
2000
-
T(eC)
Figure 29.
E x p e r i m e n t a l l y d e t e r m i n e d E C D temperature
dependent response curve and computer generated
curve to 50 ng azulene.
Forward attachment
rate constant estimated to be 1.55 x IOcc
mol"
sec
and temperature independent.
107
ECD Response Mechanisms and Temperature
Dependence of Substituted
Nitrobenzenes and Azulene
As was pointed
response
mechanisms
cpworkers
by
(36,
these
37, 41).
This
electron
by
negative ion.
from
the
the
least
in
was
shown
as
Reaction
gas .phase
analyte molecule
to
but
can
Wentworth
of
molecules
mechanism
responses
region
Equation
form
and
be
and
as fast
Che n
reacting
having
would
2 where
by
to the
the
is
forming
decreases
increase
occurring,
in
a molecular
(43)
via
electron
as 10 ' cc
concluded
this
electron
affinities of at
(Figure 2)
be temperature independent.
corresponds
Nitrobenzene
response
4
I
capture of
kinetic relationship
recombination
rate
its
a
with
indicating
and
a
molecule
that
molecular
the
200°
to
dissociative
electron
most
C
mass
(k+ )
.
i o n . . Its
temperature
processes
capture
shown
certainly
negative
300°
The
constant
is greater than the detachment rate constant (k_^).
responds
and
mechanism
18.5 kcal/mol would be in the beta region
the
beta
Wentworth
Rate constants for this process vary widely
data
capture
by
The first, designated
molecule to molecule,
that
proposed
is a reversible,
mol""-*- s e c - "*" (62).
and
were
investigators,
Introduction.
an
out in Introduction, four possible ECD
are
not
spectrum
(Figure
24)
indicates
any
abundance
C.
Its
18.5
at
the
electron
kcal/mol
only the
molecular ion is formed
relatively
affinity
level
of
high ■temperature
22.1
where
the
kcal/mol
exceeds
the
response
should
be
lie in
the
independent
beta region.
The
ECD response of 3.13 ng nitrobenzene was
r e corded
the
author
This
can
data is
be
in
Figure
identified
on
this
of
80°
range
significantly
conclusion
of
30.
to
and
should
range
of
Three
distinct
plot.
The
This
Chen
data
(43)
temperatures.
first
130° C where
increasing.
Wentworth
data
a wide
shown
temperature
is
at
the
300°
temperature
by
and
of
in
regions
is
the
in
response
conflicts
that
the
the
with
electron
capture process should be temperature independent for this
mechanism,
for
this
recorded
here
and,
in
molecule,
using
fact,
shown
a tritium
disagrees
in
Figure
source,
were
obtained
with
previously
described
in
with
31.
6I
Ni
Introduction
data
Their
FFP-ECD,
the
their
data were
while
source,
and
(37)
the
data
CCP-ECD
Experimental.
The second region identified in Figure 30 is indeed one of
temperature
240°
C.
Any
independence,
variations
due to experimental
and
temperature level,
the
recombination
with
occurring
in
response
random
errors.
negative
positive ions
ions
before
between
here
13 0°
are
and
probably
Up to the 240° C
are
undergoing
detachment
of the
KHz-minx10
/mol
109
Figure 30.
E C D temperature dependence
3.13 ng nitrobenzene.
response
curve of
H O
I/TCIO
Figure 31.
Ln
versus 1/T plot
nitrobenzene
obtained
Wentworth and Chen (37).
for ECD response of
e x p e r i m e n t a l l y by
Ill
electron ■can
30
occurs
The
occur.
between
electron
hig h
to
Finally,
the
third region
th e
analyte
temperatures.
electron
molecule
This
in
temperatures of 240° and
affinity of the molecule is not
prevent
original
the
from
and
350° C.
sufficiently
detaching
electron
detachment region
Figure
to
at
the
these
corresponds
to an
alpha kinetic relationship (Figure 2 and Equation 9) where
the
detachment
recombination
for
the
the
alpha
rate
rate constant.
author's
data
detachment
Chen's prediction
for
their
mechanistic
250°
C
and
constant
and
region
data.
it
appears
absent
The
is
32, indicating
Wentworth
that
they
temperature
the
the
1/T plot
formulating
a maximum
this
from
detector
for
than
versus
in Figure
and
had
larger
In KT"^^
is shown
experiments
proposals
A
is
and
used
their
limit of
temperature
where
detachment of the nitrobenzene negative ion begins.
Nitrobenzene
these
results
temperature
nitrotoluene.
low
were
and
not
the
only
obtai n e d .
dependence
curves
molecule
Figure
of
the
33
three
for which
shows
the
isomers of
All have a region of increasing response at
temperatures,
region
was
a
significantly
followed
region
with
of
by
a temperature
response
temperature.
tha t
independent
decreases
A g a i n , the
most
interesting feature of this data is at the low temperature
1 1 2
1.5
1.7
1.9
2.1
2.3
2.5
2.7
2.9
IZT(K) xIOOO
Figure 32. Ln K T 3//2 versus 1/T plot for ECD response
nitrobenzene, data obtained by author.
of
113
ortho
KHz-min x 10"
Gi
meta
O
£
para
Figure 33.
E C D temperature dependence curves for 100 pg
o-nitrotoluene, I ng m-nitrotoluene, and I ng
£-nitrotoluene.
114
levels
where
it
is
obvious
that
the
electron
capture
process is not temperature independent.
The
fluo r o n i t r o b e n z e n e s
isomeric
the
variations
absolute
ECD
can
make
response
a
function
isomers.
The
temperature
This
ort h o
molecule
temperature
These
two
negative
these
mechanisms
ion
to
at
are
very
be
electron
independent.
a
difference
in
mechanism
by
at
four
to
of
be
corresponding
The
two
response
C , again
show i n g
is
the
capture rate constant is not temperature
Here,
detachment.
independence,
two-step
differentiated
the
r e c ombination
of
the
negative ions with positive ions is occurring
than
an d
a molecular
a
molecule's
2 0 0°
response
Figures' 5-7.
an d
cannot
regions.
energy
temperatures.
and
three
peculiar
via
shown in
formation
high
the
distinct
potential
were
The
of
rather
temperatures
similar
15 0°
each
responding
whose
experiments.
from
for
has
involve
lo w
response
with
curves
ionization
increasing
forward
curve
IV,
dependence
dissociative
m echanisms
or
example of how
Figure 34 shows the response
isomer
a p pears
III
the
temperature
dependence
mechanism
by
of
an
a significant
and
which the analyte responds.
as
provided
the
After
a
response
to the alpha
brief
period
decreases,
detachment
due
molecular
more rapidly
of
temperature
to
detachment,
region of Figure 6.
115
ortho
„01X UIUi-ZI-W
0)
O
E
meta
t i i I I I I i I i I I I i I
para
Figure 34. ECD temperature dependence curves for 4.53 ng
£ - f luoronitrobenzene, 2.67 ng m - f luoronitrobenzene, and 3.76 ng £-fluoronitrobenzene.
116
Here the detachment process is occurring more rapidly than
both
the
recombination
reaction
because
kcal/mol)
is
the
of
the
begins
response
this
dissociating
A*.
of
This
the
e l ectron
shown
in
spectra
th a t
at
300°
C,
is
the
Figure
2
0
0
formed
NC
> 2
mass
°
C , .the
in the
ion
at
300°
at
C,
temperature.
ion
B ‘ or
ABB~
is
and
gamma region
occurs
more rapidly
-
positive
is supported
for
be
seen
by
from these
negative
abundance,
dominates
ion
o^-f Iuor o nitro­
molecular
greatest
(m/e = 46)
electron
with
scheme
It can
(23.5
to the
io n
spectra
35.
the
an d
dissociation
This mechanistic
benzene,
(m/e = 141)
A
negative
capture
affinity
negative
corresponds
the
molecular
recombination.
the
either
dissociation
Finally,
again
molecular
form
where
6
retain
to increase
the
to
to
the
electron
magnitudes.
dissociation
Figure
than
these
point
and
molecule's
insufficient
temperatures
At
reaction
the
ion
while at
spectrum.
Therefore, the terminal ion formed in the ECD dissociative
response
mechanism
is
probably
explanation
beginning
300° is that the molecule is now
via
mechanism
bimolecular
(Reaction
1
0
II
(Figures
dissociative
) which
does
molecular negative ion.
3
the
NO^
alternative
at
for
the
and
increase
4), which
electron
not
of
An
response
responding
is a single,
capture
involve the
ion.
react i o n
formation of a
117
10 0-1
m/e-
Figure 35. Electron capture mass spectra of 4.53
o-fluoronitrobenzene at 200° and 300°.
ng of
118
The
t e m p e rature
benzene,
shown
relatively
expect
that
curve of m-fluoronitro-
34, increases
with
continuously
temperature.
molecule
is
under g o i n g
capture
process,
as
this is favored
in
and
might
dissociative
by
higher
However, the electron capture mass spectra,
Figure
something
molecular
a
One
the
temperatures.
shown
Figure
consistently
that
electron
in
dependence
200° C and
else
is
(m/e=141)
300° C, lend
occurring.
At
evidence
200°
C
the
is the
only ion of consequence.
The 100° C temperature increase
does increase the degree
of
ion
36 at
fragmentation,
but
the molecular
by far the most abundant.
negative
ion
is still
What must be happening here is
that resonance electron capture (mechanism I) is occurring
throughout
forward
the
t e m p e ra t u r e
electron
independent
electron
range
investigated,
and
the
capture rate constant is not temperature
but
is
affinity
increasing
with
is sufficiently
temperature.
high
The
(27.5 kcal/mol) to
prevent detachment even at 350° C.
The
te m p e r a t u r e
fluoronitrobenzene
appearance
benzene,
of
this
with
a
dependence
is
curve
section
temperature,
a
region
responses
where
also
shown
is
in
similar
where
t empera t u r e
response
Figure
to
response
i n dependent
decrease with
curve
that
for £-
34.
The
of nitro­
increases with
region,
temperature.
and
a
The
119
m/e -
Figure 36. Electron capture mass spectra of 2.67
m-fluoronitrobenzene at 200° and 300°.
ng of
120
difference
between
nitrobenzene
commences
C ) for
temperature
and £-fluoronitrobenzene
at
a higher temperature
£ - f luoronitr o b e n z e ne
affinity
(24.3
explains
the
the
the
para
versus
temperature
and
therefore,
kcal/mol
k cal/mol
me t a
the
para
versus
due
is that
to
of
has
27.5 kcal/mol)
fairly
easy
to predict
nitrobenzene.
(m / e = 1 4 1 ) to
300°
C,
One
be
and
the most
this
The
is
Figure
37.
para
toward
fragmentation
in
the
and
lower affinity
detaches
expect
spectra of p-fluoro-
case,
isomer also has
than
does
the election
molecular
ion at both
the
(24.3
At this point it
the
abundant
fact
between
f l u o r o nitrobenzene,
the mass
would
This also
differences
more readily at the higher temperatures.
is
200°
its higher electron
22.1 kcal/mol).
isomers
of
detachment,
(300° C versus
dependence
isomer
dependence
as
anion
200° and
is shown
in
a greater tendency
the
meta
isomer,
evidenced by the greater abundance of the N O
2
as is
ion in the
mass spectrum of the para isomer.
Azulene,
which
has
an
electron
affinity
of
17.3
kcal/mol, is a molecule that should have data in the alpha
and
beta region,
Wentworth' and
according
Chen
(43)
to the
conclusions reached
on electron
capture mechanism
by
I.
The temperature dependence plot obtained by the author was
shown
previously in
Figure
29.
The response does indeed
1 2 1
m/e-
Figure 37. Electron capture mass spectra of 3.76 ng of
£-fluoronitrobenzene at 200° and 300°.
1 2 2
decrease
reaches
130° C z but
response
until
substantially
resonance
Again,
begins,
electron
temperature
the
detector
t emperature
no temperature independent
is observed.
detachment
once
the response
indicating
capture
independent
but
is
of
is increasing
that
rate
region
t h e 'forward
constant
is
increasing
not
with
temperature.
A molecule whose response increases with temperature,
but
for
Figure
a
38
different
is
the
temperature.
ECD
The
fIuoronitrobenzene,
reason,
response
curve
but
is
is
£ - b r omon i t r obenze ne.
curve
as a function
similar
to
a different mechanism
occurring. . Electron capture
mechanism
described
is
electron
The
in
Introduction,
attachment
electron
that
leading
capture mass
m-
is probably
II, previously
a single,
immediately
of
of
bi molecular
to dissociation.
spectra of £ - bromonitrobenzene,
shown in Figure 39, seem to suggest that this mechanism is
responsible
for
the
ECD
temperature
The
of
the
ion
C.
These ions are observed in a nearly 50:50 ratio due to
the
nearly
equal
observed
and
dependence.
Br-
only ions
response
(m/e = 79,81)
at
abundance
of
variations of this ratio appear
to
a
possible
detector
mass
are the two isotopes
both
79
Br
2 0 0°
and
in the mass
bias.
The
C
8
]
and
30 0°
Br.
Some-
spectrum
due
increase
in
ECD
123
500-1
400-
300-
200
Figure 38.
-
E C D temperature dependence response
128 pg of £-bromonitrobenzene.
curve of
124
Figure 39.
E l e c t r o n c a p t u r e m a s s s p e c t r a of 128
£-bromonitrobenzene at 200° C and 300° C.
pg
125
response
with
temperature
is
expected
because
the
dissociative process is favored by higher temperatures.
It appears
capture
least
that the forward resonance electron
rate constant
at
low
temperature
where
then,
the
is not
temperature
temperatures.
independent.
responses
It
may,
There
independent,
in
fact,
never
be
due,
of nitrobenzene
detachment,
however,
and
to
and
other molecules
a combination
recombination
be
are temperature ranges
are not affected by changes in detector temperature.
may
at
rate
of
This
attachment,
constant
magnitudes
that produce unchanging responses, and not due to a
constant attachment rate.
ECD Responses to Quinones
and Electron Affinity
Dependence of Responses
Three
responses
examined
high
39.5
determined
in the
ECD
are 1,4-benzoquinone
affinity
kcal/mol,
the
(MBQ)
and
of
these
dependence
and
of
the
TID-E^
whose
but were
(BQ), 2-methyl-
1,4-naphthoquinone '(NQ).
molecules
respectively
Wentworth
in
molecules
not
Introduction,
for
affinity
were
1,4-benzoquinone
electron
electron
(32).
Chen
ECD
is
As
(37)
was
41.7,
on
40.4 and
discussed
deduced
response
The
a
Equation
in
7
molecule's
126
electron
with
a
affinity.
high
response,
According
electron
and
to this equation,
affinity
therefore,
a
should
large
have
a molecule
a large
electron
ECD
capture rate
constant.
By
mu s t
examining
conclude
electron
that
the
these
these
functionalities
molecule.
ECD
A
via
molecule
p-d!nitrobenzene
(43.6
producing
the
similar
one
capture
there
ar e
readily
electron
however,
an
no
detach
affinities of
one would expect a
electron
electron
kcal/mol), responds
a measurement
mol at 200° C, as was
as
resonance
of
quinones,
would
that could
Because
response
processes.
ECD,
of these
molecules
molecules are so large,
large
the
structure
non-dissociatively,
electronegative
from
the
of 456
capture
affinity,
quite well in
KHz-min
X 10
shown in Table 3. Shown in Table 4
are the responses of the three quinones,
that the responses of BQ and
MBQ
and
one can see
are nearly insignificant
compared to that of £-dinitrobenzene.
The
lowest
reasonably
large response,
dinitrobenzene.
that
electron
a high
affinity
but
It is obvious
electron
affinity
quinone,
NQ,
gives
a
still far smaller than £from
this data
in Table
is not a guarantee
4
for a
127
large
ECD
magnitude
response.
is
In
reversed
fact,
compared
the
order
of
response
to the electron affinity
values.
Table 4.
ECJX responses (response per mole in KHz-min x
10 y) of 20 ng 1,4-benzoquinone (BQ), 32 ng 2m e t h y l - 1 , 4 - b e n z o q u i n o n e (MBQ) I ng 1,4naphthoquinone (NQ).
MBQ
BQ
200°C
300°C
200°C
300°C
.211
.344
.458
.729
There
where
are
several
cases
shown
NQ
200°C
30 0°C
75.2
116.6
previously in Table
3
strict dependence of response on molecular electron
affinity
do e s
not
exist.
Shown
in
Table
5 are
the
responses of several molecules that are assumed to undergo
resonance
electron
capture
because
to decrease at higher temperatures,
of
the
electron
from
the
their responses
indicating
molecular
negative
begin
detachment
ion.
The
responses listed were recorded at the temperature at which
Table 5.
Maximum responses of resonance electron capture molecules and their
calculated electron attachment rate constants.
Compound
Amt.
*
*** *
*****
-
8
2 0 . 0
170
69.3
3.13 x IO
-
8
21.4
2 1 0
53.1
2.40 x IO
-
8
170
64.3
2.90 x IO
"
8
17.3
130
34.3
1.55 x IO
"
8
41.7
300
1.5
x IO
"
1 0
4.03 x IO
"
8
o-nitrotoluene
I
m-nitrotoluene
I
p-nitrotoluene
I
azulene
5.83
. 1
2.05
.333
p-fluoronitrobenzene
3.76
24.3
240
89.1
m-chloronitrobenzene
1.77
28.1
300
285
1.29 x IO
"
7
o-dinitrobenzene
0.13
36.2
300
440
1.99 x IO
7
* Nanograms
** Electron affinity in Kcal/mmol from ref. 32.
*** Temperature at which response is.maximized
**** Molar response in KHz-min x 10
***** Resonance electron capture rate constant in cc mol
sec
"
128
1.99 x IO
2 2
2 0
Response
44.1
3.13
.4-benzoquinone
T(max)
2 0 0
nitrobenzene
1
***
* **
EA
129
the response is optimized.
molecular
electron
electron
affinities
and
capture rate constants.
calculated
The
Also listed in Table 5 are the
using
resonance
the calculated
The rate constants were
£ - f luoronitrobenzene
electron
forward
as. the reference.
attachment
rate
constant
of
£-fluoronitrobenzene was calculated from electron transfer
equilibria
electron
of
this
attachment
these
molecule
rate
calculated
electron
fr o m
of
affinities
this
figure
constant
rate
in
necessarily
curves
of
Significant
with
and
decreases
increasing
in
molecular
negative
electron
due
to
ion
the
(73).
whose
A plot
the molecular
40.
One
between
can
see
electron
capture rate coefficient
The
temperature
NQ
are
ECD
response
temperatures,
(32),
4
versus
a correlation
observed.
BQ, MBQ
7
Figure
affinity and the forward electron
is not
c F ]_
is known
constants
is shown
that
with
shown
is formed,
molecule's
it does
high
in .Figure
are
indicating
dependence
41.
not observed
that once
not
detach
electron
the
the
affinity.
This would seem to indicate that the ECD response is small
due
to
a
slow
electron
capture
rate
constant
and
not
because of detachment.
The electron capture mass spectra of BQ, MBQ and NQ,
shown
in
Figure
42,
lend
further
evidence
that
the
130
10
20
30
40
E A (K c a I/m o l )
50
Figure 40. Plot of calculated resonance electron capture
rate c o n s tants v e r s u s molecular electron
affinity.
Nitrobenzene
= I; o , m , £ nitrotoluene
= 4, 5,
, respectively;
£ - f l u o r o n i t r o b e n z e n e = 9; m - c h l o r o n i t r o benzene = 11; o-dinitrobenzene = 19; azulene =
37; 1,4-benzoquinone = 38.
6
131
MBQ
Figure 41.
ECD temperature dependence response curves for
20 ng BQ, 32 ng M BQ, and I ng NQ.
132
electron
attachment
electron
affinity
rate constants
molecules
BQ
are small for the high
and
M BQ.
For a molecule
undergoing resonance electron attachment, one would expect
the
molecular
The
only
molecular
which,
ion
one. of
negative
to be the
the
three
In
quinones
fact,
also
are much
greater
in
more
the
only
complex.
mass
(m/e = 108,122)
are
that
generated
th a n
seen
in
for these two molecules.
chemistry is occurring
(m/e=158),
largest
ECD
in the
ions, most of them
respective
the
the.
The spectra of BQ and
Numerous
the
produces
is NQ
the
spectrum.
ion of significance
spectrum of NQ was the base peak.
MBQ
in the
ion as the base peak
incidentally,
response.
base peak
negative
molecular
ion mass
ions
spectra
It is obvious that some complex
within the ion
source, possibly on
the source walls, when these two compounds enter the high
temperature
the
electron
atmosphere of an
attachment
rate
ECMS
constant
molecular negative ion from
BQ and
small,
electron
large.
even
thou g h
for formation
rate,
of a
MBQ must be relatively
affinities
are quite
The order of molecular ion abundance for the three
quinones is the
in
their
source. . At any
reverse
same as the
order
of
the
ECD response magnitudes, and
molecular
electron
affinities.
133
i l<i
150
m/e->
Figure 42. Electron capture mass spectra of 10 ng each BQ,
MBQf and NQ at 200° C.
134
From
this
affinity
data,
is
Another
not
a
molecule
kcal/mol)
constant
systems
during
it is obvious
(31,
guarantee
that
32)
(71)
is
has
but a
NC^.
of
a
a
just
a high
large
ECD
high
BQ
undergo
ionization
electron
response.
electron affinity
small electron
Possibly
like N O 2, or may
negative
that
and
capture
MBQ
to
rate
are rigid
large geometry
leading
(53
changes
a small electron
capture rate constant and hence small ECD response.
Comparison of the
Detection Techniques
At
quite
first
glance,
similar
in
electronegative
signal
in
by
different,
employing
this
compounds
the
ECD
a heated
study
was
to
and
both
point
is
that
the
are
ECD
are
produced,
using
solid
beta
surface.
find out
it
is
appear
selective
to
both generate an analytical
if
non-radioactive substitute for the
this
and
2
gas phase negative ions.
ions
the
TID-N
that t h e y
producing
which
the
not,
The manner
however,
radiation,
the
One of the
the
TID-N
BCD.
but
is
quite
TID-N
objects of
could
2
2
be
a
The conclusion at
can
be
applied
more
successfully than the ECD in certain situations.
For
the
substituted
molecules
discu s s e d
nitrobenzenes,
the
ECD
here,
is
most
the
of
the m
superior
GC
135
detection device.
in
the
the
ECD
for almost all the molecules
dinitro
from
to compound
device,
less
species
Response
parameters
selective
and
compounds.
compound
physical
The absolute detection limits are lower
well
are
samples,
each.
often-times
in
a
the
and
TID-N
of
with the
is
2
a more
well to dinitro-organics
compounds.
contaminants
selective
complex
a few
varies widely
detector
The
very
halogenated
for
linearity
in each
responding
to
and
compounds
of
except
in
Halogenated
environmental
detection
matrix,
the
TID-N
of
nitro
2
would
the
E CD,
be
superior to the ECD.
The
that
T I D-N
is
has
2
it
is
another advantage
less
susceptible
contaminants, especially oxygen.
that
the
the
TID-N
detector
s e rious
was
2
in
removal
of
this
gas
amounts
of
oxygen
to
carrier
ECD-GC
from
the
and
gas
The data in Table 2 show
not drastically affected
gas atmosphere.
problem
over
Oxygen
contamination
analyses
system
in the TID - N 2-GC
by varying
(35,
63,
is a
64)
is critical.
and
Trace
system
is a trivial
in this study
that the use
problem at worst.
It has also been
of
chlorinated
noticed
solvents
is quite
feasible
analyses while nearly impossible for
turning
off
the
source
heating
during
TID - N
ECD experiments.
current
of
2
By
a TID-N
2
136
during
and
elution of the solvent, analyte signal interference
baseline
responds
CHCl^,
so
for
unacceptable
disturbance
strongly
example)
for
long
are
to
avoided.
these
that
E CD, however,
solvents
the
periods
The
(CCl^
baseline
following
may
and
be
injections,
preventing accurate analysis of the analyte signal.
137
CONCLUSION
Both
detectors
have
proven
to be
successful
for the analyses of substituted nitrobenzenes.
the
more
T ID - N
universal
is
2
detector
very
of
the
selective
displays
a
two
strict
The
detectors,
to certain
devices
while
molecules.
respo n s e
ECD is
the
Neither
dependence
on
the
molecular electron affinity.
Little
generated
TID-N^
was
in
known
the
about
TID-N
candidate
at the
2
r e spo n s e
a
mechanisms
varied
simple
source
to the analyte to more, complex
species.
a negative
Two
ion was
onset of the study.
These
inorganic
from
the way
electron
response
were
Six
considered.
transfer
from
the
reactions involving
mechanisms,
both
involving thermal decomposition of the analyte followed by
association
best
with
emitted
inorganic
explanations of the results.
the
two
species
when
mechanisms
occurs
anomalous
high
are where
(gas phase
ionization
inorganic
of
an
occurs
species).
responses
mass
of the
ions
in
plugging of the APIMS
extensions
of
the
These
gas-solid
or after
AP IMS,
and
the
between
inorganic
interface)
and
association with
mechanisms
aperture.
second
differences
nitrotoluenes,
the
provide
association with
versus
(before
The
species
explain
th e
the
observation
the
consistent
These two mechanisms are
proposed
mechanism,
thermal
138
decomposition
attachment,
that
of
the
analyte
followed
by
electron
and are consistent with the results supporting
mechanism.
The
details
of
precisely
what
is
occurring at the source surface when an analyte enters the
detector
cell
are
still
not well understood.
Perhaps
a
better explanation of the response dependence on the bias
voltage
would
shed
light
on
the
occurring at the source surface.
that increases
ionization
T ID - N
in the field
or the
emission
definitely
2
molecule's electron
decomposition
strength
not
affinity,
product,
either assist in the
The
depend
but
such
processes
It appears at this point
of ions.
does
fundamental
response of the
upon
a high
the
analyte
electron affinity
as N O 2, most
likely assists
in the negative ionization process.
It
the
was
ECD
either
correlating
the
previously
the
electron
response
known
that molecules
dissociatively
ECD
response
capture
mechanisms
mass
or non-dissociatively.
temperature
the working
range of the detector were elucidated.
not
have
attachment
a
temperatures.
independent
constant,
The
By
compounds,
temperature
It was found that
via resonance electron capture do
temperature
rate
in
dependence with
spectra of certain
throughout
molecules which respond
responded
rate
at
least
constant
may
forward
at
low
never
be
electron
detector
constant
139
over
the
the
working
right
temperature
combination
of
range of the
detachment,
detector,
and
recombination,
and
attachment rate constant magnitudes may be responsible for
temperature
of
independent
molecules
kcal/mol
original
observed
can
with
electron
indeed
molecules
before
response
curves.
affinities
Molecular
higher
than
ions
18.5
detach their electrons to re-form the
and electrons.
due
to
the
This may
low
not
have
temperature
been
limit of a
tritium source BCD.
The
response
dependence
quinones
ability
the
on electron
clearly
to
form
strongly
with
electron
affinity
capture
of
rate
the
BCD
affinity,
indicated.
a
does
these
negative
magnitude of the
does
not
constant.
bear
a
strict
as the responses of the
For
molecular
not
BCD
guarantee
Possibly
a
molecules,
ion
the
correlated
response.
large
High
electron
molecular rigidity
and
geometry changes during ionization may deter the formation
of a molecular negative ion.
Computer
modeling
of
reso n a n c e
electron
capture
processes showed that the forward attachment rate constant
determines
the
temperatures
res p o n s e
magnitude
while
th e
levels at high
correlated
well with
of
BCD
electron
affinity
temperatures.
simulated
data
response
at
determines
Experimental
for ions
lo w
that
data
detach
140
the
electron
predicted
at
that
higher
detector
temperatures.
Modeling
ions of molecules with electron
affinities
higher than 30 kcal/mol should never detach throughout the
working temperature range of an E CD.
While both
negative
the
ions,
detectors function
they
substituted
do
not
generate
n i t r o b e n zenes.
selective to dinitro compounds,
universal device.
absolute
certain
by p r o d u c i n g .gas phase
The
detection
situations,
similar responses
The
while the
TI
is
to
very
ECD is the more
ECD is the superior device for the
of the
the
substituted
T ID - ^
nitrobenzenes.
is superior
but it is not a replacement for the E CD. .
to the
In
E CD,
141
LITERATURE CITED
1.
A. J. James and A. J. P . Martin, Analyst,
(1952).
2.
A. Karmen, A n a l . Ch e m . , 36 , 1461 (1964),.
3.
A. Karmen and L . Guiffrida, Nature (London), 201,
1204 (1964).
4.
L . Guiffrida, J . Ass. Offic. A q r . Chem., 47, 293
(1964).
5.
G . R . Verga and F . Po y , J . Chromatoqr., 116, 17
(1976) .
6
.
V. V. Brazhnikov and E . B . Schmidel,
122, 527 (1976).
77, 915
Chromatogr.,
7.
B . Kolb, M . Auer and P . Pospisil, J . Chromatogr.
Sci., 15, 53 (1977) .
8.
P . L . Patterson, J . Chromatogr., 167, 381 (1978).
9.
P . L . Patterson and R . L . Howe,
1_6, 275 (1978).
10.
G . R . Verga,
11.
D . R . Coahran, Bull. Envir. Contam. Toxicol., I, 141
(1966) .
12.
E . Cremer, H . Moesta and K . Ha'blik, Chem. Ing . Tech. ,
38_, 580 (1966).
13.
P . L. Patterson, Chromatographia, 16 , 107 (1982).
14.
B . Kolb and J. Bischoff,
(1974) .
15.
Roy Greenhalgh, Jjl Chrpmatogr. Sci .,. 16 , 9 (1978).
16.
C . H . Burgett, D . H. Smith and H . B . Bente, J .
Chromatogr., 134, 57 (1977).
17.
P . L . Patterson, U.S. Patent 4,524,047.
Chromatogr. Sci.,
Chromatogr., 279, 657 (1983).
Chromatogr. S c i ., 12, 625
(1985).
142
18.
P. L . Patterson, U. S. Patent 4,203,726.
19.
B . J. Erlich, Ind. Re s . Develop., 107, Apr,
20.
I . B . Rubin and C. K . Bayne, Anal. Chem., 51,
(1979).
21 .
M . Thewlis (Editor), Encyclopaedic Dictionary of
Physics, V o l . 7, Pergamon Press, New York, 1962,
p. 252.
22.
V. V. Brazhnikov, M. V. G u r 1ev and K . I . Sakodynsky,
Chromatogr. R e v .., 12 , I (1970).
23.
D . J . David, Gas Chromatographic Detectors, Wiley,
New York, 1974, C h . 5.
24.
J . Sevcik, Chromatographia
25.
W. C. Michaels (Editor), The International Dictionary
of Physics and Electronics, Van Nostrand, Princeton,
New Jersey, 2nd e d ., 1961.
26.
P . L . Patterson, R . A. Gatten and C . Ontiveros, J .
Chromatogr. Sci., 2 0 , 97 (1982).
27.
P . L . Patterson,
28.
P . L . Patterson, J^_ Chromatogr-. Sci., 24 , 41 (1986 ).
29.
C . M . White, A. Robbat and M . H . Randall, Ana l .
Chem. , 5_6 , 232 (1984) .
6
(1980).
(1980)*
541
_, 139 (1973).
Chromatogr. Sci., 24 , 466 (1986).
30.
C . S . Jones and E . P . Grimsrud, J_^ Chromatogr . , 387 ,
■ 171 (1987).
31.
E . Herbst, T . A. Patterson and W. C . Lineberger, J .
Chem. Phys., 61, 1300 (1974).
32.
E . P . Grimsrud, G . Caldwell, S . Chowdhury and P .
Kabarle,
Am. Chem. S oc. , 107, 4627 (1985).
33.
J . E . Lovelock, Institute Petroleum, GaS
Chromatography Discussion Group, Oxford, England,
May, 1957.
34.
E . S . Goodwin, R . Goulden and J . G . Reynolds,
Analyst,
(1961) 697.
8 6
143
35.
J . E . Lovelock, An a l . Che m ., 35 , 474 (1963).
36.
W. E . Wentworth, E . Chen and J. E . Lovelock, J jl Phys.
Chem., 70_, 445 (1966) .
37.
W. E . Wentworth and E . J. Chen, ^
5, 170 (1967).
38.
P . G . Simmonds, D . C . Fennimore, B . C. Pettit, J . E .
Lovelock and A. Zlatkis, Anal. Chem., 39, 1428
(1967) .
39.
R . J. Maggs , P . L . Johnes, A. J . Davies and J . E .
Lovelock, Anal . Chem., 4 3 , 1966 (1971).
40.
P . L . Patterson, J_j_ Chromatogr,, 134, 25 (1977).
41.
R . S . Becker and W. E . Wentworth,
84:, 2210 (1963).
42.
W. E . Wentworth and R . S . Becker, J_j_ Am. Chem . Soc.,
Gas Chromatogr.,
Am. Chem . Soc ,,
84, 4263 (1962) .
43.
W. E . Wentworth and E . C. M . Chen, in A. Zlatkis and
C . F . Poole (Editors), Electron Capture Theory and
Practice in Chromatography, Elsevier Scientific
Publishing, New York, 1981, C h . 3.
44.
R . S . Becker and E . C. M . Chen, J . Che m . Phys., 45,
2403 (1966).
LO
E . P . Grimsrud and S . W. Warden, Ana l . Che m ., 52,
1842 (1980).
46.
E . P . Grimsrud, Anal. Chem., 50, 382 (1978).
47.
E . P . Grimsrud, S . Chowdhury and P . Kebarle, Int. J .
of Mass Spec, and Ion Proc.,
, 57 (1986) .
6
8
C O
K . Levson and H . Schwarz, Agnew. Chem. Int. Ed.
Engl., 15, 509 (1976).
49.
F . R . McDonald and U . Christian, A m . L a b ., 97, No v .,
(1976) .
50.
T . Fuiii and H . Arimoto, J. Chromatogr., 355, 375
(1986) .
144
51.
J . A. Lubkowitz, J . L . Glajch, B . P . Semonian and L .
B . Rogers, Jjl Chromatogr. , 133, 37 (1977).
52.
F . L. Bayer, J_j_ Chromatogr. Sc i ., 1 5 , 581 (1977).
53.
D . F . McMillen and D. M . Golden, An n u . R e v . Phys.
Chem., 33, 493 (1982).
54.
Handbook of Chemistry and Physics, CRC Press, Boca
Raton, Florida, 62nd edition, 1981, p. E-2.
55.
W. Lindinger and D . L . Albritton,
3517 (1975).
56.
N . I . Shteinboc, The Application of Radioactive
Radiation in Measurement Techniques, Moscow, USSR,
1960, p. 24.
57.
D . Bombick, J . D . Pinkston and J . Allison, Anal.
Chem. , 56_, 396 (1984) .
58.
J . J . Sullivan and C . A. Burgett, Chromatographia,
176 (1975).
59.
J . E . Lovelock and A. J . Watson,
123 (1978) .
60.
E . P . Grimsrud and W. B . Knighton, Ana l . Ch e m ., 54
565 (1982).
61.
W. B . Knighton and E . P . Grimsrud, A n a l . Chem,, 5 5 ,
713 (1983) .
62.
L . G . Christophorou, A d v . Electron. Electron Phys.,
46, 55 (1978) .
63.
G . G . Guilbault and C. Herrin, Anal. C h e m . Acta, 3 6 ,
255 (1966).
64.
F . W. Karasek and D . M . Kane, Anal. Che m ., 45,
(1973).
Chem. Phys., 62 ,
8
_,
Chromatogr., 158,
576
IO
LO
E . P . Grimsrud, S . Chowdhury and P . Kebarle, J^ Chem
Phys., 83, 3983 (1985).
145
6
6
.
67.
6
8
.
F . Bruner, in A. Zlatkis and C . F . Poole (Editors),
Electron, Capture Theory and Practice in
Chromatography, Elsevier Scientific Publishing, New
York, 1981, C h . 11.
J . E . Lovelock, Nature (London), 230, 379 (1971).
P . G . Simmonds, in A. Zlatkis and C. F . Poole
(Editors), Electron Capture Theory and Practice in
Chromatography, Elsevier Scientific Publishing, New
York, 1981, C h . 11.
69.
A. Zlatkis and C . F . Poole (Editors), Electron
Capture Theory and Practice in, Chromatography, New
York, 1981, p . XI.
70.
J. Vessman, in A. Zlatkis and C. F . Poole (Editors),
Electron Capture Theory and Practice in
Chromatography. Elsevier Scientific Publishing, New
York, 1981, C h . 13.
71.
H . Shimamori and H . Hott,
(1986).
72.
E . W. McDaniel and E . A. Mason, The Mobility and
Diffusion of Ions in Gases, Wiley, New Yor k , 1973,
p. 146.
73.
E . Alge, N . G . Adams and D . Smith, J. Phys., 17, 3827
(1984) .
Chem. Phys., 8 5 , 4480
MONTANA STATE UNIVERSITY LIBRARIES
1762 1011 3470 6