I I . I w 1, I _ I I I I 5.9.1.5 .
. . . I I . I . I. I . I III-
. I. .. I . I . . _ I
I . . . I I
.
.I . .
I .., 4 I 5 , I
. . . ._ . I . I I I
I . . I I I
_ I. I I. .... . .
.I I 3 I
. I . I I I . I ... . I
l . I I I. .I
I . . . . . I I
I. . . . . I .. . . I . . . .
I I . . . I I I .
_ I . .I _ .. . I
I I I I .I . . . . .
I I; I A . «I I . I . I .
I . I .I r, . . . .
I V o . v v .I .
I I I . t. . .. ... ...I I
I I .I I . . . . . I. I . . . . . .. .
.I .. .. . ..r . . I I I
. . . a. . I I. I . . . .I .
. _. I . I r: .I I. ..
I . I : I . . I I I I .
v . . .,I l . . I I, . II
vl . I I I I I I . I 7
. . I I I v . I .
.1 . I II I . I e . .II
. . .
I . I I . . . >. .I .. I
x L r I I I
I. . ( .I .. I I
.I . I. I I I. I
. I I I . . I .. .
.I . . . . . I I. I . I I . I I
. . I I . .. . I . . I .it.v
7 I . I ,II .I .I I . . I .I I . . ...-...].
I . . . I I ... I I I
I . . .. I ...
. I. . I I . I
I
I . - d .
. ., . I .
. . _ . l . I I .
I I .
. I I.
. . I . . I. f . .
. . . I . . .I . .
. . . I
I _ w I .5 .
v I . 0 .
I. I .
I g I I
. . I . I . . .
. n n .I. . . . II . .
. . I I. .
_ I . I
.I . I I .
I . .I . I . I.
. . an . .
_ I. . .
I . . I . I .. . .
. . . I .
o «I. . I. .
I I . . . ..I. I I
. . I .
. I .
I I . . .
.. . I .
. . I.
. I ..
. . I I
I I I A, . 9
_ I. I
_ I . I I I.
. II . I . I .
. . . . .
. I . I .
. I . I I . I I. . . .I
I . I. .
A II I I I r
A . I. I . I I .I .
I . .... I I l
_ I I. I I I I .. I
I .I r . I I .I I
. I . I . I I
. . . . . .
I . I .
I . I
. I . . I . .
I I II . .I . I I I. . . I I I ..
I . I . .. .I .. .I . . I I I
I .. I . . J . I I.
. I I. I .
I . I I . )' I‘. c
. . I 1...!!!
W I. .I I . .I l...
v I I 3 . G
r I I I I. . I I
. I I I . I . .I‘
32!: £7.11]
IIIIL...I.Ic!II.
..u. I ...I .|.I ..r...t.I,..} I. up? w.. ...I.. ...I..I,.Ia I ...7..I..
. .I . . I A a. “IIIII

 

I a ...mafimmfimggfiméflwfii ..

I ..I II! I .-.,

 

231%(9 5660

H NSTATE UNIVERTYV LIBRARIES ‘H

HHHHHHHH HHHHHH |HIIIHHH

Michigan State 3 1293 00592 13

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

The Characterization of Gas Chromatography-
Mass Spectrometry Using Ion Flight Time
and Time -Array Detection

presented by

Eric Douglas Erickson
has been accepted towards fulfillment

of the requirements for

Ph . D . degree in Chemistry

W222

Major professor

 

Date JUIY 28, 1989 /

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

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

419” .7 ”ml H

 

 

H
.1 I -‘.

2 ..g. “L. .4; ~
| ' . H
...L \ ~ J , ‘ H

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

|J__________L_____H

H H

MSU Is An Affirmative Action/Equal Opportunity Inuitution

 

 

 

THE CHARACTERIZATION OF GAS CHROMATOGRAPHY -
MASS SPECTROMETRY USING ION FLIGHT TIME
AND TIME-ARRAY DETECTION

By

Eric Douglas Erickson

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

Department of Chemistry

1989

@030b34

ABSTRACT

THE CHARACTERIZATION OF GAS CHROMATOGRAPHY-
MASS SPECTRONIETRY USING ION FLIGHT TIME
AND TIME-ARRAY DETECTION

By
Eric Douglas Erickson

Sample concentrations in capillary gas chromatographic effluent
change too rapidly for scanning mass spectrometers to keep pace, which
prevents the acquisition of full mass spectral data on each eluting
component. A time-array detection scheme has been developed using time-of-
flight (TOF) mass spectrometry and an integrating transient recorder (ITR)
to increase the sampling frequency of mass spectrometry and overcome this
incompatibility. This work describes the application of time-array detection
to gas chromatographic effluent, using a conventional, gas-phase TOF mass

spectrometer.

Time—array detection involves the monitoring of all ion signals
produced by each ion source extraction pulse. Conventional, gas-phase TOF
instruments do not provide mass-independent temporal focus, thus limiting

the utith of time-array detection. The severity of this mass-dependence has

been examined by means of a computer simulatiogrll‘f‘nfawfi WE

resolution in TOF mass spectrometry. Windows of usable m/z values have
been defined as those with tolerable levels of resolution and intensity
degradation. The effect of ion focus parameters on window size and position
has been determined. As few as three windows are shown to adequately
cover the mass range from 50 to 700 Daltons with a maximum of 10%
distortion in relative peak intensity.

Time-array detection schemes were shown to provide several
advantages over scanning techniques. The high sampling frequencies
possible with the ITR provided the capacity to accurately reproduce a
chromatographic profile and avoided problems . of mass spectral skew due to
changing source concentrations. In addition, enhanced S/N ratios were
observed and it was demonstrated that the chromatography could be
optimized for speed of analysis. For example, a gasoline analysis can be

completed in two minutes.

The high mass spectral scan file generation frequencies that are
possible with time-array detection could result in an information overload. A
series of algorithms was examined in which the degree of fragmentation of a
molecule could be used as the basis for selecting data for the reconstruction of
a chromatogram in order to minimize the number of spectra that need to be
interpreted. These algorithms were successfully used to discriminate against
aliphatic components in the chromatographic efiluent and emphasize the

aromatic species.

In loving memory of my father,

Commander Douglas Leon Erickson.

iv

Acmowmenms

While the work described in this document is the result of my efl’orts,
no major undertaking can occur without the contributions of many others,
and this is no exception. I would like to take this opportunity to single out
the efforts of other investigators who provided significant contributions and
thank them for their tolerance of my idiosyncrasies.

Consultation and moral support have come from everyone that I have
encountered in the Chemistry Department at Michigan State University. I
would be remiss not to recognize the companionship that I have received fi'om
members of the Enke, Watson, and McGufin groups over the last five years.

Bruce Newcome designed the electronics for the integrating transient
recorder and provided assistance in the early days of interfacing this
instrument to the time-of-flight instruments. Mike Davenport was always
eager to assist me in keeping electronics on both the ITR and the TOFMS
operating. Programs on the ITR were written and revised to my demanding
specifications first by Russell Rogers and later by Kevin McNitt. These
individuals were usually available at all hours to assist in software
modifications. Gary Schultz, Mel Micke, Ron Tecklenberg, and Ron Lopshire
were always available to serve as an extra pair of eyes or hands, or to act as a
critical audience for my many hairbrained schemes. Linda Doherty, Ellen
Yurek, Kathleen Kayganich, and Chris Evans were always willing to impart

V

vi
some of their chromatographic wisdom on me and occasionally even supply

me with needed components.

Without the assistance of George Yefchak, the instrument simulation
probably would never have reached fi'uition. Earlier versions of this program
were loosely based on one of George’s simulations. On later versions, George
provided bountiful assistance in program debugging. Thanks to his excellent
tutelage and patience, I learned all that I know about programming in
FORTRAN and C. Consultations with Stanley Crouch and Tom Atkinson
were also very useful in debugging the program.

Funding for this work has come from several sources. Development of
the integrating transient recorder and five terms of research support for my
efforts was provided through a Biomedical Research Technology Program
grant (No. DRE-00480) from the National Institutes of Health. Funding for
the MicroVAX computers used for simulations and development of the
"degree-of-fragmentation” algorithms was provided in part by a grant from
the Office of Naval Research (contract No. N00014-81-K-0834) under the
auspices of John Michalski. In addition, I would like to acknowledge receipt
of a long term training fellowship fi'om the US. Naval Weapons Center,
China Lake, California. This fellowship provided support for the first two
years of my stay at Michigan State University. In the mean time, the Naval
Weapons Center has had to survive without their mass spectrometrist for the

last five years.

Finally, I would like to take this opportunity to thank my loving wife,
Barbara, and our children, Ivy and Benjamin. Their support and
encouragement has made it possible to survive even the toughest of times in

graduate school. It has not been easy to return to a life of poverty, but they

vii
have made the best of it. I regret that my research took so much of my time

and I hope that I will soon be able to make it up to them.

TABLE or CONTENTS

Table of contents ..................................... viii
List of figures ........................................ xi
List of Tables ........................................ xvi
Chapter 1: Overview and Perspective ......................... 1
A. Introduction ................................ 1
B. Detection of Gas Chromatographic Eflluent ........... 2
C. Chromatographic Requirements on Spectral
Detectors ................................. 4
D. Scanning Mass Spectrometric Detectors ............. 8
E. Array Mass Spectrometric Detectors ............... 9
F. Limitations of Time-of-Flight Mass
Spectrometry with Time-Array Detection .......... 11
G. Chapter 1 References ......................... 16
Chapter 2: Mass Dependence of Time-Lag Focusing in
Time-of-Flight Mass Spectrometry ........................ 19
A. Introduction ............................... 19
B. Ion Focusing in Time-of-Flight
Mass Spectrometry ......................... 19
B. Theoretical ................................ 23
1. Flight-Time Calculation .................... 26
2. Resolution Calculation ..................... 29
3. Intensity Calculation ...................... 31
4. Simulation ............................. 32

viii

ix

C. Results and Discussion ........................ 34
1. Efl'ect of Time-Lag ........................ 34
2. Effect of Extraction Grid Potentials
on Ion Focus .......................... 44
3. Effect of Ramp Time ...................... 52
D. Conclusions ............................... 54
E. Chapter 2 References ......................... 57
Chapter 3. Application of Time-Array Detection
to Capillary Column GC-TOFMS ......................... 58
A. Introduction ............................... 58
B. Experimental .............. ‘ ................ 58
1. Gas Chromatography ...................... 60
2. Mass Spectrometry ....................... 60
3. Integrating Transient Recorder ............... 64
C. Results and Discussion ........................ 65
1. Mass Spectral Representation ................ 65
2. Mass Spectral Calibration ................... 73
3. Representation of the Chromatography .......... 7 5
4. Speed of Analysis ........................ 82
D. Chapter 3 References ......................... 86
Chapter 4: Reconstructed Chromatograms Based
on Mass Spectral Degree of Fragmentation .................. 87
A. Introduction ............................... 87
B. Experimental .............................. 90
C. "Generic Sigma" Algorithm ..................... 95
D. "Number of Peaks" Algorithm .................. 101
E. Combined Algorithm ........................ 108
F. Conclusions ............................... 110
G. Chapter 4 References ........................ 115

x
Apfiendix I: Programs Used to Simulate the CVC 2000

strument ....................................... 116
Apfiendix II: Programs Used to Calibrate the CV C 2000
strument ....................................... 138

Appendix III: Programs Used for the Degree of
Fragmentation Algorithm ............................. 148

Appendix IV: Degree of Fragmentation Database ............... 153

 

her or FIGURES

Eigu_re Base

1.1. The effect of samp ° rate on the reconstruction
of chromatograp 'c profiles. (a) A simulated
chromatogram. (b) Reconstructed chromatogram
from three points across the peak. (c) Reconstructed
chromatogram from three points across the peak, not
synchronized to the elution profile.
(d) Reconstructed chromatogram from five points
across the peak ................................. 5

1.2. The effect of sampling rate on mass spectral
quality. (a) A h othetical mass s ctrum
collected underi eal conditions. ) The
concentration gradient across which spectra
are obtained. (c) The mass spectrum obtained
from the rising edge of the concentration profile.
(d) The mass spectrum obtained at the top of the
concentration profile. (e) The mass spectrum
obtafriilned on the falling edge of the concentration 7
pro e ........................................

2.1. The mass-dependence of time-lag focusing in time-
of-flight mass spectrometry. Shown are peaks at
m/z 502 and 503 from the s ectrum of
erfluorotributylamine wi the value of time
ag optimized for (a) m/z 502 and (b) m/z 28 ............. 21

2.2. A schematic diagram of the CV C 2000 Time-of-Flight
Mass Spectrometer. Typical voltages and
distances are included in Table 2.1 ................... 24

2.3. The relationship between initial velocity and
ion flight time for m/z 500 ......................... 35

 

xii

2.4. Simulated peak shapes for m/z 500 at different
values of time lag ............................... 36

2.5. The dependence of peak height on time lag
and mass ..................................... 38

2.6. The calculated effect of time lag on resolution ............... 40

2.7. The effect of time lag on resolution.
Calculated values for m/z 500 are plotted as a
continuous curve. Points indicate experimental
values for the m/z 502 ran from
perfluorotributylamine ........................... 41

2.8. Working curve for acce table signals. Ions to
the left of the reso ution curve give unit-mass-
resolution with less than a 10% valley. Ions
between the intensit curves lose no more
10% of their peak height. The curve in the
center gives the optimum time lag for each
m/z value ..................................... 43

2.9. Resolution as a function of the ion focus
potential ..................................... 45

2.10. Interdependence between resolution, time lag,
and ion focus .................................. 47

2.11. Working curves for ion focus values of .
(a) 75 V. (b) 77 V, (c) 80 V, and (d) 85 V.
The V2 voltage used was 215 V ...................... 48

2.12. Resolution as a function of time lag and
V2 potential ................................... 50

2.13. Working curves forV values of (a) 150 V,
(mag 200 v, (c) 225 vi! and (d) 250 V. The 1011
focus voltage used was 80 V. ....................... 51

2.14. Effect of ramp time on optimum time lag .................. 53

.0.

2.15. TOF spectra of perfluorokerosene collected
using time-array detection with (a) time lag
focused for m/z 85 and (b) three mass
windows centered around m/z values of 70,
175, and 320 .................................. 55

3.1. Schematic diagram of the time-array

detection system ................................ 59
3.2. Calibration curve for the time lag pot ..................... 61
3.3. Calibration curve for the ion focus pot ..................... 6 2

3.4. The mass spectrum of n-decane collected at
(a) the top and (b) on the shoulder of the
chromatographic elution profile. Spectra were
collected by summing 1000 transients ................. 67

3.5. The mass spectra of n-decane collected by
summing (a) 1. (b) 10, and (c) 10,000 transient
spectra. This corresponds to scan file
generation frequencres of 10000, 1000, and
1 Hz, respectively ............................... 70

3.6. The relationshi between the signal-to-noise
ratio and e number of transients summed ............. 72

3.7. An FID chromatogram of charcoal lighter fluid
injected onto a 22 m by 0.25 mm fused silica
capillary column coated with 0.25 um SE-54.
The GC was temperature programmed from 100°C
to 150°C at 10°C min ............................. 77

3.8. Reconstructed chromatograms of lighter fluid
from scan file generation fre uencies of
(a) 1 Hz and (b) 5 Hz. The was temperature
programmed from 100°C to 150°C at 10°C/min ........... 78

3.9. Reconstructed chromatogram of 11 ng injection
of toluene collected at a scan file generation
frequency of 5 Hz ............................... 81

xiv
3.10. Reconstructed chromatogram of the toluene
peak in unleaded gasolene. Spectra were

collected at 20 Hz. The collection of each
spectrum is represented by the vertical bars ............. 85

4.1. Mass spectra used in the data set for
testing degree of fragmentation algorithms.
S ctra include (a) acrylic acid, (b) benzene,
(c anthracene, (d) octylmercaptan,
and (e) air .................................... 92

4.2. Total ion intensity reconstructed chromatogram
of the test data set. Chromatographic peaks
are (A) acrylic acid, (B) benzene,
(C) anthracene, and (D) octylmercaptan ................ 93

4.3. DIFFerence reconstructed chromatogram of the
test data set. Chromatographic peaks are the
same as listed in Figure 4.2 ........................ 94

4.4. Reconstructed mass chromatogram of the test
data set for m/z 77. Centroids of the
chromatographic peaks are labeled as in
Figure 4.2 .................................... 96

4.5. Generic sigma reconstructed chromatogram of
the test data set. Centroids of the
chromatographic peaks are labeled as in
Figure 4.2 .................................... 99

4.6. Total ion intensity reconstructed chromatogram
of the test data set in which only compounds in
which the generic sigma value is between 30 and
60 percent are plotted. Labels marking the
centroids for chromatographic eaks are for
(A) a lie acid, (B) benzene, ( ) anthracene,
and (D octylmercaptan .......................... 100

4.7. Number of peaks reconstructed chromato am with
a 5% of base peak intensity threshol . Labels
marki the centroids for chromatographic peaks
are for A) acrylic acid, (B) benzene,
(C) anthracene, and (D) octylmercaptan ............... 103

XV

4.8. Number of peaks reconstructed chromato with
a 5% of base peak intensity threshol . Labels
marking the centroids for chromatographic peaks
are for (A) acrylic acid, (B) benzene,
(C) anthracene, and (D) octylmercaptan ............... 104

4.9. Total ion intensity reconstructed chromatogram of
the test data set in which only compounds in
which the number of peaks (5% threshold) is
between 3 and 15 percent are plotted. Labels
marking the centroids for chromatographic peaks
are for (A) acrylic acid, (B) benzene,
(C) anthracene, and (D) octylmercaptan ............... 106

4.10. Total ion intensity reconstructed chromatogram
of the test data set in which only compounds
in which the number of peaks (5% threshold)
is between 2 and 10 ercent are plotted. Labels
marking the centroi s for chromatographic peaks
are for (A) acrylic acid, (B) benzene,
(C) anthracene, and (D) octylmercaptan ............... 107

4.11. Reconstructed chromatogram of the test data
set using the combined algorithm ................... 109

4.12. Reconstructed chromatograms of unleaded
gasoline collected by injecting 0.5 ul gasoline
onto a 22 m section of 0.25 mm fused silica
tubing coated with 0.25 pm SE-54 and using a
mass spectral generation rate of 10 spectra
per second. Figures represent (a) the total
ion intensity reconstructed chromatogram and
(b) the total ion intensity reconstructed
chromatogram from only those spectra with
generic sigma values between 30 and 60 percent
of the base peak intensity ......................... 113

LIST or TABLES

Table Page
2.1 Instrumental Parameters ............................. 25
2.2 Equations of motion ................................ 27
2.3. Experimental Optimum Time Lag ....................... 37

2.4. Mass Spectral Intensities of
Perfluorotributylamine ........................... 56

3.1. Gain Voltages Applied to the Rear Surface
of the CEMA .................................. 63

3.2 Percent ion intensities relative to m/z 43
for successive scan files of n-decane
under dynamic and steady-state conditions .............. 68

CHAPTER 1:
OVERVIEW AND PERSPECTIVE

INTRODUCTION

Changes in source concentrations during the elution of species from
capillary column gas chromatographic columns occurs on the same time scale
as the time required for a scanning mass spectrometer to collect a mass
spectrum. This greatly reduces the amount of mass spectral information
available for each sample component. An attempt to rectify this situation has
been developed based on the use of time-of-flight mass spectrometry
(TOFMS) and time-array (TAD) detection [1], which greatly enhances the
rate of mass spectral scan file generation. Conventional TOFMS was
expected to have serious shortcomings for its application to TAD due to the
difficulties in focusing ions of all masses at the detector for every pulse of ions
from the source. The principal goal of the research described in this
document is to assess the capacity of conventional time-of-flight (TOF) mass
spectrometric instrumentation to obtain complete mass spectra of individual
components of complex organic mixtures compatible with their elution peak
widths from capillary gas chromatographic columns using TAD. This work
was performed through an examination of the temporal focusing

requirements for ions in conventional TOF instrumentation and an

2
experimental evaluation of the advantages of time-array detection over time-

slice detection strategies.

Complex organic mixtures typically contain large numbers of
individual species in widely varying concentrations. Component analysis of
such mixtures requires a chromatographic separation to isolate each
constituent [2]. Gas chromatography is the most common separation device
used for low boiling organic species because of its speed and efliciency.
Providing the highest separation efliciency and often the fastest separations,
capillary columns have gradually become the separation tool of choice in gas
chromatography and are ideal for use in the analysis of complex mixtures [3].

DETECTION OF GAS CHROMATOGRAPHIC EFFLUENT

Various detection techniques have been used with chromatographic
separations to increase the information content of the analytical process.
These techniques include both selective and non-selective detectors.
Identification of individual species using a non-selective detector is based on
the retention time (or index) of the eluting compounds and the stationary
phase used in the separation [4,5,6,7]. These data can be used to obtain
chemical information about structurally similar components of the mixture
when advance knowledge of the mixture’s composition is available [8,9]. A
major advantage of non-selective detectors, such as flame ionization [10,11]
and thermal conductivity detectors [12], is the ability to detect all
components in the mixture, however, little additional information about

eluting species is obtained.

Selective detectors are used to differentiate among the many eluting
species based on a desired chemical characteristic. Selective detectors such

3
as the nitrogen-phosphorous [13] or electron capture detectors [14] provide

little response to eluents that do not exhibit the particular characteristics
sensed by the detector. For example, the nitrogen-phosphorous detector has
high response factors for organic amines and phosphonates but has relatively
low response factors for (and thus discriminates against) normal alkanes.
Mass spectrometers can be used as selective detectors by using selected ion
monitoring (SIM). This approach involves setting the mass filter to pass only
ions of a single m/z value which is characteristic of the class of compounds to
be detected [15]. This added degree of selectivity provides more information
concerning the detected compounds, but distinctions among detected species
are still restricted largely to the regime of chromatographic retention time.
This technique increases sensitivity to a single molecular structure at the

expense of information about all other components.

The ideal universal detector would respond to all eluents but in a
manner that is distinctive for all possible components, thereby ensuring that
all components are detected and eliminating the reliance of identification on
retention indices. This requires the simultaneous detection of diverse
qualities of eluting species, hence the use of multichannel detectors.
Multichannel detectors that have been used for such purposes include several
spectroscopic [16,17] and mass spectrometric [18] detectors, with the latter
being more predominant as a gas chromatographic detector in analytical labs.
These detectors provide a distinct spectrum for most eluting species,
providing non-specificity by responding to most eluents and increased
specificity in the spectral domain at the same time. Both optical and mass
spectrometric detectors can offer many channels of structural information.
While many electronic transitions are available for optical detection, practical
transitions are often limited by available wavelengths of impingent radiation.

4
Mass spectrometers rely on the ability to remove (or add) electrons to the

analyte and hence are slightly more universal detectors than are
spectroscopic detectors. Both techniques use similar detection strategies, so
sensitivity in these detectors is largely a function of the cross section of the
analyte to photons or electrons. The capability of these types of detectors to
provide multiple channels of information and a high degree of sensitivity
produces more information intensive data than other chromatographic
detectors.

CHROMATOGRAPHIC REQUIREMENTS ON SPECTRAL DETECTORS

Detectors which collect spectra must perform their data collection on a
time scale much shorter than that of the chromatographic elution peak. Peak
area (first moment) calculations require around 30 to 60 spectra per peak
elution profile while higher order moment analyses often require sampling
frequencies in excess of 100 spectra per peak [19]. An inadequate sampling
frequency can result in serious errors in reconstructed chromatograms. This
fact is demonstrated in Figure 1.1, which compares a hypothetical elution
profile and reconstructions of it from different numbers of discrete data
points. Using a sampling fiequency of about 3 samples per peak, the
reconstructed chromatogram of Figure 1.1b is obtained. This figure shows an
adequate qualitative representation of chromatographic components, but the
quantitative information is degraded by significant changes in peak height
and area. The situation is even worse for Figure 1.1c where the same
sampling fi'equency was used, but there was a phase-shift in the sampling
frequency versus th GLC peak relative to that of Figure 1.1b. Qualitative
aspects of the data are affected as well. Increasing the sampling frequency to

5 samples per peak, as in Figure 1.1d provides a more accurate reproduction

 

 

 

 

 

 

 

 

B:
i
i
i
a: time
Fig. 1.1a
a a
.5 .5
i i
a time 8 time
Fig. 1.1b Fig. 1.1c
B:
i
i
E
.3
i3 time
Fig. 1.1d

Figure 1.1. The effect of samp ' rate on the reconstruction of
chromatographic profiles. (a) simulated chromatogram. (b)
Reconstructed chromatogram from 3 points across the peak. (c)
Reconstructed chromatogram from 3 (points across the peak, not
synchronized to the elution rofile. ( ) Reconstructed chromatogram
from 5 points across the peak.

6
of the chromatographic profile. Because it is not possible to synchronize

chromatographic elution times with data collection times for scanning
instruments, it is necessary to use high sampling frequencies to avoid loss of

qualitative and quantitative information during the data collection process.

Capillary column chromatographic elution peak widths can now be
produced that are on the order of one second. Such high performance
chromatography requires sampling frequencies of 30 Hz or higher. It is
necessary to sample all avaliable windows of information at this rate. Array
detectors permit the simultaneous collection of all windows of information,
permitting high data acquisition rates [1]. However, instruments that are
presently used to scan a spectrum multiplex the detector among the many
windows, reducing the detection time per window by the number of available
windows. This process can lead to an additional degradation of the
qualitative nature of the data when the scanning time is long relative to the
rate of change of sample concentrations, as illustrated in Figure 1.2 . Figure
1.2a is a simulated steady-state hypothetical mass spectrum with three peaks
of equal intensity. Since spectral intensity is proportional to sample
concentration; sampling the hypothetical species over the concentration
gradient of Figure 1.2b would result in the skewed spectra of Figures. 1.2c
through 1.2c, depending on which portion of the concentration profile was
used during the spectral collection interval. This effect increases the
difficulty of spectral interpretation when relative intensities of the various

windows are involved in the component discrimination process.

Because signals within a window of information are not measured
continuously by scanning detectors, much Signal is lost resulting in a raising
of the detection limits. Lower detection limits can be achieved through the

3.9191429.
Concentration

 

 

 

d
c e
m [2 time

Fig. 1.2a Fig. 1.2b

£111.29

:3 1.2.1—
mam
ii

 

HE

W

E

Figure 1.2. The effect of samplinfilrate on mass spectral quality. (a) A
hypothetical mass spectrum co ected under ideal conditions. (b) The
concentration gradient across which spectra are obtained. (0) The mass
gleam obtained from the risi edge of the concentration profile. (d)

e mass spectrum obtained at e to of the concentration profile. (e)

Themass spectrum obtained on the f ng edge of the concentration

profile.

8
use of a detector which continuously collects information in a channel, as in

array or specific detectors. In addition, a multiplex advantage is realized for
array detectors over scanning detectors that increases the signal-to-noise

ratio by a factor of the square root of the number of contributing elements
[20].

SCANNING MASS SPECTROMETRIC DETECTORS

The detrimental effects of scanning the spectrum that were described
in the previous section only become significant when the detector scan time is
significant with respect to the chromatographic elution peak widths.
Scanning methods in mass spectrometry typically require 0.5 to 2 seconds per
scan to cover the entire mass range and obtain sufficient ion statistics. This
spectral generation rate is clearly inadequate for the narrow
chromatographic peaks that can be produced using capillary gas
chromatographic columns.

A quadrupole-based mass spectrometer has been built for the analysis
of systems in which concentrations change rapidly, such as the detection of
thermal detonation products [21,22]. This instrument continuously scans the
spectrum at a rate of 0.3 ms per Dalton over a 1 to 200 Dalton mass range.
This instrument permits the acquisition of a 100 Dalton mass window in 30
ms, which is adequate for the sampling requirements of a gas
chromatographic detector. However, it has a limited mass range and poor
mass resolution. Additionally, the investigators had difficulty storing the
resultant data at the requisite high rate. They ultimately resorted to high
speed photography of an oscilloscopic image.

9
ARRAY MASS SPECTROMETRIC DETECTORS

Array detectors permit the acquisition of spectral data at a much
higher frequency than scanning detectors. Several mass spectrometric array
detection schemes have been developed that could prove useful for this
application.

An electrO-optical ion detector has been developed as an array detector
for magnetic sector instruments [23,24]. This device takes advantage of the
fact that magnetic sector instruments disperse the mass spectrum in the
spatial domain. Through the use of a microchannel plate electron multiplier
connected to a photoplate and a diode array detector via optical fibers, the
spatial distribution of ions at the detector is transformed into an array of m/z
values. Problems associated with the use of this type of detector include
spreading of peak intensities into adjacent channels limiting both the
resolution and dynamic range of the detector, fluctuations in the gain from
one channel to the next, and a limited mass range. While recent advances
have extended the mass range of this technique [25], there is a trade-off

between mass range and resolution or signal quality.

Fourier transform mass spectrometers (FTMS) use the precession of
ions in a magnetic field to obtain a spectral array in the frequency domain.
This technique has been used to monitor the emuent from capillary gas
chromatographic columns [26]. Spectra can be obtained in the time frame of
tens to hundreds of ms [27], which is adequate for keeping pace with the
chromatography. However, the technique requires that very few ion
collisions occur. This limits pressures in the analyzer region to no more than
10‘8 torr. Such pressures are difficult to maintain when the instrument is
interfaced to a gas chromatograph. Other problems with FTMS include a

10
memory effect, the requirements of uniform magnetic fields in the analyzer

cell, and a limited dynamic range because of space charging within the cell.

A third array detection technique for mass spectrometry, and the one
used for this study, has been developed based on time-of-flight mass
spectrometry (TOFMS) and an integrating transient recorder (ITR)
[28,29,30]. The TOFMS transforms the mass spectrum into the time domain
by giving all ions nearly equivalent kinetic energies in the direction of the
detector. Ions of difl‘erent masses have different velocities and thus different
flight times to the stationary detector. For ions of the same mass to arrive
Simultaneously requires an ion bunching technique to convert the ion velocity
distribution into a distribution of arrival times at the detector. Bunching can
be achieved by pulsing the source, deflecting the ion beam, or sinusoidal
modulation of the source and the detector [31]. Wiley-McLaren style
commercial instruments use a pulsed source with an extraction frequency of
10 KHz [32]. TOF instrumentation has the unique advantage of producing a
complete mass spectrum for each extraction pulse from the source [33]. It is,
therefore, conceivable that 10,000 complete mass spectra could be collected
each second. Hence, TOFMS has the potential to deliver mass spectral
sampling frequencies more than high enough to meet the requirements of
capillary chromatography. This feature attracted Gohlke to use TOFMS as
the first mass spectrometric detector for gas chromatography [34].
Unfortunately though, his data collection system was limited to photographic
reproductions of a Techtronics oscilloscope. So, while TOFMS offered the
requisite high mass spectral generation frequencies, it was limited until
recently by data storage techniques. With the development of time-array
detection and its ability to rapidly store data to a disk, TOFMS becomes a

viable tool for keeping pace with rapid changes in source concentrations

11
observed when measuring effluent from capillary gas chromatographic

columns.

The distribution of ion arrival times that is produced at the detector
from a single source extraction pulse is transformed into a transient
electrical signal of about 100 us duration. This transient signal contains all
the information necessary to reconstruct the entire mass spectrum. In its ‘
conventional time-slice (scanning) detection mode, however, 2 s are required
to collect information from the entire mass range. Time-array detection
(TAD) uses the integrating transient recorder to sum successive transients
and store the resultant summed spectrum to a disk as a scan file, permitting
the collection of up to 66 scan files (complete mass Spectra) each second. This
spectral generation rate is presently high enough to meet the needs of
capillary column gas chromatography.

LIMITATIONS or TIME-or-FIICRT MASS SPECTROMETRY WITH TIME-ARRAY DETECTION

Use of the ITR to perform time-array detection permits the collection of
information within all discriminating channels of data available by mass
spectrometry at a sampling fiequency which is adequate for gas
chromatography. However, several problems remain with the TOF-TAD
system when used for gas-phase analyses. Among the problems are the poor
mass resolution of TOFMS when gas-phase sources are used and the massive
quantities of data that are generated when high mass spectral collection

frequencies are used.

Initial energy, spatial, and velocity vector distributions in the source
limit resolution in conventional TOFMS. Velocities of ions leaving the source '

are a function of their kinetic energy. Most of this kinetic energy is acquired

12
as the ions pass through the potential fields in the source. However, the total

ion kinetic energy is the sum of all sources of kinetic energy, and any initial
distribution of ion energies will result in a distribution of ion velocities and
hence a distribution in ion arrival times at the detector. This source of error
is usually minimized by using high extraction potentials, on the order of 3
kV, which greatly reduces the relative contribution of the thermal energy to
the ion’s total kinetic energy.

The initial energy dispersion can also be reduced through energy
filtering [35] or focusing [36] of ions after they leave the source. Energy
filtering will remove all ions that lie outside of a small window of energies
while focusing will result in isomass ions of all energies arriving at the
detector simultaneously. While these techniques minimize peak spreading
from energy dispersion in a mass-independent manner, they do not
compensate for other sources of poor mass resolution in TOFMS and may still

result in a mass—dependent focus.

The potential field experienced by an ion in an electric field can be
compared to a hill, with the top of the hill being the maximum potential
obtainable within the field. The amount of energy gained by an ion going
down this bill is a function of the starting point on the hill. A distribution in
spatial positions in the source results in ions gaining different energies in
leaving the source, and thus a distribution in the kinetic energy of the ions.
In this manner, the spatial distribution is converted to an energy
distribution. The problem gets worse ‘ as the extraction potential increases.
Any distribution in the kinetic energy of the ions is transformed into a
distribution of ion arrival times at the detector. Gas-phase sources have

large spatial distributions of molecules in the source region, increasing the

13
probability that a large dispersion of initial ion positions will occur. The

distribution of ions in the source is commonly assumed to be confined to the
volume defined by the beam of ionizing electrons [37]. This electron beam
passes through a narrow slit, minimizing the thickness of the ion packet in
the source region. In 1955, Wiley and McLaren designed a two-stage source
which helps compensate for this initial spatial distribution of ions through
the creation of a plane at which isomass ions that originate from different
positions in the source with the same initial kinetic energy arrive
simultaneously. Placing the detector at this space focus plane ensures
optimal focus of ions with initial spatial distributions.

The final source phenomenon that affects resolution is the initial
distribution of ion velocity vectors. Ions with velocity vectors pointed away
from the detector acquire the same kinetic energy in the source as do ions
headed towards the detector, but they must turn around in order to leave the
source. The time needed to perform this act , the "turn around time", results
in ions leaving the source with the same velocities but at difl'erent times. The
magnitude of this effect is reduced by increasing the extraction potential.
Along with their two-stage source, Wiley and McLaren developed the
technique of time-lag focusing in which a delay time between ion formation
and ion extraction from the source is used to permit ions to move in the
source as a function of their mass and initial velocity. In this manner, the
initial velocity distribution is transformed into a spatial distribution which is
compensated for by placing the detector at the space-focus plane. This
technique permits a mass-dependent correction for the "turn around" effect.

Mass-dependent focus is inconsequential when a single m/z value is

monitored for each extraction of ions from the source. In this case, the focus

14
can be readjusted before each ion extraction pulse for the ion being

monitored. However, when using an array detection scheme which monitors
all ions produced by each ion extraction pulse from the source, a mass-
dependency to the focus is detrimental to the resultant spectrum. N 0 one
value of time lag or other focusing parameters will suffice to keep the entire
spectrum in focus.

Ion mirrors (reflectrons) [38] can be used to increase the obtainable
resolution of TOFMS instrumentation. These instruments permit the
acquisition of a complete mass spectrum in 25 us with no resolution
degradation [33]. Therefore, these instruments have the potential to proVide
the needed resolution for gas chromatographic analyses while still providing
high mass spectral scan file generation frequencies. However, reflectrons
have the capacity to correct for either the spatial or energy dispersions in the
source, but not both [39]. For this reason, they are seldom used for gaseous
sources and are used most fiequently in applications with planar sources
such as 2520:: plasmas [40], molecular beams [41], SIMS [36], and LAMMA
[36]. Since gas chromatography requires a non-planar, gas-phase source, the

reflectron is not currently appropriate for use in this application.

Attempts to modify conventional TOFMS instrumentation to improve
resolution and alleviate the mass-dependency problem have included beam
modulation [42,43], energy filtration followed by beam modulation [44], post
source pulsed focusing [45], and time-dependent potentials applied to ion
acceleration grids [46,47]. Until such a mass-independent means of focusing
TOFMS spectra is obtained, it would be useful to identify regions of the
spectrum in which acceptable signal quality is obtained for preset mass-

dependent focus parameters. In this manner, the focused region of the

15
spectrum could be changed for successive ion source extraction pulses, with

focused regions patched together to obtain the complete spectrum. Chapter 2
of this thesis covers an investigation into this possibility.

The use of the TAD system has several advantages over scanning
systems. These include the capability to accurately reproduce the
chromatographic elution profile, to obtain unskewed spectra of gas
chromatographic eluents, to improve the signal to noise ratio, to optimize the
chromatography for speed of analysis, and to collect all the information
generated during the analysis. These advantages are discussed in detail in
Chapter 3 of this thesis.

With the high chromatographic sampling frequencies possible using
array detection methods in conjunction with TOFMS instrumentation, comes
a large quantity of data that must be reduced in order to complete the
analysis. A halfhour gas chromatographic run with a mass spectral scan file
generation frequency of 20 Hz produces 36,000 mass spectra. Generally, only
a limited number of these spectra are analytically useful. It is therefore
necessary to develop a strategy that will rapidly permit the identification of
analytically useful spectra with a minimal interaction from the analyst. One
such strategy is examined in Chapter 4 based on the degree of fragmentation

observed in the mass spectrum.

10.
11.
12.

13.
14.
15.

16.

17.

18.

19.
20.

16
CHAPTER 1 REFERENCES

. Holland, J. F.; Enke, C. G.; Allison, J.; Stults, J. T.; Pinkston, J. D.;

Newcome, B.; and Watson, J. T.. Anal. Chem, 55 (1983) 997A.

. Ettre, L. 8.. Introduction to Open Tubular Columns. Perkin-Elmer

Corporation: Norwalk, Connecticut, 1979.

. Freeman, R. R.. High Resolution Gas Chromatography. Hewlett-Packard

Company: Palo Alto, California, 1981.

. Kovats, E.. Helvetica Chimica Acta, 41 (1958) 1915.

McReynolds, D.. J. Chromatogr. Sci, 8 (1970) 685.

Sfigvard, Ellen M. and Pitzer, Edward W.. J. Chromatogr. Sci, 26 (1988)

. Hayes, P. C. and Pitzer, E. W.. J. Chrom. Sci, 22 (1984) 456.

Gupta, P. L. and Kumar, Pradeep. Anal. Chem, 40 (1968) 992.
Chorn, L. G.. J. Chrom Sci, 22 (1984) 17.
McWilliam, I. G. and Dewar, R. A.. Anal. Chem . , 29 (1957) 925.
Tong, H. Y. and Karadek, F. W.. Anal. Chem, 56 (1984) 2124.

Ashbury, G. K.; Davies, A. J .; and Drinkwater, J .W.. Anal. Chem, 29
(1957) 918.

Karmen, A.. J. Chromatogr. Sci, 7 (1969) 541.
Lovelock, James E.. Chemtech, 11 (1981) 531.

Sweeley, C. 0.; Elliot, W. H.; Fries, I.; and Ryhage, R.. Anal. Chem, 38
(1966) 1549.

Griffiths, Peter R.; de Haseth, James A.; and Azarraga, Leo V.. Anal.
Chem, 55 (1983) 1361A.

3%, Edward S. and Synovec, Robert E.. Anal. Chem, 58 (1986)

McFadden, W. H.. "Mass Spectrometric Analysis of Gas-
Chromatographic Eluents" in Advances in Chromatografihy, Vol 4, J. C.
gsirzldings and R. A. Keller ed.. Marcel Dekker: New or (1967), pp. 265-

Chesler, Stephen N. and Cram, Stuart P.. Anal. Chem, 43 (1971) 1922.

Ingle, James D. and Crouch, Stanley R.. spectrochemical Analysis.
Prentice Hall: Englewood Cliffs, New Jersey (1988) pp. 159-161.

21.
22.

23.
24.
25.

26.

27.

28.

29.

30.

31.

32.
33.

34.
35.

36.

37

17
Mohan, V. Krishna and Tang, Tong B.. J. Chem. Phys, 79 (1983) 4271.

Tang, Ton B.; Swallows, G.M.; and Mohan, V. Krishna. J. Sol. State
Chem. 55 1984) 239.

Hedfiall, Bjorn and Ryhage, Ragnar. Anal. Chem, 51 (1979) 1687.
Hedfiall, Bjorn and Ryhage, Ragnar. Anal. Chem, 53 (1981) 1641.

Hill, J .A.; Biller, J .E.; Martin, S.A.; Biemann, K.; and Ishihara, M.. "An '
Array Detector for High Performance Tandem Mass Spectrometry."
Presented at the 1989 Pittsburgh Conference & Exposition on Analytical
Chemistry and Applied Spectroscopy, March 6-10, 1989, Atlanta, GA.

Ledford, Edward B. J r.; White, Robert L.; Ghaderi, Sahba; Wilkins,
Charles L.; and Gross, Michael L.. Anal. Chem, 52 (1980) 2450.

gielfls, Charles L. and Gross, Michael L.. Anal. Chem, 53 (1981)

Newcome, B.; Erickson, E.; Yefchak, G.; Davenport, M.; and Holland, J ..
"An Integrating Tra£sient Recorder for Time-Arm Detection."

Presented at the 34 ASMS Conference on Mass pectrometry and Allied
Topics, June 1986, Cincinnati, Ohio.

Holland, J. F.; Erickson, E. D.; Eckenrode, B. A.; and Watson, J. T..
"Time Array Detection: Mass Spectromfitry’s Answer to High Resolution
Chromatography." Presented at the 35 ASMS Conference on Mass
Spectrometry and Allied Topics, May 1987, Denver, Colorado.

Davenport, M.; McKnitt, K.; Boling, R.; Enke, C. G.; and Holland, J. F..
"Integrating Transient Recorder or Continuous High-Speed Data
Collection." Presented at the 14 Annual FACSS meeting, October 1987 ,
Detroit, Michigan.

Knorr, Fritz J .; Ajami, Massoud; and Chatfield, Dale A.. Anal. Chem, 58
(1986) 690.

CVC 2000 users manual

Price, D. and Milnes, G.T.. Int. J. Mass Spectrom. Ion Processes, 60
(1984) 61.

Gohlke, R. 8.. Anal. Chem . , 31 (1959) 535.

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

Gohl, W.; Kutscher, R.; Lane, H. J .; and Wollnik, H.. Int. J. Mass
Spectrom. Ion Phys, 48 (1983) 411.

Studier, MH.. Rev. Sci. Instrum, 34 (1963) 1367.

38.

39.
40.

41.

42.
43.

45.

46.
47.

18
Karatev, V.I.; Mamyrin, B.A.; and Shmikk, D.V.. Sov. Phys. Tech. Phys,
16 (1972) 1177.
Yang, Mo and Reilly, James P.. Anal. Instrum, 16 (1987) 133.

McFarlane, R. D. and Torgerson, D.F.. Int. J. Mass Spectrom. Ion Phys,
21 (1976) 81.

Opsal, Richard B.; Owens, Kevin G.; and Reilly, James P.. Anal. Chem,
57 (1985) 1884. '

Bakker, J .M.B.. J. Physics E: Sci. Instrum, 6 (1973) 785.

Shultz, G.A.; Tecklenburg, R.; Holland, J .F.; Watson, J .T.; and Allison, J ..
"GC-MS by Time-of-Flight Mass pectrometry with Time-Array
Detection." Presented at the 37 ASMS Conference on Mass
Spectrometry and Allied Topics, May 21-26, 1989, Miami Beach, Florida.

. Pinkston, J. D.. Studies in Time-of-Flight Mass Spectrometgi' Improved

Resolvi Power and Versatility, and Mass Spectrometry by me
Resolve Ion Kinetic Energy S ctrometry, Michigan State University:
East Lansing, Michigan (1985 .

Kinsel, GR. and Johnston, M.V. "Post Source Pulse Focusing - A Simple
Method to Achieve Improved Resolration in Time-of-Flight Mass
Spectrometry" Presented at the 37 ASMS Conference on Mass
Spectrometry and Allied Topics, May 21-26, 1989.

Muga, M. Luis. Anal. Instrum, 16 (1987) 31.

Yefchak, G. E.; Enke, C. G.; and Holland, J. F.. Int. J. Mass Spectrom.
Ion Processes, 87 (1989) 313.

CHAPTER 2
MASS DEPENDENCE OF TIME-LAG FOCUSING IN
TIME-OF-FLIGHT MASS SPECTROMETRY

INTRODUCTION

Time-array detection uses the signals from all m/z ions produced fiom
each extraction of ions fi'om the TOFMS source. If ion focus at the detector is
dependent on mass, resultant spectra will have poorly resolved peaks in
regions of the spectrum that are distant from the m/z value that is in focus.
Focusing of TOFMS ion packets in conventional, gas-phase instrumentation
involves the adjustment of source parameters. Unfortunately, these
deviations in source parameters are not mass independent, and conventional
instrumentation does not provide the opportunity to further refine the focus
after ions leave the source. The work described in this chapter was initiated
in order to ascertain the severity of this mass dependence when conventional

instruments are used to monitor large regions of the mass spectrum.
ION Focvsmc IN TIME-or-FIICHT MASS SPECTROMETRY

Temporal focusing of isomass ions having difi‘erent initial energies is
obtained in conventional time-of-flight mass spectrometry (TOFMS) by
means of a delay time (time-lag [1]) between ion formation and extraction
from the source and by adjustment of the extraction grid potentials. During

19

20
this time-lag, the initial kinetic energy of the ions causes them to be

displaced from their incipient positions. Delay times are selected such that
ions which at the time of extraction are farthest from the detector catch up to
ions whose positions at the extraction time are closest to the detector.
Extraction grid potentials are adjusted so that this coincidence occurs at the
detector surface. The Optimum delay time for the extraction pulse is a
function of the mass-to—charge ratio of the ions being focused. Figure 2.1
illustrates the influence of this mass-dependence for the m/z 502 and 503 ions
of perfluorotributylamine. With the proper focus, seen in Figure 2.1a, the
peaks at m/z 502 and 503 are well resolved. The spectrum in Figure 2.1b was
collected with the focus optimized for m/z 28. In Figure 2.1b, the peak at m/z
503 appears as an extension of the shoulder of the 502 peak, reducing the
qualitative and quantitative information available in this region of the
spectrum.

As discussed in Chapter 1, the mass-dependence of this focusing
technique is inconsequential when only one m/z ion is monitored for each
extraction pulse, as in time-slice detection [2]. In this case, focus parameters
are adjusted for successive extraction pulses and thus the optimum focus is
obtained for the m/z ion being monitored. However, this scanning data
collection technique results in the loss of information concerning all other
ions in the transient mass spectrum because these ions are not being
observed [3]. AS shown in Chapter 1, it is desirable when monitoring systems
in which concentrations are changing rapidly, such as chromatographic
effluent, to avoid the loss of chromatographic or mass spectrometric
information by using high sampling frequencies. These high sampling
frequencies can be achieved using time-array detection [3], but this technique

21

 

 

 

 

 

 

I H 502
>- H.
t
(f)
(a) E ‘
j...
Z
"" . H 503
69:18 ' H : t ; 69.56
Flight Time (as)
E V H 502
m
(b) E .
*- l
-2- HH [\ 503
Hy,
n L .4 “W
71.93 72.31

Figure 2.1. The mass-dependence of time-lag focusing in time-of-flight mass
spectrometry. Shown are m/z ions 502 and 503 from the spectrum of
perfluorotributylamine with the value of trme lag optimized for (a) m/z

502 and (b) 28.

22
requires that a large range of m/z ions be monitored from each extraction

pulse, making mass-dependent focusing undesirable.

Several instrumental modifications have been examined in an attempt
to eliminate the mass-dependency of ion focus in TOFMS. Among these are
beam modulation [4], energy filtering in combination with beam deflection
[5], post source pulse focusing [6], and dynamic-field focusing [7]. Each of
these approaches requires modification of the TOFMS instrumentation, with

the latent complications of additional ion optics and instrument electronics.

The approach taken during this investigation was to investigate the
use of conventional instrumentation with a succession of time-lag settings,
each providing a mass window of acceptable focusing without prohibitive
signal degradation. By using conventional instrumentation over smaller
mass windows, TOF-TAD could be put to immediate use while more eloquent
solutions to the ion focus problem can be developed. However, this approach
requires that the number of sequential time-lag settings used to cover the
instrument’s mass range be small or the time required to collect full mass
spectra will be too long to provide full mass scans at the rates required by
high-performance chromatography. Use of a multi-time-lag approach
necessitates the capacity to predict the boundaries of such mass windows.
This chapter is the result of efforts to determine the severity of the resolution
degradation caused by time lag, and to characterize the effects of other
instrumental parameters on the temporal focus of ions in a commercial,

linear, gas-phase TOFMS, the CV C 2000.

23
THEORETICAL

Sources of poor resolution in TOFMS include the initial spatial,
energy, and velocity distributions of ions in the source as well as velocity
distributions fi'om metastable decompositions after ions leave the source.
Due to the difficulty of finding molecules which have fragment ions
throughout the mass range of the instrument and which do not undergo
metastable decompositions, a computer simulation of the CV C 2000 TOFMS
instrument was developed. This simulation permitted the characterization of
the instrument and examination of factors that limit performance; studies
that would have been dificult or impossible using data generated exclusively
from the instrument. When possible, simulated data were compared with

experimentally obtained data.

A schematic representation of the source and ion drift: regions in the
CVC 2000 instrument is presented as Figure 2.2. This instrument has a 4-
grid ion source with a 2 m field-free drift tube. Electrons are pulsed into the
ionization region of the source for a period of time controlled by the Operator.
The first extraction grid is held slightly positive with respect to ground while
the backing plate is held at ground to provide a confining potential for ions in
the source region located a distance 31 from the exit grid for the first region.
The second grid is also held at ground during this time period. While the
electron beam is passing through the source, ions are confined to the
potential well formed by this electron beam [8]. After the electron beam has
been deflected away, the ions’ thermal energy causes them to drift within the
ionization region. Upon expiration of the time lag, a square wave pulse of
magnitude V1 and ramp time t, is applied to the first grid to initiate

extraction of ions fi-om the source. A second square wave pulse of magnitude

 

:H W

 

 

§ ./

0 V2 V3 V
l . l .
t t
I’ I'
Figure 2.2. A schematic di of the CV C 2000 Time-of-Flight Mass

Spectrometer. Typical vo tages and distances are included In Table 2.1.

25
V2 is simultaneously applied to the second grid. Remaining grids in the

source are held at steady-state potentials to ensure that ions do not
experience changing fields after leaving the source. Instrumental
parameters used on the CV C 2000 were measured and are listed in Table 2.1.
Values in Table 2.1 were used in simulations of instrument performance.
Focus parameters that have been examined as a part of this work include the
duration of the time lag, the magnitude of voltages applied to the first two
grids, and the ramp time needed to achieve the maximum voltage on the first
two grids. Those factors that have been shown to have the largest effect on
ion focus are the time lag and the V1 potential.

Table 2.1. Instrumental Parameters

Mater V_al_eu

x1 0.0037 In
x2 0.0017 In
x3 0.006]. m
x4 0.0056 m
x5 2.1 m
81 0.0020 m
V i +0.44 V

f -6 to -140 V
V1,. 0.0 V

f -150 to -250 V
V2 -1400 V
vi -2700 V

t, 30 ns

26
Flight time calculation

The acceleration, a, of a charged particle in an electric field is
described by Equation 2.1:

a = qV/ mx (2.1)

where q is the charge on the particle, V/x is the potential field experienced by
the particle, and m is the particle’s mass. The velocity, v, and distance
traveled, 3, within each region can be calculated through integration of
Equation 2.1 with respect to time. The total flight time of an ion is the sum
of the flight times through each of the individual regions:

tor: t1+t2+t3+t4+t5 (2.2)

Each individual flight time can be Calculated from the solution of quadratic or

cubic equations.

Potentials on the first two grids in the source are pulsed between
initial Wu and V2,) and final (VIf and V2,.) voltages as listed in Table 2.1.
Since ideal square waves cannot be obtained, it takes a certain time, t,, to
reach the final potential. Hence, the possibility that ions leave a region
before t, has elapsed should be considered. This results in several possible
flight times through the first three regions of the source. Table 2.2 contains
the equations of motion for ions in each of the 5 regions in this instrument.
In the development of these equations, the voltage rise was approximated by
a linear ramp between the initial and final voltages applied to grids 1 and 2.
Using this approximation, the voltages on these two grids before time t1. is

27

Table 2.2 Equations of motion for cases where tr expires while ion is

before, after or within each region.

Amer-Regen

 

fin l:

unr
l l
"1 ' WI” WV" V..) + -...-.1—+ v.
1 6mxlt, 1” 15 mel 0 1

 

W

4‘2
v: '_mxa (sz- V”) + vi

2

9‘2
8 8 ——(V2f-V1)+vlt a x

2 mea 2

v =——
2 mat,

a a—(V -V

...:
(Vw- Vlf+ V“) -

thtz

mx

 

2 l

3
9‘
2 2 + V .) -
‘ 6mg, 2/ If 15
2
tht2

20:12

 

+ vlt2 a :2

 

m
9‘s

03 a —mx3 (V3-V2I) + 02
2

7‘s
83 a 2mxa (V3— V2) + v2:3 s x

 

3

2
2V3ta _ qufa + v
3 "‘33 2mxat, 2

 

2 a
qu‘a 4V2}:

 

8 a - +02t83x3

3
20:33 6mg,

 

Wonk

 

 

 

(1‘ qt
v‘ a 37;:(V‘-V3)+v3 v4 - 75% (V‘-V3) +03
2 2
4 4
s4:- 2 4(V4-V3)-r-vat4 a :4 84 3217234 (V‘-Va) + v3t4 a x‘
5:
v s v

 

28

Tabb 2.2 (W Within m

 

17......

 

Q_V_l( qr, V + V“.
”I ' mxl(1‘+‘r) mxl[ 1’2 ‘r “’0

3
81 BRUI-t) +[mxl[ 2 V..-DO (t l-s-mT-tr)+ 11(V WV)+vt

 

 

W

qV
v2 - m1: —(V -Vu)(t1+r2-¢,)+—— 2m: 2" (Vii-VII1+V‘.,12--)(r-—r)--Fx-lzi(r,—¢l)+v1

‘2.W(V2f -12Vu)(t+t-tr)2+vlrlz(t-t)+[m:‘—2—tr (sz MV+VXt,12-t)-

 

qvh- q 3 q !£ 2
m—xzar- t!) + UIJUI + ta-tr) + WW1— v1f+ VnXtr- ‘1) - 2mx2 (tr-£1) I :2
mi
(1(Va — V

qV qV ,
03- ""‘a slug“ -¢,)+— (r-t day—10,4142) “’2

 

 

 

 

q<V,-V,,J 2 9V.
83 .—2-m_x—3—-(tl+t2+t3-t’) + (tl+t2+t3-tr) m—xa(t,-tl-tz)-
qu 2 9V3 2 ‘IV a
2mx_(t'-3t-t1-t2) +02] + m(t'-tl-t2) -W3%(t'—tl-t2) 4-
020,414; a :3
We
9'4
0‘ 8 —mx‘ (V‘-V3) + 03
2
9‘4
34.2," (V-V3)+vt ‘
Won&
05 I v‘

29
reached are given by:

where t is the time expired since the initiation of the extraction pulse.

Resolution calculation

Using Equation 2.2 and those of Table 2.2, the flight time of any ion
having a known initial position and velocity can be calculated. Particular
distributions of initial ion positions and velocities in the source lead to a
distribution of ion arrival times at the detector, and ultimately, a simulated
TOFMS peak shape. In this work, normal (Gaussian) distributions of initial
position and velocity were assumed. New positions and velocities were

calculated for ions that moved due to the trapping field and the time lag.

Resolution, R, in mass spectrometry is defined by the following
relationship:

R=m/Am (2.5)

where m is the mass of the ion and Am is the peak width. The peak width is
usually determined at 10% valley for magnetic mass separators and at 50%
valley for time-of-flight instruments. In TOFMS, the above relationship can

be rewritten as:
R=t/2At=t/2(t2-t1) (2.6)

where t is the mean flight time for ions of mass m, At is the peak width in
units of time, and t1 and t2 are flight times for ions that appear on the

. 30
shoulders of the peak. Wiley and McLaren [1] have shown that the flight

time in the field-free region is:
t = 1.02 mm“5 D / 200-5 (2.7)

where D is the length of the field-free region and U is the total kinetic energy
of the ion. Assuming that mass is not converted to energy, and flight times in
the acceleration region are negligible compared to that of the field-free
region, equations 2.6 and 2.7 can be reduced to:

R = (U1U2)0-5 / 200-5(U20-5 - U105) (2.8)

where the subscripts represent the kinetic energies needed to produce the
required arrival times on the shoulder of the peak. Notice that there is no
mass dependency in Equation 2.8. . Hence, in TOFMS this definition of
resolution is independent of mass. However, the difference in ion arrival
times between adjacent ions is a function of mass as can be discerned by the

equations in Table 2.2.

Another means of relating the ability to separate ions in mass
spectrometry is to identify the mass at which the height of the valley between
peaks of equal intensity originating from adjacent integer m/z values exceeds
a threshold percentage of the maximum intensity. All masses for which the
threshold is not exceeded are said to be at least unit-mass-resolved for the
specified definition of percent valley. This approach has been taken in the

course of the current investigation.

In determining the percent valley between two adjacent m/z peaks it is
not necessary to calculate the entire TOFMS peak shape, but only the
relative intensities of the peak maximum and the valley. Flight times for

3 1
these positions on the arrival time peak shape distribution can be calculated

from equations in Table 2.2. The sum of the products of the probabilities of
all positions, P(si,tof), and velocities, P(vi,tofl, that result in a specified flight
time yields the probability that ions have that arrival time.

P(tof) = Z P(v,-,tof) P(si,t0f) - (2.9)

This probability is directly related to the intensity observed at the
detector for a particular flight time.

Intensity Calculations

The peak height of a particular m/z value is afl'ected by the adjustment
of focus parameters in a mass-dependent fashion. One reason for this mass
dependency is that low m/z ions, which have higher thermal velocities than
their high m/z counterparts, can actually leave the source region prior to the
application of the ion extraction pulse. This causes a reduction in the
number of ions available for detection. In addition, ions for which optimal
focus has not been provided will result in arrival time peak shapes that have
lower peak marn’ma and broader peak widths relative to ions that are in
focus. Because the optimal focus is mass-dependent, broadening of the
arrival time peak shape is also mass-dependent.

As a consequence of the mass-dependent relationship between focus
parameters and the peak intensity, a skew in the mass spectral peak
intensities can be observed when large ranges of m/z values are monitored.
This is especially apparent when using algorithms based on peak height
rather than area. In the definition of useful m/z windows, it is therefore
necessary to include a threshold for the loss of peak height as well as one for

the loss of mass resolution.

32
Ions that are calculated to be outside of the ionization region before or

after the application of the time lag are ignored. In this manner the
contribution of ions that have left the source prior to extraction have been
considered. Other intensity calculations are based on the peak height, or the
probability at the most probable arrival time.

Simulation

The simulation used for this work was written in FORTRAN-77 on a
MicroVAX II. The code consists of a main program, LIMRES, and nine
subprograms and functions. A source code listing of this program is included
as Appendix I of this document. Several of the algorithms used in this
program are modifications to programs obtained from George Yefchak or
from reference 9. The source of these algorithms has been noted in the code.

The main program is used to set the parameters used in the
calculations and to loop through variables. These parameters include
voltages and distances in the various regions of the instrument, ion m/z
value, time lag, ramp time, and desired limits of resolution and intensity.
Through subroutine and function calls, this program calculates flight times,
resolution, and intensities of two adjacent m/z value peaks of equal
probability. The main program is also used to write the results of the
calculations to a file in Cricket Graph [10] format. This program was altered
frequently to incorporate features that were desired in the output. Appendix
I contains a version of this program which looped through mass, time lag,

and ion focus in acquiring data.

LIMRES performs its functions through calls to CALTOF and
MAXMIN. CALTOF is the subroutine in which ion flight times are

33
calculated. This routine is based on Equation 2.2 and the solutions with

respect to time of equations in Table 2.2. This requires the evaluation of
quadratic or cubic solutions using QDSOL or CUBSOL, respectively. The
routine accepts variables for the initial position, initial velocity, and m/z

value to calculate flight time in us.

MAXMIN is the subroutine that is used to find the peak maximum and
the height of the valley. This routine receives boundary flight times and
returns intensities for the valley and peak heights. The peak shape
distribution resulting fi-om Gaussian distributions of initial positions and
velocities is not necessarily a Gaussian distribution, and hence, this routine
was written to examine several points across the distribution of arrival times.
This is done through calls to TPROB and SIMPS. The latter fimction
calculates the area of a peak through a Simpson’s rule approximation.

The subprogram TPROB is used to calculate the probabilities or
intensities of a particular ion arrival time. This routine loops through
possible spatial positions in the source and calculates the velocity needed to
achieve the desired flight time through a call to VELCAL. Initial positions
and velocities are then calculated to compensate for movement during the
time lag. Through calls to PRBLTY, the probabilities of the initial positions
and velocities are calculated. Equation 2.9 is then implemented to determine
the total probability as the sum of the products of individual probabilities of

initial positions and velocities.

PRBLTY is the function that returns the probability of occurrence of
an event. It assumes a normal (Gaussian) distribution of events.
Normalization of the function was not included so that the maximum
intensity obtained is 1.

34
VELCAL is the subroutine that is used to calculate the initial velocity

of an ion from its arrival time and m/z value. A plot of initial velocity versus
ion flight times to the detector, such as the one presented as Figure 2.3, is
approximately linear with the intercept being a function of the ion’s initial
position and m/z value. Deviation from linearity is observed to occur as the
initial velocity differs significantly flour 0 m/s. VELCAL uses a linear
approximation between small intervals of velocity to determine the initial
velocity. This is done with the aid of function calls to LINE which is used to
compute the slope and intercept of a line. The interval approach was used
instead of root finding algorithms to speed up the calculations while
minimizing deviation from the correct result that would be observed if a
single interval were used. In regions of significant deviation from linearity,
the probability of occurrence for the. initial velocities is small enough that
weighting factors make these errors insignificant. For example, the
probability that m/z 10 would have a velocity of 3000 m/s fi'om thermal
energy is 0.000024. This probability drops further as the value of m/z is
increased. The weighting of such a small probability with those of smaller
velocities makes this value and any inherent errors insignificant as long as a

sufficiently large number of samples are obtained.

RESULTS AND DISCUSSION
Efi’ect of time lag
The effect of time lag on the peak shape is illustrated in Figure 2.4.
This figure includes some of the contribution to the peak intensity from a
peak of equal intensity at the next higher integer m/z value in order to
determine the height of the valley between the two peaks. AS the time lag
approaches its optimum value, the peaks get narrower and more intense with

a decrease in the height of the valley between adjacent masses. Additionally,

Flight-Time (us)

35

 

 

615 . I
~5000 -3000

I T

I I ' I ' fl
-1000 1000 3000 5000

Initial Velocity (m/s)

Figure 2.3. The relationship between initial velocity and
ion arrival time at the detector for m/z 500.

36

00
02
"4

 

 

 

 

10"

ASE: brags 5835

53.68

53.66

53.64

Flight Time (usec)

53.62

53.60

53.58

Figure 2.4. Simulated peak shapes for m/z 500 at

difi‘erent values of time lag.

37
as the time lag is increased beyond the optimum, the peak shape is skewed

towards longer flight times.

The effect of time lag on the peak intensity is demonstrated in Figure
2.5. In this figure, the relationship between peak intensity and time lag is
plotted for ions of several different m/z values. The lightest ions exhibit the
largest intensity. This is an artifact of the narrower peak widths found for
the lightest ions. The Optimum intensity for each ion occurs at higher values
of time lag as the m/z value increases. This phenomenon has been observed
in data from the instrument as Shown in Table 2.3. These results are also in
agreement with those of Wiley and McLaren [1] which demonstrated that the
optimum time lag is proportional to the square root of the ion’s mass. Data in
Table 2.3 fit the equation:

t,“ = 0.0956 (m/z)°-5 + 0.139 (2.10)

where tlag is the time lag in usec. The regression coefficient for fitting these
data to Equation 2.10 is 0.988 and is due in part to a dificulty in reading
accurate time lags from an oscilloscope. Constants in this equation are a

function of the voltages applied to the extraction grids.

Table 2.3. Experimental Optimum Time Lag

ELZ. Time Lag
69 0.3 usec
100 1.1 usec
131 1.3 usec
181 1.4 usec
231 1.6 usec
281 1.6 usec
331 1.9 usec

38

0.3 -

Units)

0.2 "‘

0.1 ..

Intensity (Arbi

 

 

 

00 v I I I I I l j
0 1 2 3 4 5
Time Lag (usec)

Figure 2.5. The dependence of peak height on time lag and mass.

39
A plot of time lag versus percent valley for a particular m/z ion passes

through a minimum at the optimum time lag for that m/z value. An example
of such a plot for several values of m/z is included as Figure 2.6. As
previously noted, the optimum time lag increases as a function of m/z. This
figure can be used to identify the values of time lag at which ions will have at
least unit-mass-resolution. For example, at least unit-mass-resolution is
achieved at 10% valley for m/z 300 at all values of time lag below 3.8 usec.
Unit-mass-resolution is achieved for ions of m/z 700 only for values of time
lag between 2.5 and 3.2 usec. Minima in Figure 2.6 also correspond well with
experimentally observed optimum time lags in Table 2.3, agreeing to within
0.15 usec.

Figure 2.7 is a comparison plot of time lag versus percent valley for
experimentally observed values. from the m/z 502 ion in
perfluorotributylamine relative to calculated values for m/z 500. Except for
one point, the calculated curve lies within the error of the experimental
measurements. Excellent agreement between the Simulation and the
instrument is observed for those values of time lag which result in sharp
peaks (adequate focus). The calculations result in a systematic difference in
the height of the valley relative to experimental values as the resolution is
Significantly degraded. This may be an artifact of the ability to accurately
measure peak and valley heights in the experimental data when significant
overlap occurs. Difficulties in correlating the experimental data to the model
at low values of time lag may be due to errors in the model’s ability to mimic
ion events occuring immediately upon the application of the extraction pulse.
In addition, the model does not account for ion repulsion or fringing fields

which do occur in the instrument.

‘~ \ —— m/z = 300
‘~ \ ~-----~ m/z = 500

Percent Valley

 

 

 

 

Time Lag (usec)

Figure 2.6. The calculated effect of time lag on resolution.

41

100-

 

 

 

 

 

 

 

 

o . 1 2 a ' 4 5
TimeLag(usec)

Figure 2.7. The effect of time lag on resolution. Calculated values for
m/z 500 are plotted as a continuous curve. Points indicate
experimental values for the m/z 502 peak from
perfluorotributylamine.

42
The useable m/z windows are bounded by the intensity and resolution

limits required to obtain satisfactory signal. The definition of these
boundaries on a plot of time lag setting versus. m/z value will identify the
usable mass range for a given time lag setting or the appropriate time lag
settings required to cover a given mass range. Such a plot is included as
Figure 2.8. In this figure, a resolution threshold of 10% valley and an
intensity threshold of 90% were chosen as acceptable limits. All m/z values to
the left of the 10% valley boundary have at least unit-mass—resolution. At
low values of m/z, the arrival times between integer m/z values is so large
that unit mass resolution is always obtained. All points between the
intensity boundaries have no greater than a 10% loss in the peak height. The
area defined by these three boundary curves gives the useful mass window
for each value of time lag. The curve between the intensity boundary curves
of Figure 2.8 provides the optimum time lag as a function of the ion’s m/z

value.

It is now possible to use Figure 2.8 to determine the window of
acceptance for monitoring multiple ions from a single extraction pulse of a
TOFMS source. For example, to collect a Spectrum from m/z 50 to 700 would
require the use of three different time lags with their inherent functional
mass windows. Values of time lag used would be 1.0, 1.5, and 2.8 us and
would cover m/z ranges between 50 to 150, 150 to 300, and 300 to 700
Daltons, respectively. Multiplexing these windows together fiom successive
extraction pulses would produce a complete mass spectrum with each peak in
adequate focus. The price of multiplexing these 3 windows relative to using
an array detection scheme in which all ions are in focus is a factor of 3 loss in
sensitivity and a factor of 1.7 loss in signal-to-noise (S/N) ratio. Compared to
a scanning detection method with a 5 ns window, the use of these 3 windows

43

 

 

5 -
1
Optimum Focus
4 .
3 90% Intensity Boundaries
O a
Q 3
3-
E
i 2'
1
1 H
l 10% Valley Boundary
o ‘ I ' T r I ‘ I ' I
O 200 400 600 800 1000

M

Fi e 2. 8. Working curve f0 acceptable signals. Ions to the left of the
gur reso lutron curve 1give unit-mass-resolution with less than

a 10% valley. Ions between the intensity curves loose no
more than 10% of their peak height. The curve in the
center gives the optimum time lag for each m/z value.

44
gives a factor of 4000 increase in sensitivity and a factor of 63 increase in the

SIN ratio.

Efl'ect of extraction grid potentials on ion focus

Isomass ions with identical initial energies that are in different
positions in the source when the extraction pulse is applied will acquire
difi'erent energies, and hence, velocities as they leave the source. Ions that
travel farthest in the source region will have the highest energy and velocity,
but will be the last to leave the source. They catch up to the slower ions at
the Space focus plane. The distance of this space focus plane from the source '
is a function of the potential applied to the first grid, the ion focus potential.
Ideally, this plane is located at the detector surface. When it is not located at
the detector surface, broadening and a distortion of peak shapes can be
expected. Experimentally, improper adjustment of the ion focus voltage
presents itself as a loss of peak height, a broadening of the peak, and a
distortion of the peak shape. .

A plot of ion focus voltage against percent valley should have a
minimum at the optimum value of the extraction potential. Such a plot is
included as Figure 2.9 for a m/z value of 500. This figure also shows that the
optimum ion focus voltage is a function of the time lag, with larger time lags
requiring a shift to lower ion focus voltages. In Chapter 1 the energy
obtained by an ion as it leaves the source was visualized as the height of the
ion on the potential hill. The ion focus potential sets the maximum height of
this bill and hence its slope. Large values of ion focus potential require
smaller delay times to permit ions to migrate to the requisite new positions
for proper spatial focusing than do small potentials. This results in large ion
focus potentials requiring small values of time lag while small ion focus

45

 

100

 

 

Ion Focus (V)

ion focus potential for m/z 500.

Figure 2.9: Resolution as a function of the

46
potentials require large values of time lag. This phenomenon is illustrated in

Figure 2.10, where time lag is plotted against the percent valley for several
difl‘erent values of ion focus. In this figure it becomes apparent that the
window of acceptable time lags is smaller for the larger ion focus voltages
while the smaller voltages require time lags larger than are readily available
on the commercial instrument. Additionally, high values of the time lag will
increase the probability that low m/z ions have left the source prior to the
application of the extraction potential, which would result in an undesirable

skew in the mass Spectra.

The selection of an optimum ion focus potential is a trade ofi' between
the size of the mass window at a particular time lag value and the desired
total mass range for the analysis, as illustrated in Fig 2.11. This figure
Shows the efi'ect of ion focus potential on the shapes of the resolution and
intensity limiting boundary curves for the same threshold values as used for
Figure 2.8. Increasing the value of the ion focus potential causes the
optimum focus to occur at lower values of time lag. This causes a desirable
leveling off of the curves which define the intensity window, making the
degradation in signal intensity less prominent across wider mass ranges. At
the same time, however, the increase in ion focus potential causes a shifling
of the resolution curve to lower values of time lag and m/z. For example, at
an ion focus potential of 75'V, a time lag of 2.5 usec has a practical mass
range between about m/z 150 and 350 for a 200 Dalton window size. Setting
the ion focus voltage to 77 V, this mass range changes to roughly 200 to 500
with a window size of 300 Daltons. Both of these voltages cause the window
size to be determined only by the intensity threshold. Using an ion focus
potential of 80 V, the intensity limits at a time lag of 2.5 usec provide an
adequate signal intensity over the 750 Dalton window between m/z values of

47

100 - ._,__._.-.-.-.-.

 

     

V1 .60V

I
j I ...”... V1-7OV
0' '
I I
. ° I """"" VISBOV
>3 604 .‘ 3’ I,
% \-. 0.’ , ----- V1=90V
'\
a H \H’. .......... v1.1oov
v .
g 4.. .,
24

 

 

0 V I T I 4 I r I ' _I
O 1 2 3 4 5
Timc Lag (usec)

Figure 2.10. Interdependence between resolution,
time lag, and ion focus

LagThmeQunc)
to o

 

51

t

u

q

1‘
200 400m 600 800

O 1000

 

q
41
1

31
1

2d

Ll‘1‘fllflhln)

‘1

 

 

0 200 400m 500 1000

 

lagThneQulfl
n or

 

 

0 200 400M,z 800 1000

5- (d)

IagThnewuu»

 

 

 

I V I v I ' T ' l
0 200 400 600 300 1000
DUI

Figure 2.11. Wor ' curves for ion focus values of (a) 75 V. (b) 77 V, c 80
V, and (d) 85 V. e V2 voltage used was 215 V. ( )

49
225 and 950. However, the resolution threshold has moved down into this

window, restricting the upper limit to a m/z value of about 700. The window
size is thus only about 475 Daltons. At an ion focus voltage of 85 V and a
time lag of 2.5 usec, the resolution cutoff occurs before entering the

acceptable intensity range, leaving all ions outside of the acceptable limits of
signal degradation.

The ion-accelerating pulse is applied to. the second grid concurrently
with the ion focus pulse applied to the first grid (V1 and V2 in Figure 2.2).
The efi'ect of the potential applied to the second grid on ion resolution is
demonstrated in Figure 2.12. This potential has a minimal efi'ect on low m/z
ions Since they require small values of time lag for optimal focus. Ions that
require larger values of time lag will benefit from the increased window of

acceptable time lags from using higher V2 potentials.

The effect of the V2 potential on the size of the mass windows for
selected values of time lag and a constant value of the ion focus potential is
illustrated in Figure 2. 13. This figure was generated using the same
resolution and intensity boundaries as used to produce Figures 2.8 and 2.11.
The observed effect is similar to that of the ion focus voltage. Lower V2
voltages lower the slope of the optimum value and intensity limiting curves
(dotted lines) while also causing the 10% valley limiting curve (solid line) to
be pulled towards the plot origin. A trade off is therefore needed to get the
largest m/z range and the widest m/z window sizes. This effect is not as
severe as that caused by the ion focus potential, requiring much larger
changes in potential to produce noticeable effects.

50

7O

'_ V2=150V
W V2=180V

V2=200V
V2=230V

 

 

60-1

 

Time Lag (usec)

Figure 2.12. Resolution as a fimction of time lag arérl V potential.

51

Iag'l‘ime (usec)

 

 

f I t I V V V

V I U I I r 1' U V '
200 40m~ 600 800 1000

Lag'lime (pace)
O ?W

 

Lag'l‘ime (usec)

 

' fT I V

I ' ' r . ' ' I
0 200 40 600 800 1000

(d)

I f l j

 

Lag'l‘ime (usec)

 

V V V t

o ' v v I r t v

' I ' V ' I ' I
0 200 40¢‘M/z 600 800

Figure 2.13. Wor ' curves for V values of (a) 150 V. (b) 200 V, (c) 225 V,
and (d) 250 V. e ion focus vgltage used was 80 V.

52
Efl'ect of ramp time

Use of digital simulations instead of instrumental experiments permits
the examination of the effects of changing parameters that are dificult to
manipulate on an actual instrument. One such case is the examination of the
efl'ects caused by non-ideal extraction pulses. Experimental examination of
this effect would require instrument modification. AS described in the
theoretical section of this chapter, this case is accommodated within the
simulation through the assumption of a linear extraction potential ramp in
the derivation of the equations of motion of ions in the instrument. The

duration of the ramp, t,, is an adjustable parameter in the simulation.

Figure 2.14 shows the effect on resolution of using different ramp
times. Increasing the ramp time of the ion extraction pulse produces a minor
increase in the resolution for low values of time lag. Since this effect is
similar to that of increasing the time lag, longer values of tr require lower
values of time lag to obtain the same focus. It is interesting to note that this
phenomenon results in an apparent enhancement in resolution as the rise

time of the ion extraction pulse increases.

The effect of the ramp time on the plot used to determine the size of
the mass windows is minor. Increasing the ramp time causes the curves that
delimit resolution and intensity thresholds to maintain their shape and slope,
but shift to lower time lags. This outcome is a result of the fact that at higher
values of the ramp time, ions spend more time in the ionization region of the
source and hence have more time to be displaced from their original positions
as a result of their thermal velocities. More time is required for the
extraction potential to overcome the initial ion energies and direct the ions
towards the detector. This ion displacement is similar to that occurring as a

53

 

if- 008

 

"8500118

Percent Valley

20"

10"

 

 

0 . I . . .
0 1 2 3 4 5

Time Lag (us)

Figure 2.14. Efi'ect of ramp time on optimum time lag for m/z 500.

54
result of the time lag, and therefore, requires lower values of time lag to

obtain the necessary ion separation distance in the source.

CONCLUSIONS

In TOFMS, mass windows of acceptable Signal quality for particular
settings of time lag can now be defined for use with detection schemes that
involve monitoring more than one m/z value from each extraction pulse. The
boundaries of these mass windows are dependent upon many instrumental
parameters and are instrument specific. Parameters that are most likely to
afi‘ect the size of the windows are the time lag and the ion focus potential
(V1). N onideal extraction pulse shapes have limited efi‘ect on the size of the
mass windows, being manifested instead as a slight diminution to the

requisite time lag.

The width of functional mass windows is limited by intensity
boundaries throughout the low mass range. Resolution boundaries only
become a factor at high masses. This fact can be used to help generate plots
similar to those of Figs. 2.8, 2.11. and 2.13 for other TOFMS instruments and
parameters. It is possible to derive the intensity boundaries by optimizing
time lag for a particular peak followed by adjustment of the time lag until the
peak intensity has been reduced by a factor of 10%. Performing this
operation for several m/z values will result in the collection of sufficient data
to generate the intensity boundaries. As long as the resolution boundary is
not exceeded, these boundary curves will define the useful mass ranges.

An examination of data used to generate the working curves in Figs.
2.8, 2.11, and 2.13 has revealed a means to estimate the size of the functional
mass window around a selected m/z value. For those cases where resolution

is not the limiting factor and the ion focus voltage is close to its optimum

55
value, the upper limit on the window is a factor of between 1.55 and 1.70 of

the m/z value that is in optimum focus. The lower limit is between a factor of
0.65 and 0.75 of the optimum m/z value. This approximation fails for m/z
values below 100 and for cases where resolution determines the upper
boundary. Intensity thresholds other than 90% of the optimum intensity will
also alter this approximation. In addition, these estimates may not be valid
if instrumental parameters are not the same as those listed in Table 2.1.

The TOFMS-time-array detection system was used to collect the
spectra of perfluorotributylamine listed in Table 2.4. A time lag value of 0.8
as was used for Case A, which provided an optimal focus for m/z 69 and a
mass window of acceptable intensities from m/z values of 40 to 110. Using
time lag values of 0.9, 1.5, and 2.0 as and windows of 50-100, 100-300, and
225-575 respectively, intensifies for Case B were put together to obtain a
Spectrum in which all peaks are focused within acceptable limits of intensity
and resolution degradation. The Case B spectrum has a slight decrease in
intensities at low masses and a significant increase in intensities at high

masses relative to the Case A spectrum.

56

ctral Intensities of
rotributylamine

S
.53.

Pa

Table 2.4. Mass

........ 003. LJJD.J9J20L41AJ902L361
0000000000001 00000 000 00000 000

B87175184 817 .1617 4469398 .8188220029 .4116511
i e23800020 mm .8 owiaasnsi
0 000 0 0

O 00300L110 0&JJ0141 0501 02 002 1

0000000000001 00000 00 0000000000 00

l A19385194 81740.1617 880815fiflfl9534m82M55 324 2.
3.380002 14

mm
28
31
32
40
41
42
43
44
50
51
55
57
58
69
70
71
81
85
93
95
mo
m1
1m
1w
1M
H9
B1
B2
M5
M0
m2
m4
m9
U6
m1
B6
2M
2m
%0
n6
%1
%4
«M
W2
ms

57
CHAPTER 2 REFERENCES

. Wiley, W. C. and McLaren, I. H., Rev. Sci. Instrum, 26 (1955) 1150.
. Katzenstein, HS. and Friedland, S.S., Rev. Sci. Instrum, 26 (1955) 324.

. Holland, J .F.; Enke, C.G.; Allison, J .; Stults, J .T.; Pinkston, J .D.;
Newcome, B.; and Watson, J .T., Anal. Chem, 55 (1983) 997A.

. Bakker, J .M.B., J. Physics E: Sci. Instrum, 6 (1973) 785.

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

. Kinsel, G. R. and Johnston, M. V. "Post Source Pulse Focus' - A Simple
Method to Achieve improved Resolu‘gon in Time-of-Flight ass
Spectrometry" Presented at the 37 ASMS Conference on Mass
Spectrometry and Allied Topics, May 21-26, 1989.

. Yefchak, G.E.; Enke, C.G.; and Holland, J .F Int. J. Mass Spectrom. and
Ian Processes 87 (1989) 313.

. Studier, M.H., Rev. Sci. Instrum, 34 ( 1963) 1367.
. Press, William H.; Flannery, Brian P, Teukols S,aul A.; and Vetterling,

William T.. Num ri in :Th Art 0 en In
Cambridge University Press: Cambridge (1988).

10. Cricket Graph is a software product of Cricket Software, Inc., Malvern,

PA..

CHAPTER 3:
APPLICATION OF TIME-ARRAY DETECTION
To CAPILLARY COLUMN GC-TOFMS

INTRODUCTION

As previously discussed, time-array detection should have several
advantages over scanning detection schemes. Among these are the abilities
to accurately reproduce the chromatographic elution profile, to obtain
unskewed spectra of rapidly changing source concentrations as in gas
chromatographic efiluent, to improve the signal to noise ratio of the mass
spectral signal, to optimize the chromatography for reduced analysis times,
and to collect all the information generated during the analysis. This chapter
demonstrates that the time-array apparatus at Michigan State University
has these advantages and, hence, meets the demands placed on mass
spectrometers when used as detectors for high resolution capillary gas
chromatography. .

EXPERIMENTAL

The time-array detection (TAD) system used in this research is Shown
schematically in Figure 3. 1. It consists of a gas chromatograph, a timeoof-
flight mass spectrometer, and an integrating transient recorder.

58

59

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.1. Schematic diagram of the time-array detection system.

60
Gas Chromatography

A Hewlett-Packard 5790 gas chromatograph was used in the split
injection mode. A 22 m length of 0.25 um I.D. fused silica column coated with
0.25 mm SE-54 was used for the separations. The column was directly
interfaced to the mass spectrometer. Column temperature programming was
optimized to obtain the desired information in the minimum time.

Mass Spectrometry

The instrument used for this work was a CV C 2000 time-of-flight mass
spectrometer equipped with a conventional 2 m linear flight tube. This is the
same instrument that was modeled in the previous chapter. The signal fi'orn
the detector was amplified by a Comlinear Corp. E220 preamplifier prior to
processing by the ITR. The original potentiometers controlling the values of
time lag and ion focus voltage were replaced with 10-turn potentiometers to
improve the precision with which these values could be set. Calibration
curves for these potentiometers are included as Figures 3.2 and 3.3.

The CV C 2000 was originally equipped with a magnetic electron
multiplier (MEM) detector. This detector exhibited poor sensitivity and was
subject to noise and ringing fi'om the gating anodes. Removing the gating
electronics and grounding all extraneous plates improved the noise problems,
but sensitivity was still inadequate. It also became dificult to obtain
expendable components for these detectors. The detector was replaced with a
Galileo ETD-2003 channel plate electron multiplier (CEMA). A new flange
was ordered to fit the flight tube and modified to pass the required voltages
to the CEMA with 2 extra electrical feedthroughs for future use. Threaded

holes were included on the inner surface of the flange in case supports would

61

 

 

 

0 r I
0 200

j

I ' I ' I ' I
400 600 800 1000
DialSetting

Figure 3.2. Calibration curve for the time lag potentiometer.

62

 

 

i

 

I ' I ' 1 ' I
400 600 800 1000

BMW

- I
0 200

Figure 3.3. Calibration curve for the ion focus potentiometer.

63
be needed for future modifications. The voltage applied to the front surface of

the CEMA was obtained by running the high voltage line applied to cathode
number 1 (pin 13, -2837 V) from the old detector to the CEMA. Voltages for
the back surface of the CEMA were pulled from the line that had previously
supplied cathode number 2 (pin 11, -1000 to -1500 V) on the old detector. In
this way, the CV C electronics were able to provide power and an adjustable
gain to the new detector. This also permits control of the detector gain using
controls already on the instrument control panel. However, the gain control
now works in reverse with a setting of 11 providing the least gain and a
setting of 1 providing the maximum gain. Voltages applied to the rear
surface of the CEMA are listed in table 3.1.

Table 3.1. Gain Volta es Applied to the Rear
Surface of the EMA.

Earnings W
-995
-992
-1039
-1085
-1133
-1179
-1226
-1273
-1320
-1366
-1414

HI—I
Hommqmmswmu

Replacement of the MEM detector with the CEMA necessitated a
change in the mechanism for scanning the mass spectrum. The scanner
circuitry was disconnected and replaced with a model 162 boxcar averager
from EG&G Princeton Applied Research. Use of this boxcar enabled the
acquisition of time-slice (scanned) spectra.

The integrating transient recorder provides its own start pulse in order

to synchronize the data collection. It was necessary to modify the mass

64
spectrometer’s electronics (pulse 2 card) to accept such a signal and use it to

pulse the electron beam and extraction of ions from the source. An opto-
isolator was used in the circuitry to minimize noise on the trigger pulse that
could be carried between the instrument and the ITR. It was also desirable
to permit the instrument to occasionally operate fi'om its own clock when
performing time-Slice detection, so a switch was placed on the pulse 2 card to

permit this operation.

As mentioned in the previous chapter, ion focusing in time-of-flight
mass spectrometry is a function of mass. Selection of the appropriate focus
parameters will result in mass ranges in which acceptable ion signals can be
obtained. Unless stated otherwise, data in this chapter were obtained with
an optimum focus for m/z 71 (a time lag of 0.9 us). This value of time lag
provides acceptable signal for ions between 50 and 120 Daltons. Compounds
analyzed were generally low molecular weight species for which this mass

range was adequate.

Integrating Transient Recorder

The ITR was designed and constructed at Michigan State University.
As the ITR has been described elsewhere [1,2,3], it will be discussed only
briefly here. Signals from the mass spectrometer are sampled every 5 us by a
LeCroy TR8828B 200 MHz AID converter, dividing the spectrum into 16,000
time bins. This permits collection of the entire mass spectrum with suficient
separation of the 20 ns wide mass Spectral peaks. High speed emitter
coupled logic (ECL) circuitry is used to sum and store bettveen 10 and 30,000
successive transient mass spectra into one of two memory banks to collect

each spectrum. Meanwhile, the other bank passes data from the previous

65
summed scan to the disk. When the first bank has finished collecting data,

the roles of these two banks are switched, ensuring that all data generated
are collected.

Microprocessors on a VME rack are used to handle data transfer from
the ECL circuitry to a Priam SD10? 300 Mbyte hard disk, as well as operator
interaction with the ITR. The step which currently limits the mass spectral
production fiequency is the process of writing to the disk. Transferring
information from 5000 time windows to the disk limits the maximum scan
file production rate to 25 summed spectra per second. Through the use of
peak finding algorithms on parallel processors, the quantity of data written
to the disk is reduced, increasing the maximum scan file generation rate to
60 summed spectra per second. The operator has control over the number of
scans summed and therefore, the scan file generation rate. In addition, the
operator can adjust the instrument trigger frequency and hence cover large

mass ranges when needed.

REsUL'rs AND DISCUBSION

Mass Spectral Representation.

Qualitative information in GC-MS analyses is gained from the mass
spectra. It is therefore necessary to ensure that the quality of the mass
spectrum is preserved. Two features of data collection can influence this
quality. The first involves ion counting statistics, while the second occurs
from changes in source concentrations during acquisition of the mass
spectrum.

66
The mass spectrum is derived from concentrations of ions in the source

upon the application of the ion extraction potentials. Formation of ions in the
source is a function of the instantaneous concentration of molecules in the
source. Ion formation occurs over a short period of time, up to a few
microseconds of the operational duty cycle of the TOFMS instrument (0.1
ms). During the interval of ion formation, ions are confined to the potential
well formed by the electron beam [4]. The ion concentrations in the source
upon the initiation of ion extraction are therefore an integration of all
instantaneous ion concentrations during the ion formation process. During
the time required to extract ions from the source, ion concentrations do not
change perceptably. Hence, each transient signal from individual source
extraction pulses represents an unskewed mass spectrum. The spectrum
contained in each scan file produced by the ITR is a linear sum of these
unskewed transient spectra. Thus mass spectra collected by the TAD process
from any point on a chromatographic elution profile are identical within the
limits of noise. This is illustrated by data. in Figure 3.4 in which the
spectrum of n-decane is shown as acquired at the apex (Figures 3.4a) and at
the side of the elution profile (Figure 3.4b) at a scan file generation rate of 10
scans per second (1000 summed transients). An air background can be
observed in these figures at 28 and 32 Daltons. Except for this air
contaminant, mass spectral intensifies in these two figures agree within 13%.
Much of the error can be attributed to the increase in the background
interference in the spectrum collected on the side of the elution profile,

caused by the lower partial pressure of n-decane in the source.

The consistency of consecutively recorded mass spectra was assessed
under conditions of dynamic partial pressure of n-decane as well as under

conditions of constant sample pressure. The ratio of peak intensities at m/z

67

100 n

(a)

Intensity

l
l
50- f
i
I

Relative Intensity (%)

 

 

40 60

100 ‘ I

Relative Intensity (%)

 

 

 

 

 

 

40 60

 

 

 

 

 

Elution Time

 

 

 

 

Elution Time

71
86
99 142
ll “3 i

1 I i Fur I l
80 100 120 140
m/z

3.4. The mass (Emu-um of n-decane collected at (a) the top and Cb) on the
side of the matographic elution profile. Spectra were collected by

summing 1000 transients.

68
29, 71, and 142 relative to that at m/z 43 was determined in each of 14

consecutive summed spectra (1000 transients summed) collected across the
chromatographic elution profile of n—decane. The ratio of these peaks was
also determined in each of 14 consecutive summed spectra obtained from n-
decane at a constant source pressure. These ratios are presented in Table
3.2. Intensity ratios were consistent within less than 10% relative standard
deviation. Smaller values of relative standard deviation were observed for
intensities collected under conditions of constant sample pressure than from _
the chromatographic eluent. This is a result of decreased values of S/N as
the partial pressure is decreased. Mean relative intensities from these two

tests were within 1% of each other.

Table 3.2. Percent ion intensities relative to m/z 43
for successive scan files of n—decane
under dynamic and steady-state conditions.

WW WM
22 L1 1_42 2.9 7.1 .142.

18.61 19.62 5.15 19.53 20.71 5.74

17.83 20.47 5.62 17.92 20.37 5.88

18.60 21.40 5.06 17.53 19.30 5.30

18.76 19. 19 5.14 19.07 21.86 5.71

18.88 19.65 4.57 19.12 20.42 5.27

17.56 19.01 5.13 18.56 18.67 5. 12

17.03 21.58 4.94 19.86 19.38 5.19

17.65 20.18 4.98 20.03 18.17 5.31

19.21 21.33 4.80 18.19 19.59 4.67

18.25 19.46 5.44 17.88 17.48 5.88

18.29 19.96 5.29 18.26 19.58 4.49

18.85 19.74 5.22 17.81 18.42 5.22

20.03 20.41 5.14 17.48 20.86 5.84

20.06 21.47 5.43 18.30 20.31 5.55

mean 18.54 20.25 5.14 18.54 19.65 5.36
RSD 4.7% 4.4% 5.3% 4.6% 6.1% 8.0%

69
Spectra in Fig. 3.5 were obtained by summing different numbers of

transients while the pressure of n-decane in the source was held at 51:10'6
torr. These spectra illustrate the improvement in precision obtained by
summing additional successive transients. Quantization noise is apparent in
the spectrum from a single transient (Fig. 3.5a). This spectrum contains
many of the features of the reference n-decane spectrum, but there are
deviations in relative peak intensities due to the low overall signals; in fact,
at low analyte concentrations, a signal at a given m/z value may be missing
in any given individual transient. For example, peaks at m/z 29, 53, 99, and
127 are more intense than they should be, while peaks at m/z 32, 83, 84, and
98 are missing entirely from this spectrum, obtained fi'om a single transient.
With as few as 10 transients summed (Fig. 2b), the spectrum is noisy and not
all relative intensifies are consistent with the reference spectrum, but it has
all peaks in the reference spectrum are present. This spectrum corresponds
to a scan file generation frequency of 1000 Hz. In summing 10,000 transients
(Fig. 2c), features of the spectrum are not significantly altered but a clean
spectrum with a very high signal-to—noise ratio is obtained.

The improvement in signal-to-noise ratio should be proportional to the
square root of the number of transients summed as long as the noise is
random. A quantitative measure of the S/N ratio was performed by collecting
fifty repetitive spectra of n-decane at different numbers of summed
transients. The signal intensity was determined for several values of m/z.
The noise was determined as the variance in the signal intensity. Results
from these experiments are presented in Figure 3.6. While the value of SIN
increases significantly with the number of transients summed, it does not
follow the predicted square root relationship. This non-ideality stems from
the fact that the prototype version of the ITR produces a non-random noise

70

p
8
‘1
8

Wu [nun-it! (5)

 

 

 

 

 

 

 

 

 

Am. (c) Ff“
L i
i . "- ,. i

 

 

 

3.5. The mass spectra of n-decane collecmd b s ' (a) 1. (b) 10, and (c)
10,000 transient . This correspon to scan e generation
frequencies of 1 , 1000, and 1 Hz respectively.

71
component of the signal that is synchronized to the sampling of the AID

converter. This component of the noise, like the signal, is enhanced with
each summing. The signal-to-noise ratio can be expressed algebraically as:

SIN= A (two-5 / (B + 0 (mo-5 (3.1)

where A is the proprotionality constant for the signal, B is the proportionality
constant for the random component of the noise, C is the proportionality
constant for the non-random component of the noise, and n is the number of
transients summed. At low numbers of summed transients, white noise is
the major contributor to the noise in the signal. As the number of transients
are increased, the random contribution increases as a factor of the square
root of the number of transients summed while the synchronous contribution
increases proportional to the number of transients summed. When 2000 or
more transients are summed, the synchronous component becomes the major
source of noise in the data and the SIN ratio becomes constant. Hence,
summing more than 2000 transients with the current version of the ITR does
not provide any advantage in the signal-to-noise ratio. This can be observed
in Figure 3.6 as the leveling off of the curve at large numbers of transients

summed.

When both array and scanning systems are limited by white noise or
shot noise of the same magnitude, and data are collected over the same range
and at the same spectral generation frequency, a multiplex (Fellget’s)
advantage is obtained by array detection systems over scanning systems (5).
The multiplex advantage states that an increase in the SIN ratio proportional
to the square root of the number of resolution elements is observed for array
detectors relative to that obtained by scanning detectors as a consequence of
the fact that in the array system all resolution elements are being

72

 

 

0 1

 

" V U V

I r 1
6000 8000 10000

Number of Transients Summed -

I 1
0 2000 4000

Figure 3.6. The relationship between SIN and the
number of transients summed.

73
continuously monitored. Array and scanning detection systems using the

CV C 2000 TOFMS monitor the same signal from the preamplifier and hence,
the noise is comparable. Assuming that readout noise can be neglected, the
5000 time windows used in TAD to cover a mass range of 20 to 160 Daltons
would provide a SIN improvement of a factor of 71 over TOFMS instruments
which use a boxcar integrator with a 5-ns window to scan the spectrum.
Assuming that readout noise can be neglected, the 16,000 time-bins used in
TAD would provide a SIN improvement of a factor of 126 over scanning
TOFMS instruments which use a 5 ns window.

Mass Spectral Calibration

The capacity to collect spectra that are the linear sums of unskewed
transient spectra opens new avenues for data reduction algorithms. It is now
possible to subtract the contribution of one spectrum from another when
overlapping chromatographic peaks are encountered, without having to

account for skewed spectra caused by difl‘erences between steady-state and

dynamic spectra.

Interpretation of a mass spectrum includes the assignment of m/z
values to mass spectra. Output from the ITR lists ions by their flight times
rather than m/z values. It was necessary to write a program which could be
used to reduce ITR data to massointensity pairs. The program CV CMASS
was written in FORTRAN on a PDP-11I43 and later on a MICROVAX-2 to
perform this function. A listing of this program is included as appendix II of
this document. This program was developed from the following linear
_ relationship:

74
tof= k (m)0-5 + C (3.2)

where tof is the ion flight time to the detector, m is the ion’s m/z ratio, 12 is a
collection of constants, and C is an ofi‘set constant. CV CMASS can be used to
calculate the calibration parameters from data in a run file, or just to
transform the data to mass/intensity pairs using predetermined values of the

calibration parameters.

In the process of calculating the calibration parameters, CVCMASS
requires the input of three approximate flight times for known m/z values.
The program then searches the source file for all occurrences of these flight
times, within 30 ns windows, to determine the mean values of flight time for
each of these ions. Ion intensities below a threshold of 50 are ignored to
minimize noise. Values of k and C are then calculated from the solution of
simultaneous equations. Mean and standard deviation determinations are
made for each of these parameters and reported to the terminal. If the
standard deviation of either value is too large, an error was made in the
initial assignment of the m/z value. Shifting values of k over a period of time
are indicative of changing conditions in the source, such as contaminated
grids or unstable voltages. Shifling values of C are indicative of changes in
the pulsing of the source and can originate from either the trigger pulse on
the ITR, or one of the pulsing boards on the CV C 2000.

Once the calibration parameters have been determined, CV CMASS
uses these parameters to convert the flight times in the source file to mIz
values and writes the data to the destination file in a format that can be read
by other data reduction programs written during this research. While
working on this program, it was noticed that at random intervals mass

assignments of peaks in the mass spectra were too low and flight times for

75
these erroneous assignments were always multiples of 80 ns low. Once a

shift occurred in a mass spectrum, all peaks at higher m/z values in that
spectrum were also shifted. CVCMASS was modified to print a warning
message when peaks were outside of a 0.3 Dalton window centered at integer
values. Using this feature of the program it was determined that this shift
occurs randomly in 2 to 18 percent of the spectra collected. No correlation
was observed between the rate at which data were written to the disk and
the probability of a shift: in the flight time. The problem is most severe when
large mass ranges are stored to the disk. The source of this problem appears
to arise fiom a counter in the ITR which is used to locate the next time bin to
read out to the disk. It appears that this counter is being incremented
randomly. Attempts to observe an extraneous incrementing pulse have thus
far been fruitless.

Representation of the Chromatography.

Information desired from a GC-MS analysis is usually centered around
the identification of chromatographic eluents. A reconstructed
chromatographic profile can be used to locate the best spectra to represent
eluting species. The chromatographic representation also contains some
quantitative information about the eluents. Insufficient sampling
frequencies result in a distortion of the reconstructed chromatographic profile
relative to the true elution profile and hence, a loss of desired information [1].
Mathematical evaluation of the chromatographic profile requires that as
many as 100 samples be collected for each chromatographic peak (6)
depending on the degree of accuracy desired.

76
The capacity to accurately reproduce the chromatographic profile from

a limited number of mass spectra is illustrated in Figures 3.7 and 3.8. These
figures were collected using difi‘erent gas chromatographs under similar
chromatographic conditions. The FID chromatogram of a charcoal lighter
fluid is shown in Figure 3.7 and represents the analog chromatogram,
unrestricted by the bandpass of the data collection system. The time-array
detection system permits the selection of sampling frequency by providing
control over the number of transients to sum. Reconstructed chromatograms
were obtained from separate injections of charcoal lighter fluid with difi‘erent
numbers of transients summed per scan file. These reconstructed
chromatograms are included as Figure 3.8. Difi‘erences between Figs. 3.7 and
3.8 are due in part to the non-specificity of the flame ionization detector (Fig.
3.7) while mass spectrometers are a little more specific. In addition, subtle
differences in the chromatographic conditions between the two instruments
can alter the chromatographic separation process. Chromatographic peak
widths range from 3 to 4 seconds at baseline. The apparent chromatographic
resolution increases as the sampling frequency increases from 1 to 5 to 10 Hz
(3 to 15 to 30 samples across the peak profile). Relative chromatographic
peak heights do not significantly change when mass spectral generation
frequencies above 5 Hz are used. In addition, the signal level in the
reproduced chromatogram decreases as the mass spectral generation
frequency is decreased, which causes an associated decrease in the SIN ratio.
For each different chromatographic condition, it is possible to optimize the
number of transients summed per scan file to provide the maximum SIN in

the mass spectra while maintaining adequate chromatographic resolution.

77

 

 

 

 

 

 

 

r I I T r I
0 1 2 3 4 5
Elution Time (min)

3.7. An FID chromatogram of charcoal lighter fluid injected onto a 22 m by
0.25 mm fused silica capillary column coated with 0.25 pm SE-54. The
G0 was temperature programmed fi'om 100°C to 150°C at 10°C/min.

78

 

 

 

 

 

 

 

 

 

 

 

(I)
5‘
1 animal? ' ' I no
1.0 go 3.0 ‘0 (b)
magnum
(c)

1 aluminum 460 900
1.0 2.0 3.0 4.0

 

 

 

 

1 Sunfloflubcm mm
1.0 go an 40
”The“

3. 8. Reconstructed chromatograms of lighter fluid fi-om scan file generation
frequencies of (a) 1 Hz and (b) 5 Hz. The GC was temperature
programmed from 100°C to 150°C at 10°C/min.

79
Problems associated with inadequate sampling frequencies when using

scanning instruments can be avoided by operating in the selected ion
monitoring (SIM) mode in which ion current at only one (or a few) mIz values
is monitored. Of course, when SIM is used, informafion about ion current at
all other m/z values is sacrificed. Because fime array detecfion permits the
collection of complete mass spectra, all mass spectral information generated
is retained. The ITR integrates the ion current in each 5-ns fime window
from successive transients in much the same way as do conventional TOFMS
instruments when the boxcar integrator is not scanned (SIM mode), thus,
detection limits for complete spectra obtained by TAD should be the same as
those otherwise achievable in TOFMS only by SIM.

The method of chromatographic reconstruction used in this work
differs from that used in conventional total ion chromatograms.
Convenfionally, the sum of all ion intensifies in each scan is used to create a
single point on the reconstructed chromatogram. The algorithm used in this
work saves fime in the data reducfion by plotfing the difference between the
highest and the lowest values in each scan file. The resulfing "DIFFerence
plot" is sfill usefiil in locafing the desired scans, but relafive peak intensifies
are not as analyfically useful as in the convenfional technique.

In order for TAD to become a viable detecfion method for GC-MS,
detecfion limits by TAD need to be comparable to those found using other
mass spectrometers. This could be a problem using the CV C 2000 since a
pulsed source is used in which ions are only formed for 1 to 4 us out of the
100 us duty cycle. In addifion, ions are only extracted out of the source once
every duty cycle. This is in contrast to mass spectrometers with confinuous

ionizafion and extraction and will result in a decrease in signal intensity.

80
To increase the sensifivity of TOFMS, the electron control slit was

grounded permitfing confinuous ionizafion in the CV C 2000 source. This
experiment should have increased the number of ions in the extracfion
packet since ions are trapped in the electron beam [4]. The response at the
detector should increase by as much as a factor of 20. In addifion, the efi‘ect
of metastable decomposifions after ions leave the source should be reduced
since decomposifions and ion-molecule collisions are likely to occur while the
ions are being stored [7]. Signal intensifies were determined to be a factor of
4 higher than found using a 4 us ionizafion time. This indicates that some
ion storage did occur, but also that ions were able to leave the extracfion
volume prior to applicafion of the extracfion pulse.

Serial dilufions of toluene in hexane were used to determine detecfion
limits while using confinuous ionizafion. A "DIFFerence plot" of an 11 ng GC
injecfion of toluene is included as Figure 3.9. At a SIN rafio of 2, the
detecfion limit was determined to be 3 ng based on measurement of the peak
intensity at m/z 91. This is comparable to detecfion limits found by scanning
a quadrupole instrument (Hewlett-Packard 5985A) at 2 scan files per second.
Operafion of the TAD system at the same scan file generafion frequency as
the quadrupole instrument is performed by increasing the number of
transients summed, and hence, would result in lowerdetecfion limits by TAD
for similar sample generafion frequencies. In addifion, the sensifivity of
quadrupole instruments has improved over many years of instrument
evolufion, while comparitively little work has been done on pulsed TOFMS
sources towards such optimizafion. The sample-use duty cycle of the pulsed
TOFMS source is often so low that the detecfion limits achieved are worse
than those of confinuous beam mass filter mass spectrometers. One way to
improve the signal-use duty cycle is to store ions created between extracfion

Intensity (AID Counts)

81

251

14c

 

 

 

 

 

4 MA

1 Scan Number 75 150
70 85 Elufion Time (s) 100

Figure 3.9. Reconstructed chromatogram of 11 ng injecfion of toluene
collected at a scan file generafion fi-equency of 5 Hz.

82
pulses. This inifial attempt resulted in a four-fold improvement in sensifivity

over pulsed ionizafion. Further work should produce sfill better results,
potenfially to much better levels than exhibited by scanning filter mass
spectrometers. Note that since Figure 3.9 is a DIFFerence plot the intensity
is that for the most intense peak in the spectrum, m/z 91, not the total ion
intensity, so informafion from the rest of the spectrum is sfill available. TAD
ofi‘ers the opfion to search the data for just the m/z 91 peak in the spectrum
of toluene, reducing the background noise and lowering the detecfion limit.
Scanning instruments suffer from insufficient sampling frequencies when
using this quanfitafion method, but such a problem does not eifist with TAD
since data collected represent rapid changes in source concentrafions.

Speed of Analysis.

When the scan file generafion rate is not a limifing factor, the
chromatography can be opfimized for speed of analysis, greatly reducing the
fime needed to perform an analysis. The shorter analysis fime results in a
corresponding increase in the chromatographic peak heights, compensafing
for the loss in S/N from the requisite higher scan file generafion frequencies.
Typical GC-MS analysis fimes for these types of mixtures are on the order of
12 to 60 minutes (8). The fime of analysis can be reduced by altering any of
several variables, such as the temperature, column length, or carrier gas flow
rate. The reconstructed chromatograms of Fig. 3.8 were collected from a
charcoal lighter fluid sample in less than 4 minutes. These reconstructed
chromatograms were collected by temperature programming the GC oven
from 100° to 150°C at a rate of 10°C/min on a 60 m SE-54 column.
Chromatograms of gasoline have been obtained under similar condifions
through the trimethylbenzenes in less than 2 minutes. While these

83
condifions are not advisable for the separafion of small hydrocarbons, the

xylenes and other aromafic isomers are adequately separated. In addifion,
the use of cold trapping inlets with capacitafive heafing has been reported to
separate nine of the major components in gasoline in as little as 2.5 seconds

(9).

An example of the scan file generafion rate possible with TOFMS-TAD
is illustrated in Fig. 3.10, which is a segment of a reconstructed total ion
chromatogram representing the analysis of gasoline by capillary column GC-
MS. The major peak in Fig. 3.10 is due to toluene and corresponds to an
injecfion of about 4 ug into the mass spectrometer source. Scan files used to
generate this profile were collected at a rate of 20 spectra per second.
Fourteen scan files were collected during the elufion of toluene, all of which
are readilly recognizable as toluene mass spectra. This peak is only 0.7
seconds wide at baseline. The clean peak shape of the reconstructed
chromatographic profile shows that peak elufion fimes and areas can be
accurately determined at this scan file generafion frequency. Without the
high sampling frequency offered by TAD, it would have been dificult to
collect accurate qualitafive and quanfitafive informafion for this component
by mass spectrometry in a single GC run.

Reducing the fime of analysis increases the probability of occurrence of
overlapping chromatographic peaks. While this increase is undesirable,
many applicafions exist in which it can be tolerated. Addifionally, mass
spectrometry offers the possibility of deconvolufing overlapping
chromatographic peaks when there are unique ions in the spectrum of the
individual components (10). Inconsistent mass spectra, as in skewed spectra,

complicate deconvolufion algorithms by requiring that correcfion factors or

84
tolerances be considered along with the mass spectral intensifies. Because

TAD offers the advantage of unskewed mass spectra across the enfire
chromatographic elufion profile, better success in the applicafion of
deconvolufion and pattern recognifion algorithms can now be expected.

85

1195 "'

 

 

 

 

 

 

 

 

 

 

 

 

 

E?

i

O

t: 598 -

.53

9°
1 ‘ . a
90 Scan Number no 130
645 65.5 Elufion Time (s) 66.5

3. 10. Reconstructed chromatogram of the toluene peak in unleaded gasolene.
chtra were collected at 20 Hz. The collecfion of each spectrum is

represented by the verfical bars.

86
CHAPTER 3 Rumors

. Holland, J .F.; Enke, C.G.; Allison, J .; Stults, J .T.; Pinkston, J .D.;
Newcome, B.; and Watson, J .T.. Anal. Chem. 55 (1983) 997A.

. Allison, J .; Holland, J .F.; Enke, C.G.; and Watson, J .T.. Anal. Instrum, 16
(1987 ) 207 .

. Holland, J .F.; Tecklenburg, R.; et.al.. Submitted to Rev. Sci. Instrum..
. Studier, M.H., Rev. Sci. Instrum, 34 (1963) 1367 .

. Ingle, James D. and Crouch, Stanley R.. Spectrochemical Analysis.
Prenfice Hall: Englewood Cliffs, New Jersey (1988) Pp. 159-161.

. Chesler, Stephen N. and Cram, Stuart P.. Anal. Chem, 43 (1971) 1922.

. Hezgd, AA. and Harrison, A.G.. Int. J. Mass Spectrom. Ion Phys., 4 (1970)
41 .

. Holzer, G. and Bertsch. Amer. Lab, (Dec. 1988) 15.

Lanning, L.A., Sacks, R.D., Mouradian, R.F., Levine, SR, and Foulke,
J .A.. Anal. Chem., 60 (1988) 1994.

10. Sharaf, Muhammed Abdallah and Kowalski, Bruce R.. Anal. Chem, 54

(1982) 1291.

CHAPTER 4:
RECONSTRUCI‘ED CHROMATOGRAMS BASED ON
MAss SPECTRAL DEGREE-OF-FRAGMENTATION

INTRODUCTION

Time-array detecfion permits an accurate reproducfion of the
chromatographic elufion profile, but only when high scan file generafion
frequencies are used. However, these high frequencies result in large
numbers of mass spectral data files for interpretafion. For example, a 1hr
gas chromatographic separafion sampled at a frequency of 20 Hz results in
the producfion of 72,000 mass spectra; only a limited number of these contain
informafion that is analytically useful. The fime needed to interpret these
spectra could be reduced through the use of computer algorithms which filter
data based on a common characterisfic of the species of interest prior to
interpretafion. In this manner, only spectra of interest for the analysis at
hand will be flagged. All collected data are retained in case they are needed
for other purposes.

The most commonly used prefiltrafion algorithm involves the
production of a reconstructed total ion chromatogram[1]. This algorithm
involves passing the sum of all ion intensifies in the spectrum to the plot file.

In this manner, a non-selecfive response is obtained for each

87

88
chromatographic eluent. If the enfire mass range is collected, if the electron

cross secfions of all neutral molecules is idenfical, and if all ions produce
idenfical gains at the mulfiplier, this algorithm will generate chromatograms
in which the molar response for one compound, as indicated by the intensity
of its chromatographic peak, can be used to quanfitate all other components
based on their corresponding peak intensifies. When these condifions are
approximately valid, relafive peak intensifies can sfill provide a rough
quanfitafion of chromatographic eluents.

The prefiltrafion algorithm used on the ITR, DIFFerence plots,
involves passing the difi'erence between the most and least intense points in
the spectrum to the plot file. This algorithm responds to all chromatographic
eluents, but in a manner that is a funcfion of the eluent concentrafion and
the number of peaks in the spectrum. This method enables the rapid locafion
of eluent spectra in the data base, but sacrifices quanfitafive informafion in
the reconstructed chromatogram.

It is often beneficial to obtain more specificity than can be obtained
from either of these algorithms in order to further reduce the number of mass
spectra that need to be interpreted. The algorithm most commonly used to
provide this added specificity involves the producfion of a reconstructed mass
chromatogram [2]. This method involves a search of the data field for a
specified m/z value. When such a value is found in a spectrum, the intensity
of that m/z value is passed to a chromatographic plot file. If the selected m/z
value is not in the mass spectrum, a 0 is passed to the plot file. This
algorithm is useful for monitoring classes of compounds with common
fragment ions, such as phthalate esters (m/z 149) or alkanes (m/z 43 and 57).
Selection of ions for the search can be made from tabulafions of structural

89
correlafions [3]. This algorithm is also used to locate spectra of

chromatographic eluents with a parficular molecular weight by monitoring
the ion current at the m/z value that correlates to the desired molecular ion.

Now that some tandem mass spectrometric methods can be performed
on the fime scale of capillary chromatography [4], an alternafive selecfive
filtrafion method is becoming feasible which is based more on ion chemistry
than on ion mass. The GC-MS/MS data space can be searched for
characterisfic m/z ion daughters or losses. Intensifies of qualifying ions can
then be passed to the chromatographic plot file. In this way, chlorinated
hydrocarbons could be idenfified by their characteristic losses of m/z 36 (HCl)
or m/z 70 (C12). These filtrafion methods provide an advantage over
reconstructed mass chromatograms in that they bestow addifional chemical
informafion about elufing species. 4 ’ However, they require specialized

instrumentafion as well as advance informafion about the sample matrix.

Situafions often arise in which few common features are available in
the mass spectra of chemically related compounds. Examples include the
analysis of aromafics [5] and polyaromafics [6,7] where the electron impact
mass spectra are sparsely populated and the limited features that are
present have few structural associafions. One feature these spectra do have
in common is their paucity of peaks, a direct result of the strength of

resonant bonds in aromafic species [8].

The lack of common spectral features in mass spectra of highly
aromatic species limits the usefulness of either of the two previously
menfioned specific filtrafion algorithms. A desire has been expressed for
analyfical methods which amplify signals from these structures while
eliminafing signals that interfere with their detecfion [4]. An example of

90
where these methods are needed is in the analysis of combusfion residues

where aromafic species make the largest contribufion to the toxicity [5,6] or
can be used to idenfify the accelerant used to inifiate combusfion [7,9]. In
each of these cases, only a few of the hundreds of chromatographic peaks
represent aromafic consfituents of the sample matrix. It could prove
beneficial for these applicafions to use an algorithm which examines mass
spectral data based on the profusion of peaks and their distribufion in each
spectrum. Such a selecfive filter would permit mass-independent
discriminafion of species that have a large number of fi'agment ions, such as
alkanes, from those with only a few fragment ions, such as polycyclic
aromafic hydrocarbons (PAH), inorganics, and molecules that easily fragment
via a single pathway to a stable ion, such as nitroaliphafics.

Two algorithms have been examined as part of this effort for use in
filtering GC-MS data based on the degree-of-fragmentafion of ions. The
foundafion for the first algorithm is the fact that highly aromafic compounds
have most of their mass spectral intensity confined to only a few peaks. The
second algorithm is based on the number of peaks in the spectrum, which is
low for aromafic compounds [10] and high for aliphafics.

EXPERIMENTAL

Algorithms written to generate reconstructed chromatograms are
included in the program GCSIM.C in Appendix III. This program was
written in C on a MicroVAX II. It takes output from CV CMASS.FOR
(Appendix II) and creates a data plot file for reconstructed TII, DIFF, mass,
and degree-of-fi-agmentafion chromatograms. The resultant data plot file is
compatible with CricketGraph [11].

91
A data set was created from normal (Gaussian) distribufions of five

different mass spectra, representafive of a GC-MS data field, to test the
filtrafion algorithms. Mass spectra chosen are presented in Figure 4.1.
Benzene and anthracene (Figures 4.1b and 4.1c, respecfively) were chosen as
representafive of aromafic eluents. Spectra of acrylic acid and
octylmercaptan (Figures 4.1a and 4.1d, respecfively) were selected as
representafive of spectra with much fi'agmentafion. Fifty successive spectra
(scans) were created using evenly spaced distribufions of varying amplitudes
and variances fi'om the mass spectra of these four compounds to simulate
data from a GC-MS analysis. Variances for individual spectra were selected
to produce two separated and two unseparated peak shapes, as shown in the
T11 reconstructed chromatogram (Figure 4.2). In addifion, an air
contaminant (Figure 4.1e) was introduced throughout the data set to
simulate a high background by selecfing a large variance for the distribufion

of this spectrum.

The dependence of the DIFFerence plot algorithm on the number of
peaks in the spectrum is illustrated in Figure 4.3. Signal intensity is reduced
relafive to the TII plot (Figure 4.2) because fewer m/z intensifies contribute
to the filtered signal intensity. Responses of acrylic acid (A) and
octylmercaptan (D) are attenuated relafive to those of benzene (B) and
anthracene (C). This is a consequence of the fact that the peak intensity in
the attenuated species is distributed among many mass spectral peaks, while
in the aromafic compounds it is distributed among only a few peaks. The
reconstructed TII chromatogram (Figure 4.2) produces peaks whose area is
approximately proporfional to concentrafion. The DIFF reconstructed
chromatogram contains peaks whose intensity is proporfional to the

concentrafion of the eluent, the number of peaks in the mass spectrum, and

92

100'
80‘
60"

40"

 

 

20‘

 

 

o- ‘vv‘rfiIvvvv' vvvvvvvv

10 35 60

80"

40-

WM
0- ' I

 

 

 

10 30 50 70

(c)

100-
80"

6°-
4

{.5‘
E
.5
E
.s
a?
i
3
5°”
i
i
'5
E

 

 

 

 

 

 

40" '
1
20-
d
i I l

v I v I ' l 'W ‘
sum/20 130 150 170 190

 

 

 

3 1o 35 so as mIz° 135 185
d
3 ..., < )
i “-
.5 “‘1
0 4o-
.3 .
.3 2°“
a t
01 .1.-,...
10 35 60 85 W20 135 160 185
i “-
.5 6°‘
0 4o;
.3 1
.23 2°“.
3% o 'anmfiqnv. ..
185

10 35 60

4.1. Mass spectrausedinthedatasetfortes

as m/z° 135

d

egree-of-fragmentafion

a1 rithms. Spectra include (a) acrylic acid, (b benzene, (c) anthracene,

(d octylmercaptan, and (e) air.

93

 

 

5000 -

4000 T
.3’
g 3000 -‘
i
g 4
.9.
a 2000 .
c
E!

1000 '-

A B c D
o I r I I I I
0 10 20 30 40 50
Sean Number

4.2. Total ion intensity reconstructed chromatogram of the test data set.
Chromatographic peaks are (A) acrylic acid, (B) benzene, (C) anthracene,
and (D) octylmercaptan.

 

 

 

 

 

1000 F

800 '-
A
c:

e: 600 -
v
III!
I
”:3

E 400 "
v
H

zoo -

A B C D
o I f I ‘ I I I
0 1 0 20 30 40 50
Scan Number

4.3. DIFFerence reconstructed chromatogram of the test data set.
Chromatographic peaks are the same as listed in Fig. 4.2.

95
the distribufion of intensities among the mass spectral peaks. Quanfitafive

informafion is not as readily available through the use of this algorithm.

The reconstructed mass chromatogram algorithm was used to search
the test data set for monosubsfituted phenyl rings (m/z 77). The resulfing
chromatogram is presented as Figure 4.4. The benzene peak (B) is passed
through the filter while all other components are discriminated against.
Anthracene has three benzene rings, but there are no common ions between
benzene and anthracene that would permit both compounds to pass through
this filter.

"GENERIC SIGMA" ALGORITHM

In much of the older mass spectrometry literature, the percentage of
total ionizafion, %2, was used as a measure of the intensity of individual
peaks [12]. This value is the ion’s peak intensity relafive to the sum of all
peak intensifies above a selected m/z value reported as a percentage. As the
sum of all m/z peak intensifies approaches the intensity of the peak of
interest, the percentage of total ionizafion approaches 100. When all m/z
values in the spectrum are considered, those compounds that produce most of
their intensity at only a few m/z values will have high values for percent total
ionizafion at those m/z values while compounds with much fragmentafion

will have low values for the percent total ionization.

By rafioing the intensity of the most intense peak in the spectrum, the
base peak, to the total ion intensity of the spectrum, a value can be obtained
that is indicative of the degree-of-fragmentafion of the molecular species.
This rafio has been designated the "generic sigma" value (2g) because of its
similarity to the percent of total ionizafion axis. In addifion to giving

 

 

 

 

300 -
200 -
g
i
r~
r~
N
E 100-
A B C D
0 ' ' ' I ' I ' I ' i
o 10 20 so 40 so
ScanNumber

4.4. Reconstructed mass chromatogram of the test data set for m/z 77.
Centroids of the chromatographic peaks are labled as in Fig. 4.2.

97
informafion concerning the number of different species that can be formed in

energefically stabilizing the ion, the generic sigma value would also provide
informafion on the stability of major fragment ions. In this way, some
structural informafion can be obtained.

A database of more than 2600 compounds from Co to 012 has been
derived from published spectra in reference 13 in an attempt to demonstrate
the ufility of the algorithms developed in this chapter. This database is
included as Appendix N. The database includes calculated values of generic
sigma (28) from reference spectra, and is sorted on the number of carbon
atoms in the molecule and the generic sigma value. As predicted, species
that are highly aromafic have high values of generic sigma while those with
largely aliphafic characterisfics have low values of generic sigma. Reference
spectra used to generate this database were accumulated fi-om a wide variety
of mass spectrometers. Work performed by D. Guido [14] has demonstrated
that spectral intensifies obtained using the cold, open source in the CV C 2000
TOF mass spectrometer differ from spectra obtained using heated, confined
sources on other instruments. In addifion, the intensity dependence of TOF
mass spectra on the fime lag that was demonstrated in Chapter 2 of this
document can cause a shifiing of the generic sigma values in Appendix IV
that is dependent on the focus parameters used. These facts limit the ufility
of this database for TAD data, but the database can sfill prove useful in
determining relafive values. I

A plot the generic sigma value rather than the total ion intensity
against scan number would provide a means of prefiltrafion which would be
independent of the chromatographic intensity; Such an algorithm would be
useful in locating both trace and major components in the chromatographic

98
data space, at the expense of any quanfitafive informafion. However, when

dealing with trace components, it is likely that porfions of the spectrum
would be of low enough intensity that they would not be recorded. This
would increase the value of generic sigma calculated from the GC-MS data
file. In addifion, a background of air, everpresent in vacuum systems, would
result in high values of generic sigma for background and lower values for

eluting components.

The generic sigma algorithm was used to produce the reconstructed
chromatogram of the test data set in Figure 4.5. Intensifies of the base peaks
and total ion intensifies were determined for each scan file and used to
calculate a generic sigma value. Air, which has a high value of generic
sigma, is present in the background and causes a high baseline from which
depressions indicate the presence of eluents. A broadening of the acrylic acid
peak (A) is observed as a result of the gradual increase in the number of
peaks as the acrylic acid components grow into the air spectrum. This efi'ect
is also responsible for the wings present on the benzene peak (B). The
apparent chromatographic resolufion is significantly degraded as a result of
the overlap of chromatographic components as seen in the response to species

B, C, and D.

One means of reducing the contribufion to peak broadening from trace
contaminants in the mass spectrum is to use threshold values of the generic
sigma as a window in which to pass the ion intensity to the plot file, in much
the same manner as is done with the reconstructed mass chromatogram
algorithm. Figure 4.6 was obtained by passing the total ion intensity to the
chromatographic plot file for all scans where the value of generic sigma was
between 30 and 60 percent. As with the reconstructed mass chromatogram

99

 

 

 

 

 

 

 

80 n
60 -
q; C
€13
>
.3 40 -
(I)
.0
5 .
g
20 -
0 v I V I ' l ‘ I ' l
0 10 20 30 40 50

Scan Number

4.5. Generic sigma reconstructed chromatogram of the test data set.
Centrords of the chromatographic peaks are labled as in Fig. 4.2.

100

 

 

 

 

 

 

 

 

2500 -
2000 -'
g 1500 -
i
::
£3
3 10001
e
E1
500 -
A B c D
0 ' I ' I ' r ‘ I I I
o 10 20 30 40 so
Scan Number

4.6. Total ion intensity reconstructed chromatogram of the test data set in
which only compounds in which the generic sigma value is between 30
and 60 percent are plotted. Labels marking the centroids for
chromatographic peaks are for (A) acrylic acid, (B) benzene, (C)
anthracene, and (D) octylmercaptan.

101
algorithm, values outside of the preselected range resulted in a 0 being

passed to the plot file. Under these condifions, the aromafic components are
passed to the chromatographic plot file while the aliphafic components are
totally discriminated against. The skewed chromatographic peak for
anthracene is a result of the overlap with the octylmercaptan peak which
results in a decrease in the value of the generic sigma, pushing it outside of

the acceptable window.

Problems seen fi-om Figures 4.5 and 4.6 have demonstrated several
requirements for the successful use of the generic sigma algorithm. Complete
chromatographic separafion of the sample matrix needs to be obtained.
Mixture spectra from overlapping components will reduce the contribution of
the base peak to the total ion intensity, thus lowering the generic sigma
value. Because this algorithm is based on intensifies, the mass spectral
integrity needs to be maintained throughout the chromatographic elufion
profile or else a wider acceptance range of generic sigma values will have to
be used. Such a requirement necessitates the use of TAD. However, the
dependence of relafive intensifies in convenfional TOFMS on the focus
parameters limits the use of this algorithm to data that have been collected
under the same experimental condifions as the spectra used for the data

base.

"NUMBER or PEAEs" ALGORITHMS

Many of the dificulfies with the generic sigma algorithm are due to its
dependence on mass spectral intensifies. To get around this dependence, a
second algorithm was examined which was a funcfion of the number of peaks
in the spectrum. Entries have been added to the database in Appendix IV to

102
incorporate the number of peaks above 5% (NP5) and 25% (NP25) of the base

peak intensity.

Simply counfing the number of peaks in the spectrum will provide
informafion as to the degree-of-fragmentafion, but will not eliminate the
difliculfies caused by noise that were observed using the generic sigma
algorithm. Instead, to minimize this problem, a peak threshold was used
which required peak intensifies to exceed 5% of the base peak. In this
manner, contribufions to the spectrum from minor components and noise

were reduced while only minor peaks in the spectrum were ignored.

A reconstructed chromatogram based on the number of peaks in the
spectrum of each scan is included as Figure 4.7. Little difference can be
observed in this figure among the responses of the first three compounds.
Only octylmercaptan stands out from the rest as having many more peaks in
its spectrum. By providing a small window of acceptable number of peaks in
the spectrum, the octylmercaptan could be discriminated against, thereby
reducing the number of components that require interpretafion.

The wing on the left shoulder of the octylmercaptan peak in Figure 4.7
is due to the increased number of mass spectral peaks observed when
significant quanfifies of anthracene are present. Likewise, wings and
broadening in the other peaks can be associated with the overlap of elufing
component and traces of air, contribufing to a larger number of peaks in the
spectrum. This problem is most severe when low concentrafions of elufing
species are present and hence their contribufion is limited to the shoulders of
the peaks. Figure 4.8 illustrates that the selecfion of the threshold for peak
definifion is crifical. In this figure, a threshold of 25% of the base peak

intensity was used. The aromatic species can now be easily disfinguished

103

so}

fly

40‘

 

tsinHflneSkxxirunn
run
a

fin
(Rneaterthmuififltfhuxafkuflilnme

0

hhnaflxnnoff’

 

 

 

0 v I ' l ' ' ' ' ' I
(T 10 20 3° 4° 5°
Ekwurhhnaflxnr

4.7 . .N umber of peaks reconstructed chromatogram with a 5% of base peak
intensity threshold. Labels marking the centroids for chromatogaphic
)

peaks are for (A) acrylic acid, (B) benzene, (C) anthracene, and (
octylmercaptan.

104

50-

40-

e Spectrum
Peak Intensity

Number Of Pomts 111 th
Greater than 5% Base

30-

 

 

 

 

I I I I I I I I ' fl

Scan Number

' f base peak
. . be of eaks reconstructed chromatogramwrth a 5% o .
4 8 irlllgnmsit; thrpeshold. Labels marking the centrords for chromatographic

peaks are for (A) acrylic acid, (B) benzene, (C)‘anthracene, and (D)
octylmercaptan.

105
from the aliphafic consfituents, but it is difficult to discern the presence of

anthracene from the background.

As was done with the generic sigma algorithm, it is possible to pass the
total ion intensity to the plot file for those scans in which the number of
peaks fall within a predetermined window. This was done to generate data
for Figure 4.9. A lower threshold of 3 peaks was used to eliminate the
contribution from the air background. An upper threshold limit of 15 peaks
was used to permit aromafic species to pass through the filter while
discriminafing against non-aromafic components. All scans that did not fall
into this range resulted in a 0 value being sent to the plot file. Since the
overlap region of anthracene and octylmercaptan result in spectra that
exceed the permissible number of peaks, the anthracene peak is skewed.
This algorithm discriminates against compounds with many peaks in the
mass spectrum, such as octylmercaptan. In this way, a concentrafion-
dependent means can be achieved to filter out some of the undesired
informafion. However, Figure 4.9 shows that acrylic acid, which has a
similar total number of peaks in its spectrum to that of the aromafic
components, is passed through the filter along with the aromafic components.

Care must be taken in the selecfion of threshold values when using
these algorithms based enfirely on the number of peaks in the spectrum.
This is illustrated by Figure 4.10. With a minimum threshold of 2 peaks, the
air background is not discriminated against. A maximum threshold of 8
peaks was insumcient, resulfing in a discriminafion against all elufing
components. Only the shoulders of eluting components pass through the
filter.

106

 

 

 

 

 

 

 

 

2500-
2000-
g 1500-
I:
a 1
3 1ooo~.
o .
5..
500-
A s C D
0 ' I ' I ' I V I ' ““I
0 10 20 30 40 50
ScanNumber

4.9. Total ion intensity reconstructed chromatogram of the test data set in
which only compounds in which the number of peaks (5% threshold) is
between 3 and 15 percent are plotted. Labels marking the centroids for
chronmtographic peaks are for (A) acrylic acid, (B) benzene, (C)
anthracene, and (D) octylmercaptan.

107

 

 

 

 

 

 

 

400-
A
300-
:5
33
200'
5
H
:g M
e J
100“
s c D
o . , . I r I , ' . fl
0 10 20 30 4O 50
ScanNumber

4.10. Total ion intensity reconstructed chromatogram of the test data set in
which only compounds in which the number of peaks (5% threshold) is
between 2 and 10 percent are plotted. Labels marking the centroids for
chromatographic peaks are for (A) acrylic acid, (B) benzene, (C)
anthracene, and (D) octylmercaptan.

108
COMBINED ALGORITHM

Algorithms that have been examined to this point have shown some
promise in discriminafing against species which undergo much
fragmentafion under electron ionizafion condifions, but problems sfill exist.
The generic sigma algorithm suffers from too much dependence on the ion
intensifies to be useful for convenfional TOFMS. While the "number of
peaks" algorithms are promising, they do not difi‘erenfiate between spectra
with one major peak and many smaller peaks, as in aromafic compounds, and
spectra that contain many large peaks, as the acrylic acid. The best features
of both algorithms have been combined to develop an algorithm which will
meet the objecfive. This combined algorithm calculates the rafio of the
number of peaks whose intensity is greater than 5% of the base peak to the
number of peaks whose intensity is greater that 25% of the base peak
intensity. Compounds that undergo much fi'agmentafion will have rafios
around 1. Species that do not have many fragment ions in their spectra will
have high rafios.

The rafio calculated as part of this combined algorithm has been
included in Appendix IV under the heading R for rafio. Figure 4.11 is a plot
of a reconstructed chromatogram of the test data set using this combined
algorithm. Only the aromafic components show significant response in this
chromatogram. Wings are sfill present on the shoulders of the peaks. As the
concentrafion of contaminant species start to increase, the number of peaks
in the spectrum will go through a corresponding increase, but the number of
peaks above 25% of the base peak will not change significantly unfil the
contaminant becomes a major component. This will cause an increase in the

calculated ratio. For those compounds with many peaks in the spectrum, this

109

 

 

 

 

Scan Number

4.11. Reconstructed chromatogram of the test data set using the combined
algorithm.

110
increase will be insignificant. However, for compounds with only a few peaks

in the mass spectrum this phenomenon is easily observed. The increase
becomes noficeable when a species with only a few peaks is contaminated by
one with many peaks as is demonstrated by the anthracene peak in Figure
4. 11. The use of background subtracfion algorithms should minimize the
occurrence of this phenomenon fi-om background contaminants. Overlapping
chromatographic peaks require more drasfic measures ranging from peak
deconvolufion to reanalysis under different chromatographic condifions.

CONCLUSIONS

Without a pure spectrum, results from a degree-of-fragmentafion
algorithm will be quesfionable. Difliculfies have been observed when using
all the degree-of-fragmentafion algorithms when a contaminant is present in
the form of a high background or an overlapping peak. These problems
present themselves as a broadening of the apparent chromatographic
resolufion or as spikes on the shoulders of chromatographic peaks.
Background subtracfion algorithms are available on most mass spectrometer
data systems and should be used to minimize the contribufions from this
error source. Alternafively, one could sacrifice the capacity for having a
concentrafion independent method of locafing peaks and pass the total ion
intensity or the DIFF intensity to the plot file. Overlapping chromatographic
peaks present a more serious problem. Two solufions to this problem are
available. The first is to collect the data again under different
chromatographic condifions which will eliminate the co-elufion of
components. The second solufion involves the use of cleanup algorithms such

as that of Biller and Biemann [15].

1 11
An interesting phenomenon can be observed in the reconstructed

chromatographic profiles of Figures 4.5 and 4.7. When producing plots of
scan number against any of the degree-of-fragmentafion measurements, a
flat-topped peak is observed for each eluent. This is due to the fact that near
the chromatographic peak maximum the mass spectrum is less suscepfible to
significant changes in peak number or intensity from trace contaminants or
noisy signals than it is on the shoulders. Those scans that result in the flat-
topped chromatographic peaks are also the ideal scans to use for spectral
matching. The length of the flat porfion of the peak is a funcfion of the
number of scans collected across the peak, with high mass spectral
generafion fi'equencies producing many clean spectra across the elufion

profile.

The algorithms that have been developed as part of this work are all
based on the degree-of-fi-agmentafion, a close examinafion of Appendix 1V
shows that they are not very well correlated with each other. In the
extremes, compounds with high values of 2.8 are found to have low values of R
as would be expected. However, much fluctuafion can be found in the middle
of the range. This is due in part to the fact that the value of generic sigma
has a much larger dynamic range than does the number of peaks in the
spectrum. In addifion, the value of generic sigma will disfinguish between
compounds that undergo one major fragmentafion pathway with a few minor
pathways and compounds that readily fragment by a few major pathways.
This disfincfion cannot be discerned using any algorithm based exclusively
on the number of peaks in the spectrum.

Algorithms discussed in this chapter should be used with caufion.
They are based on the degree-of-fragmentafion and not on the aromaficity of

112
the molecule. Species that give a high response to these algorithms are not

necessarily aromafic. The air contaminant introduced into the test sample is
a prime example of a non-aromafic species with a strong response to this
algorithm. Other examples are tabulated in Appendix IV. This tabulafion
gives a good esfimate of the predicted degree-of-fi-agmentafion response
factors, but these values should be determined for each species under the

condifions of analysis.

The algorithms that have been discussed in this chapter have been
written to discriminate against those species that produce mass spectra with
large amounts of fragmentafion. They also can be put to the opposite use.
For example, by inverfing the combined algorithm rafio strong responses will
be found for those eluents that have large amounts of fragmentafion and
weak responses will be found for those species with little fragmentafion.
This could ba useful if the sample matrix is heavily contaminated with
phthalate esters (plasficizers that produce spectra composed of little more
than peaks at m/z 149 and 163) while the desired informafion regards

aliphafic components of the sample matrix.

The real test of any algorithm is to observe its performance under
nonideal condifions. Figure 4.12a is the T11 reconstructed chromatogram of
unleaded gasoline collected at 10 spectra per second. Data from this
chromatographic separafion were passed through the generic sigma filter,
with TII being passed to the plot file to obtain the reconstructed
chromatogram in Figure 4.12b. Gasoline is a mixture of alkanes, alkanes,

113

4

....“ dE
p
I.

“0.:

huh-Once
I
i
l l
I'I'I

.‘O. -

“0'. --0

J N
moon .9”

 

 

 

 

 

 

 

J..“
0
“O.
1,“...
3.4.5.
“0”
noes YAL:J - 1-- A
J. ----,-v--§-:v:§ -:-§;A.:;4
'“ ‘. .l ‘C It: ..

4.12. Reconstructed chromatograms of unleaded gasoline collected by
injecting 0.5 pl gasoline onto a 22 m section of 0.25 mm fused silica

tubing coated with 0.25 um SE-54 and using a mass spectral generation
rate of 10 spectra per second. Figures represent (a) the total ion intensit
reconstructed chromatogram and (b) the total ion intensity reconstructedv
chromatogram from only those spectra with generic sigma values
between 30 and 60 percent of the base peak intensity.

114
and light aromafics. As expected, species passed through this algorithm

include the benzene, toluene, xylenes, and trimethyl benzenes as was
determined through interpretafion of the mass spectra. Aliphafic
components in the T11 chromatogram were discriminated against in the

generic sigma plot.

10.

11.

12.

13.

14.

15.

115
REFERENCES

. Hites, Ronald A. and Biemann, K.. Anal. Chem., 40 (1968) 1217.

Hites, RA. and Biemann, K.. Anal. Chem., 42 (197 0) 855.

McLafi‘erty, Fred W. and Venkatara taraghaven, Rengachari. W
W- American Chemical ociety Washington, DC. (1982).

Eckenrode, B.A.; Victor, M.A.; Watson, J .T.; Enke, C.G.; and Holland, J. F.'.
"Complete MSIMS Data Field Acquisifion on the Chromato aphic Time
Scale by Time-Resolved Ion Momentum Spectrometry with e-Array
Detecfion," submitted to Anal. Chem”

. Holzer, G. and Bertsch, W.. American Lab., (December 1988) 15.
. Johnson, Joseph H.; Erickson, Eric D.; Smith, S. Ruven; Knight, David J .;

Férie, Dwight A.; and Heller, Carl A.. J. Hazardous Materials, 18(1988)
1 .

Erickson, Eric D. ,Johnson, Joseph H. ,Smith, S. Ruven; Cordes, Herman
Fgg‘éeelb 113mghtA.; and Knight, David J.. J. Hazardous Materials, 21
1

McLafi'erty, F.W.. mtg -- retafignofliassjmtra. University Science
Books: Mill Valley, C 'i ’ornia (1980).

Blefich, W.; Sellers, C.S.; Holzer, G.; and Babin, K.. LOGO, 6 (1988)

 

 

 

 

 

Lee, M. L.; Novotny, M. V.; and Bartle, K.D.. mm
W Academic Press: New York (1981).

gAiCket Graph is a software product of Cricket Software, Inc., Malvern,

Watson. J Throck WW Raven Press:
New York (1985) pg. 6.

Stenha en, E.; Abrahamsson, S.; and McLafl'erty, F.W.. Aflagngass
Spmfflata. John Wiley & Sons: New York (1969).

Guido, D ..M ___ca_'o_o_£a_'__9_F_ligthodifi t1 n 'I‘tme- f- quafineclrmrelam
Wflmgtographig Effluents, 'ic igan State University:

East Lansing, Michigan (1985).
Biller, J .W. and Biemann, K.. Anal. Lett., 7 (1974) 515.

 

 

 

 

 

Appendix I:

Programs Used to Simulate the CVC 2000 Instrument

APPENDIX I
Program CALTOF.FOR

The program in this appendix is used to calculate the intensifies and
unit-mass-resolufion in the CV C 2000 TOFMS. Subprograms and funcfions

include:

CALTOF.FOR -- This subroufine calculates the flight-fime of ions of given
inifial spafial posifion, inifial velocity, and mass.

CUBSOL.FOR -- This subroufine is used to find the solufions to a cubic
equafion. It is used in the calculafion of flight-fimes.

LIMRESFOR —- The main program used to set up parameters and cycle
through loops of the variable parameters. This is also the porfion of the
program that writes data files and hence, is used to calculate the percent
valley and other perfinent resolufiOn informafion.

LINE .FOR - Subroufine to calculate the slope and intercept of a line. This
roufine is used in the calculafion of inifial velocifies.

MAXMIN.FOR -- Subroufine to find the maximum and minimum intensifies
of a peak shape. The TOF peak shape does not have a Gaussian
distribufion and it is not safe to assume that mean posifions and

velocifies in the source will result in the mean flight-fime to the detector.

116

117
This roufine was included to ensure accuracy in the assignment of the

proper flight-fimes.

PRBLTY.FOR - This funcfion call gives the probability of a value given the
mean and variance of the distribufion. A normal (Gaussian) distribufion
is assumed. This function is used to calculate ion intensifies.

QDSOL.FOR - This subroufine is used to find the solufion to the quadratic
equafion. This is needed in the calculafion of flight-fimes.

SIMPS.FOR -- This funcfion performs a Simpson’s rule approximafion to
integrate the area under a curve. It is used in the normalizafiOn of

intensifies.

TPROB.FOR -- This funcfion corrects for ion movement fi'om fime lag and
then calls the proper subroutines for calculafing the probability of ion
arrival fimes. This is the roufine that sums the products of posifion and
velocity probabilifies.

VELCAL.FOR -- Given values of inifial posifion, mass, and flight-fime, this
fimcfion will calculate the inifial velocity that was required. This is

based on a series of linear approximafions.

118

mLJLLm+LAAALA thm++k+ALA+L

1 'v T v TTV‘TTT - v w v rm TTTTTTT

"' LIMRESFOR *

lLLLAA-LLLL-LJ LLLmnmLL-L
'-

'Wf‘v'jj'IT‘Vv writ-T-rfw

Eric D. Erickson
8/21/87

Edited 7/6/88 to incorporate delay in reaching max
extracfion voltage using a model containing 2
ideal square wave extracfion pulses.

Edited 7/14/88 to loop through velocifies rather than
spatial posifions

Edited 7/21/88 to calculate velocities assuming a
linear relationship between the inifial velocity
and the final flight-fime. This assumption
appears to be valid when the inifial velocity is
small, as is the case when the velocity is caused
by k’l‘ energy.

Edited and included commentary 12I23/88

Edited 12I27/88 to incorporate shift in initial
posifion caused by time-lag and new value for
ssigma after fime-lag.

Edited 12I29/88 to correct for calculafion of inifial
posifion after fime-lag. REMEMBER: S0 AND SI ARE
DISTANCES TRAVELED IN THE FIRST REGION!!! The
spatial fime-lag correcfions must therefore be
subtracted rather than added to the calculafions.

Edited 1/16/89 to correct for the fact that the flight-
fime distribufion at the detector is not
necessarily Gaussian. This means that the mean
velocity and position in the ionizafion region do
not necessarily produce the maximum signal
intensity at the detector. It was necessary to
use a root finding algorithm to locate the flight-
time that yields the maximum intensity.

Edited 1/18/89 to permit different filenames.

Edited 2/28/89 to include accurate determinafion of

valley tof.
Edited 3/21/89 to find max and min of tof peak shape
through iterafion.

This is the main roufine used in the calculation of
flight-fimes and unit mass resolufion in the CVC 2000
TOFMS. This version of the program calculates the mass
of unit mass resolufion in time-of-flight mass
spectrometry from fundamental principles. Unit mass
resolufion is defined as that point where the
contribufion of the M ion to the point midway between M
and M+1 is less than a threshold intensity relative to
the most probable flight-fime for ion M. This is done
assuming a normal (Gaussian) distribution of initial
velocity and position in the source. Taking a point on
the spafial distribufion, and its probability of
occurrence, the velocity needed to give the desired

**************************{'1’**********************

119

flight-fimes (maximum and valley points) and its
probability are calculate. The product of these two
probabilifies is summed over many spafial positions to
obtain probabilifies, representafive of intensifies at
these two flight-times for adjacent peaks of equal
intensity.

This program needs to be linked to AUX.FOR, which
contains all the necessary subprograms.

**********

PROGRAM LIMRES

nLL.LL.A..A.;A.-AL-LL-‘LH-ALA-ALJA-nn.-...-..Jn-AL-IAA-L-J‘L

vr-v IV‘TUT'TII'I'rTT'Vf'v-‘ITTTT

n

- "T‘Iv‘wivv‘vvvv-vw'vvv‘v'
II..L-tALLLLLA—AAAA-.LL_‘ALLLLL‘LLAAAAALAL‘AL‘A—IALLJAAAA‘kA‘A‘
I'Vjv‘IUVV‘V‘V w vv

’V'U'v‘v" V‘v‘vvw urvr1111v‘v1-TwTv'w1V-ucurt—V'T

PARAMETER (COUELE = 1.6021892D-19)

COMMONI VTIMEI TR ! Time to maximum extn. voltage

COMMONIPOTDIFIVDIF1,VSUM1,VDII'2,VDIF28,VDIF3,VDIF4
! Grid voltages

COMMONI POT/ V1, V1, V2, V3

COMMONI POS/ D1,D2,D3,D4,D5 ! Grid distances

COMMONI PARl/ SSIGMA, TLAG, SO

COMMONI PAR2/ ITER
COMMONI PAR3/ MASS, TEMP
REAL‘B D1 ! Distances of various regions
REAL*8 D2 ! in m
REAL*8 D3
REAL*8 D4
REAL*8 D5
REAL*8 EM ! End mass
REAL*8 ETL ! End fime-lag
REAL*8 EV 1 ! End ion focus
REAL*8 IFSTEP ! Ion focus step
REAL*8 M ! Mass in kg
REAL‘S Mhi
real*8 mlo
REAL*8 MASS ! Mass of the ion
REAL’B MSTEP ! Mass step
REAL*8 MTLAG !Time-Lag in microseconds
REAL‘B SO ! Center posifion for space
! distribufion
REAL*8 SI ! New center position due to tlag
REAL*8 SINIT(2) ! Inifial position of the ion
REAL*8 SM ! Start mass
REAL*8 SSIGMA ! Std dev of initial posifion
REAL*8 STL ! Start fime-lag
REAL*8 SVl ! start ion focus
REAL*8 TMAX
real‘8 thi

REAL‘S TMIN

120

READS TEMP ! Source temperature in Kelvin
REAL‘8 TLAG ! Time-lag
REAL*8 TLSTEP ! Time-lag step size
REAL*8 TOF ! Flight-time of the ion
REAL*8 TOFHI ! Flight-time for M+1 ion
real*8 toflo
REAL‘S TR ! Time to maximum extn. voltage
REAL*8 V1 ! Ion focus voltage

! Range: 6 to 140
REAL‘8 VI ! Voltage on grid 1 B4 extracfion
REAL'8 V2 ! Voltage on grid 2

! Range: 150 to 260
REAL*8 V3 ! Voltage on grid 3
REAL*8 V4 ! Voltage on grid 4
REAL*8 VSUMI
REAL‘8 VDIFl
REAL*8 VDIF2
REAL*8 VDIFZB
REAL*8 VDIF3
REAL*8 VDIF4
READS VINIT ! Inifial velocity, corr for tlag
REAL*8 VSIGMA ! Std dev of velocity
REAL*8 TPROB ! Flight-time Probability
REAL*8 PROMAX ! Probability of t(m)
REAL*8 PROMIN ! Probability of t(m+.5)
REALM RATIO ! PROMIN/PROMAX
REAL*4 RATLIM ! Maximum acceptable valley
REAL*4 PREV ! Previous rafio
INTEGER IIF ! Ion focus increment
INTEGER IM ! Mass increment
INTEGER ITER ! Number of spatial iterafions
INTEGER ITL ! Time-lag increment
INTEGER ICNT ! Integer counter
CHARACTER*20 DEST ! Desfinafion filename

L-I‘AnJ-‘LLL-LAm-LAAA‘A..--tLL-ALAALL

T wT-waw—T

.MfidiéiYeeiklei---_ -‘f‘

DUI-1rvvvvvvr'vvvwr vvvvvvvvvvvvvv

D1 = 0.365449D-2
D2 = 0.17018D-2
D3 = 0.60706D-2
D4 = 0.56 13413-2
D5 = 2.10

SO = 0.2E-2
TEMP = 500.
SSIGMA = 0.01D-2
Vi = -0.44

V2 = 215.

V3 = 1417.

V4 = 2556.

! Distances in meters

! In meters

! In Kelvin

! Meters

! Negative due to repulsion
! Grid 24 final voltages

121

VDIF3 = V3 - V2
VDIF4 = V4 - V3

.LLLLAJL-ILAILJLJLL+IIIIQLLAIIAJLLAJLLIIIL-IILII...IL

-v-v~--w-1-v—varvrv'fwuututti-iiv-w-wwvvvwnun--

- I
. .LAAAALLLLA-L-L.L‘AALAAJllL-ALALL-nL-‘LLLL.LLLLLJLLL-IL
I w. r- ‘V‘VUVVVV'V'V

ITT'I'I‘VV Iva—1V vw-vrw‘ kuvfi'v’vv-vv'r’vT-uvvv-T

TYPE *, ’Welcome to LIMRESFOR. This program ’
TYPE *, ’calculates the unit mass resolufion for ions ’
TYPE *, ’in the CVC 2000 instrument, based on the ’
TYPE *, ’inifial spafial and velocity distribufions of
TYPE *, ’ions in the source.’

TYPE *, ’ ’

AALLLALA-AIA- LLAALL-LAL-LAAJ .1.-n LLLLAL‘LA‘LL‘LJ-‘AAILLA

'vr'v—wrrv vTvi‘v‘rvv TivT-v—W'Viwv uvwv~11v1wfi11wvvvvv

* Query for variable quantifies *

TwrvvvrwvwvaWv—TVTV vvawTwTwTwva-Wirvv

TYPE *,’What is the desfinafion filename?’
READ (5,5) BEST
5 FORMAT(A20)

TYPE *,’ ’

TYPE *,’Enter ranges for variables to be used in ’
TYPE *, 'loops.’

TYPE ‘,’ ’

TYPE ‘,’What is the maximum acceptable valley ’

TYPE *,’height?(%)’

ACCEPT *, RATLIM ! The contribufion of both peaks
! are calculated in this version.

* Ion focus is the voltage applied to the first
"' (extraction) grid in the CV C2000 instrument. This
* value ranges from 6 to 140 Volts.

10 TYPE ‘, ’What is the start ion focus? (V olts)’
ACCEPT *, SV1
TYPE *, ’What is the end ion focus? (volts)’
ACCEPT *, EV 1
IF (SV1 .EQ. 0.0) THEN
V1 = 79.
VSUM]. = V1 + Vi
VDIFl = V1 - Vi
VDIF2 = V2 - V1
VDIF2B = V2 - VDIFl
GO TO 30
ELSE IF (SV1 .LT. 6.0 .OR. SV1 .GT. 140.) THEN
20 TYPE *,’Outside of range. Legifimate values range ’
TYPE *,’from 6 to 140 Volts. Try again.’
GO TO 10
ELSE IF (EV1 .LT. 6.0 .OR. EV1.GT. 140.) THEN
GO TO 20
ELSE IF (SV1 .EQ. EV 1) THEN

***§******

03
O

i****

45

122

IFSTEP = 1.
V1 = SV1
VSUMI = V1 + Vi
VDIFI = V1 - Vi
VDIF2 = V2 - V1
VDIMB = V2 - VDIFI
GO TO 30

ENDIF

TYPE *,’What is the ion focus step? (V olts/Step)’
ACCEPT ', IFSTEP
HI" = INT((EV1 - SV1)/ IFSTEP)

TOFMS instruments essentially have no maximum mass
range. They permit the detecfion of any ion with

enough energy to knock an electron off of the detector.
Unless the ion’s flight time is within a fime window
defined by the extracfion frequency however, it may be
dificult to actually assign a mass to a detector

event. For the purposes of this program, negafive
masses are discarded and posifive ones are unlimited.
The pulse frequency of the instrument is 10 KHz, but
this limitafion is not considered in these calculations.

TYPE ",’ ’
TYPE *, ’What is the start mass? (Daltons)'
ACCEPT ‘, SM
TYPE *, ’What is the end mass? (Daltons)’
ACCEPT *,EM
IF (SM .LT. 1.0 .OR. EM .LT. 1.0) THEN
TYPE *,’Mass out of range. Negafive values are not’
TYPE *,’permitted. Try again.’
GO TO 30
ELSE IF (SM .EQ. EM) THEN
GO TO 40
ENDIF
TYPE *,’What is the mass step? (Daltons)'
ACCEPT *,MSTEP
IM = INT((EM - SM) I MSTEP)

Time-lag is a delay fime between ion formation and ion
extracfion out of the source that is used for energy

focusing on convenfional TOFMS instruments. The CV C-2000
has an adjustable fime-lag of between 0.0 and 5.0
microseconds.

TYPE *,’ ’
TYPE *,’What is the inifial fime-lag? (microseconds)’
ACCEPT *,MTLAG
STL = MTLAG] 1.E6
TYPE *, 'What is the final fime-lag? (microseconds)’
ACCEPT *, MTLAG
ETL = MTLAG/ 1.E6 ! Convert to seconds
IF ((STL .LT. 0.0) .OR. (ETL .LT. 0.0)) THEN
TYPE ‘,’Time-lag out of range. The permissible range’

50

*****

53

55

ww'I—w'rrw

123

TYPE *,’is between 0.0 and 5.0 microseconds. ’
TYPE *,’Try again.’
GO TO 40

ELSE IF ((STL .GT. 5.0) .OR. (ETL .GT. 5.0)) THEN
GO TO 45

ELSE IF (STL .EQ. ETL) THEN

GO TO 50
ENDIF
TYPE *, ’What is the fime-lag step? (microseconds)’
ACCEPT *, MTLAG
TLSTEP = MTLAG/ 1.E6 ! Convert to seconds

TTL = INT((ETL - STL) ITLSTEP)

This program uses an iterafion of spafial position
around a mean posifion, and a calculafion of the
velocity needed to have an ion start at that posifion.

TYPE *,"
TYPE *,’How many posifion iterations are desired? (>=2)’
ACCEPT *, ITER

The flight-fime calculations take into considerafion
that an ideal square wave extracfion voltage is
impossible to obtain. The operator is therefore able
to set a fime for a linear ramp to the maximum
extracfion voltage.

TYPE *,’ ’

TYPE *,’How much time to maximum extraction voltage?

1 (nsec)’

ACCEPT *, TR

IF (TR .LT. 0.0) THEN
TYPE *,’Negative values do not make sense. Try again.’
GO TO 53

ENDIF

TR = TR * 1.E-9 ! Convert to seconds

TYPE *, ’ ’
TYPE *,’The calculations have begun’

OPEN (UN IT: 1, FILE=DEST, STATUS='NEW’)
OPEN (UNIT=2, FILE='LIM.DAT', STATUS=’NEW’)

ILAL‘LAL-AILALL-IL-IAILlL-JL-lL-Al...-L...-

TTTITTTTww—r—vaIWWjvw'wr—VT—Vj’v‘v ITr' v-v—v

'I'I‘VI'II'VII'I"VIVIIVVI‘U'IUI'W'II‘UUIV'IITI‘IIW

IF ((SVl .EQ. EV1) .OR. (SV1 .EQ. 0.0)) GO TO 60

DO 2000 I = 0, IIF ! Start ion focus loop
V1 = SV1 + FLOAT(I) * IFSTEP ! Set ion focus
VSUMI = V1 + Vi
VDIFl = V1 - Vi
VDIF2 = V2 - V1
VDIF2B = V2 - VDIFl

60

70

80

110

120
121

510
1000
490

124

IF (SM .EQ. EM) THEN ! Check for single mass
MASS = SM
GO TO 70

ENDIF

DO 500 J = 0, IM ! Start mass loop

MASS = SM + FLOAT(J) * MSTEP! Set mass

Mhi = (MASS + 1.D0) * 1.6605655d-27
Mlo = (MASS - 1.D0) "' 1.6605655d-27
PREV = 100.
ICNT = 0
IF (STL .EQ. ETL) THEN ! Check for single time-lag
TLAG = STL
GO TO 80
ENDIF

Assumpfion: Ions stay put unfil the electron beam is
turned off.

DO 1000 K = 0, ITL ! Start fime-lag loop
TLAG = STL + FLOAT(K) * TISTEP ! Set fime-lag
SI = SO-((COUELE*Vi*TLAG*TLAG)/(2.*eri*D1))
VINIT = COUELE*TLAG*Vi/(Mhi“D1)
CALL CALTOF (SI, VINIT, Mhi, TOFHI) !TOF for M+1
SI = SO-((COUELE*Vi"TLAG*TLAG)/(2.*Mlo*D 1))
VINIT = COUELE*TLAG*Vi/(Mlo*D1)
CALL CALTOF (SI, VINIT, Mlo, TOFlo) ! TOF for M-l

CALL MAXMIN (TOFLO, TOFHI, PROMAX, PROMIN)
IF (PROMAX .EQ. 0.0D0) GO TO 1000
! Avoid division by 0
RATIO = 100. * PROMIN / PROMAX ! Percent
WRITE (2,121) MASS,TLAG*1.E6, PROMIN, PROMAX, RATIO
IF ((RATIO .GT. PREV) .AND. (TLAG .GT. 2.E-6) .AND.
1 (RATIO .GE. RATLIM» THEN
WRITE (1,110) MASS, TLAG’1.E6
FORMAT (2X, F5.0, ’ ’, F7 .2)
ICNT = 0
GO TO 490
ELSE IF (ICNT .EQ. 0) THEN
IF ((RATIO.LE.PREV) .AND. (RATIO.LE.RATLIM)) THEN
WRITE (1,110) MASS, TLAG*1.E6
ICNT = 1
ENDIF
ENDIF
FORMAT (2X, F5.0, 4(’ ’, F7 .2))
FORMAT (2X, F5.0, ’ ’, F7.2, 3(’ ’, e12.4))
PREV = RATIO
IF (STL .EQ. ETL) GO TO 490
CONTINUE ! end fime-lag loop
TYPE *, ’MASS’,MASS,’DONE’
PREV 3 100.
IF (SM .EQ. EM) GO TO 1010

500
1010

2000 CONTINUE

125

CONTINUE ! End mass loop
IF (SV1 .EQ. EV1) GO TO 2010
! End ion focus loop

2010 CLOSE (1)

L

L

L.

CLOSE (2)

TYPE *, 'Your data have been stored in ’, DEST
END

A‘LLALA-flAthL-AALL-IILJALALL-LQALLLL-LLLLA‘LAL

1va'wUVWT-wvvvvv-Iuv'vIrv-Iuv-v-vuvvvv-vvviuvr’vi’VT

' ...

*************************§**************

AIL

. T925793“

A. AL-ILA-.J-ILL‘J-ILLIAAILLAILL

Eric D. Erickson

6/29/88

Edited 7/2/88

Edited to model 2 square waves 7/6/88
Math correction 7/7/88

Math correction 7/28/88

Math correction 8/8/88

Math correction 12/22/88

This program was written due to my frustrafion with
earlier versions of the CV C simulafion. This version
calculates the flight-fimes of an ion of mass M (in kg)
given its inifial posifion (in m) and velocity (in mIs)

in the CV C 2000 source. Other versions of this program
have been written and adapted from George Yefchak’s
simulafions, but this one is a significant variafion
from those programs which are based on inifial ion
energy rather than velocity as in this program. In
part, this program was written to see if I could get it
to work (i.e. if I understand all the concepts

correctly) and in part it was written to simplify the
incorporafion of the fime that it takes for the
extraction pulse to reach its maximum voltage.

This derivafion is based on the following fundamental
relafionships:

a = dv/dt
a = qV/md
v = dx/dt

Derivafions of the individual expressions used in this
program are in my notebook dated 7/28/88. The
correction of time needed to reach maximum voltage is
done using a model consisting a linear ramp between the
inifial and final voltages. This version now includes

the remote possibility that ions may enter the third
region of the source before the maximum voltage is
obtained. The model used for this program will permit

‘l‘

126

the entry of 0 as if a single ideal square wave

extracfion pulse is desired.

SUBROUTINE CALTOF(SINIT, VINIT, M, TOF)

PARAMETER (COUELE = 1.6021892D-19)

COMMONI VTIMEI TR

COMMONI POTDIF/ VDIF1,VSUM1,VDIF2,VDIF2B,VDIF3,VDIF4
COMMON/ POT] Vi, V1, V2, V3

COMMONI POSI D1,D2,D3,D4,D5

REAL*8 V1 ! Final voltage on grid 1
REAL*8 Vi ! Inifial voltage on grid 1
REAL*8 V2

REAL*8 V3

REAL*8 VSUM]. ! V1 + Vi

REAL*8 VDIFl ! V1 - Vi

REAL*8 VDIF2 ! V2 - V1

REAL*8 VDIFZB ! V2 - VDIFl

REAL*8 VDIF3 ! V3 - V2F

REAL*8 VDIF4 ! V4 - V3

REAL*8 A ! Variables for quadrafic solufion
REAL*8 B

REAL*8 C .

REAL*8 C82 ! Cubic solufions

REAL*8 CS3

REAL*8 M ! Ion mass in kg

REAL*8 SINIT ! Inifial ion posifion in m
REAL*8 VINIT ! Inifial ion velocity in m/s
REAL*8 TMP1 ! couele*q/m*d

REAL*8 TMP2

REAL*8 TMP3

REAL*8 TMP4

REAL*8 D1 ! Grid distances

REAL*8 D2

REAL*8 D3

REAL*8 D4

REAL‘8 D5

REAL*8 RTIME

REAL*8 T1 ! Time spent in various regions
REAL*8 T2

REAL*8 T3

REAL*8 T4

REAL*8 T5

REAL*8 TOF ! Time-of-Flight for the ion
REAL*8 TR ! Time for voltage ramp to grids 1 and 2
REAL*8 XTMP ! Second solufion to quadrafic eqn.
REAL*8 VELl ! Velocifies leaving regions
REAL*8 VEL2

REAL*8 VEL3

REAL*8 VEIA

INTEGER I,J

! Counters

*i'l'i'

10

127

INITIALIZE VARIABLES

ASSUMPTION: All ions are singly charged.

Ifnot, then the TMP variables need to be mulfiplied by
the number of charges on the ion.

TMP1 = COUELE / (M "‘ D1)
TMP2 = COUELE / (M "' D2)
TMP3 = COUELE / (M * D3)
TMP4 = COUELE / (M "' D4)

IN REGION 1

IF (TR .EQ. 0.0) then ! Avoid division by 0
T1 = 1.
GO TO 10

ENDIF

tr greater than t1

A: 3. ‘TR*V1/VDIF1

B=6. *TR*VINIT/(TMP1*VDIF1)

C = -(6. "‘ SINIT "‘ TRI(TMP1 * VDIF1)) - SINIT

CALL CUBSOIIA, B, C, T1, CS2, CS3) ! Calculate t1

IF (Tl .LT. CS2) T1 = CSZ ! Take highest solufion

IF (T1 .LT. CS3) T1 = CS3

VEL1 = CTMPlWi*T1) + (TMP1‘VDIF1*T1*T1I2.*TR) + VINIT

tr<= 11

IF (TR .LE. T1) THEN
A = TMP1 * V1 / 2.
B = (TMP1 * VSUMl * TRI 2.) + VINIT
C = (TMP1*TR"'TR*(V 1+2*Vi)/6.) + VINIT“TR - SINIT
CALL QDSOL(A, B, C, XTMP, T1)
IF (T1 .LT. XTMP) T1 = XTMP! Take highest soln
VEL1 =(2. *A*T1)+ B
T1 = T1 + TR
IF (Tl .LT. TR) THEN ! Remove impossible times
T1 = 0.0130
VEL1 = VINIT
ENDIF
ENDIF

IN REGION 2
tr < t1

IF (TR .LE. T1) THEN
A=TMP2* VDIF2/2.0
CALL QDSOL(A, VEL1, -D2, XTMP, T2)
IF (T2 .LT. XTMP) T2 = XTMP
VEL2=2.*A*T2+VEL1

128
ELSE

tr>t1+t2

RTIME = TR - T1
A = -3.*Vi"‘TRIVDIF28
B = 6."TR*VEL1I(VDIFZB*TMP2)
C = —6."‘TR*D2I(VDIF23“'TMP2)
CALL CUBSOL(A., B, C, T2, C82, CS3)
IF (T2 .LT. 082) T2 = CS2 ! Take highest soln
IF (T2 .LT. CS3) T2 = CS3
VEL2 = (TMP2‘VDIF2B“I‘2*T2/(2.*TR))-
1 (V i*TMP2*T2)+VEL1

t1<tr<=t1+t2

IF (TR .LE. T1+T2) THEN
A = TMP2 * VDIF2 I 2.
B = (TMP2*RTIME*(VDIF2-Vi)/(2.‘TR)) + VEL1
C = (TMP2*(VDIF2-(2.*Vi))*RTIME*

1 RTIME*RTIME/(6."TR))-
2 (TMP2*Vi*RTIME*RTIME/2.)+(VEL1*RTIME)-D2
CALL QDSOL(A, B, C, XTMP, T2)
IF (T2 .LT. XTMP) T2 = XTMP ! Take highest soln
VEL2 = (TMP2‘VDIF2‘T2)+(TMP2*(VDIF2Vi)*RTMEI
1 (2.*TR))+VEL1
T2 = T2 + TR - T1
ENDIF
ENDIF
IN REGION 3
tr < t1 + t2

IF (TR .LE. T1 + T2) THEN
A = TMP3 * VDIF3 / 2.
CALL QDSOL(A, VEL2, -D3, XTMP, T3)
IF (T3 .LT. XTMP) T3 = XTMP ! Take highest soln
VEL3=2.*A*T3+VEL2
ELSE
RTIME = RTIME - T2

tr>t1+t2+t3

A = -3.*TR*V3IV 2

B = -6."TR*VEL2/(TMP3"V2)

C = -B*D3IVEL2

CALL CUBSOL(A, B, C, T3, C82, CS3)

IF (T3 .LT. 082) T3 = CS2 ! Take highest soln

IF (T3 .LT. CS3) T3 = CS3

VEL3 = (TMP3‘V3‘T3) - ('I'MP3‘V2"'I‘3‘T3/(2."'I‘R)) + VEL2

129
"‘ t1+t2<tr<=t1+t2+t3

IF (TR .LE. T1+T2+T3) THEN
A = TMP3 "' VDIF3 / 2.
B = (TMP3*V3*RTIME)-
1 (TMP3*V2*RTIME*RTIME/(2.*TR))+VEL2
C = (TMP3*V3*RTIME*RTIMEI2.)-
(TMP3*V2*RTIME*RTIME*RTIMEI
(6."‘TR))+(VEL2*RTIME)-D3
CALL QDSOL(A, B, C, XTMP, T3)
IF (T3 .LE. XTMP) T3 = X'TMP ! Take highest soln
VEL=(2. *A*T3)+B
T3=T3+TR-T1 -T2
ENDIF
ENDIF

NH

* REGION 4
A = TMP4 * VDIF4 I 2.0
CALL QDSOL(A, VEL3, -D4, XTMP, T4)
IF (T4 .LT. XTMP) T4 = XTMP
VEL4=2.*A*T4+VEL3
* REGION 5: DRIFT TUBE
T5 = D5 / VEL4
* TOTAL FLIGHT TIME
TOF=T1+T2+T3+T4+T5 !Inseconds
* QUIT SUBROUTINE

1000 RETURN
END

....;A.-.L;.-.A..LL-...-LannnLLL-LA-n_n_4.‘_
1

IvIII-'Iwruvvu-u-vur'v'v‘vuuuivw—w vvvvvvvvvv

Eric D. Erickson
8/8/88

This roufine takes 3 values, A, B, and C, and
calculates the real roots of the cubic equation:

0 = x*"'3 + A*X**2 + B*X + C
The algorithm used is derived from one given on page

157 of NUMERICAL RECIPES IN C. All variables must be
REAL*8.

*************

*‘I‘I‘I’i‘lfi"

130
SUBROUTINE CUBSOL (A,B,C,X1,X2,X3)

REAL*8 A, B, C

REAL*8 X1, X2, X3

REAL*8 Q, R, THETA

REAL*8 THIRD, TMPQ, TMPR, P12

THIRD = 1. / 3.
P12 = 6.283185307

Q=(A-3.*B)/9.

R = (2.*(A**3) - 9.*A*B + 273-0) / 54.
TMPQ=Q*Q"'Q

TMPR=R*R

IF (TMPQ .GT. TMPR) THEN ! There are 3 real roots
THETA = DACOS(R/ SQRTCTMPQ»
X1 = -2.*SQRT(Q)*DCOS(THETAI3.) - AI3.
X2 = -2.*SQRT(Q)*DCOS((THETA+PI2)/3.) - A/3.
X3 = -2.*SQRT(Q)*DCOS((THETA+2.*PI2)I3.) - A/3.

ELSE ! There is only 1 real root
TMP = (SQRT(TMPR-TMPQ) + ABS(R))*"'THIRD

X1 = -DSIGN(1.D0,R) * (TMP + Q/TMP) - A/3.

ENDIF .

.-LALALL-LLLLLAL- LLLLL- . - LL;J..A.4_A_-xn_._g_n_‘.

Eric D. Erickson
12I29/88

Subroufine to calculate the slope and intercept of a
line.

SUBROUTINE LINE (X1,Y1,X2,Y2,M,B)

REAL*8 X1 ! Coordinates of 2 points on the line
REAL‘8 X2 .

REAL*8 Y1

REAL*8 Y2

REAL’8 M ! Slope

REAL*8 B ! Intercept

M=(Y1-Y2)/(X1-X2)
B=Y2-X2*M

RETURN
END

A.;LAALJAL-L-IOAAALJ LAAAILAl-QA-A-ALQAL

Tijw-rTwervvrTTwiww—‘va'vwww-vv v VIVVI'VTT

"' MANINFOR

.LAAAALA- AAA—ALA LLAALLLAALLL+IIALLLIL

.-
ww-Ijv—w-vw-vIT’VITjr—r'VT11I-II

Eric D. Erickson
3/21/89

Routine to find the maximum and minimum intensifies of
a peak shape. This roufine was written due to problems
with the search algorithms previously used. This
algorithm searches for the limits using a bin filling
approach.

Edited 4/5/89 to permit the normalizafion of peak
areas.

Edited 4/7/89 to include a Simpson’s appron'mafion of
the area

*‘Ii‘l-‘I'I'i-fl'ififi‘l‘l'fl'i

SUBROUTINE MAXMINCXI, X2, RMAX, RMIN)

PARAMETER(POINTS = 50.)

REAL*8 X1, X2 ! Limiting flight-fimes
REAL*8 TMAX ! Flight-time for peak maximum
REAL*8 RMAX, RMIN ! Probabilities of max and min
REAL*8 INCR ! Flight-fime increment
REAL*8 TIME ! Flight-fime

REAL‘8 ION PROB(50) !TOF probability
REAL‘8 TOTPROB ! Total probability
REAL*8 SIMPS ! Simpson rule approximafion of area

* lnifialize variables

TOTPROB = 0.0

INCR: (X2 -X1)/ POINTS
RMAX = -10.0

RMIN = 1000.0

TMAX = 1.

* Find TOF with the max and min probability
D0 10 I=1,POINTS
TIME = X1 4» FLOAT(I) * INCR
IONPROBG) = TPROB(TIME)
IF (IONPROBG) .GT. RMAX) THEN
RMAX = IONPROBG)
TMAX = TIME
ENDIF
IF ((IONPROB(i).LT.RMIN).AND.(TIME.GT.TMAX))
1 RMIN = ION PROB(i)

132

IF ((IONPROB(I).GT.RMIN).AND.(TIME.GT.TMAX))
1 GO TO 15
10 CONTINUE

* Normalize intensities

15 TOTPROB = SIMPS(1, I, IONPROB)
IF (TOTPROB .EQ. 0.0) THEN
TYPE *, ’Divide by zero error occurred in area.’
GO TO 20
ENDIF
RMAX = RMAX / TOTPROB
RMIN = RMIN / TOTPROB

20 RETURN
END

.LAL-LL-AL-nnn._A__-_L-.--_n.n--nnn_-nnn-

u'v-rv—T'v—vw-vr’u—V'1‘u1—v— Iw-Ivr1111'vuvu

* PRBLTY.FOR

'TTT'V‘VV‘VTI‘TTWU‘I T'TTT'VT'VVV

Eric D. Erickson
1129/88
Changed to a funcfion call 7/14/88

This subroufine calculates the probability of an
occurrence, assuming a normal (Gaussian) distribufion of
events. This uses the relafionship:

P(x) = exp((x-u)**2/2(sigma)**2)

***********'-

FUNCTION PRBLTY (X, SIGMA, MEAN)

REAL‘8 DEV ! Deviation

REAL*8 MEAN ! Mean posifion
REAL*8 SIGMA ! Std dev about mean
REAL‘S X ! True posifion

REAL‘8 PRBLTY ! Probability

DEV = X - MEAN
PRBLTY = EXP(-(DEV * DEV)I (2. * SIGMA * SIGMA»

RETURN
END

***§************

ALI-.A-l-AI-AlJ-LALLLIIIIIA-A-IAIAL nnnnnnnn

I‘UII‘W'II'I’VI"I'VII‘UI‘UII’I'TI'I'II‘V‘U'I’I‘U’V

I'vvvuur-uIT‘IIv-vuwvrlvivrv'wvvfivvvvvvuvwvvv

Eric D. Erickson
6/30/88

This subroufine finds the two solufions to the

quadratic equation and returns them as X1 and X2. All
variables must be declared as REAL*8. I have not taken
the fime to fix this for imaginary numbers, so

imaginary numbers are passed as -1.

Edited to reduce possible errors on 7/8/88. The method
used is that described on page 156 of NUMERICAL
RECIPES IN C.

Edited to eliminate division by zero 1I17I89.

SUBROUTINE QDSOHA, B, C, X1, X2)

REAL*8 A ! All symbols are commonly used
REAL*8 B ! ones in general Algebra
REAL*8 C .
REAL*8 Q
REAL*8 ROOT ! Temporary value to check for real
! roots of the equation
REAL*8 X1 ! Two solufions
REAL*8 X2

ROOT=B*B-4.*A*C
IF (ROOT .LT. 0.0) THEN

X1 = -1.D10

X2 = -1.D10

RETURN
ENDIF

Q = -0.5 "‘ (B + DSIGN(1.0D0, B) "' SQRT(ROOT))
X1 = Q I A
IF (Q .EQ. 0.0) THEN

X2 = -1.D10
RETURN
ENDIF
X2 = C IQ
RETURN

END

134

ItL-LL-LLLLn-‘A.LA-ILQL-Alfi-AA-

‘I'UV'VT‘V u T'U'VT’I’I’I I'v-w-uvvv-Vi

I'V‘II'VWjIW‘I‘WT'I'I'UU'WI'V‘

Simpson’s rule approximafion

*
:I:
*
* This funcfion is a modification of SIMPSO from George
* Yefchak dated April 1985. It was modified and input by
a:

*

Eric D. Erickson on 4/7/89.

FUNCTION SIMPS (A, B, FUN)

REAL*8 FUN(50) ! Array of peak heights
REAL*8 SUM ! Sum of peak areas

REAL*8 SIMPS ! Simpson’s area
INTEGER A ! Counters

INTEGER B

SUM = 0.0

IF (MOD((B - A), 2) .NE. 0) THEN ! If not even
SUM = SUM+3.0*FUN (B) ! Add rectangular area from
B = B - 1 ! last interval and make it even
ENDIF
DO 10 I = A+1, B-2, 2
SUM = SUM + 4.0 * FUN(I) + 2.0 “' FUN(I+1)
10 CONTINUE

SUM = SUM + FUN(A) + 4.0 * FUN(B-1)+ FUN(B)
SUM = SUM / 3.0

SIMPS = SUM

RETURN

END

LLALLJLLLLAAIAAIA‘A‘LL‘Ll‘L-LAA-‘AAA
uv

* TPROB.FO

.Ax-AAAALLLLAALLAH.-L-A-LALAMLLLLALIAMAL

I‘V'II‘VII‘U'TTTT‘V‘T

vrvvuv-riVvaIIvT-vuvvv-I

*

Eric D. Erickson
1/16/89

Funcfion to calculate probability of a parficular
flight fime.

SINIT is the ion’s posifion when the fime-lag ends.
VEL is the ion’s velocity at the end of the time-lag.
XINIT and VINIT are these values immediately upon
cessafion of ion formafion (the electron beam is turned

ofi).

*******‘I****

*******

135

Modified 2/28/89 to permit the calculation to include
adjacent peak contribufions. This was done to enable
the search for the true position of the valley.

Modified 4/5/89 to loop through equidistant spafial
posifions in the source.

FUNCTION TPROB(TOF)
PARAMETER (COUELE: 1.6021892D- 19, BOLTK=1.380662D-23)

COMMONI PAR]! SSIGMA, TLAG, SO

COMMONI PAR3/ MASS, TEMP

COMMONI POSI D1,D2,D3,D4,D5 ! Grid distances
COMMONI PAR?! ITER

COMMONI POT/ V1, V1, V2, V3

REAL*8 SIGMAS ! SSIGMA corrected for fime-lag
REAL*8 Vi ! Inifial voltage on grid 1
REAL‘8 V1 ! Final voltage on grid 1
REAL*8 V2 ! Final voltage on grid 2
REAL*8 V3 ! Grid 3 voltage
REAL*8 TOF ! Flight fime

REAL*8 M ! Mass in kg

REAL*8 SO ! Inifial mean posifion
REAL*8 SI ! Mean position
REAL*8 VSIGMA ! Velocity sigma
REAL*8 SINIT ! Inifial Position
REAL*8 TLAG ! Time-lag in seconds
REAL*8 MASS ! Mass in Daltons

REAL*8 TEMP ! Temperature in Kelvin
REAL*8 D1 ! Region 1 distance
REAL*8 D2 ! Region 2 distance
REAL*8 D3 ! Region 3 distance
REAL*8 D4 ! Region 4 distance
REAL*8 D5 ! Region 5 distance
REAL*8 PIN C ! PIObability increment
REAL*8 VEL ! Velocity

REAL*8 VELCAL ! Calculated velocity
REAL*8 XINIT ! Corr init position
REAL*8 SSIGMA ! Position Std. Dev.
REAL*8 VINI ! Corr init velocity
REAL*8 TPROB ! Probability of fime tof
REAL*8 SPROB ! Posifional probability
REAL*8 VPROB ! Velocity probability
INTEGER ITER

TPROB = 0.0

DO 200 N = 1,2

M = (MASS + FLOAT(N-1)) "' 1.6605655D-27
SI = SO-((COUELE*vI"TLAG*TLAG)/(2.*M*D1))
VSIGMA = SQRT(BOLTK * TEMP / M)

DO 100 L = 1, ITER

! Start spatial loop

136

SINIT = 0.0005 + FLOAT(L) * (D1 - 0.0005)/
1 FLOAT(ITER)
IF ((SINIT .LT. 0.0D0).OR.(SINIT .GT. D1)) GOTO 100
VEL = VELCAL(SINIT, M, TOF)
VINI = VEL - COUELE * Vi * TLAG/(M * D1)
XINIT = SINIT+COUELE*Vi*TLAG*TLAG/(2.*M*D 1)
1 +VINI*TLAG
IF ((XINITLT.0.0DO) .OR. (XINIT.GT.D1)) GO TO 100
! Throw out ions outside of the source
SPROB = PRBLTYOIINIT, SSIGMA, SO)
VPROB = PRBLTYWINI, VSIGMA, 0.0D0)
TPROB = TPROB + SPROB * VPROB

100 CONTINUE ! End probability loop
200 CONTINUE

RETURN

END

. .
. .. cw..ww-w-1.-u..
‘LJ-“-‘A-LJL-AAAA-IIAAJ.“A-LL-JII
vvrvv www-vrw

v‘v'vv‘ writ-vu-

Eric D. Erickson
12/28/88

This program calculates the velocity needed to achieve
a given flight-time from known initial positions and
masses. A linear approximation is used.

Edited 1/5/89 to speed up calculations of outliers. If
the velocity is greater than 3000 m/s or less than -
3000 m/s, the program now comes to a guess that has
more error than smaller velocities, but since the
probability of these high velocities is so small to
begin with, (<2.4e-5 for m/z 10), these velocities
would receive a much smaller weighting in the main

program anyway.

*i***************“

FUNCTION VELCAL (POS, M, TOF)

REAL*8 POS ! Distance traveled in the first

! region of the source.
REAL*8 M ! Mass of the ion in kg
READS TOF ! Expected flight-time
REAL*8 VELCAL ! Calculated velocity
REAL*8 TOFl ! TOF of ion with velocity -1000
REAL*8 TOF2 ! TOF of ion with velocity 1000
REAL*8 INTER ! TOF intercept
REAL*8 SLOPE ! Slope of the line
REAL*8 X1 ! Temporary positions
REAL*8 X2

INTEGER N ! Counter

100

200

210

300

137
Initialize variables

SLOPE = 0.0D0
INTER = 0.0D0
X1 = 0.0D0
X2 = 5.0D2

Calculate flight time for 2 points on the line and
calculate the slope and intercept of that line.

CALL CALTOF (POS, X1, M, TOFl)
IF (TOF .LE. TOFI) THEN
CALL CALTOF (POS, X2, M, TOF2)
IF (TOF .GE. TOF2) THEN
CALL LINE (X1, TOF1, X2, TOF2, SLOPE, INTER)

GO TO 300
ELSE
N = N + 1
IF (N .GT. 6) GO TO 210
X1 = X2
TOF]. = TOF2
X2 = X2 + 5.D2
ENDIF
GO TO 100
ELSE
X2 = -5.E2

CALL CALTOF (PCS, X2, M, TOF2)
IF (TOF .LE. TOF2) THEN
CALL LINE (X1, TOF1, X2, TOF2, SLOPE, INTER)
GO TO 300
ELSE
N = N + 1
IF (N .GT. 6) GO TO 210
X1 = X2
TOF]. = TOF2
X2 = X2 - 5.D2
EN DIF
GO TO 200
ENDIF

Calculate the velocity.
VELCAL = (TOF - INTER) / SLOPE

RETURN
END

Appendix II:

Programs Used to Calibrate the CVC 2000 Instrument

APPENDIX 11
Program CVCMASS.FOR

Programs used in TAD calibration.

. CV CMAS - This is the main program in the calculation of m/z values fiom
flight-times and handles operator interaction. Options are ofl‘ered for
computer calculation of linear parameters from a data file, or operator
entry of critical parameters.

. FILEIO - This subroutine prompts for source and destination filenames
and passes them back to the calling program.

. CALIB - This subroutine calculates the linear calibration parameters.

. FILCAL - This subroutine is used to derive the calibration parameters and
rform a statistical analysis of them. It uses approximate flight-times
or three different "known" m/z values and searches the data file for all
occurrences of these values within a 15 nsec window, ignoring ions with
intensities below 50 counts. Values of k and C as well as their sample
standard deviations are output to the terminal.

. FILDAT - This subroutine takes the calibration parameters and uses them
to reduce time/intensity pairs in the source file to mass/intensity pairs
and write them to the destination file. In addition, if the mass is not
within 0.15 Daltons of a unit mass, an error message is also printed to
tslficieélestination file. This latter event makes it easy to locate the 80 nsec

138

'r'rv—Tj

at

IT-Tfiv

***********

.4 LL‘

"1

139

ALLQLALALAL

"jW‘UT'V' .1~'—

CVCMASSFN "'

ALLJQILLL-LJ

Version: July 3, 1987
Edited 8/24/88 to include changes to ITR output format.
Eric D. Erickson

PURPOSE:
This program inputs a flight time and calculates the
m/z for data collected on the CV C-TOFMS. This can now

be done using a file that has been ported over from the
ITR.

PROGRAM CVCMAS

REALM TIME ! Flight time

REAL*4 K ! Slope

REAL*4 C ! Intercept

REALM MASS ! Mass to charge ratio
CHARACTER*1 A ! Space
CHARACTER'l B ! Response
CHARACTER*1 A2 ! Data source response
CHARACTER“ 15 SOURCE ! Source filename
CHARACTER“ 15 DEST ' ! Destination filename
TYPE *,’Welcome to CVCMASS.FTN’

DATA Af ’/ ! Define space

WRITE (5,1) (A, I=1,3)

FORMAT (A)

Where are the data coming from?

TYPE *,’Are the data to be input from a File ’
TYPE *, ’or the Keyboard?

READ (5,1) A2

IF(A2.EQ. ’i’)A2=’F'

IF (A2 .EQ. ’F') CALL FILEIO (SOURCE, DEST)

Is calibration routine needed?
TYPE *, ’Do you want me to calibrate the data?’
READ (5,1) B
IF (B .EQ. ’y’) B = ’Y’
IF (B .EQ. ’Y’) THEN
CALL CALIB (SOURCE, DEST, K, C, A2)
! Calibrate if answer is yes
ELSE
TYPE *, ’What is the value of k?’
ACCEPT *, K
TYPE *, ’What is the value of C?’
ACCEPT *, C
WRITE (5,1) (A, I=1,2)
END IF

140

IF (A2 .EQ. ’F') THEN
! Do calculations if files are used
CALL FILDAT (SOURCE, DEST, K. C)
GO TO 20
END IF

5 TYPE * ,'What 18 the ion flight-time 1n microseconds?
READ (5, 10, END: 20) TIME ! End input upon "Z
10 FORMAT (E 15.6)
MASS = ((TIME - CYK)‘*2
TYPE *, llThe m/z value is: ’, MASS

 

 

GO TO 5
20 END
* - SUBROUTINE FILEIO *
* This subroutine obtains the source and destination
*

filenames for V0 operations.

SUBROUTINE FILEIO (SOURCE, DEST)

CHARACTER*1 A ! Define space
CHARACTER" 15 SOURCE , ! Source filename
CHARACTER“ 15 DEST ! Destination filename
DATA Af ’/

WRITE (5,1) (A. I=1,2)
1 FORMAT (A)
TYPE *, 'What is the source filename?’
READ (5, 3) SOURCE
3 FORMAT (A15)
WRITE (5,1) A
TYPE *, ’What is the destination filename?’
READ (5,3) DEST
WRITE (5,1) (A, I=1,3)
RETURN
END

LmemL-menmmnn A.. amt-...

V'V‘UV'jUUVI‘V'jI“ vvvvvvvv

"' SUBROUTINE CALIB "‘

-Lngmm ALLLA‘L- ALLA..—

vrvvv' I v—Iw-

* This subroutine calculates k and c, the tof calibration
* parameters.

SUBROUTINE CALIB (SOURCE, DEST, K, C, A2)

DIMENSION TIME(20)
DIMENSION M(3)
DIMENSION TAVE(3)
DIMENSION FUN(3)

10

20
50

7O

75

141

REALM TIME ! Ion flight-time

REALM M ! Ion mass to charge ratio
REALM C ! Intercept

REALM K ! Slope

REALM SUMTIM ! Flight-time sum
REALM TAVE ! Flight-time mean

REALM FUN ! Function

REALM KSUM ! Sum of the slopes
REALM CSUM ! Sum of the intercepts
CHARACTER“ 15 SOURCE ! Source filename
CHARACTER"I 15 DEST ! Destination
CHARACTERM A ! Space
CHARACTERM ANS ! Correction response
CHARACTERM A2 ! File response
DATA Al’ ’/ ! Define space

TYPE *, ’This subroutine calibrates the input data.’
WRITE (5,10) (A, I=1,2)
FORMAT (A)

IF (A2 .EQ. ’F’) CALL FILCAL (SOURCE, K, C)
IF (A2 .EQ. ’F') RETURN

Input masses and flight-times for standards
D0 100 I=1,3
SUMTIM=0.0 ! initiate SUMTIM
TYPE *, ’What is mass number ’, I, ’1"
ACCEPT *, M(I)
DO 50 J =1,20
TYPE *, ’What is the flight time?’
READ (5, 20) TIME(J)
IF (TIME(J) .LT. 0.0) GO TO 70
! Negative values end data collection
FORMAT (E156)
CONTINUE

Correct input values

WRITE (5,10) (A, L=1,3)

TYPE *,’The values input for mass ’,M(I),’are:’

DO 75 L=1,J-1
TYPE *,L,TIME(L)

CONTINUE

TYPE *, 'Are there any corrections?

ACCEPT *,AN S

IF (ANS .EQ. ’y‘) ANS = 'Y’

IF (ANS .EQ. ’Y’) THEN
TYPE *,'Which value needs to be changed?’
ACCEPT *, L
TYPE *, ’What is the correct flight-time?
ACCEPT *, TIME(L)
GO TO 70

ENDIF

142

* Calculate mean flight-time
DO 80 L=1,J-1
SUMTIM = TIME(L) + SUMTIM
80 CONTINUE
TAVE(I)=SUMTIM/(J- 1)
WRITE (5,10) (A. L=l,3)
100 CONTINUE

* CALCULATE K
KSUM=0.0
DO 200 I=2,4
N=I-1
IF (N .EQ. 3) THEN
FUN(I-1)=(TAVE(N)-TAVE(1))/(SQRT(M(N))-

1 SQRT(M(1)))
ELSE
FUN(I-1)=(TAVE(I)-TAVE(N))/(SQRT(M(I))-
SQR’T(M(N)))
END IF

KSUM = KSUM + FUN(I-l)
200 CONTINUE
K=KSUMI3.0

* CALCULATE C
CSUM=0.0
DO 300 I: 1,3
FUN(I)=TAVE(I)-K‘SQRT(M(I))
CSUM=CSUM+FUN(I)
300 CONTINUE
=CSUM/3.0

"‘ PRINT RESULTS
TYPE *, ’k = ’,K, ’C = ’,C
WRITE (5,10) A

* RETURN TO PROGRAM
RETURN
END

LAALAAJL-LLLALL- ... LLLA... A.;A;

*.. WWWTWEFPPCAL *
Version: July 3, 1987
Edited for new ITR data format: 8/24/88 EDE

Eric D. Erickson

PURPOSE:
This module permits calibration of the CV C TOFMS using
the equation:

tof: k * sqrt(m/z) + C
This module operates on files imported to it from the
main program. The operator inputs the masses and

***********

‘l'

10

20

143

flight-times of three known ions and the program then
calculates mean values for k and C. A sample standard
deviation of each is also calculated.

SUBROUTINE FILCAL (SOURCE, K, C)

DIMENSION MASS(3)
DIMENSION TAVE(3)
DIMENSION FUN(3)
DIMENSION TTIM(3)
DIMENSION SUMTIM(3)
DIMENSION N2(3)
CHARACTERM A ! Space
CHARACTER“ 15 SOURCE ! Source filename
INTEGER N2 ! Number of times in ith average
INTEGER N3 ! Total number of times in ave
REALM K ! Slope
REALM C ! Intercept
REALM TTIM ! True flight-time
REALM TIME ! Flight-time
REALM MASS ! Ion mass to charge ratio
REALM TAVE ! Mean flight-time
REALM FUN ‘ ! Function
REALM SUMTIM ! Sum of flight-times
REALM KSUM ! Sum of slapes
REALM CSUM ! Sum of intercepts
REALM SDK ! Standard deviation of k
REALM SDC ! Standard deviation of C
REALM TOT ! Std dev precursor
CHARACTER‘50 TEXT
INTEGERM INT
DATA Al’ ’/
Input m/z values and approximate times
DO 201:1,3

WRITE (5,10) (A, J=1,2)

FORMAT (A)

TYPE *, ’What is the mass of ion number ’,I, ’?’
ACCEPT *, MASS(I)
TYPE *, 'What is the approximate flight-time?
ACCEPT *, 'ITIM(I)

CONTINUE

Read file for flight-times
OPEN (UNIT: 1, NAME=SOURCE, READONLY, STATUS=’OLD’,
1 DISP=’SAVE’)
DO 25 I: 1,3
SUMTIM(I)=0.0
N2(I)=O

25

30

35

40

100

110

120

300

350

400

144

CONTINUE
DO 100 J=1,10000
READ (1,30, END=110) TEXT
FORMAT (A30)
DECODE (20, 35, TEXT) TIME
FORMAT (10X, F10.3)
DECODE (30, 40, TEXT) INT
FORMAT (23X, 17)
IF (INT .LT. 50) GOTO 100 ! Weed out noise
IF (ABSCTME-TTIMH» .LT. 0.015) THEN
SUMTIM(I) : SUMTIM(1)+ TIME
N2(1)=N2(1)+1
ELSE IF (ABSCTIME-TTIM(2)) .LT. 0.015) THEN
SUMTIM(2) = SUMTIM(2) + TIME
N2(2)=N2(2)+1
ELSE IF (ABS(’HME-TTIM(3)) .LT. 0.015) THEN
SUMTIM(3) = SUMTIM(3) + TIME
N2(3):N2(3)+1
END IF
CONTINUE

Calculate mean flight-time
DO 120 I=1,3
TAVE(I) : SUMTIM(D/N2(I)
CONTINUE \
CLOSE (1, DISP:’SAVE’) ! Close file

Calculate k
KSUM : 0.0
D0 300 I=2,4
N=I-1
IF (N .EQ. 3) THEN
FUN(I-1)=(TAVE(N)-TAVE(1))/(SQRT(MASS(N))-

SQRT(MASS(1)))
ELSE
FUN(I-1):(TAVE(I)-TAVE(N))/(SQRT(MASS(I))-
SQRT(MASS(N)))
END IF
KSUM = KSUM + (FUN(I-1)* N2(I-1))
CONTINUE
K:KSUM/(N 2(1) + N2(2) + N2(3))
TOT=0.0
DO 350 I=1,3

TOT=TOT + ABS((FUN(I)'K)*N 2(1))
N3=N3 + N2“)

CONTINUE

SDK=SQRT(TOT/(N3~ 1))

Calculate C

CSUM:0.0

D0 400 I=1,3
FUN(I):TAVE(I)-K*SQRT(MASS(I))
CSUM:CSUM+(FUN(I)"‘N2(I))

CONTINUE

145

C=CSUM/N3
TOT : 0.0
DO 450 I=1,3
TOT=TOT + ABS((FUN(I)-C)*N2(I))
450 CONTINUE
SDC:SQRT(TOT/(N3-1))

* Print results to the screen
WRITE (5,500) K,SDK,C,SDC
500 FORMAT (’ K: ’,F6.3,3X,’ SDK: ’,F6.3,3X,’ C:
’,F6.3,3X,’ SDC:',F6.3)
RETURN
END

ALmL-L-LLAAALAA LQLLLLAA

Iii-II ’I’V’IT1IV‘TWITT

‘ _ SUBROUTINE FILDAT *

AA. LAALALLL++LLLAJL

' 'I

i
1.
i
4r

11"! vv‘v‘rT‘v—TTTww—v—v—Tvr

Version: July 3, 1987
Eric D. Erickson

PURPOSE:

This program converts time/intensity data to
mass/relative intensity data for the CV C TOFMS. The
results are printed to file DEST. Required inputs are
the I/O filenames, and the values of the slope and
intercept for the time conversion to mass.

This program has been modified to incorporate changes
in ITR data format. The new format has 10 digits for
each of the scan number, flight-time, and intensity
respectively. - EDE 8/24/88

*§*************

SUBROUTINE FILDAT (SOURCE, DEST, K, C)

REALM K ! Slope

REALM C ! Intercept

REALM M(200) ! Mass of the ion
REALM TOF(200) ! Flight time
INTEGER SCAN ! Scan number counter
INTEGERM INTEN(200) ! Ion intensity
INTEGERM IMAX ! Max ion intensity
INTEGER SCN(200) ! Scan number

INTEGER MA ! FRACTIONAL MASS
INTEGER I,II,N

CHARACTER’ 15 SOURCE ! Source filename
CHARACTER“ 15 DEST ! Destination filename
CHARACTER*60 TEXT ! File contents

CHARACTER"I 15 MSG(200) ! Error in data calibration

10

25

26

28

140

150

146

OPEN (UNIT: 1, FILE=SOURCE, READONLY, STATUS=’OLD’,
1 DISP=’SAVE’)
! Open source file
OPEN (UNIT=2, FILE=DEST, STATUS:’NEW’, DISP=’SAVE’)
! Open destination file
L=O
SCAN = 0
IMAX = 0
DO 1001:1,10000
L=L+1
READ (1, 10, END:110) TEXT! Read line as text
FORMAT (A30)
DECODE (10, 25, TEXT) SCN(L)
FORMAT (6X, I4)
IF (SCN(L) .EQ. 0) GOTO 100 ! Ignore blanks
DECODE (20, 26, TEXT) TOF(L)
FORMAT (10X, F10.3)
M(L) = «TOF(L) - C)/K)"‘*2 ! Calculate mass
Determine if mass is within reasonable calibration
MA = M(L)
X : ABS(M(L) - MA)
IF (X .GT. 0.15 .AND. X .LT. 0.85) THEN
MSG(L) = ' Out of range’
ELSE
MSG(L) = O 9
ENDIF
DECODE (30, 28, TEXT) INTEN(L)
FORMAT (24X, 16)
IF (SCN(L) .NE. SCAN) THEN
H=0
DO 150 N = 1,L-1
IF (II .EQ. 1) THEN
H=0
GOTO 150
ENDIF
IF (N+1 .LT. L) THEN
IF (M(N+1)-(N) .LE. 0.3) THEN
IF (INTEN (N) .LT. INTEN(N+1)) GOTO 150
IF (INTEN(N) .GT. INTEN(N+ 1)) II = 1
EN DIF
ENDIF
WRITE(2,140) SCN(N). TOF(N). M(N), INTEN (N ),
1 100.*FLOAT(INTEN(N))/FLOAT(IMAX),MSG(N)
FORMAT (I7,2X,F6.3,2X,F5. 1,2X,
1 I7,2X,F7.2,2X,A15)
CONTINUE
SCAN : SCN(L)
SCN( 1) = SCN(L)
INTEN( 1) = INTEN(L)
MSG( 1) = MSG(L)
TOF( 1) = TOF(L)
M( 1) = M(L)
L = 1
TYPE *,’WORKIN G ON SCAN ’,SCAN

147
IMAX = INTEN( 1)

ELSE
IF (IMAX .LT. INTEN(L)) IMAX : INTEN(L)
ENDIF
100 CONTINUE
110 CLOSE (1, DISP:’SAVE’) ! Close source file
DO 350 N : 1,L
WRITE (2, 140) SCN(N).TOF(N).M(N),INTEN(N),
1 100.*FLOAT(INTEN(N))/FLOAT(IMAX),MSG(N)

350 CONTINUE
CLOSE (2, DISP:’SAVE’)

RETURN
END

Appendix III:

Programs Used for the Degree of Fragmentation Algorithm

APPENDIX III

PROGRAMS USED FOR THE DEGREE-OF-FRAGMENTATION ALGORITHM ‘

This appendix contains the algorithms used to generate degree of

fi-agmentation reconstructed chromatograms.

/*

#*
it
**
*1!
13*
**
ti
**
**
**
tt
**
it
ti
**
it
**

*/

GCSIM.C
Eric D. Erickson
7/3/89

C version of GCSIM. Program to produce TII, DIFF, GS,
NP, and reconstructed mass chromatogram plot files.
This version uses CRICKET GRAPH output format. This
version was written to get around formatting problems
with data reduced on the ITR which were constantly

seen when using the FORTRAN version of this program.
A C version of CV CMASS (MASCAL.C) was written also and
this program has only been tested with output from that
version. I am not sure how this program will handle

the different tags used in the two calibration

programs. The tags in the C version of the calibration
program are "Good Calib" and "Bad Calib". Other than
this glitch, the output should be compatible between

the two programs.

#include <stdio.h>
#include <math.h>

mainO

{

char source[20], /"' Source filename */
dest[20], /"' Destination filename */
text1[10], f" good and bad markers */
text2[10]; f“ good and bad markers */

148

149

FILE *fp1,*fp2; /* File pointers */
float rithr, f" Relative intensity threshold */
ibase[2000], l" Intensity of the basepeak */

tiil2000], /" Total ion intensity */
intmas[2000], /* Intensity for selected mass */

rint, I" Real intensity */
tmp, /"‘ temporary variable */
gensig[2000], /"' Generic Sigma value "7
gslo, /* Minimum generic sigma value */
gshi, f" Maximum generic sigma value */
mass, /* m/z ratio */
srchm, /"' search mass */
delay, /‘ Chromatographic delay time */
time, /* Chromatographic elution time */
rate, /" Spectral generation rate */
nprat[2000], /" Number of peaks ratio */
relint; /"' Relative intensity */
long int inten; f" Intensity */
int i; /* counters */ ‘
int difiI2000], /"‘ Difference plot data */
scanl, f" Previous scan *I
scan, /" current scan number */
numpk[2000], /* Number of peaks *1
npk25[2000],/* Number of peaks >: 25% base peak "7
numplo, /"' Minimum number of peaks */
numphi; /* Maximum number of peaks */

/* Query for needed information */

printfl'What is the source filename?\n");
gets(source);

fpl : fopen(source, "r”);

printtI'What is the destination filename?\n");
gets(dest);

fp2 : fopen(dest, "w");

printf("What is the chromatographic delay time?");
printfl" (sec)\n");

scanfl"%f', &delay);

printfl'What is the mass spech generation rate?");
printfl" (Spectra/sec)\n");

scanfl"%f', &rate);

printt("What is the desired relative intensity");
printfl" threshold? (Percent)\n");

scanfl"%f', &rithr);

printf("Which mass would you like to search for?\n");
printf("\tEnter 0.0 to skip this search\n");
scanf("%f", &srchm);

printf("What is the minimum acceptable value for ");
printf("generic sigma? (Percent)\n\tEnter a ");
printf("negative value to skip this calculation.\n");

150

scanf("%f", &gslo);
if (gslo >: 0.0)
{
printtI'What is the maximum acceptable value ");
printfl"for generic sigma? (Percent)\n");
scanf("%f', &gshi);
)
printf("What is the minimum number of acceptable ");
printf("peaks?\n\tEnter a negative value to skip ");
printf("this calculation.\n");
scanf("%d", &nump10);
if (numplo >: O)
{
printf("What is the maximum acceptable number");
printfi" of peaks?\n");
scanfl"%d", &numphi);
)
printf("\n\nStarting data reduction.\n");

/* initialize variables */

scan1 : 0;

for (i=0; i<2000; i++)
{
difili] = 0;
intmas[i] = 0.0;
ibase[i] : 0.0;
gensig[i] : 0.0;
tii[i] = 0.0;
numpk[i] : 0.0;
npk25[i] = 0;
l

/* read data */

while ((fscanfifpl,"%d%f%flbd%f%s%s", &scan, &trnp,
&mass, &inten, &relint, &text1, &text2)) !: EOF)
{
if (scan !: scanl)
[
printf("Processing scan #%d\n", scan);
scanl : scan;
I
rint : (float)inten;
if (difllscanl < inten)
l
difiIscan] = inten-
(0.01*rithr*((float)inten));
l

if (fabs(mass-srchm) < 0.1)
intmas[scan] : rint;
if (relint > rithr)
[

tii[scan] +: rint;

151

if (ibase[scan] < rint)
ibase[scan] : rint;
numpk[scan]++;
if (relint > 25.) npk25[scan]++;
)
l

/* calculate ratios */
for (i=0; i<scan; i++)

{
/* correct for division by 0 */
if (tii[i] == 0.0)
{
gensig[i] : 100.;
nprat[i] : 1.;
l
I“ calculate the value of generic sigma and
*"‘ number of peaks ratio */
else
(

gensig[i] : 100. "' ibaseli] / tii[i];
nprat[i] = 100 "' ((float)npk25[i])/

((floathumpklil);
l
l
close(fp 1);
/* print out results ‘/
fprintflfp2,"*\nscan\ttime\t");
fprintflfp2,"tii\t");

fprintflpr,"difi\t");
if (srchm !: 0.0) fprintflfp2,"mass\t");
if (gslo >= 0.0)
[
fprintflpr,"gsig 1V3");
fpfintflfpzftsigmt");
}

if (numplo > 0)
l
fprintfifpz."np5\t");
fprintflfp2,"npti\t");
fprintflpr,"np25\t");
fprintflfp2,"inprat\t");
fprintfifpz,"nprat\n");
l

for (i=0; i<scan; i++)
[
time = delay + (((float)i)/rate) + 1./(2.*rate);
fprintflfp2," %d\t%f\t", i, time);
fprintflfp2," %f\t", tii[il);
fprintflfp2," %d\t", diffIil);
if (srchm !: 0.0)

l

152
fprintflpr," %t'\t", intmas[iD;
}

if (gslo >= 0.0)

l
if (gensig[i] == 100.)
{
fprintf(fp2," \t\t");
l
else
{
fprintflfp2," %t\t", gensig[i]);
if ((gensig[i]>:gslo)&&
(gensig[i]<:gshi))
{
fprintflfp2." 96M". tii[i]);
1
else
[
fprintflfp2, " 0.0\t");
}
l
1
if (numplo > 0)
{

fprintflpr." %d\t". numpk[i]);

if ((numpk[i]<:numphi)&&(numpk[i]>:numplo))
{
fprintflfp2," %f\t", tii[i]);
}

(
fprintflpr," 0.0\t");

}
fprintflfp2," %d\t", npk25[il);
if (nprat[i] :: 1)

{
fprintflfDZ." “3 \n");
1

else

else
(
fprintflfp2," %f\t", NOW/nprat[i]);
fprintflfp2," %f\n", nprat[i]);
l

l
close(fp2);
printf("Your data have been stored in %s.\n", dest);

}

Appendix IV:

Degree of Fragmentation Database

APPENDIX IV
DEGREE 0F FRAGMENTATION DATABASE

Symbols used in the following tabulation include:
C The number of carbon atoms in the molecule.

:3 Generic sigma value calculated as the base peak intensity divided by the
g'total Ion intensity expressed as a percentage.

NP5 The number of peaks in the spectrum with intensifies greater than 5%
of the base peak Intensity.

NP25 The number of peaks in the spectrum with intensifies greater than
25% of the base peak intensity.

R The ratio of NP25:NP5 expressed as a percentage.

153

HHHHHHHHHHHHHHHHHH HHHHHH rat-I HHHHHHHOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOO

E EEEEEE {it 5533553552‘."éfififiééfifififitfifitfitfififiéfifi 255E323?

anneawwwreeswroesrereorss» MUFNI‘NIF are alum-F@bfiHt—‘HHHWHHHN‘ONNG‘“NNQGNIOQGOIFWW IFIOOGIFGOIE

2

40.0
50.0
60.0
50.0
50.0
zoo
25.0
75.0
25.0
50.0
50.0
50.0

Phosphorous Oxychloride
Chlorodifluoroamine
Sulfur Dichlorlde
Hydrazine

Hydrogen Bromide
Sulfuryl Chloride Fluoride
Phosphorous Oxychloro
Difluoride

Sulfuryl Fluoride
trans-Difluorodiazine
Silane

Difluoroamine
Hydrosoic Acid
Phosphorous Oxifluoride
Thionyl Fluoride
Oxygen Difluoride
Hydrogen Sulfide
Sulfur Tetrafluoride
Ammonia

Nitrogen Trifluoride
Phosphine
cis-Difluorodiasine
Tetrafluorohydrasine
Sulfur Dioxide
Nitrogen Dioxide
Silicon Tetrafluoride
Nitrous Oxide

Sulfur Hexafluoride
Hydrogen Chloride
Sulfur Oxytetrafluoride
Air

Water

Nitric Oxide

100.0 Nitrogen

36.4
36.4
66.7
57.1
310
50.0
625

25.0
27.3

44.4
22.2
50.0
15.4
40.0
18.7

33.3
37.5

Trifluoromethanethiol
Chlorobromomethane
Carbon Tetrachloride
Formic Acid

Methyl Bromide
Ammonium Carbamate
Trifluoromethyl-
iminosulfurdifluoride
Carbonyl Fluoride
Trifluoromethanesulfenyl
Chloride

Fo rmamide

Methyl Mercaptan
Dichloromethane
Bromodichloromethane
Nitromethane
Difluorochlorobromo-
methane
Bromofluoromethane
Methyl-N,N-Difluoroamine

100.0 Methylamine

40.0
sec
27.3
40.0

Cyanogen Bromide
Chloroform
Chlorofluoromethane

Difluoromethane

100.0 Formaldehyde

42.9
18.2
10.0
60.0
429
«10
80.0
40.0
40.0
75.0

Methylene Chloride
Fluorochlorobromomethane
Difluorobromomethane
Trifluoromethane
Carbonyl Chloride Fluoride
Methanol

Carbonyl Chloride
Trichlorofluoromethane
Fluoromethane

Methane

NNNNfllONNNNNNNNNNN NNNNNNNN NNNN ”NNNNNNNNNNNN NNNNNNHHHHHHHHHHHHH HHHO

bmuaabmuuhfinmauaa ”fiflfllduafl *GGQ QGGQQOIIFQGGOOQ QQQOQNHHHHHO—‘HNHNNNN ”ION:

R Carla-n!

40.0 Dichlorodifluoromethane

33.3 Dichlorofluoromethane

ass Trifluoromethylsulfirr
Trifluoride

ass Carbonyl Sulfide

50.0 Methyl Nitrite

50.0 Cyanogen Chloride

40.0 Methyl Nitrite

14.3 Bromotrifluoromethane

50.0 Iodomethane

16.7 Chlorodifluoromethane

16.7 Carbon Disulfide

20.0 Chlorotrifluoromethane

33.3 Tetrafluoromethane

50.0 Hydrogen Cyanide

25.0 Carbon Dioxide

100.0 Carbon Monoxide

25.0 Acetyl Bromide

41.2 Dichloroacetic Acid

sao 1,1,2-Trichloroethane

38.9 1,1-Dimethylhydrasine

39.1 Ethanedithiol

41.2 2-Chloro-1,1,2,2-
Tetrafluoroethane

44.4 Trichloroethylene

34.8 Dichloroacetyl Chloride

30.0 2-Meroaptoethanol

37.5 Ethanethiol

30.0 Chloral

38.8 Bromoacetic Acid

18.0 Dimethylsulfone

29.4 Chlorodifluoroacetic Acid

45.5 Tetrachloroethylene

50.0 Methyl Sulfate

54.5 Chloroethane

54.5 Methyl Carbamate

46.7 Chloromethyl
Dichloromethyl Ether

58.3 sym-Dimethylhydrazine

27.8 2-Bromoethanol

42.9 Acetamide

25.0 1,2-Dichloro-1,2-
Difluoroethylene

45.5 Trichlorofluoroethylene

62.5 Glyconitrile

27.8 Chloroacetaldehyde

38.9 1,1,1,2-Tetrafluoroethane

33.3 Methylthiocyanate

8.4 Chloroacetonitrile

18.2 Ethylene Bromohydrin

8.3 1,1,2-Trichloro-2—
Fluoroethane

60.0 2-Thiapropane

54.5 Acetic Acid

35.7 Fluoroacetic Acid

429 Dichloroacetylene

54.5 Dimethyl Sulfide

3.3 1,2-Dichloro-1-Fluoroethane

40.0 Diazoethane

25.0 Chloroacetylchloride

41.7 1,2-Dichloroethane

33.3 1,1-Dichloro-1-Nitroethane

7L4 Bromoethane

sos 2,3,4-Trithiapentane

41.7 4-Amino-1,2,4-Triasole

41.7 Chloropentailuoroethane

27.3 1,1,1-Trichloroethane

320 Dimethyl Sulfoxide

1,2-Dichloroethylene

ESBBEQ°EEEECKKCB ESSEEE‘N‘S 335E EHflKfi’SREBEBBE 58535““*’°“°‘°’°’*“°‘**“ 4°“;

i

RONNIONNNNNNNNNNN NNNNNNNNNNMNNN NNNNNN N NIONNNNNNN NNNNNN NN NNNNNNNN NO

25.6

25.6
258

3.5
39
27.0
27.0
27.4

27.6
$1

$8
@.1
39
30.0
$3

$3
305
31.1
31.3
31.4
31.7
319

33.1
33.1
33.7
34.6
348
349

35.3
35.6
359
3.2
$3
372
37.4
37.4
379
379
38.7
3.1
39.7
3.9

40.3
40.4
409
41.1
41.1
41.2
41.2
41.3
41.4
41.6
419

43.0

“HwawaWWNO-‘flufl NGWhNNOfiNO‘WCfiNNN ”WWWIFQ £0 weebb-burooae WWN’CDIF Nth hfiwbfimmm DIE

E E§§§§§E§ Eu

16.7

EEEfiE

310
50.0
16.7
40.0

50.0
57.1

57.1

E

57.1

14.3
60.0

CW
1-Chloro-L2,2-
Trifluoroethylene
2,3-Dithiabutane
Vinyl Fluoride
1,1-Difiuoroethane
Trifluoroethylsne
Trichloroacetic Acid
Trifluoroacetic Acid
Ethylene Imine
1,1,1-Trifluoro-2-
Chloroethane
Dimethyl Peroxide
1.1-Dichloro-2,2-
Difiuorosthans
2-Mercaptoethanoic Acid
Dimethylamine
Ethylene On'de
Bromo-1,2-Difluoroethylene
Dichlorodimethylsilane
1, 1-Dichloro-2,2-
Difluoroethylene
1,1-Dichloroethylene
1-Chloro-2-Fluorosthylene
Ethanolamine
1,1-Dichloroethane
Vinylidine Chloride
1-Chloro-2-Bromosthans
Ethyl Iodide
Methyl-i-Thiocyanate
1,2-Dichloro-1,1-
Difiuoroethane
1-Chloro-1,2-
Difiuoroethylsns
Iodoacetylsne
1,1-Chlorofiuoroethylene
Ethyl-N,N-Difluoroamine
Ketene
Clorodifluoroacetaldehyds
1,1-Dichloro-2,2,2-
Trifluorosthane
Acetyl Chloride
N itrosthane
Glycolic Acid
Ethyl Nitrate
Dimethyl Sulfite
Tetrafluoroethylene
Trichloroacetonitrile
Ethanal
Ethylamine
Chloroacetic Acid
Methyl Formats
Chloroacetylene
Vinyl Chloride
1-Chloro-2,2-
Difluoroethylene
Fluorosthane
Chloromethyl Ether
Glycerolsldehyde
Chloral Hydrate
Ethanol
Oxalyl Chloride
Azomethane
1,2-Difiuorosthans
Ethylene
Vinyl Bromide
Dimethyl Ether
Trifluoroacetonitrile
Bis Chloromethyl Ether
2,2,2-Trifiuoroethanol
Diclorofluoroethylene

wamwwwwwwwwuwce awesome:mauwwwuwwmwwwwumwwwauwwmwumw “NNNNNNNNNNNNNNNNNNNO

z‘mstw Bo. Carpool!

432
45.2
456
46.6

46.7 i

468
48.2
48.6
49.6
50.6
£1
56.6
569
58.4
58.5
588
62.6
NA
76.6
10.2

118
129
13.2
13.3
13.5
14.2
14.2
14.5
15.2

158

16.0
16.3
16.5
16.7
16.7
168
169
17.0
17.4
17.5
17.6

mhbbmmmaanGQG thereon-normed».manquqaqbaqummoqmmwsogmm BHHHHIOHHNHHNHNNWNHNN

KEKSSEHEBBl-Tfiififi Efififimfififififififififififififififi85886585“88558 IS“"‘°°°'°"“°*“‘°’“°°°‘°°“‘“;

28.6
14.3
25.0

60.0 Eth

33.3
33.3
14.3
33.3
14.3
25.0
50.0
16.7
25.0
40.0
20.0
3&3
210
33.3
37 .5

20.5
3.6
66.7
$0
45.5
429
47.1
30.3
46.7
47.4
34.8
20.7
18.5
$8
$8
18.2
46.7
37.5
41.2
”.4
58.3
22.7
368
50.0

35.7

01-Bromo-2-Chloroethylene
Acetonitrile
2,2,2-Trichloroethanol
Ethylene Glycol

ans
1,1-Chlorofiuoroethane
Methyl Chloroformate
Glyoxal

Oxalic Acid
2-Fluorosthanol
1,1,1-Trifluoroethane
Hexafluoroethane
Fluoroacetylene
1,2-Ethanediamine
Ethanesulfonylfluoride
2-Chloroethanol
Dichlorfluoroacetonitrile
Chlorodifluoroacetonitrile
Acetylene
1,1,1-Trifluoro-3-
Chloropropans
1,1,1-Trichloropropanone
1-Chloro-2-Propanone
Methyl Bromoacetats
1-Propanethiol
Trimethylhydrasine
1-Chloro-2-Bromopropane
Ethyl Carbamats
2,3-Dichloropropionic Acid
2-Cbloropropionic Acid
1,3-Propanedithiol
2,3-Dichloroacrylic Acid
1,1-Dichloropropanone
2,2-Dichloropropionic Acid
1,2-Dichloropropane
Methyl Chloroaoetats
1,1-Dichloropropene
3,3,3-Trifiuoro-1-Propens
3-Chloro-1,2-Propanediol
Propanoic Acid
1,3-Propanediol
3-Bromopropionitrils
1,2-Dichloropropens
Epichlorohydrin
1-Chloro-3-Bromopropane
1,2,2-Trichloropropene
Acrylamide

Allyl Alcohol
2-Propanethiol
1,1-Dichloro-1-Fluoropropane
1-Bromo-2-Propanol
Vinyl Formats

Glycidol

Propylene Sulfide
Formaldehyde Dimethyl
Hydrazone

2-Thiobutane

Propylene Oxide
Imidazole
1,2-Epoxypropane
1,2,3-Trichloropropane
Dichloromalonitrils

N, N -Drmethylformamide
Dimethyl Carbonate
Ethyl i-Thiocyanate
2,3-Dichloro-1-Propanol
Allylamine
1,2,2-Trichloropropane
1-Nitropropane
Trithlane

W (O aseessceseece«maneuver-recessoeseoseceeesccescceeteasascceses sotemcoeeesoreeeeececeeeeeooeooeeoeeeteoscececceeaccurate “O

u N unnanmbunuuewuwuwuuuuueaawumawhen «hwwahhabbhfiewmmeaauqqambmwahu mg
5 3 656m5m°353§¢5“@K°¢55°°°°°3Km§°q55 3°”Emexxfigfinfirfi“5°5535“336555“ S;

53 En

E§E§§§ EEEEEE§E§§E§E t

g?

EEEEéEE

27.3
21.4

82

30.0

25.0
27.3
20.0

$5

375
20.0
375
210
16.7
16.7

Cancun!
3,3,3-Trifiuoro-1,2-
Chloropropens
Propionaldehyde
3-Chloropropanenitrils
Glycerol
2,3-Dichloropropene
Cyclopropane
1,2-Dichloro-2-Fluoropropane
Acrylic Acid
Acrolein
n-Propyl Nitrite
Propiolactone
Hexafluoropropene
Ethylene Carbonate
Acrylonitrile
Ethyl Formats
2-Chloroethylchloroformate
Carbon Suboxide
3,3,3-Trichloropropene
1,3«Dichloropropanone
Epibromohydrin
2-Hydroxypropanenitrils
2oNitropropane
1,1,2-Trichloropropane
Trichloropropionitrils
3-Chloropropionic Acid
2,2-Dichloropropane
1,3-Dioxylane
Chloromethoxy Acetic Acid
Propionyl Chloride
1,3-Dichloropropene
2,2-Dichlor0propionyl
Chloride
3-Chloropropene
2-Chloropropene
Propanal
Pyruvaldehyds
Propane
Methyl Carbonate
Trimethylamine
3-Chloroacrylonitrile
Propylene
Chlorotrimethylsilane
Lactic Acid
Acrylyl Chloride
1-Bromopropane
3,3,3-Trifiuoropropyne
Propionamide
2-Fluoropropene
l-Chloropropene
3-Chloro-1-Propanol
1.1.2.2,3-Pentafluoropropane
1-Bromo-2-Propyne
Allyl Fluoride
N-Methylethylenimine
Dihydroxy Acetone
Pyruvonitrile
2-Propyn-1-ol
Trimethylene Oxide
1-Fluoropropane
3,3-Dichloropropene
Propylene Glycol
1,3-Dichloropropane
l-Chloroo2-propanol
Trimethylenediamine
3,3,3-Trifluoro-1-
Chloropropens
Ethylene Glycol
Monoformate
Propargyl Alcohol

I5 fifibbhfllfiflhbbhb AAAAA-I-Aeseeecoeeeeeosccwuosccteeeceeeeo “flute?“ wouneeeeeeoeeoceceoeocce essenceweoeuecon

Z‘MNB R W

35.5 3 9 33.3 l-Chloropropane

358 2 10 21.0 Propanenitrile

$9 3 8 37.5 2-Chloropropane

3.5 4 4 100.0 Aliens

$8 3 10 31.0 2-Bromopropane

37.7 4 7 57.1 Malonic Acid

38.0 1 10 10.0 Pymvic Acid

38.0 4 9 44.4 2-Methoxy-1-Ethanol

385 2 6 33.3 1,3,3-Tr'lfluoro-3-
Chloropropens

$3 3 10 31.0 2-Amino-1-Propanol

40.1 4 5 81.0 Propyne

40.3 2 7 28.6 Thiacyclobutane

40.4 3 7 42.9 Methyl Dichloroacetate

41.9 3 8 37.5 Propargyl Bromide

43.0 3 4 75.0 Vinyl Methyl Ether

43.2 2 7 28.6 Propargyl Chloride

43.7 3 5 60.0 s-Trioxane

43.7 4 5 81.0 Allyl Bromide

44.1 3 5 60.0 1,3-Dichloro-2-Propanol

44.2 1 8 125 1-Amino-2-Propanol

44.7 3 4 75.0 Dimethoxymethane

448 1 9 11.1 1,1,2-Trifluoro-2oChloroethyl
Methyl Ether

45.0 3 7 429 1,1,1,3,3,3-Hexafiuoropropane

45.4 2 6 33.3 1,2-Diaminopropane

478 1 7 14.3 Acetol

49.0 2 6 33.3 Methoxyacetic Acid

49.4 1 6 16.7 3-Brorno-3,3-Difluoro-1-
Propane

50.0 1 10 10.0 n-Propanol

50.1 2 6 33.3 1,3,5-Trioxane

51.0 1 8 12.5 1,2-Propanediol

51.4 1 7 14.3 l-Chloro-2,2-Difiuoropropane

51.6 1 5 21.0 Trimethyl Silanol

52.3 1 7 14.3 2-Methoxyethanol

52.4 2 5 40.0 2-Fluoropropans

53.9 1 7 14.3 i-Propanol

55.0 1 6 16.7 Chloroaeetone

555 1 6 16.7 1-Amino-3-Hydroxypropane

56.7 2 5 40.0 Acetone

573 1 3 33.3 Propylenediamine

58.4 1 4 25.0 n-Propylamine

61.2 1 5 21.0 Hydroxy-2-Propanone

..7 1 5 21.0 Methyl Acetate

65.3 2 4 50.0 i-Propylamine

66.3 1 3 33.3 2,2-Difiuoropropane

10.6 11 a) $7 1,2,3,4oDiepoxybutane

11.1 12 24 50.0 3-Methoxypropylamine

122 9 27 33.3 Butyryl Chloride

12.2 10 x 40.0 1.2,4~Butanetriol

13.0 7 5 28.0 2-Methyl-2-Propen-1-ol

13.0 7 24 29.2 Butyne-1,4-diol

13.1 11 21 52.4 Tetramethylammonium
Hydroxide

13.2 3 17 17.6 1,2-Dichlorobutane

13.3 10 m 38.5 2-Methy1-1,3-Thioxalane

13.5 9 24 37.5 Diethanolamine

13.9 7 fl 318 2,3-Dichlorobutyric Acid

14.2 8 17 47.1 n-Butylnitrate

14.4 7 18 38.9 2-Butanethiol

14.9 7 E 31.4 2-Butene-1,4-diol

15.0 7 a 31.8 '1,4-Butanedithiol

15.1 8 fi 348 2,3-Dichloro-2-Butsne

15.1 8 16 50.0 i-Butyryl Chloride

15.3 6 14 429 l-Fluorobutane

15.3 9 16 56.3 2(2-Chloroethoxy) Ethanol

15.4 8 % zae 3-Chloropropyl
Chloroformate

15.4 7 21.

33.3 l-Butanethiol

Ibublbublh-hub IF Ibubtb chub-b bfiflufifihfibhbébfi hhbbbhhhbhhbfibhh fihbhfibbhhfi-h hub ububub-bblbo

2 g m5 N5 R
15.4 7 17 41.2
155 7 14 50.0
15.7 9 17 52.9
16.1 8 24 33.3
16.1 8 % 36.4
16.3 6 m 23.1
16.4 5 21 238
16.4 5 21. 238
16.4 8 M 57.1
16.5 6 2) 310
168 9 21 429
168 7 16 438
17.0 8 17 47.1
17.2 3 E 10.3
17.2 6 16 37.5
17.3 8 16 50.0
174 5 18 27.8
17.6 7 14 50.0
179 5 22 227
18.2 4 2) ”.0
18.2 5 15 33.3
18.2 7 14 50.0
18.3 7 17 41.2
18.3 7 14 50.0
18.4 6 15 40.0
18.6 6 14 429
18.6 7 15 46.7
18.7 8 12 66.7
188 2 10 210
188 4 17 23.5
19.0 4 21 19.0
19.0 5 17 ”.4
19.1 6 15 40.0
19.3 5 17 29.4
19.4 4 16 25.0
195 6 13 46.2
19.6 5 17 ”.4
19.6 5 14 35.7
19.7 6 19 31.6
19.7 6 17 35.3
198 6 15 40.0
198 6 15 40.0
199 6 18 33.3
19.9 6 17 35.3
21.2 4 2) 250
m2 5 21 m8
”.4 7 15 46.7
21.5 7 18 38.9
21.6 6 21 28.6
”.7 6 18 33.3
ms 5 2) 25.0
21.8 5 16 31.3
21.0 4 16 25.0
21.0 5 18 278
21.1 4 19 21.1
21.1 5 15 33.3
21.2 5 14 35.7
21.3 4 12 33.3
21.4 6 11 54.5
21.5 4 14 28.6
21.5 5 16 35.7
21.6 4 19 21.1

Diethyl Sulfide
2-Chlorobutans
2-Methyl-1-Propanethiol
Hydroxybuteric Acid
Cyclopropyl Carboxylic Acid
Methyl 2,3-
Dichloropropionate
3,4-Dithiahexane
1,3-Dichloro-2-
Methylpropene
Butadiene Dioxide
Methyl Vinyl Carbinol
Methyl Hydracr'ylate
Crotonic Acid
Aldol
2-Chlorocrotonaldehyde
2,3-Dithiahexane
i-Propyl Formats
3,4-Dithiahexane
n-Butyl Nitrate
1,1,3-Trichloro-2-
Methylpropene
1,3-Butansdiol
Butenal
Allyl Formats
2-Ethylethylenimine
Butanal
2,5-Dihydrofuran
Methyl Allyl Ether
l-Butanol
Morpholine
4-Bromobutyronitrile
1,4-Dichloro-2-Butene
2-Chlorobutyric Acid
1,4-Dithian
1,4-Butanediol
i-Crotonic Acid
1,2-Bis(Methyl Mercapto)
Ethylene
Ethyl Vinyl Ether
3-Chloro-2-Methylpropene
Dioxane
2-Aminopyrimidine
Divinyl Sulfide
3-Chlorobutyric Acid
Chloroacetal
Diethyl Sulfate
Vinyl Glycol Ether
1,1-Dichlorobutane
2,3-Dichloro-1,4-Dioxane
Butyrolactone
Methyl Mercapto
Propionaldehyde
2-Buten-1-ol
Vinyl-2-Chloroethyl Ether
(2-Hydroxyethyl)-Ethyl
Sulfide
2-Fluoro-1,3-Butadiene
Ethoxyacetic Acid
3,3,3-Trichloro-2-
Methylpropene
2,3-Dichloro-2-
Methylpropionaldehyde
3-Butanoic Acid
Butyramide
Methyl Vinyl Ketone
1,3-Butadiene
Trichlorobutane
3-Methyl-2-Oxazolidinone
Allyl i-Thiocyanate

ubib‘hlbub bhbtfihbfih bbhbhbhbh lb fifibibhbbbh #bbbbbbbbb bbhbbbh‘bbbhb rhubabab hbc

1'- s
21.7
21.3

22.1
22.1
22.1
223

22.3
22.3
m
£7
228
22.3
22.3
3.0
23.1
3.3
23.4
$4

23.4-

23.5
3.7
$7
23.7
$9
24.0
24.1
24.1
24.2
24.3

24.3
24.4
24.5
24.7
248
249
25.0
5.2
3.2
5.7

5.7

$8
59
3.2
ms
85
8.7
3.7
88
$9

$9
39
27.2
275
27.6
27.6
27.7
278

3.6
8.7
$9
3.1
8.1

ummbb wnbhbbmm nuanauuah w umttduAAGG bhowbbmaaa thummmaaummbq mama QGE
68666 “$553353 EEEEEHE”: 5 :‘66555565 KESK°25555 5555535353555 5355 54

8.4
53.8

188

35.7
35.7

46.2
58.3
222
33.3
429
25.0
35.7
40 0
57 1
37 .5
40 0
188
m8
m8
46.2
38.5
46.2
46.2
44.4
44.4
21.4
25.0
25.0
3.6
38.5
50.0
23.7
28.7
25.0
58.3
40.0
25.0
55.6
21.4

$1
8.6

55.6Pyrro

31.0
45.5
50.0
38.5
50.0
21.0
13.3

45.5
60.0
$8
m
303
33.3
41.7
25.0

sea
40.0
273
2&1
273

Canard
3-Methylpyrssole
2-Methyl-1, 3.
Dioxyacetylpentane
2-Methylpropenal
Diethylamine
Methyl-3-Chloropropionate
2-Chloro-1-Methyl
Chloroformate
cis-2,3-Epoxybutane
Tetramethoxysilane
Methyl Propyl Sulfide
2,3-Dichlorobutane
2-Thiophensthiol
Tetrahydrothiophens
Methyl i-Propenyl Ether
Piperasins
3-Methoxy-1,2-Propanediol
n-Buteric Acid
3,4-Dichloro-1-Butsne
Methyl 2-Chlor0propionats
Methyl 2,2
Dichloropropionate
i-Butylsne Oxide
Hexafluoro-2-Butyne
l-Butyne
trans-2,3-Epoxybutane
Cyclobutane
Bis(2-Chloroethyl) Ether
Methacrylic Acid
Methyl Cyanoacetate
Dichlorotetrahydrofuran
1,2-Bis(Methylmercapto)
Ethane

Acetone Cyanohydrin
Ethylene Diformats
2-Thiapentane
1,4-Thioxane
1,1-Dichloro-3-Buten-2-one
Acetaldazine

Malsic Acid

Glycerol Monomethyl Ether
1,2-Epoxy-2-Methylpropsne
1,2—Dichloro-2-
Methylpropene
2-Amino-4-(Chloromethyl)-
Thiazole
2-Methoxypropene

1e
1,4-Dichloro-2-Butyne
Vinylidine Cyanide
N,N-Dimethylacetamide
Vinyl Acetonitrile
1,2-Butadiene
Fumaronitrile
2,2-Dichloro-4-Hydroxy-
butyric Acid Lactone
1,1-Dimethoxyethane
Dimethylketene
2-Bromoethyl Ethyl Ether
Ethyl Sulfite

Butadiene Monoxide
Thiophene

2-Butene
2-Chloro-1,2-Difluorovinyl
Ethyl Ether
2-Hydroxy-i-Buteric Acid
2-Bromobutane
1-Bromo-2-Methylpropans
2-Msthylpropanenitrile
1,2-Butanediol

IFIIIhO-Ib chubcb abhlbblbblhub-bb hbfilfibhhufiubnbhabit-blink##hhhbbfihbbbhhhfiblfib hhbbb-fi fifihfibbn

QHNNt-A H0300 HfiHNRth-IION ”uanNNNwNO-InwwwwwwNthwIF-Olbbhnbbh0909‘s uncommon-Is common-bung
“menace was antenna-evade: madafiuusdggmomxscgmqasmumaumsusm064 scwfifis fiufiagfia

R
33.3
33.3
21.0
25.0

27.3
25.0

30.8
23.1
41.7
33.3
33.3
40.0

57.1
30.0
33.3
50.0
40.0
36.4
21.0
44.4
36.4
318
55.6
36.4
37.5
40.0
66.7
3.6
25.0
25.0
27.3
31.0
21.4
37.5
33.3
25.0
7.1
16.7
429
21.0
15.4
18.2
8.3
50.0
$6
50.0
22.2

40.0
$6
11.1
57.1
40.0
28.6
222
11.1
222
8.3

30.0
60.0
16.7
11.1

11.1
41.7

cm
Ethylene Cyanide
Butanoic Acid
1,3oDichlorobutane
1-Chloro-2-Butenol
t-Nitrobutane
1,1-BMMethylmercapto)
Methyl Sulfide
l-Chlorobutane
3-Butyn-2-ol
Thiadimethyl Acetal
Malic Anhydride
2-Chloro-2-Methylpropane
1,1-Bie(Methylmercapto)
Ethane
t-Butylbromide
Pyrrolidine
Acetoin
2-Ethoxyethanol
2-Butyne
Diethyl Peroxide
Methyl Acetyl Carbinol
Methylmalonitrile
3,4-Epoxy-1-Butene
Phthalic Anhydride
N-Butane
l-Butene
Methyl Propionate
Methyl i-Pmpyl Ether
Trimethyl o-Formate
Diethyl Ether
3-Hydroxy-2-Butanone
l-Bromobutane
1,4-Dioxane
2-Methylpropene
3-Butyn-1-ol
1-Chlom-2-Methylpmpane
N-Methylol Acrylamide
Ethyldifluoroacetate
1,4-Dichlorobutane
Ethyldimethylamine
Succinic Acid
Chlorobutyronitrile
Diketene
Tetramethylenediamine
2-Amino-1-Butanol
n-Butonitrile
Furan
1, 1, l-Trifluom-Z-Butanone
2-Chloro1nethyl-1,3-
Dioxolane
Ethyl 'h'ifluoroacetate
Perfluorocyclobutene
2-Butanol
1-Buten-3-yne
Succinic Anhydride
2-Chloroethyl Acetate
Methyl Acrylate
aec-Butanol
Ethyl Dichloroacetate‘
2-Methyl-2-Amino-1-
Propanol
Ethylene Glycol
2-Methylpropane
2-(Dichloromethyl)-1,3-
Dioxolane
2,3-Butanediol
Diethylene Glycol
Perfluoro-l,3-Butadiene
aec-Butylamine
Chloroprene

OIOIOIOIOIOIOIGOIOIGOIOIDIG!OIOIGGGGOIOIOIGGOIOIOIGOIOIGOCGGOIOI OI Gamflufihhbbfi fibbbbbfibbéfilfihhblfi-bo

ZIMNB R Capt-In!

474
47.4
47.8
48.0
48.3
49.0
493
”5
50.8
51.3
514
53.1
534
539
54.1
56.3
57.6

58.0
eon
63.3
645
0.8
729
9.5

9 7

10.3
114
11.5
12.5
12.8
129
13.0
13.1
13.3
13.6
13.7
14.1
14.2
14.2
14 7
148
14.8
14.9
149
149
15 1
15.1
15.2
15.4
154

15.6
15.7
15.8
159
159

16.0-

164
16.4
16.4
16.5
16.5
16.7
16.7
17.4
17.4
17.5
17.6

oodcocaoaauooeoncameramanomqmmmmoqocmcsmcpamsmoo cc figsaxHI-‘HHHH HnoeHHHHHHHHIeh-Hwto»

KEEEEEK’SSSEEK885688358858868885385236912 8 88898“°"‘*°‘°‘ ““‘°‘°°S°“°““°"°°°’°

22.2
33.3
50.0
11.1
14.3
33.3
14.3
14.3
11.1
20.0
10.0
16.7
16.7
21.0
75.0
40.0
25.0

2110
21.0
25.0
25.0
21.0
33.3
56.0
54.2
38.5
48.1
27.6

33.3

37.5
$0
421
83.3
3.0
31.8
50.0
56.3
36.4
50.0
39.1
421
45.0
38.1
41.2
429
320
47.1
33.3
32.0
35.0
«21
52.9
39.1
52.4
57.1
45.0
38.9
56.3
57.1
41.2
57.1
31.3
37.5
31.6
47.4
38.9
57.1

Dicyanoacetylene
2-Methoxy-1-Propanol
Pyrizine

Methyl 2-Hydroxypropionate
l-Methoq-Z-Propanol
t-Butanol

Methyl n-Propyl Ether
2-F1uoro-2—Methylpropane
2-Hydroxyethyl Acetate
Tetramethyleilane
Ethyl Acetate

Methyl Ethyl Ketone
t-Butylamine
2,3-Butanedione
2-Vinyloxyethanol
1,3-Butadiene
4-Chloro-4,4-Difluoco-2-
Butanone

n-Butylamine
3-Chloro-2-Butanol
Diacetyl

2-Chloropropyl Methyl Ether
Vinyl Acetate

Acetic Anhydride
3-Penten-1-ol

Valeryl Chloride
2-Chloro-2-Methylbutane
3-Methylthiacyclopentane
1,1-Dichloro-2,2-
Dimethylcyclopropane
1,3-Bie(Methylmercapto)
Propane

n-Butyl i-Thiocyanate
t-Amyl Chloride
l-Chloropentane
3-Methyl-1-Butanol
Furfuryl Alcohol
2«Methyl-3,4-Dithiaheaane
2-Methyl-1-Butanethiol
aec-Amyl Chloride
eec-Butyl i-Thiocyanate
2-Pentanethiol
3-Thiahexane
2-Methylpyrrolidone
i-Butyl Formate

Ethyl 3-Chloropropanoate
3-Methylbutanal

i-Amyl Nitrate
2-Methyl-3-Thiapentane
1,4-Pentadiene
Dihydropyran

Ethyl i-Propyl Sulfide
2-Methyl-1,3-Butadiene
1,3-Pentadiene
Tetrahydropyran
Cyclopentanethiol
3-Methyl-1-Butanethiol
2-Methoxyethyl Ethenyl Ether
4-Pentenal

3-Furoic Acid
4-Methyl-n-Dioxane
1-Pentanol
3-Chloropentane
3-Pentanethiol
2-Methylbutanoic Acid
Dimethyl Malonate
2-Thiahexane
3-Methyl-2-Butanethiol
l-Pentyne
2-Methyl-1-Butanol

OIOIGGGOIGOI OI mama:cacao-atone!amummmummmmauamuuummmummmmmmm GUIOIOICROIOIOIOIOIGOIOIOIOIOIOIOIOIO

weq¢wqwa a phmoamabm¢upa.Aaawamana$auammaqahaamqan mag¢ammqmaqqmmuqmqa§
3585855” I fififlsfififififififififlfififiEMEEKQEEHKEGKKBSGKEKEEE SG5KE$BEGE§KKKBBEEEZ

h

61.5
25.0
57.1
57.1
50.0
318
47.1

31.0
28.6
”.4
37.5

37.5
57.1

70.0

Omani
2-Pentyne
2-Methyltetrahydrofuran
N,N-Diethylformamide
Valerolactone
3-Methylfuran
3-Methylhutyle Nitrite
i-Amyl Nitrite
2-Methoxyethyl Vinyl Ether
2,3-Pentadiene
N-Valeraldehyde
2-Methyl-2oButanethiol
3-Methyl-1,2~Butad.iene
3-Methyl-2-Butenal
N-Ethylacrylamide
i-Valeraldehyde
Spiropentane
2-Furoic Acid
2-Methylfuran
4-Hydroxy-2-Pentenoic Acid
Lactone
3-Ethoxypropionaldehyde
i-Valeric Acid
N ,N -Dimethyl Acrylamide
l-Pentanethlol
3-Ethoxy-1-Propanol
Tetrahydrofiirfiiryl Alcohol
Dimethylmalonitrile
2-Chloropentane
Piperidine
Valeronitrile
N-Methylpyrrolidine
1,5-Dichlompentane
eec-Amyl Nitrate
2-Methylbutyraldehyde
2-Methylpentane
Diethyl Carbonate
2,2-Dimethoxypnopane
Ethyl Lactate
1,1-Dimethoxypropane
Methylenecyclobutane
2,2-Dimethyl-1-Propanethiol
3,5-Dimethylpyrrazole
Senecioic Acid
2-Methyl-1-Buten-3-yne
Dimethylpmpiolactone
Tiglaldehyde
1,3-Dioxepo5-ene
2-Pentyne
N-Methylpyrrole
2-Chlorovaleric Acid
Diethoxy'methane
2-Methylthiacyclopentane
2-Chloroethyl Pmpionate
Cyclopentanone
Methyl i-Propenyl Ketone
2-Furaldehyde
i-Amyl Chloride
2-Methyl-2-Butenoic Acid
3-Chloro3-Methylhutyro-
nitirle
Hydroxyvaleric Acid
Lactone
n-Pentylbromide
1,1-Metho ane
1,2-Dimethylcyclopropane
3,4-Pentanedlol
3-Methyl-1-Butyne
2-Bromopentane
l-Bromofl-Methylbutane
1,5-Pentanediol

OIGRDIGOIOIG OIOIOIGMGGOIGQGOIQG GGGQGGOIOIQOI manque: QGQGQGGGQGGGG UGOIOIOIOIGGUUOIQGOIOGO

“whwhwa wawhmwhabbwwah unambbmbwm watt“ unmanaawauqau ommw+uuuwaautaub§
5&55353 555¢pfifiosmfi°5° 5¢°55°3§33 ‘wfififi Sm‘fiEEESHESSE afiufiafifioggxsxafixz

41.7

m
Methyl t-Butyl Sulfide
Ethoxy Propionltrile
Butanediol Formal
3,3-Dimethyl-3-Thiabutane
Difluoroallyl Acetate
341oThiaethyl) Thiophene
Methyl Vinylidine Cyanide
3-Bromopyridine
3-Methyl-3cButen-2-one
3-Methcaybuteric Acid
3-Penten-1-yne
2-Chloropyridine
1-Bromo-3-Methylbutane
l-Bromopentane
2,2-Diethyl-1-propanol
4-Hydroxy-3-Methyl-2-
Butanone
2-i-Propoxyethanol
1-Chloro-3-Methylbutane
1-Pentene
Methyl Crotonate
2-Methvl-2-Butanol
3-Butemyl i-Thiocyanate
Ethylcyclopmpane
i-Amyl Bromide
Methyl Hydrogen Succinate
Vinyl Acrylate
Ethyl Cyanoacetate
Methyl Cyclopropyl Ketone
1,1-Dichloro-2-Vinyl-
cyclopropane
Furi’uryl Mercaptan
1,3-Cyclopentadiene
Trimethylacetaldehyde
l-Ethoxy-z-Propanol
1.1-BMEthylmenapto)
Methane

2-Methyl-1-Butene
Glutaronitrile

Methyl Methracrylate
2-(2-Methoxyethoxy) Ethanol
2-Bromo-2-Methylbutane
Pentanoic Acid
2-Methyl-1,3-Butanedlol
3-Chloropyridine
2-Bromopyridine
2,3-Dithia-4,4-Dimethyl-
pentane
2,3-Pentanedione
2-Methyl-2-Butene
2-Ethoxy-1-Propanol
2-Methyl-3-butyn-3-ol
Pentane

2-Pentene
3-Bromopentane

Methyl i-Butyrate
n-Butyl Formate
Cyclopentene

Methyl i-Butanoate
i-Propenyl Trifluoroacetate
Chlorocyclopentane
2.2-BMMethylmercapto)
Propane
2,2-Dimethyl-1-propanol
Ethyl

Pyridine

Valeric Acid
Trimethylacetic Acid
Cyanovinylacetate

2-Methylthiophene

 

GGGGG OGOIO! GO! OI“GOGGOGGOIGOIGOIOIQOIGGOIGGGOIGOI OI OIGGOIGOIG GOIOIOIOIOUOI OIOIOIOIOIOIGOIOIOIGO

47.1

“mp-:50 Ext-Ira HH HHHHHHHHHNHHHNNNHHHwHunmww H HMHNJINOD tee-tenant» automate-bahcecea
88383 game: one: .ateennoocmodsqoqmmqssasooofis s Odwddum 58587105“ mfigususcgwma

E

50.0
18.2
57.1
8.6
11.1
8.6
11.1

10.0
16.7

50.0
10.0
50.0

10.0
14.3

11.1
14.3
10.0

11.1
11.1

11.1
11.1

mlo

R
375
33.3
33.3
66.7
40.0
18.2
23.1
27.3
15.4
18.2
37.5
37.5
8 0
33.3
8.6
so 0
33.3
18.2

18)

Candid
Furfural
Ethyl Propanoate
Trimethyl-o-Acetate
3-Methyl-1-Butene
Methyl t-Butyl Ether
2-Chloroallyl Acetate
Allyl Vinyl Ether
N-Methylmorpholine
Cyclopentanol
Cyanobutadiene
2-Methyl-2-Hydroxybutan-3-
one
3-Methylthiophene
3-Pentanol
2,2-Dimethylpropanoic Acid
Vinyl Propionate
1,3—Dimethoxy-2-Propanol
Cyclopentadiene
Dimethylamine
Propionitrile
2,4-Pentanediol
2-Methylpyrizine
Cyclopentane
2-Bromo-3-Pentanone
3-Methyl-1-Butyno3-ol
3-Chloropropyl Acetate
l-Methoxy-2-Methyl-2-
Propanol
3-Dimethylamino
Pro pylamine
Ethyl i-Propyl Ether
2-Pentanone
Serine Ethyl Eater
2,4-Pentanedione
2,2-Dimethylpropane
3-Methyl-2-Butanol
2-Methoxyethyl Acetate
Methyl Acetoacetate
Methyl i-Butyl Ether
Ethyl Acrylate
Methylcyclobutane
Levulinic Add
2-Cyanofuran
i-Propenyl Acetate
3-Methoxy-2-Butanol
2-Pentanol
Bromocyclopentane
i-Propyl Acetate
n-Propyl Acetate
Glutaraldehyde
Methyl Propyl Ketone
Methyl 3-Ketebutyrate
Acetin
Allyl Acetate
N-Amylamine
Tetramethyldiamino-
methane
n-Butyl Methyl Ether
2-Chloro-1-Cyanoethyl
Acetate
Alanine Ethyl Eater
Propargyl Acetate
Thiacycloheptane
(2-Hydroxypropyl) n-Propyl
Sulfide
Hexanal
3-Methylthiacyclohexane
2,4-Hexadien-1-ol
2-Methylcyclopentanethiol
2,4-Furandicarhoxylic Acid

GGG “GOOOOGGGGGGGGGOOOGOGQGOOGOGG OGOGGOOGGOOGO QGGO @GGGGOG OGQOGGGGiGO

E:

EEEE EEEEEEEE

555
um»:

5555

mG-d oa-aqqqouecoo-aammqqmssmmocuqmgmqa QBOSBSOQQOO:Q ddss “6551335 Eubxsgsgfii

E
558 85585858858588555555555888888 5588888885888 8858 8888588 888888888;

5
b

14.0
14.0
14.1
14.1
14.1
14.1
14.2
14.2
14.3
14.3
14.5

14.6
14.7
14.7
148

15.0
15.1
15.3
15.3
15.4
155
155
155
15.6
15.7
159

16.1
165
165
16.7
16.7
16.7
168
168
169
169
17.1
17.1

17.1
17 .1
17.1

47.6

47.1

31.4
348
421

85.3
227

8.1

318
41.2
8.7

31.8
40.0
421

CW
Cyclohexene Sulfide
1,6-Hexanediol
Cyclohexene Oxide
1,2-Cyclohexanediol
Hexamethylene Glycol
3-Methylpiperidine
2,5-Furandicarboxylic Acid
3-Methyl-1-Pentanethiol
2-Vinyl-4-Methylol-1,3o
Dioxolane
2-Methyl-3-Pentanethiol ,
7-Thiabicyclo(2,2,1)Heptane
Di-n-Propyl Sulfide
Cyclohexanethiol
3-Thiaheptane
Ethyl-n-Butyl Sulfide
2-Methyl-3,4-Dihydroxy-
tetrahydropyran
Methyl-n-Pentyl Sulfide
l-Fluomhexane
Butynediol Dil'onnate
3-Methyl-1,2-
Cyclopentanediol
2-Hexyne
3-Hexanethiol
Methyl-n-Amyl Sulfide
Pyridine-3-Aldehyde
4-Methyl-2-Pentanethiol
3-Chloro-3-Methylpentane
Adipic Add
2-Hexanethiol
2-Methylthiacyclohexane
5-Methyl-2-Thiahexane
2-Hexenal
i-Amyl Formate
Chloroacetaldehyde Diethyl
Acetel
3-Hexyne
Kojic Acid
Cyclopentyl-1-Thiaethane
2-Methoxypyridine
Phenylhydrazine
Cyclopentylmethanol
3,6-Dithiaoctane
l-Hexanethiol
Methyl i-Pentanoate
I-Hexene
Vinyl Chloroprene
n-Propyl-i-Propyl Sulfide
2-Methyl-3-Thiahexane
l-Hexanol
n-Butyl Chloroacetate
(1-Thiaethyl)-Cyclopentane
Ethyl Hydroxy-n-Butyrate
Hexanenitrile
2,4-Hexadienal
Ethyl Butanoate
2,3-Dimethyl-1,3-Butadiene
p-Phenylenediamine
2-Ethyl-1-Butene
l-Hydrobenzotriazole
Butanediol Diformate
6-Methyl-3,4-Dithiaheptane
i-Butyl Vinyl Ether
4-Methyl-3-Penten-2-one
2,4-Diamino-2-
Methylpentane
l-Methylcyclo pentanethiol
Ethylon-Butanoate
1,1,1-Trimethoxypropane

GGQ¢¢OQ¢ O QG$O®QGQGGQ QGwOOGOO QGGGGQGOQG camacacacaazaaeacacoczcacaca O) Q¢GQGGGQQGOQQQ

23512516511
172717412
172716436
176519266
176722316
176715467
174620300
175670300
175622364
175621361
177621256
17.7617356
177716369
176522227
179615400
179715467
160617471
160713536
161620300
161615566
162712566
164516313
164616375
164717412
164715457
165421174
165315200
165716369
165917529
166614426
16.7416250
167719366
166519266
169716466
169919474
190423174
191519266
191716436
192716466
192715457
196617471
194615566
196615566
196712566
19.7616375
197614571
196616336
199716436
200319156
200210200
200617356
201521266
202521236
206520250
206622.276
206516313
204615400
204714500
205514357
206615400
207324125
207419211
207416222
207615400

161

4-Methyl-1-Pentyne
4-Methyl-3-penten-2-one
5-Methyl-2-Furaldehyde
4-Methyl-3-Thiahexane
n-Butyl Vinyl Ether
3-Methyl-3-Pentanethiol
2-(Methylallyloxy) Ethanol
Ethyl-i-Butyl Sulfide
5-Methyl-3-Thiahexane
Dipmpargylamine
Butanediol Diformate
1,2-Bin(Ethylmercapto)
Ethane

2-Ethyl-4-Methylol-1,3-
Dioxolane

Ethyl Ethaxyacetate
Di-i-Propyl Sulfide
2-Methylpiperazine
Hexane
Methoxyethoxypropane
3-Methyl-1-Pentene
2,2-Dimethylbutane
3-Hexen- 1-01
2-Ethylhutanoic Acid

Vinyl i-Butyl Ether
3-Methyl-1-Pentanol
1,3,3-Trimethoxy-1-Propene
3-Hexene

l-Hexyne
l-Methyl-l-Ethyl-
cyclopropane
Chlorocyclohexane
2,3-Dimethylthiophene '
6-Methyl-3,4-Dithiaheptane
4-Thia-1,6-Heptadiene
1-Chlorohexane
3-Hexen-1-ol
3-Cyclohexen-1-ol
2,5-Dimethylol-1,4-Dioxane
2-Methanol
Tetraahydropyran
2-Methyl-1-Pentanethiol
1,5-Hexadiyne

3-Hexanol
Methyl-n-Pentanoate
2-Hexene
3,3-Diethyl-2-Butanol
4-Methyl-2-Ethyl-1,3-
Dioxolane
2-Methyl-1-Pentene
Dimethyl Sucdnate
2,2-Di1nethyl-1-Butanol
2,5-Dimethylthiophene
2-Methyl-3-Pentanol
2-Chlorocyclohexanone
l-Methyl-z-Pyridone
2-Methyl-1,3-Pentadiene
m-Dichlorobenzene
2-Ethyl-2-(Hydroxy1nethyl)—
1,3-Propanediol
2,5-Dimethylthia-
cyclopentane
2-Methyl-2-Pentenal
Cyclohexane
2,4-Dimethyl-3-Thiapentane
2-Methylpentanal
l-Fluorocyclohexene
l-i-Propoxy-Z-Propanol
2-Hydroxycyclohexanone
Triethanolamine

O: Q @QQGG GQOQOQGQQG @QGGGOOQQGGGOGGGQGQGGOGQQGQGOG@O GOGGQG OQGQGOGGGG

a m hemmb unau¢ufiemu Abaawebwmqamamambammmawfihfiqamaaa hataah phaauaqq¢§
5 8 $5888 888868388: KEESSEEEGEKKKSKEBfifiSBK8585fi58¥58 8‘8555 88:8858553

ROW

%.7
41.2
43.8
60.0
158
37.5
42.9
20.0
21.1

$5
40.0
38.5
21.0
42.9
31.8

31.3
46.2
42.9
33.3
41.2
63.6
23.5
”.7
m8
13.6
42.9
25.0
”A
45.5
545
m8
278
35.7
37.5
35.7
57.1
42.9
43.8
”.4
20.0
257
257
17.6
60.0
37.5
3.7
”.6

35.7
31.3
31.8
35.7
15.8
21.1
15.8
3.3
35.7
£10

21.1
375
41.7
222
25.0

41.7

Amyl Formate
Cyclohexanone
4.5-Dimethyl Dioxane
3-Hexanone

Benzenethiol

Vinyl n-Butyl Ether
l-Bromohexane
2-Hexen-4-yne
1,4-Bie(Methylmercapto)
Butane

Allyl Acrylate ,
2-(2-Ethoxyethoxy) Ethanol
Bie(2-Hydroxypropyl) Ether
Vinylthiophene
Caprolactone
3,3-Bie(Hydroxyxnethyl) 2-
Butancne

i-Propyl 3-Chloropropionate
1-Bmmc-3-Methylpentane
Allylethyl Carbonate
2-n-Butoxyethanol
4-Methylpentanenitrile
Dimethyl Sulfolane
o-Fluorophenol
Hydroxyethyl Methacrylate
Ethyl Methacrylate
3,3-Prim Iminobiapropyl
3,3-Dimethyl-2-Thiapentane
Diallylamine
2.6-Dimethylol-1,4-Dioxane
2-Methyl-3-Pentanone
4-Methyl-2-Pentanone
Tetraxnethyl Ethylene Oxide
3-Methylpyridine
1,6-Hexadiene
1,3,5-Hexatriene
1,2-Dimethylene Cyclobutane
2-Ethyl-1-Butanol
Dipropargyl Ether
2-Ethyl-n-Butanol
Cyclohexene
p-Fluoroanaline
2,4oDimethylthiophene
Dipropylene Glycol
gamma-Picoline
3-Methylpentane

Hexanoic Acic
2-Methyl-2-penten-1-al
4-Hydrcxy-4-Methyl-2-
pentanone

Naphthodioxane
2,5-Dimethylpiperaaine
Di-l-Propylamine
3,4-Dimethyl-2-Thiapentane
i-Propenylcyclopropane
2-Ethyl Thiolane
2-Methylpyridine
Fluorocyclohexane
1-Chloro-3-Nitrohenzene
3,7,9-Trioxabicyclo (3.3.1)
Nonane

2-Hexen-1-ol

Cyclohexanol
Methylcyclopentane
2-Amino-6-Methylpyridine
2-Hydroxy-3-Methyl-2-
Cyclopenten-l-one
6-Hydroxyhexanoic Acid
Lactone
2-Methyl-1-Pentanol

 

OGQGGQQQOGOQQGGGQGGQGGGOGGGOQQG03 GOO GGQG05$GGQGGGGGGQQGGGGQQQOGGQGOQOGO

abwuwwauwaamw5wnw§umucbhunuwuhnu a» hauawbu¢+N¢whmhw5nuh5555555555855E
535$:mi:553'91353535'985335553855555515383 '35 KKGQK5°EK§KK3KEKEEEES‘SfiEKESfiEfiEEa

146

17.6

17.6

1%

w
m-Chloroanaline
1,4-Cyclohexadiene
Hydroquinone
2-Methylpentanoic Acid
Hexamethyl Diailoxane
EthyloioButyrate
Bmmobenxene
o-Dichlorohenzene
Bis(2-Methoxyethyl) Ether
2-Methylvaleraldehyde
Allyl Propionate
2-Vinylpyrrolidone
2.Ethyl-1-Butanol
2-Methyl-2-Pentene
1,5-Hexadien-3-yne
Acetone Aside
1,3-Cyclohexadiene
4-Methyl-1-Pentanethiol
2,2-Di1nethyl-3-Thiapentane
2,3-Dimethyl-2-Butene
2oMethcxyethyl Acrylate
3,3-Dimethyl-l-Butene
1,2-BidVinyloxy) Ethane
2,3-Dichloroanaline
Ethyl-t-Butyl Sulfide
6-Hydroxyhex-1-ene
Quinone
l,2-Hexanediol
Acetylhuterolactone
1,2-Diethcxyethane
1,2-Diethoxyethane
4,4-Dimethyldioxane
5-Methyl-1,3-
Cyclopentadiene
o-Nitrophenol
3,3-Di1nethyl-4-Hydroxy-2-
Butanone
2,6-Dichlorophenol
2,4-Dimethylpymle
Triethylamine
2-Amino-3-Methylpyridine
3-Methyl-1-Pentyn-3-ol
3-Bro1nohexane
3,4-Dichloroanaline
Nitrobenzene
2,4-Hexadiyne
2,5-Dimethylpyrazine
n-Propyl Propionate
2-Ethynyl-2-Butanol
4,5-Dithiaoctane
Ethylcyclohutane
Butadiene Acetylene
Allyl Ether
1, 1-Dimethylbutanol
1-Acetoxy-2-Butanone
i-Propyl Propionate
2-Methyl-2,4-Pentanediol
2-t-Butoxyethanol
Adiponitrile
Diglyme
2,3-Dimethylpyruine
Ethyl Acetoacetate
Furfuryl Methyl Ether
2-Bromohexane
2,3-Dimethylhutane
Catechol
Methylene Cyclopentane
1 , 1,2-Trimethylcyclopropane
2,2-Dimethyl-4-Methylol-1,3-
Dioxolane

QGGQGGQGGOGGGQ 90 ca OGGO G GQGGO¢OOG GQGGGG GGQOGGGOGGQGQQGGQGOGGGGGO¢a

3L1

emwoeumsxuusms “a m °5°s 7 °desus°<° 5°sucs 76°n=°*s"sus°°°nnsssss”u°sa

B
30.0
62.5
27.3
22.2
40.0
40.0
25.0
23.1
27.8
27.3

BEE E

”.0
20.0

E E 582% 5 SEE

50.0

53

375
54.5

14.3
11.0
25.0
27.3

25.0
37.5
60.0

cm
o-Fluorochlomhenaene
Methyl t-Butyl Ketone
2-Methylthio-5-Methylmran
2,5-Dichlorophenol
Methyl Furoate
2-Methylpentane
1.1-Diethcxyethane
Phenol
2,6-Dimethylpyraxine
Ethyl t-Butyl Ether
3-Methyl-1-Pentyn-3-ol
Vinyl Methacrylate
t-Butyl Acetate
2-Ethylbuteraldehyde
2-Methyl-3-Hydroxypyrrone
2,5-Hexanediol
m-Chlorophenol
3-Ethylthiophene
3,3-Dimethyl-1-Butyne
Vinyl Butyrate
p-Dichlorohenzene
3-Methyl-2-Pentanone
Methyl Propyl Keytone
3,3-Dimethyl-1-butyne
An aline
Bie( l-Methyl-Z-
Hydroxypropyl) Ether
n-Butyl Acetate
3-Methylcyclopentene
4-Methylcyclopentane
Butadienyl-4-Acetate
coChlonophenol
2,4-Dimethyl-2-Methylol-1.3-
Dioxolane
2,3-Dimethyl-2-Butancl
2-Ethylthiophene
Methyl i-Butyl Ketone
Di-2-Propylamine
l-Methylcyclopentene
2-Methyl-2-Pentanol
4-Methyl-2-Pentanol
2-i-Propoxy-1-Propanol
1,1-Bie(Ethylmempto)
Ethane

2,2,4-Trimethyl- 1,3-
Dioxolane

Triethylene Glycol
p-Chlorophenol
2,4-Dichlorophenol
2.5-Dimethyl.3,4-
Dithiahexane
2-Diethylaminoethanol
Hydrochloride
4-Methyl-2-Pentanone
2-(2-HydroxyprOPOIY)-1-
Propanol
4-Chlorocyclohexanol
Pyrocatechol
2-Methyl-2-Pentanethiol
4-Methyl-2-Pentanone
Cyclohexylamine
Cyclohex-2-en-1-one
Methallyl Acetate
2-Hexanone
Benzofurazan
Chlorobenzene
Di-i-Pmpyl Ether
N-Ethyl-N-Butyl Amine
t-Butyl Acetate
Propanoic Anhydride

QQQQQQQQQQQQQQ QQQQQQQQ QQGGOQQGGOGOGGGQGQOQQ O) GOGQOO @GO GQOOGQOQOGGO

163

E g NIB N5 B Cancun!

40.6 2 8 25.0 Bromocyclohexane

408 1 9 11.1 i-Butyl Acetate

41.0 2 9 222 Diacetone Alcohol

41.7 1 11 9.1 Ethyl 3-Ketohutyrate

41.7 3 7 429 Paraldehyde

418 1 9 11.1 2-I-Iexanol

41.9 1 9 11.1 Di-n-Propyl Ether

41.9 3 8 37.5 4-Methyl-1-Pentyn-1-ol

423 1 8 12.5 p-Fluorophenol

42.3 2 8 25.0 N ~Methylpiperidine

43.0 2 8 25.0 2,4,6-Trimethyl-1,3,5-
Trioxacyclohexane

435 2 5 40.0 Furyl Methyl Ketone

43.6 2 8 25.0 Diacetonearnine

44.4 2 8 25.0 Methyl 2-Methyl-3-
Ketobutyrate

44.6 1 8 125 Reeorcinol _

44.6 1 8 125 2-Butylacetate

44.6 3 5 60.0 Hexamethylene Tetramine

45.2 1 8 125 Benzene

46.3 2 7 26.6 2,2-Bie(Dioxolanyl-1,3)

46.4 1 7 14.3 2-(2-Vinyloxyethoxy)
Ethanol

468 1 6 16.7 Three-3-Chloro-2-
Acetoxybutane

46.9 1 8 125 Hexamethylene Diamine

47.1 1 7 14.3 5-Hexen-2-one

478 1 7 14.3 4-Methyl-2,3-Pentanedione

48.1 2 6 33.3 Ethyl Hydroay-i-Butyrate

49.9 1 6 16.7 2,5-Hexanedione

51.0 1 8 125 2.3.3.2-Thi0phenothiophene

51.6 1 6 16.7 Diethyl Oxylate

51.7 1 7 14.3 Di(Acetyl Cyanide)

51.9 2 3 66.7 Dimethyl Fumarate

52.4 2 6 33.3 2,5-Hexanedione

52.4 2 5 40.0 2-Acetylfuran

52.6 1 5 200 Methoxyacetic Anhydride

53.6 1 6 16.7 Fluorobenzene

58.4 2 3 66.7 Hydroxyadipaldehyde

$2 2 3 66.7 Methyl 'I'hiofumate

59.9 2 5 40.0 Methyl Furan-2-Carboxylate

81.7 1 5 200 2,3-Di1nethyl-2,3-Butanediol

63.2 1 4 25.0 1,2-Ethane Diacetate

m 1 3 33.3 N-Hexylamine

8.4 11 42 m2 2-Thiabicyclo (2.2.2) Octane

8.7 13 41 31.7 2-Thiahicyclo (3.3.3.0)
Octane

9.2 13 37 35.1 Methyl-4-Hexenoate

9.3 12 8 429 n-Heptanal

95 12 3 41.4 2-Heptenal

9.5 16 27 59.3 2-Methyl-3-Thiaheptane

9.8 11 27 40.7 Heptaldehyde

9.9 13 34 38.2 2,4-Heptadienal

10.0 8 49 16.3 1,2-Propanediol Diacetate

10.0 11 38 28.9 2-Hydroxycyclohexane—
carboxylic Acid

10.0 13 E 44.8 3-Heptyne

10.1 13 5 520 l-Heptanethicl

10.2 10 3 25.6 2-‘l'hiabicyclo (3.3.0) Octane

10.3 12 24 50.0 n-Heptyl Alcohol

105 11 :5 31.4 8-Thiabicyclo (3.2.1) Octane

10.6 1.2 41 29.3 3-Thiabicyclo (3.3.0) Octane

10.6 9 8 31.0 5-Methyl-2-Hexyne

10.7 11 a 50.0 Cycloheptane

10.7 14 5 56.0 i-Propyl n-Butyl Sulfide

10.8 13 31 41.9 2,3-Dimethylpiperidine

10.9 10 30 33.3 3,7-Dithianonane

11.2 10 31 323 Pimelic Acid

11.2 12 9 41.4 2-Thiooctane

11.2 11 25 44.0 n-Propyl i-Butyl Sulfide

aqqqqquqq 4545555555 4 «a dqeqaqqaq qqaqaaaqqqq «5555444455544 4445440

EEE:

(panacea-acumen aomooqommm a OS Samoan-ems amasgsomuoo EEOEGcQOuwmgOO 6‘36“me

8855555558 8558885858 8 85 8558855585 88588858888 88888888888888 588888;

11.7
11.7
11.7

E

12.1

13.7

14.0
14.0
14.1
14.1
14.1
14.3
14.4
14.4

14.5
14.6
14.7
14.7
14.9
14.9
14.9

15.2
15.3

15.6

15.6
15.6
15.7
15.7
16.0
16.0
16.0
16.1

16.4

165
165
16.7
16.7
16.7
16.7
16.9
16.9
17.2

17.2 '

555555855555

57.1
31.0
375
m8
41.7
50.0
”.6

17.6
50.0

40.9
421

320
19.4

38.1
60.0

40.9
40.9
47.1

20.0

47.1

Cow
2,6-Dimethoxypyridine
3-Methcxy-1,3,4-Heaatriene
Methyl n-Hexyl Sulfide
Cycloheptene
n-Propyl-n-Butyl Sulfide
2-Methyl-2-Pmpyl-L3-
Propanediol
Methyl Sorhate
Cycloheaane Carhoxylic Acid
Methyl Hex-2-Encate
Methyl 5-Heaenoate
3-Hepten-1-ol
l-Heptyne
2-Heptyne
2-Ethoxypyridine
Methyl i-Hexanoate
Methyl N icotinate
2-Methylcyclohexanol
2,5-Dimethyl-3-Thiahexane
1-Chlcmheptane
2,2-Di1nethyl-3,4-
Pentadienol
Methyl-3-Hexenoate
Ethyl i-Valerate
Ethyl Ethcvpropicnate .
1-eec-Butoxy-2-Propanol 1
1-Fluonoheptane
Toluquinone
3-Methyl Cyclohexanol
Ethylcyclcpentane
5-Methylhexanol
2-Methyl-2-Heaanethiol
3-Cyclohexene-1-
Carboxaldehyde
2-Methyl-1.5-Hexadiene
p-Methoxyphenol
3-Methyl-1-Heaanol
3-Methyl-1-Hexanol
o-Creeol
n-Heptanenitrile
2-Heptene
3-Methyl-2-Ethyl-1-Butene
Dimethyl (Vinylethinyl)
Carbinol
l-Heptene
2,2-Di1nethyl-3,4-
Pentadienal
7,7-Dichlorohicyclo (4,1,0)
Heptane
5-Methyl-1-Hexyne
3-Methyl-1-Hexene
Heptanoic Acid
1.3-Dimethylcyclopentane
Ethyl n-Valerate
2,4-Dimethyl-3-Thiahexane
i-Propyl e-Butyl Sulfide
2-Chloroheptane
1.6-Heptadiene
1,3-BMEthylmercapto)
Propane
Benzyl Alcohol
1,1-Dimethylcyclopentane
Methallyl Pmpyl Ether
3-Ethyl-2-Pentene
Butyl Lactate
3,4-Dimethyl-1-Pentanol
Acetone Methyl Propyl Acetal
Norbornylane ‘
Methylcyclohexane
Allyl Acetothioacetate

 

44 444444 444444444444 444 4444444 44444444444444444444444444444444444o

164

231425145 R Omani

17.2
17.3
17.4
17 .4
17.4
17.4
175
175
17.6
17.6
17.6
17.7
179
17.9
18.0
18.0
18.1
18.1
18.2
18.2
18.3
18.5
18.6
18.6
18.6
18.7
18.7
18.7
18.8
188
188
19.0
19.0
19.1
19.2

19.3
19.3
19.3
19.3
19.3
19.3
19.4

19.4
195
195

195
19.6
19.7
198
198
19.9
199
199
21.0
200
Z10
21.1

21.2
202
205
m8
817
117

m8
am

no 8mn4ww a4a¢4mb4au44 mac mmaammm a4omma¢mamaaa4m4aomoaoa4w4amaomm4am
88 888888 888888888888 888 8888888 88888888888888888888888888888888888

57.1
33.3
41.2
47.1
50.0
56.3
24.0
36.4
25.0
38.9
44.4
26.9
37 .5
56.3
251
529
30.0
52.9
Z10
35.0
23.8
33.3
30.0
33.3
37 5
29.4
29.4
35.3
16.7
37 .5
429
3.3
33.3
304
m5

227
31.3
35.3
375
37.5
57.1
27.8

30.0
29.4
33.3

38.9
41.2
227
37.5
41.2
30.8
38.5
38.9
810
35.3
41.2
m

18.8
35.3
50.0
31.3
3.3
28.6

%.7
31.3

2,3-Dimethylpentane
5-Methyl-2-Hexene
2,2-Diethyl-1,3-Propanediol
5-Methyl-1-Hexene
Cycloheptanone
1,1-Diethoxypropane
3-Methylcyclohexanol
Furfuryl Acetate
4-Methyio2-Hexyne
Dimethyl Glutarate
Methyl Capmate
m-Methoxyphenol
m-Nitrobenzaldehyde
n-Hexyl Formate
3-Fluoroealicylic Acid
1,2-Dimethylcyclopentane
Hydroxypropyl Methacrylate
3-Methyl-1-Cyclohexene
Chlorophenyl i-Cyanate
l-Methylcyclohexene
2,6-Dimethylpiperidine
2,4-Dimethyl-1,3-Pentadiene
3-Ethylpyridine
3-Methylcyclohexanone
n-Propyl n-Butyrate
o-Nitrotoluene
p-Nitrohenaaldehyde
Cyclohexyl Formate
Cyclohexanecarboxaldehyde
2-Vinylpyridine
4-Methyl-1-Hexene
1-i-Butoxy-2-Butanol
3-Methyl-3-Hexene
3-Methyl-1-Cyclohexene
5-Hydroxymethyl-4,5-
Dimethyl-1,3-Dixoane
Phenyliaocyanate
Benzoic Acid
3-Acetylpyridine
3-Methylhexanal
m-Chlorobenzoic Acid
3-Methyl-3-Hexanol
3-Cyclohexene-1-Carboxylic
Acid

t-Butyl Trioxane
4-Chlom-o-Creeol
1.1.2.2-Tetramethylcyclo-
pmpane

3-Heptene
p-Methylcyclohexanol
m-Creeol
4,4-Dimethyl-2-Pentene
Methyl Hexanoate
m-Hydroqbenzoic Acid
m-Chlombenzaldehyde
3-Methyl-2-Hexene
2,3,4-Trimethylthiophene
2,3-Dimethyl-1-Pentene
3-Heptanol
1,5-Bie(Methylmercapto)
Pentane

Methyl Analine
4-Methylcyclohexanone
2,3-Dimethyl-3-Pentanol
1,2-Heptanediol

Benzyl Amine
4-Hydroxymethyl-4,5-
Dimethyl-1,3-Dioxane
p—Chlorobenzaldehyde
Benzaldehyde

44444444444444 444444444444 444444 444444444444 44444444444444444444449

M
33%-

E558 E§EEEEEEEEEEEEEEEEE

bwaunawwmppwmn hb$wwub8aaab banbbh Aubu¢¢¢ahaaa anh4amahhbaaaum5aahawm§
88888888888888 888888088888 888888 888888888888 88888888888888888888883

17.4

17.6

10.5

21.1
8.7
8.7
35.7
17.6
”.0
41.7
10.5
50.0

Command
3-Methylheaane
2,3-Dimethylpyridine
4-Ethylpyridine
5-Chloroealicylaldehyde
6~Thiabicyclo (3.2.1) Octane
1-Heptene-4-ol
2-Methylcyclohexanethiol
2-Methyl-3-Hexanol
Cyclohexanemethanol
1-n-Butoxy-2-Propanol
3-Vinylpyridine
Cylcoheptanol
p-Chlorobenxoic Acid
o-Chlorobenzaldehyde
3,4-Dimethylpyridine
4-Methylcyclohenene
4-Vinylpyridine
3-Ethyl-3-Pentanol
n-Heptane
2,5-Dimethylpyridine
n-Propyl n-Butyl Ether
Ethoxyacetaldehyde Methyl
Vinyl Acetal
Ethyl Methyl Dionne
3-Ethyl-1-Pentene
3,3-Dimethyl-1-Pentene
3,5-Dimethylpyridine
1,5-Heptadien-3-yne
i-Propyl Butyrate
(l-Thiaethyl) Benzene
2-Methyl-2-Hexene
2,6-Dimethyl-4-Pyrone
2,3,3-Trimethyl-l-Butene
Ethyl 2-Methylhutyrate
l-Cyclohexene-l-Carboxylic
Acid
Allyl Methacrylate
i-Propyl Crotonate
m-Aminohenzoic Acid
Diethyl Malonate
m-Hydroxybenzaldehyde
4,4,5-Trimethylo5-Hydroxy-
1,3-Dioxane
Allyl n-Butanoate
2,2,3-‘l‘rimethylhutane
2,4-Dimethyl-3-Pentanol
2-Methyladiponitrile
Di-i-Propylcarbinol
4-Chloro-2-Fluoroaniaole
Methionine Ethyl Eater
p-Nitrotoluene
Orcinol
o-Chlorobenzoic Acid
2-Methyl—5-Ethylpyrazine
o-Phenylene Cyclic
Carbonate
Ethyl Invulinate
p-Fluoroaniaole
m-Toluidine Hydrochloride
Methyl Phenyl Ether
Benzoyl Fluoride
2,4-Dimethyl-1-Pentene
3-Methylprocatechol
Di-n-Propoxymethane
2,6-Dichlorotoluene
o-Toluidine
2,4-Dimethylpentane
Ethenylcyclopentane
i-Propyl t-Butyl Sulfide
3-Ethyl-1-Pentyn-341

 

4454454 4544545544444445444555454454454444454454545 554455 4 455444450

165

23585185 R Canard

38
3.0
3.0
3.0
3.0
3.1
3.1
3.6

3.7

3.7
38
39
27.0
27.0
27.2

27.3
273
27.4
275
27.6
278
279

3.1
3.2
3.3
34
3.7
38
38
3.0
3.3
3.4
3.4
35
38
38
3.0
3.6
31.6
3.6
3.6
3.7
303
31.1
31.1
31.2
31.3
31.4
315
315
315
31.6
31.6
31.6
31.7
319
319

32.4
32.4
325
33.2
3344
33.4
335

W‘WGHNN IFNIhW§N$NNN§WNHQ$WG§NNI§$NWC¢N“IF#O‘OI-FwflIFGBDNWNOBNN IOCONcD-CDOI uh “$IFNMNWG

8 75.0
3.0
22.2
3.3
66.7
30.8
40.0
21.4

28.6

50.0
17.6
3.4
16.7
25.0
13.3

9.5
11.1
3.0
16.7
3.0
3.0
22.2
27.3
3.4
23.1
17.6
33.3
55.6
31.3
40.0
33.3
33.3
18.8

6“”8863 88°8888888888888°8888”$838”888°88888°88888B 888888 8 £68“8‘°8

2,2-Dimethylpentane
2,2-Dimethyl-1-Pentanol
o-Fluoroaniaole
p-Creeol

Benzaldehyde Oxime
2,4-Dichlorotoluene
Benzotrifluoride
1,2-BMEthylmercapto)
Propane

Triethylene Glycol Methyl
Ether

1,1,1o'l‘riethoxy Methane
2,6-dimethylpyridine
Neopentyl Acetate
Salicylaldehyde
4-Chloro-m-Creaol
Pyruvaldehyde Diethyl
Acetate
2,4-Dimethylpyridine
Methyl 2-Methylvalerate
Diethylaminopropyl Alcohol
Methyl n-Hexanoate
Phenyl i-Thiocyanate
2,4-Dimethyl Pentanal
2,4-Toluenediamine
o-Toluidine Hydrochloride
3-Heptanone
6-Chloro-o-Creaol

n-Propyl Acetoacetate
n-Amyl Acetate
2-Methylhexane
1-Butoxy~2-Methoxyethane
Benzoyl Chloride -
3,4-Dichlorotoluene
1,1-Dipropoxyethane
Benzothiazole
o-Methoxyphenol

Diethyl Propyl Amine
Diethylaminopropyne
4-Methyl Hexanoic Acid
i-Propyl Carbonate
3-Cyclohexene-1-Methanol
o,a-Dichlorotoluene
l-Ethoxy-l-Propoxyethane
3-Ethylpentane

Induole

i-Amyl Acetate
Ethylidinecyclopentane
Cyclopentylacetate
Dicyclopropyl Ketone
n-Propyl Capronate
n-Propyl Acetothioacetate
Vinyl t-Amyl Ether
2-Furanacrylene
Bicyclo(2.2,1)-2,5-Heptadiene
3-Ethyl-3-Pentanol
Tetrahydrofurfuryl Acetate
5-Methyl-3-Hexanone
4,4-Dimethyl-1-Pentene
m-Chlorotoluene -
Dipropyleneglycol Methyl
Ether

p-Toluidine Hydrochloride
3,3-Dimethylpentane
2-Methyl-5-Hexanol
2-Ethylpyridine

Ethyl o-Formate
ar,ar-Dichlorotoluene
Ethyl-ZoMethyl-s-
Ketobutyrate

m m 8884 554455554444554 4 5 555544455555555555444 555 555455554 55440

g 8 EESH Hwnwnnwwwwnwwnw » H ”Huuuuuuwmnwuuwwwnuuu ”at ”annuawmw ”swag
8 a 8:88 GQQQQQEGGGQQQGE 8 o mmao5m5mu§mssw5ummm§§ K08 8888*”8m8 55:3;

7.1

3.0

25.0

75.0

11.1

Omani
4oHeptanol
l-Ethylcyclopentene
i-Pmpyl Acetoacetate
2,2-Bh(Ethylmercapto)
Propane
4-Heptanone
o-Fluorotoluene
Toluene
Trifluorotoluene
2-Propen-1,1-diol Diacetate
3-Methyl-2-Hexanol
Heptalactone
1,6-Heptadiyne
1,3-Blddimethylamlno)-2-
Propanol
n-Butyl Propionate
n-Butyl Acrylate
Methyl 2-Methyli’uran-3-
Carboxylate
Benzamide
p-Chlorotoluene
1,3,5-Cycloheptatriene
Di-n-Propyl Carbonate
2-Methyl-3-Hexanone
aec-Butyl Acrylate
Ethylene Dimethacrylate
m-Fluorotoluene
Trimethylpyrazine
2-Methyl-2-Hexanol
p-Fluorotoluene
2,3-Dimethyl-2-Pentanol
N orbornylene
p-Chlorobenzonitrile
p-Hydroxybenzoic Acid
6-Methyl-2-Hexanone
2-Heptanone
Benzonitrile
Cyclohexyl Methyl Amine
3-Ethylcyclopentene
3,3-Dimethylcyclobutane-
carbonitrile
1-i-Propoxy-2-Methyl-2-
Propanol
Methyl 3-Methylfiiran-2-
Carboxylate
Benzyl Chloride
Allyl Crotonate
2-Methylbutyl Acetate
aec-Butyl Propionate
t-Butyl Propionate
2,4-Dimethyl-3-Pentanone
2-n-Propylthiophene
2-Heptanol
o-Chlorotoluene
2-Furyl Ethyl Ketone
1.2-Dimethylpropyl Acetate
1,3-Propanediol Diacetate
Furfiiryl Thioacetate
Allylidene Acetate
2-Amino-4-Methyl-n-
Hexane
Praline Ethyl Eater
o-Methoxybencoic Acid
9-Methyl-1,3,6-Trioxadecolin
2-Ethoxy-4-Methyl-
tetrahydropyran
2-Carboethoxy-
cyclopentanone
Octylmercaptane

eemeommmeeeom mecca en mmmenmenonooee mmcoeeeo GOOGG meoeoeneoenenmooeoeneoeneoeoeneoenenoooo cameeeeeneoeeeo @O

12.7

°°°"°5°°“3I$° “’33: S3 “SHEfiBfifi” fir-”'66 lim‘HSB °°°°ISE°°°ISSSESE°I=¢ISISE§¢° 5S°KSEKS 5E

88883832358 63:? 8 SBBEBBfiSB E38388 S8588 BEEESRBEEEBBEEISEBBfiBE i386f3§85£8 8;

35.1

33.3
41.2
40.0
59.3
66.7
19.0
48.7
50.0

23.7
38.7
54.2
38.7
54.5
39.3
50.0
25.7
68.4
50.0
41.4
43.5
50.0
28.1
32.0
52.0
40.7
31.0
3.0
31.0

47.8

40.7

60.0
55.0
61.9
38.7

29.6
60.0

64.7
37.0
47.4

39.1
61.1

Z17
47.6
15.0
17.9
28.6

166

CW
3-Cyclohexene-1,1-
Demethylol
2-Octenel
Caprylyl Chloride
Ethyl 4-Methylpentenoete
Methyl o-n-Valernte
2-Amino-6-Methylheptene
Ethyl-2-Hexenoete
2-Thianonene
1-Methyl-3-Cyclohexen-1-
Carboxeldehyde
2-Ethyl-2-Hexen-1-ol
Ethyl-p-Quinone
i-Octenol
2-0ctyne
Heptyl Formate
n-Octanol
6-Methylheptenol
Ethyl Son-bate
1-Octenol
l-Octanethiol
l-Octyne
2-Methyl-2-Heptennl
1,4-Diethoxyh1tene
Venillin
4,4-Dimethyl-2-Penten-2-el
Cyclooctene
Cyclooctene
3-0ctyne
p-Methylbenzyl Alcohol
4,4-Dimethyl-1,5-Hexediene
Methyl Cyclohexnne-
cerhoxylate ‘
1,4-Dimethylenecyclohexen
2,6-Dimethylcyclehemol
2-Chlorooctene
7 -Thiebicyclo(4.3.0) Nonone
1-Methenel-4-Methylene-
cyclohexene
2-Thie-3-Methyl Octane
l-Octene
Inntin
4-Octyne
1-Methyl-2-Ethyl-
cyclopentene
4-Octanol
1,2-Dimethylcyclohexene
l-Chlorooctane
4-n-Butoxy-n-Butenol
n-Propyl-n-Pentenoete
n-Propylcyclopentnne
2-Octen-1-ol
1,1,3-Trimethylcyclopentene
2,3-Dihydro-2,5-
Dimethylpyren-2-Methylol
2,5-Diethoxy
Tetrehydrofuren
3-Ethyl-3-Hexenol
Ethyl Nicotinote
1,1-Diethoxy-2-
Methylpropene
2-Ethyl-1-Hennethiel
2-Methyl-3-Ethyl-3-Pentenel
4-Octanene
3-Thiabicyclo(4.3.0) Nonane
148-Thenyl)-2-Butene
2,4-Dimethyl-3-hexenol
1,1-Diethyoxybutnne
142-Thenyl)-2-Bntene
1,4-Benzenedicerbinol

mwmmmm ounce mommmmcbmmmebeb GOG QOOOOOGOOOGGOOO QQGQDQQOGmOGOOQQO G QQDGOO

14.4
14.6
14.6

14.8_

14.8
148

§§§§EEEEEE EEEfiEEEEfiEEEEEEfiE

dehqm 45am oqaqqaauaams amen mommcmsuqqequebeb «dampaqtgausscoamo l: melanoma
85368658 $88 ESSEB‘EERSSSE 85$ mssaeasssesfiaaa BSEfiBEEEEEBREBREE IS [88886;

47.1

47.4

21.7

Cancun!
1,8,7-0ctetrlen-5-yne
2-Cyclohexylvinylchloride
2-0ctene
o-Methoxybennldehyde
Methyl-3-Methyl-4-
Hexanoete
1-Methyl-3—
Ethylcyclopentene
2.6-Dimethyl-1,5-Hexediene
2-Ethyl-1-Hexenethiol
Ethyl-3-Hexenoete
Octalene Glycol
1, 1,1-T11ethoxyethene
1,3-Cyclooctediene
1,1,2-Trimethylcyclopentene
1,4-Diethoxy-2-Butene
1,7-0ctadiene
Benzyl Formate
Ethyl 2,3-Methylbutyrete
1,3,6-0ctetriene
1,2-Dimethylcyclohexnne
Methylfomenelide
2-Methyl-5-Heptenone
Cyclohexyl Acetic Acid
Ethyl 2-
Ketocyclopentencubexylete
p-Creeyl Methyl Ether
3~Octenone
6-Methyl-3-Heptenone
2, 5-Dimethyl-2,4-Hexodlene
3-Hepten-3-el
Butyl Methellyl Ether
Caprylic Acid
Propyl i-Valente
t-Butyl 3-Ketohutyrete
o-Toluic Acid
4-Methyl-3-Heptanol
2-Methyl-3-Heptnnone
Methyl Cyclohexyl Ketone
1,6-0ctediene
3-Cyclohexenyl Methyl
Ketone

3-Octen-1-ol
4-Methyl-3-Heptenene
Spiro (5.5) 1.3.9-
Trioxeundecene

4-Octene
l-Ethynylcyclohexonol
2-Thiahexahydroinden
2-Methyl-3-Heptene
2,3-Dimethylphenol
1,5-Cyclooctadlene
3oPropenylcyclopentene
2-Ethylhexenediol-1,3
Terephthnleldehydic Acid
2-(2-Butoxyethoxy) Ethanol
i-Propylcyclopentene
2,12,4-Trimethyl-4-Penten-1-
e

o-Vanillin

1,2-D' benzene
Methyl 2-Hydrexy-
cyclohexanecarboxylete
2,6-Dimethylphenol
2,3-Dimethyl-2-Hexene
Phenylecetonitrile
2,6-Dimethyl-4o'1‘hieheptnne
Cyclooctetetraene
3,5-Dimethyl-3-Hexanol

moooemmceenencoouenm 00555550900005me5 oneemmmcomenmmoommeomoomebeomcoeneo cemencneoao commeneoeneoonon eon

”5955:555555 m$mnm¢auhqmahq mamhqmaaahqaab¢wmafimwmm $55554 haabmmaah mg
555555555585 58855555555555 58’85555555555555555555 858555 555855558 5;

31.6

21.1

31.6

31.6
35.7
41.2

31.3
37 .5
30.8
66.7
8.7

15.0
20.0
31.3

27.3
17.6
29.4

30.8
35.7
22.7
25.0
35.7
17.8

167

W
2,3,4-Trimethyl-3-Penten-1-
ol
p-Methoxybenzoic Acid
2,3,4-Trimethyl-2-Pentene
1,3-Dimethylcyclohexene
n-Butyl-i-Butyrete
Ethyl 2-Methylvelerete
Chloroetyrene
Terephtheleldehyde
2-Ethylhexenel
2-Hydrexy-3-Methylbenzoic
Acid
2-Methyl-3-Heptenol
n-Butyl-n-Butyrete
n-Methyl-o-Toluidine
sec-Butyl Acetoacetete
2,5-Dimethyl-3-Pyrezine
Methyl 2,5-Dimethylfuren-3-
Carboxylete
p-Tolueldehyde
2,3-0ctanediol
l-Thieinden
1,1-Dimethylcyclohexene
1,4-Dimethylcyclohexene
Methyl Selicylete
3,5-Dimethyl-4-Thieheptene
1-(2-Butexyethoxy) Ethanol
3,5-Dimethylcyclohexenol
2-Ethyl-1-Hexene
Ethylidenecyclohexene
3,4-Dimethylhexene
l-MethylenH-Cyclohexene
Benzyl Methyl Ether
2,4,4-Tfimethyl-2-Pentene
o-Methylbenzyl Alcohol
Bid2-Ethexyethyl) Ether
3,4,4-Trimethyl-2-pentene
3-Octenol
Ethyl 3-Ketocapreete
m-Chloroethylbenzene
p-Phenylene Diieocyenete
p-(Hydroxymethyl)
Chlorohenzene
2,2-Dimethyl-3-Hexene
m-Aminoecetophenone
4-Ethenyl-1-Cyclohexene
Allyl i-Velerete
n-Butyl Acetoecetete
3,4-Dimethylphenol
i-Amyl Propionete
Furmryl Propionete
Phenoxyecetic Acid
3-Methylheptene
o-Fluorophenetole
2-Hexyloxyethenel
3-Ethyl-5-Methylpyridine
l-Methyl-l-
Ethylcyclopentene
Di-i-Butyl Sulfide
3,4-Dimethylchlorohenzene
p-Methoxybenzeldehyde
Cyclohexyl Acetate
m-Tolueldehyde
2,4-Dimethylhenne
o-Tolueldehyde
2-Ethyl-1,3-Hexenediol
2,3,5-Trimethylpyridine
Ethylcyclohexene
2-Cyclohexylethenol
Allyl Tiglete

OQOOOQQOOOGDGOOQOQ OOQQQOQ ummmmmmmmmmmmmm mmmmmmmmmmmmmmmmmmm “GOOGQO

whwubwwafiwhbmnm$nb “hubbWfl Embbbabwmhwmawq cahwuahbbmuaaoauahw aqbu+hi
555555555555555555 5555555 55555555555555: 5055555555555555555 555555;

17.6
10.5

14.3
318
16.7

308
12.5
303
41.7

23.1

23.1
40.0
54.5
20.0
25.0
40.0
25.0
17.6
3.7
14.3

Cancun!
1.7-Octediene
Triethyl o-Acetete
4-Methyl-4-Heptenol
2,4-Dimethyl-3-Ethylpyrrole
eec Butyl Ether
Ethyl 4-Methyl-3-
Ketevelerete
2,4-Xylenol
Methyl Heptenoete
2,3,6-Trimethylpyridlne
Phthaleldehydic Acid
i-Venillin
2-Methylheptene
Styrene Oxide
2-Ethyl-1-Hexenol
l-Ethylcyclohexene
p-Toluic Acid
2,5-Dimethylhexane
1,2,8-Trimethylcyclopentene
Piperonel
2,3-Dimethyl-3-Hexenol
1-(2-Chlorophenyl) Ethanol
p-Cyenobenzoic Acid
Styrene Glycol
Trimethyl o-Velerete
2,2,4-Trimethyl-1,3-
Pentenediol
3-Methyl-3-Heptnnol
Anieeldehyde
3,6-Dimethyl-3-Hexenol
2-Methyl-5-Ethenylpyridine
3,6-Xylenol
eec-Butyl Methacrylete
Ethyl 2-Ethyl-3-Ketobutyrete
1,2,4-Trimethylcyclopentene
1,1-Dipropoxyethnne
n-Octene
t-Butyl Acetoecetete
o-Chloroethylbenzene
2,5-Dimethyl-2,5-Hennediol
3,3-Dimethylhenne
24o-Chlorophenyl)
Ethylamine
3-Ethyl-4-Methylpyridine
2.6-Xylene]
Di-n-Butylemine
Phthelic Anhydride
2,5-Dimethyl-2-Hexene
n-Amyl Propionete
Tetrahydro-2,2,4,4-
Tetremethyl-s-Furenol
3-Ethyl-2-Methyl-1-Pentene
2-Methyl-6-Ethylpyridine
Di-i-Butylemine
n-Butyl Methecrylete
4-Methyl-6-Heptenol
2-Methylbenzothiuole
p-Chlomethylbenxene
i-Butyl Methecrylete
1,2-Octenediol
Tetramethylpyrnzine
2,3,3-Trimethylpentene
l-Methylbutyl Pmpionete
m-Toluic Acid
2,3-Dimethylhexene
3-Pyridylethyl Ketone
2-Ethyl-l-Hexanol
2-Methyl-6-Ethylpyridine
p-Anilic Acid

mommmmmmmmmmmmmmmmmmmmm GGQW menace mince mencncnoameeeuoneomonmmeoeneo mmmmmmmcbmmmeb can

168

235125145 B CW

27.4

27.4
27.5
27.6
27.6
27.6
27.7
27.7
27.8
27.9
27.9
3.0
m1

m4
m5
m7
37
as
38
as
m9
as
$2
293
3.4
3.5
”.6
3.7
$3
3.7

$8
9.9
$9

300
300
30.1
30.1

302
30.2
303
303

$4
$5
3L7
3).?
30.9
31.0
31.0
31.0
31.0
31.1
31.2
31.3
31.5
31.6
31.6
31.9

N”NH*N”N“N”NN$*$NN$GW$“ “H63” thN “a” bwwuuunwuwwuwuuwu “WWNNOII‘IFIFIBtFnF W

5“”55555355555m5m5555m5 55‘s 5°55 555 55555555°55555555 555555565555 5

16.7

36.4
40.0
30.8
36.4
40.0
36.4
38.5
18.2
125
20.0
21.4
20.0

13.3
27.3
18.2
23.1
23.1
23.1
5.9

20.0
33.3
23.1
11.8
13.3
30.0
18.2
25.0
25.0
31.8

14.3
17.6
27.3

lLfl
23.1
22.2
33.3

20.0
33.3
9.1
R1

25.0
50.0
27.3
36.5
23.5
16.7
25.0
36.4
50.0
36.4
15.4
16.2
14.3
125
30.0
16.2
31.0
15.4
36.4
7.1
22.2
42.9
15.4

1-Methanol-4-Methyl-
cyclohexane
m-Hydroxyacetophenone
4-Methylheptene

Amyl Lactone
p-Chloroecetophenone
m-Chloroecetophenene
Phenyl Methyl Ketene
2,3,4-Tfimethyl-3-Pentenol
Dimethyl Aniline
2-Methyl-5-Ethylpyridine
Benzylcyenide
Di-eec-Butyl Amine
Acetylaldehyde Dipropargyl
Acetal

m-Methoxybenzoic Acid
2-Methyl-3-Ethylpentane
3-t-Butylthiophene
o-Chloroacetophenone
o-Tolnnitrile
m-Tolunitrile
2-Amino-5-Methylheptane
1,3,7-0ctetriene
Acetophenone
po'l‘olunitrile
2-Ethyl-3-Hexen-1-ol
n-Ethyleneline
2,3,4-Trimethylpentane
i-Propyl IMroete
1,3-Dimethylbenxene
p-Fluoroetyrene
2,2,4,4-Tetramethyl-
tetrahydrofuran ~
N-Ethyl Cyclohexylamine
1,4-Dimethylbenzene
Tetrehydrophthalic
Anhydride
Bie(2-Vinyloxyethyl) Ether
a-Chlorophenetele
2-Methyl-2-I-Ieptanethiol
2,2,5,5-Tetramethyl-
tetrahydrofuran
Difluoroetyrene
N,N-Butylpyrrole
2-t-Butylthicplnne
1-(2-Ethoxypropoiyl-i-
Propanol
ar-Methoxybenzeldehyde
3-Methyl-3-Ethylpentene
2-Methylhutyl Propionate
i-Butyl Acetoecetete
2-Propyl-4-Methylfuren
2-Chloro-p-Xylene
2-Thieinden
3o0ctylexnine

Methyl Benzoete

i-Butyric Anhydride
3-Acetyl-2,4-Dimethylpyreole
Bicycle (3.3.0) Octane
6-Methyl-2-Heptanone
1,2—Dimethylbenzene
2,4,4oTrimethyl-1-pentene
Indole '

n-Butyl Crotonete
3-Ethylhexene
Di-n-Propyl Acetal

Butyl Tetrehydmthiephene
2,5-Diethylthiophene
Methyl p-Aminobenzoete
1,4-Diexeepiro(4,5) Decene

once GGQODOO OOGQQGOGOOG GOOQOOOOG 00 50055550055500 GOOGGOUQOGGOGG OGQOGOO

47.0
47.3
47.6

47.6
478

H” NNHHNHH HHNNt-nnwnwww ”countenance to“ «Huuuwawunw unuwwwwanwnwww ”weetouui
dd Qbsfllmmm EOQGGQQGGQG Smudsdmdfi as asauOSSpsm‘; tangent-3095;513:535; 0400555;

$333 855%:

21.0

11.1

11.1
10.0

40.0
20.0
10.0

14.3

Cw
Di-i-Butylene
1-Ethyl-4-Fluorobenzene
n-Hexyl Acetate
2,2,3-Trimethylpentane
m-Propyl Fumete
Tetrehydro-2,5-
Dimethylpyran-Z-Methylol
p-Chlerophenetole
Phthalide
o-Chlorephenetole
Ethyl Phenyl Ether
Hydroxy-o-Toluic Acid
2-Heaenyl Acetate
i-Leucine Ethyl Enter
2-Octanol
o-Hydroxyecetophenone
p-Ethylphenol
4,4-Dimethyl-1-Penten-2-al
2-Amino-n-Octene
2-Octanone
1-Methylbicycle(2.2.1)

e

2,6-0ctediene
2-n-Butyltetrehydroi‘uren
i-Chloro-z-Ethylhexane
Octelectone
2-Phenoxyethanol
3-Hexen-1-yl Acetate
2.3-Bemthiophene
2,5-Dimethyl-2-Hexanol
p-Hydroxyacetophenone
2-Methyl-2-Heptanol
1-Phenyl-1,2-
Dihydroxyethane
eec-Butyl Cmtonete
alpha-(Chloromethyl) Benzyl
Alcohol

n-Leucine Ethyl Eater
2-Phenyl Ethanol
l-Phenyl-z-Thiapropane
o-Ethylphenol
(p-Chlomphenyl) Acetylene
2,2,4-Trimethylpentane
Methyl p-Hydroxybenzonte
2,2-Di1nethylhexene
Ethyl-1,3-Dimethyl-
butylemine
m-Ethylphenol
Benzofuran
2.2.3.3o'1'etrunethylhutane
t-Butyl-i-Butyl Ether
p-Ethoxyphenol
Ethylhenzene
2-n-Butylthiephene
p-Fluorophenetele
2-i-Butylthiophene
Phenylecetylene
4-Hydroxyoctenoic Acid
Lectone
1.4-Dicyenobenzene
1,2-Dicyenobenzene
Phenyl Acetic Acid
n-Butyl Ether
(l-Chloroethyl) Benzene
m-Acetoxyphenol
2.2.4.4-Tetramethyl-3-
Thiapentane

Di-t-Butyl Peroxide
t-Butyl Sulfide

COO 0‘9 @@0¢9000@ 0'0 300 COQOCO @QOOQQQOQQOQQD O cocooocoeoenaommco cameo comment» mo

23mm R w

48.0 11.1 N,N-Dixnethyl
Cyclohexylamine

49.0 126 t-Octylamine

49.4 11.1 2-Amino-2-Methylheptane

49.8 14.3 2-( Chloroethyl) Benzene

602 14.3 Phenyl Acetate

60.3 Z10 Dimethyl Aminoethyl
Methacrylate

60 7 20.0 Styrene

51:2
545

Ether
66.6 33.3 1-0ctyla1nine
69.0 14.3 Furfuryl i-Propyleulfide
69.0 26.0 Phenylethylamine
69.1 60.0 Phenyl Acetaldehyde
67.8 60.0 2-Octylamine
70.7 33.3 Di-e-Butyl Sulfide
7.7 2&0 Methyl 2-Octynoate
8.2 $6 6-Nonenal
8.4 3.2 Methyl 2-Octenoate
8.6 38.6 N onadienol
8.7 34.1 2-Thia-trana-
Decahydmnaphthalene
9.3 46.7 2,6-Dimethyl-3-n-
Propylpiperazine

9.6 66.7 1,9-Nonanediol

cyclopentane

Butane

9.7 40.6 2-Nonenal
10.1 38.2 2-i-Propylcyclehexanol
10.2 33.3 3,4-Dimethyl-4olieptanol
10.2 691 1-Nonanol
10.6 44.4 6-Ethyl-2-Heptanol
10.7 41.9 Nonanal
10.7 44.4 n-Nonyl Aldehyde
11.1 34.6 2,3-Dimethyl Benzoic Add
440 2-Methyl-3-Octanol
11.6 60.0 1-Nonyne
11.9 34.6 1,8-Nonadiyne
12.0 39.3 Ethyl
12.1 27.6 Dimethylhenzyl
Hydroperoxide
12.1 370 l-Nonanethiol
12.1 46.2 Hexahydroindan
12.2 64.6 i-Butylcyclopentane
12.6 60.0 1,2,3-Trimethylcyclohexane
12.6 30.0 2-Thia-1,2,3,4-
Tetrahydronaphthalene
126 423 1,2,4-Trimethylcyclohexane
12.7 47.8 n-Butylcyclopentane
13.1 21.2 2-Thiatricycle (3.3.1.137)
Decane
13.1 34.8 4-Nonanol
13.1 42.9 1.1.3.4-Tetralnethyl-
cyclopentane
13.1 46.6 2,4-Dimethylbenzyl Alcohol
13.3 23.9 Cinnamyl Alcohol
13.3 33.3 3-Nonyne
13.4 28.6 Cinnamic Acid
13.6 41.7 l-Nonene
30.8
46.8
66.6
364
36.4
43.6

:5 “m sums"s*s °° “up °5555 “usucscucmusss s asxam--ww “NH HHHHH H
88 we 5§astae EB 888 Sfifififl maeeaesasuaees 8.33528“~**<~ an“ udawm o

Pentane

40.0 Butyric Anhydride
16.7 Bie(2-Dimethylaminoethyl)

n-Octyl Formamide
1, 1-Diethoxy-2-Methylhutane
1, 1,3,3-Tetramethyl-

2,6-Dimethylbenzoic Acid
1, 1-Diethoxy-3-Methyl-3-

1,1,2-Trimethylcyclohexane
66.0 4,4-Dimethyl-3-Ethyl-2-

commooccococcocoocooo 0 OQQOU‘DQQOQQ‘DO OQGCOQOQ$O@DOOUD 900$” 000000030690” 00

14.3

14.6
14.6
14.6
14.6
14.7
148
16.2
15.2
15.2
16.3
154

15.5
15.6
15.6
15.7

16.1
16.2

16.6
16.7
16.9

17.0
17.0
17.1
17.3
17.6
178

amouaaathmmaaaw A chamaqm+5aaau «mammomamcoaaaaq $5505 555655550055 5i.
5555558585555555 5 5585558855558 8855855855855888 85888 588885885558 8;

R
46.2
34.8
38.1
42.9
42.9
35.0
31.8
33.3
33.3

50.0

1744

31.6
40.0
50.0
17.4
40.0

”.1
”.4
14.8
14.3
50.0
57.1
31.3
22.7
25.0
50.0

37.5
31.3

w
1-Methyl-1,3-Dimethyl-6-
Cyclohexanol
4-Methyl-4-Octanol
3-Methyl-2-Propylpentanol
2,6-Di1nethyl-4-Heptanol
1,1-Diethoxypentane
n-Butyl i-Valerate
3,6-Dimethylbenzyl Alcohol
AR-Vinylbenzyl Alcohol
ar-Ethylbenzyl Alcohol
1,1-Diethoxy-3-Methylbutane
2,6-Dimethylbenzoic Acid
3,4-Dimethylbenzoic Acid

2-Cyclohexylethyl Methyl
Ether

Alpha Hydrindone
2,4-Dimethyl-4-Heptanol
2-Nonyne
n-Hexylpmpionate
1-Thia-1,2,3.4-Tetrahydro-
naphthalene

Nonaneic Add
p-Chlorocurnene

Allyl Caproate

N-Amyl Methacrylate
n-Butyl n-Valerate
2,6-Dimethyl-3-Heptanol
i-Butyl n-Valerate
3-Nonanol

1-Indanone

i-Amyl Butyrate
2,4-Dimethyl-3oHeptanol
4-Nenyne
3-Methyl-4oEthylhexene
2-Methylpmpyl i-Valerate
N ornicotine
2-Butyl-2-Ethylo1,3-
Propanediol
2,6-Nonedienal
i-Propylcyclohexane
l-Ethoxy-l-Pentoxyethane
6-Nonano ne
3,3,4-Trimethylhexane
1-Methoxy-2-Phenoxyethane
4-Nonene
3,3-Dimethylheptane
i-Prepenylbenzene
6-Nonano ne
3,8-Dimethyl-3-I-Ieptanol
l-Phenoxy-z-Propanone
1.1-Dimethylhutoxy-2-
Propanol

Methyl 4-Methylcyclo-
hexanecarboxylate
3,3,6-Trimethylcyclohexanol
Cumene Hydroperoxide
Phenyl 2-Propynyl Ether
Amyl Butyrate
3,4-Dimethylheptane
1,3,6-Trimethylcyclohexane
Phenyl Allyl Ether
o-Creeylethyl Ether
1,1-Dipmpoxypropane
m-Creeylethyl Ether
l-Chlorononane
2,4,4-Trimethylhexane
i-Butyl i-Valerate
1,1,3-Trimethylcyclohexane
2,4-Dimethylbenzoic Acid
Methyl Caprylate

no mac cccoomcmoooooooom oomoc coco oceceooc cocoa owe cooeoomwm e @000

b” #55 +wwmhuwmmmwuuawmw ”mama babe auhubuma ”5555 55+ Ahauuahhq h 555§
B5 555 555°5555555555555 55°55 5555 55555555 55555 555 555555555 5 555;

41.7
17.6
26.0

36.7
41.7
13.0
20.0
17.6

10.6

23.1
21.4
36.7

26.0
23.6

E

16.7
16.7
26.0

31.3
17.6

16.7
21.0

170

Omani
2,6-Dimethylbenzyl Alcohol
Methyl-4-Ethylcyclohexane
1-Phenyl-1-Methyl-1,2-
Epoxyethane
A-Ethyl-p-Hydroxyhenzyl
Alcohol
2,6-Dimethyl-4-I-Ieptanone
p-Creeylethyl Ether
AR-Ethylbenzaldehyde
Cinnemaldehyde
Chlorovinyltoluene
2-Thiebicyclo (4.4.0) Decane
o-Allylphenone
3-Phenyl-1-Propene
2.3-Dihydro-2-
Methylbenzofiiran
Dihydrocoumarin
3,4,4-Trimethyl-2-Hexene
242,2-Dimethylpropyl)
Thiophene
2,2-Difurylmethnne
2,3-Dimethyl-3-Ethylpentane
mPropylcyclohexane
Di-eec-Butoxymethane
2.6-Dimethyl-3-
Ethylpyridine
3,5-Dimethylbenxoic Acid
3-Hepten-1-yl Acetate
3-Cyclohexylpropionic Acid
m-Methyletyrene
Methyl p-Tolyl Ketone
Methyl o-Toluate
i-Propyl Hexanoate -
3,5,5-Trlmethylhexanoic
Acid
7-Methyl-4o0ctanone
n-Nonane
o-2-Propylphenol
3,6,6-Trimethylcyclohexen-
3-one
Ethyl Benzyl Ether
2-Methylhutyl i-Butyrate
3-Nonano ne
4-6-Propylcyclohexenol
2,6-Dimethyl-4-
Ethylpyridine
l-Methyl-I-Ethylcyclohexane
Benzyl Acetate
1-Methoxy-1-Hexoxyethane
2-t-Butyl-4-Methylfi1ran
4-Ethylheptane
2,4,6-1‘rimethylphenol
6-Methylphthalide
2-Methyl-4-Ethylhexane
6oNonanol
Vinylbenzaldehyde
n-Heptyl Acetate
o-Methylacetophenone
n-None-2,4-Dienal
p-Methylacetophenone
2-Nonano ne
2-Ethylbenzimiduole
2-(2-Methylellyloxy)-2-
Methyl-l-Propenol
3-Methyl-3-Octenol
Triallylemine
4-Butoxy-3-Methyl-2-
Butanone
6-Methylhenzo (B) Thiophene
2-Indanone

@000 $00 @QQOO@@0 @QOOG@Q@O@ 0 @OfifiGQQOOOUQQOO ooocoomoccoomoocoooccooc

31.1
31.1
31.3

31.6
31.6
31.9

32.1

mam» HhH uhuu&u»u wnmwwaa¢on » unfin+nnw¢onunau anaunwuN+NAuuaaaaubhbaa§
555'0 555 555555°5 0555555555 5 55555555°555555 55555555555555555555555;

R
46.6
21.4
26.0
m

(humans!
2.2.3.3-Tetramethylpentene
Cyclohexyl Acrylate
Coumarin
2,2,3,4-Tetraxnethylpentene
3-Ethylheptane
2-Methyloctene
3-Methyl-3-Ethylhexane
4-Nonanone
2,3,4-Trimethylhexane
o-Methylatyrene
2-Methy1-1-Octene
o-Methylphenetole
2,6-Dimethylheptene
6-Methyl-2,3-BenzothiOphene
4,4-Dimethylheptane
6-Methyl Indole
3,6,6-Trimethyl-1-Hexenol
n-Ethyl-o-Toluidlne
2,3,3-Trimethylhexane
4-Methyloctane
2-Methyl-3-Ethylhexane
4-Methyl-2,3-Benxothlophene
2,4oDimethyl-6-
Ethylpyridlne
Hydrocinnamic Acid
2-Methyl-3-Octanone
Beta-Phenylethyl Formete
6-Methyl Indole
Chloromeeitylene
2.6-Dimethylheptane
o-Chlorocumene
Methyl Octanoate
p-Creeyl Acetate
1,2,4-Trimethylbenzene
2,4-Dimethylheptene
er-Ethylbenzyl Chloride
Styrene Glycol Methyl Ether
3-Chloroallyl Benzene
l-Methoxy- l-trana-Hexene-
2-oxy Ethane
2,6-Dimethyl-2,6-Heptadien-
4-one
2-Methyl Indole
2,6-Dimethyl-3-Heptanone
2,4-Dimethyl-3-Ethylpentnne
24m-Tolyloxy) Ethanol
2,3,3,4-Tetramethylpentane
2-Phenyl Propynal
1,2,3-Trimethylbenzene
n-Aznyl i-Butyrate
Indene
p-Dimethylamino-
benxeldehyde
2,2-Dimethyl-3-Ethylpentane
l-Chloroindane
Tri-n-Propylemine
3-Methyloctene
AR-Methylacetophenone
2,3,6-Trimethylhexane
3,6-Dimethylheptane
2,4-Dimethyl-2,4-
Heptedienal
2-Nonenol
2,3-Dimethylheptane
2,6-Dimethyl-3-n-
Propylpyrezine
Methylphenylacetylene
2,2,3~Trirnethy1hexane
Methyl p-Toluate
Indane

 

owooomomomeooocmccmmwo

5

47's

waww»»»HNHHH»uhnu+N»§
5°°°mm°55555555°55555a

5
5

11.1

11.1
11.1

26.0
11.1

12.6
16.7
40.0
16.7

16.7
60.0

31.1

3.7

63.6

29.7

31.6

36.1

171

(human!
3-Methyl Indole
l-Methylhutyl i-Butyrote
Di-n-Butoxymethene
2,2-Dimethylheptene
Phenyl-Z-Propanone
Ethyl Benzoete
Methyl m-Toluete
2-Ethyl-6-n-Propylthiophene
Decahydroquinoline
Quinoline
gamma-Nonalactone
1,3,5-Trimethylbenzene
i-Quinoline
2-Methylbenzofurnn
2-i-Amylthiophene
Phenylecetone
Di-i-Butoxymethane
p-Methoxyacetophenone
ar-Vinylbenzyl Chloride
2-Phenylpropanol
3-Cyclohexene-1-
Cerboxaldehyde Dimethyl
Acetal
3-Chloropropenyl Benzene
(l-Chloroethyl) Toluene
3,6,5-Trimethylhexylemine
2,2,4-Trimethylhexene
1-Chloro-2-Propylbenzene
p-Methylphenethyl Alcohol
p-Ethyltoluene
i-Propylbenzene
3,3-Diethylpentane
6-Methquuinoxnline
p-2-Propylphenol
2,2,6-Trimethylhexene
6~Methquuinoxaline
1-Phenyl-1,2-Propened.ione
m-Ethyltoluene
l-Methoxy-l-(cieo3-
Hexenoxy) Ethane
(2-Chloropropyl) Benzene
Phenylpropioneldehyde
o-Ethyltoluene
2-Phenoxy-1-Propenol
(3-Chloropropyl) Benzene
1-Methoxy-1-cia-Hexane-3-
Oxyethane
2-n-Pentylthiophene
Propiophenene
2,2,4,4-Tetremethylpentene
Santene
p-Propoxyphenol
Phenyl n-Propyl Ether
n-Propylbenzene
n-m-Butylpiperidine
3A,4,7,7A-Tetrahydro-4,7-
Menthanoinden-lool
n-Decanel
Cinnamyl Methyl Ether
3-p-Tolylpropynal
Dipentene Oadde
2-Cyclopentylidene
Cyclopentanone
1-Decanol
a-Pinene Oxide
2-Decenal
1,2,4-Trimethyl-1-
Cyclohexene-4o
Carboxaldehyde
Cinnamyl Formate

555555555555555555555655555555555555555555 55555556555555555566 5555550

55658588888855868588385385589655588888853885 aseaeaceaaawzcaeureas 88568925;

sic

11.1

14.3

24.2
34.5
26.7
19.4
31.0
66.0
31.0
31.0
40.7
40.9
83.3
33.3
37.6
47.4
203
46.0
39.1
40.9
510
7.7

26.6

Omani
alpha-Terpineol
2,6-Dimethyl-5-Octenol
3,7-Dimethyl-1-Octanol
2,4-Decadienal
Verbenone
4-(3-Cyclohexen-1-yl)-3-
Buten-2-one
Decehydronaphthalene
2,4,5-Trimethylbenzoic Acid
l-Decyne
7,8-Dihydrolinalool .
2,2,5,5-Tetramethylhexene
p-Menthadien-1(7),8-ol-10
6-Decanone
2-(Butynyl) Cyclohuanone
2,7 Dimethyl Octanol
p-Menthen-B-ol-IO
Spiro(4,5) Decane
1-Decene
2-Decyne
Cerveol
Linelool
Pinane
3,6-Dimethyl-1-Thieindene
Alpha-Fenchene
Anebecine
3,7-Dixnethyl-2,6-Octadien-1-
ol
Camphene
i-Menthol
Neoieothujyl Alcohol
Citronellol
Myrtenal
2-Propylheptanol
2-aec-Butylcyclohexenol
1,8-Cineole
Cyclodecenone
Cyclopentylcyclopentane
4-Cyclohexyl-2-Butanol
Menthol
6-Decyne
Di-i-Amylene
3-Allyeelicyleldehyde
aec-Butylcyclohexane
6-Ethyl-3-Octanol
2.2-Dimethyl Octane]
Cumicaldehyde
Citronellol
i-Pulegol

100.0 Caren-2-ol

43.6
42.3
28.6
31.0
47.4
46.0
17.9
36.8
46.0
12.6
40.0
21.4
36.4
40.0
46.7
13.8
31.6

4,5-Dimethyl-4-Octenol
3-Decyne

4-Ethy1- 1-0ctyn-3-ol
Adamantane
2-Methyl-3-Nonene
8-Menthene
alpheoTetranol

l-Octyl Vinyl Ether
Endo-i-Camphane
Cyclopentylcyclopentanol
4-Methy1-4-Nonanol
Dipentene
3,7.Dimethyl-3-Octenol
Exoiecamphane
1-Hexyl-1,3-Butadione
Umbellulone

Fenchone

100.0 Caren-B-ol
18.5 Nerel
44.4 Fenchane

 

172

555 555555555555555555 5555555555555555555 555555555555 55555555555 556

2 3 M6 N6 R Gangland C E 3 M6 N6 B Omani!

16.6 6 2) 30.0 Ethyl l-Phenylethyl Ether 10 $.1 3 21 14.3 6-Methyl Quinoline

169 7 2) 35.0 1-Methyl-4-i-Propyl-3- 10 22.1 3 18 16.7 Myrcene
Cyclohexene 10 22.2 2 2) 10.0 Citral

169 9 $ 40.9 4-Decyne 10 $3 4 16 25.0 n-Butylcyclohexane

17.1 10 10 100.0 Caran-4-ol 10 22.4 3 2) 16.0 2-Methylindan

17.2 2 1% 6.3 i-Borneol 10 22.6 3 21 14.3 6-Methyl-2-Furfurylfuran

17.2 7 $ 36.0 i-Butylcyclohexane 10 $6 4 18 22.2 4-Ethyl-2-Octene

17.4 10 10 1000 cia-Caranone-3 10 22.6 6 13 46.2 3-Decanone

17.6 7 $ $0 D-Limonene 10 $0 2 18 11.1 2-(1-Propenyl)-6-

17.7 3 $ 13.0 p-Methylallylphenol Methoxyphenol

17.7 10 10 100.0 trans-Caranon-3 10 $0 3 2) 15.0 p-Diethylbenzene

180 7 2) 35.0 Camphor 10 $1 4 20 $0 Methyl-l-Indene

18.1 7 m 35.0 Terpinolene 10 $2 6 16 40.0 3,6-Dimethyl-3-Octanol

13.2 6 $ 27.3 2,3-Dimethyl-6-i- 10 $3 3 14 21.4 2,5-Dimethylhenzo (B)
Butylpyrizine Thiophene

184 3 $ 13.0 alpha-Tetralone 10 $3 6 16 33.3 Propyl Hexyl Keytone

18.4 7 14 50.0 Methyl ar-Vinyl Ether 10 23.4 4 18 22.2 i-Amyl Ether

186 8 22 36.4 n-Butyl Cyclohexyl Amine 10 $4 4 17 23.5 4-n-Propyl-3-Heptene

18.6 6 19 31.6 4-Ethyl-3-Octene 10 $7 3 15 $0 4-t-Butylcyclohexanone

18.7 6 19 $3 2-t-Butylcyclohexanol 10 $.7 3 13 23.1 2,7-Dimethylhenzo (B)

18.8 6 16 37.6 3-Methylindan Thiophene

189 6 22 27.3 1-Phenyl-2-Butene 10 23.7 4 16 26.0 p-i-Propyl Benzoic Acid

18.9 6 14 42.9 2-Methyl-5-Ethylheptane 10 $9 3 19 16.8 Carvotanacetone

190 6 14 429 2,6-Dimethylmethylbenzoate 10 $9 3 16 18.8 1-Methoxy-4~(1-Propenyl)

190 7 16 43.8 2,6-Dimethyloctane Benzene

19.1 6 2) 30.0 Vinyl 2-Ethy1hexyl Ether 10 24.0 3 16 18.8 2.4.6-Trimethylhenzaldehyde

19.2 7 12 58.3 2-Methyl-2,3-Dihydro-1,4- 10 24.0 3 16 $0 L-Phellandrene
Benzopyran 10 24.0 4 18 222 1-Phenylpyrrole

19.3 2 21 9.6 Fenchyl Alcohol 10 24.1 3 16 $0 Geranial

19.3 3 23 13.0 Nerol 10 24.1 4 10 40.0 2,4-Dimethylacetophenone

19.3 7 19 36.8 Benzyl Propionate 10 24.2 3 14 21.4 Sabinol

19.4 9 21 42.9 p-Menthen- 1-01 10 24.6 6 12 60.0 Cuminyl Alcohol

19.6 4 21 19.0 Terpinen-4-ol 10 24.6 3 16 18.8 m-Diethylbenzene

19.6 6 18 33.3 4-t-Butylcyclohexanol 10 24.6 3 14 21.4 2,3-Dihydro-2-Methyl—

19.6 5 19 $3 Myrcene benzofurancarhoxaldehyde

19.7 4 21 19.0 o-Diethylbenzene 10 24.8 1 2) 6.0 Cycloi‘enchene

19.8 3 22 13.6 2-Naphthalene Thiol 10 24.8 5 13 38.6 ar-Ethylphenethyl Alcohol

19.8 6 $ 26.0 Nopinene 10 249 2 19 10.6 Thymol

199 3 $ 13.6 Methylallyl Phenyl Ether 10 26.0 4 12 33.3 3-Methylindene

19.9 7 14 50.0 2,7-Dimethyloctane 10 266 6 16 33.3 4-Decanone

$0 5 17 29.4 Alpha-Terpinene 10 26.7 2 18 11.1 1,2,4,6-Tetramethylbenzene

$0 5 12 41.7 4-Phenyl-3-Buten-2-one 10 $.7 3 16 18.8 3-Menthene

$0 10 10 100.0 Myrtenal 10 268 1 21 4.6 Nicotyrine

$0 10 10 100.0 i-Menthone 10 $8 4 14 $6 2,3-Dimethyloctane

$3 4 $ 182 Diethylcyclohexane 10 $.1 3 16 18.8 3,6-Dimethylatyrene

$3 6 14 429 Dion-Pentyl Ether 10 $2 3 16 $0 o-Allyltoluene

$4 3 $ 13.0 1,2,3,4-Tetrahydro- 10 $2 7 10 70.0 Eucarvon
naphthalene 10 $4 3 19 16.8 Geraniol

$6 4 17 23.6 2,6-Dimethylatyrene 10 $6 6 16 33.3 i-Propyl Benzoate

$6 4 $ 18.2 Carvomenthone 10 $.7 3 14 21.4 2,3-Dimethylindole

$6 4 18 22.2 Trimethylbenzyl Alcohol 10 $.7 6 16 31.3 Allocirnene

$6 7 $ 31.8 4-eec-Butylcyclohexanol 10 $8 3 16 18.8 5-Methylindan

21.7 6 16 37.6 Allyl Benzyl Ether 10 $8 4 18 22.2 Carvone

$9 2 19 10.6 1,7-Dimethylindole 10 $8 5 14 36.7 p-t-Butylphenol

21.0 3 2) 16.0 6,7-Dimethylindole 10 $9 2 14 14.3 2,2-Prim-Bipyridyl

210 6 21 $8 3,8-Menthadiene 10 $9 3 11 27.3 2,2-Dimethyl-4-Ethylhexane

21.1 3 2) 16.0 1,3-Diethylbenzene 10 27.0 3 13 23.1 4-NoPentylpyridine

21.1 7 16 43.8 2-Methylnonane 10 27.1 2 16 13.3 4-Methylindan

21.3 3 19 16.8 Lavandulol 10 27.1 5 17 29.4 Gamma-Terpinene

21.4 4 18 222 Limonene 10 27.1 6 12 41.7 2,4-Dimethylmethylbenzoate

21.6 2 18 11.1 1-p-Menthen-9-al 10 27.2 4 16 26.0 Thujene

216 4 19 21.1 a-Campholene Aldehyde 10 27.2 4 16 26.0 l-Menthene

21.6 5 14 36.7 n-Decane 10 27.2 4 12 33.3 t-Butylcyclohexane

21.7 7 13 53.8 3,3,5-Trimethylheptane 10 27.6 3 12 26.0 o-t-Butylphenol

218 4 19 21.1 (2-Methylpropenyl) Benzene 10 279 2 14 14.3 1,4-Diacetylbenzene

218 4 17 23.6 1-Methyl-4-i-Propyl- 10 $2 3 16 $0 2-Cyclopentyl-1-
cyclohexane Cyclopentanone

219 6 19 $3 2,4,6-Trimethylhenzoic Acid 10 $6 2 12 16.7 t-Butylbenzene

$0 3 17 17.6 o-Methallylphenol 10 $6 4 12 33.3 2-Decanone

22.0 4 18 222 o-Allyloxybenzaldehyde 10 $.1 4 13 m8 4-n-Propylheptane

 

14‘ II III] Ill 14"

 

 

55 5555555 5 5555555555555555 55555555555 5555555555555555555555555555O

NH HNHNNNH H wanwwwuunmwnpww unwwwupnunn unwnwwwwunnwbhuwwhwnwwnwnwubE
a” 5555500 0 °555°°5555°5°555 qqq55m5m555 m5555““5555555555555555'95555a

42.9

E

16.7
7.7
7.7
26.0
7.7
37.6
$0

7.7
11.1

30.0
11.1
11.1
11.1

26.0
14.3

11.1
40.0

173

1,4-Naphthoquinene
3-Carene

m-Ethylatyrene

Di-i-Amyl Amine
1-Chloronaphthalene
i-Eugenol

Menthofurane
1,3-Dimethylindole
3-Acetylindole
l-Methylindan
3,4-Dimethyletyrene
p-Allyloxybenzaldehyde
Bicyclodihydrodipentadiene
Piperitone
2-Methyl-3-Nonanone
Alpha-Phellandrene
1.2,3.6-Tetramethylbenzene
Nicotine
1,4-Dimethyl-2-Ethylhenzene
Beta-Pinene

Tricyclene
2,6-Dimethylindole
2,6-Dimethylindele
2-Chloronaphthalene
Alpha-Pinene
1,3-Dimethyl-6-Ethylhenzene
Carvacrol
1,3,3-Trimethyl-2-
Norboranone
1,2-Dimethyl-3oEthylbenzene
n-Butylbenzene

Divinyl Benzene
m-Ethylphenyl Acetate ‘
1-(2-Phenethyl)Azridine
Methyl Cinnamate
2,2,4-Trimethylheptane
1,2-Dimethyl-4-Ethylbenzene
i-Butylbenzene
ar,ar-Diethylphenol
1-Methyl-4-i-
Propenylbenzene
p-Ethylatyrene
l-Methyl-z-i-Prepylhenzene
2-N-Pentylpyridlne
p-i-Propylaniaole
o-aec-Butylphenol
3,6-Dimethylmethylhenzoate
Ethyl Phenyl Acetate
1-Methyl-3-i-Propylbenzene
1,3-Dimethyl-2-Ethylhenzene
Azulene

p-Diacetylbenzene
p-Ethylacetophenone
1,3-Dimethyl-4-Ethylbenzene
1-Methyl-4~i-Propylhenzene
i-Piperitenone
2,6-Dimethyl-3-
Butylpyrizine
2,6-Dimethyl-3-
Butylpyrizine
1-Methyl-2-n-Propylhenzene
p-Cymene

3-Phenylfuran

o-Cymene
4-Phenyl-1-Butene
2-i-Propyl-6-Methylphenol
3A,4,7,7A-Tetrahydro-4,7-
Methanoindene
aecoButylhenzene
m-n-Propyltoluene

5555555 5555555555 5555 55555555 555 555 55555555555555555555555555555O

47.4

14.9

160
16.7
17.6
179
18.4
198
$8
$8

$8
21.7

239
24.1
24.3
$2

$8
$7
$0
$2
$6

$8

$9
27.6

$0

”wanna-ac: “wanton-henna wana #4550944“ 50:5 51:5 swat-druidHHHHHHHHHHNQNHNHN’HIOHHNHE

5°=ax°c ssssesauss 8:5: 85558588 589 888 erswrw°“*°°°°°°s“““*us“m°m°4“a

n-Decyl Amine
i-Safrole
Borneol
1-Methyl-2-n-Propylhenzene
n-Butyrophenone
1-Methyl-4-n-Propylbenzene
2-Vinylbenzofuran
p-aec-Butylphenol
2-Naphthol

n-( l-Methylpropyl) Phenol
i-Butyrophenone

1N0 p-Cymene

40.0

16.7
16.7
12.6

16.7
14.3
16.7
12.6
33.3

(ar-Vinylphenyl) Acetic Aci
Naphthalene
o-n-Propyltoluene
p-Methoxy Propiophenone
1-Decylamine
1-Fluoronaphthalene
2-Fluoronaphthalene
n-Butyl Phenyl Ether
t-Butyl Phenyl Ether
Thujone

Allyl Benzoate

Pulegone

100.0 Menthone

33.3
$0

2-Menthene
Sabinene

1N0 Beta-Phellandrene

238
$8
37.9
$8

60.0

11.1

6-Methyl-1,2,3,4o
Tetrahydronaphthalene

l Ethyl Ether
Spiro (6.6) Undecane
3,4-Dihydro-4-Methyl-1(2H)-
Naphthalenone
l-Undecene
5,6-Dimethylindan-1-one
3-Methyl-4-Phenyl-3oButen-
2-one
4,7-Dimethylindan-1-one
6-Methyldecene
4-Methyldecane
6,7 -Dimethylindan-1-one
6-Methyldecane
5,6,7-Trimethylindole
2-Cyclohexyl-2-Methylhutane
p-( 1, 1-Dimethylpropyl)
Phenol
n-Undecane
4-i-Propylacetophenone
1,2,3-Trimethylindole
2-Methyl-1,2.3,4-
Tetrahydronaphthalene
Propyl Benzyl Keytone
3,3-Dimethylindan-1-one
l-Phenylo2oMethylhutane
Neopentylbenzene
i-Butyl Propionate
1,2-Dimethylindan
2-Methyldecane
6-t-Butyl-m-Cmeol
l-Methyltetralin
1,23,4-Tetrahydro-6-
Methylnaphthalene
ar.ar-Diethyltoluene
l-Methoxynaphthalene
6,6-Dimethylindan
l-Phenyl-a-Methylhutane
6-t-Butyl-o-Creeol
1-Methylnaphthalene
1-Ethylindan

55 555 55 555 55 5555 55 555555 55 5 55555555 55 55 5 5555555555555550

$0
$.1
$.7
$6
$6
316
31.7
319
319
319
32.1
32.7

$0
33.1

33.4
33.4

34.2
34.4

$.1
$6
15.7
$9
37.1
38.1
396
416

42.7

440

456
45.7
490
14.2
16.4
16.6

176
$6

$3
$.4
$6

$0
$.1
$3
$4

$6
$0

$9
32.1

34.6
34.7

H” “N10 N66 20”” MN Math-b 034 wamwam NH H NNWHHNHN HH N“ H w»5wwwwwuwwnwww§

55 555 55 555 55 5555 58 858°q“ °°5 “ “°°‘"°5‘"°" 55 5‘0 5 555“”555555m555a

11.1

11.1
429

14.3
10.0

E

EEEEE

10.3

$6

17.6
11.1

16.4
13.3

11.8
16.4
16.7

$1
143

16.7
16.7
23.1

$0
9.1

174

CW C 2 3 M5
1-Methyl-3-t-Butylbenzene 12 346 1
4,6-Dimethylindan

4,7-Dimethylindan 12 349- 2
2-Methylnaphthalene
1-Methyl-4-t-Butylbenzene 12 35.2 2
1,6-Dimethylindan 12 35.7 1
p-i-Propenylacetophenone
2,6.Dimethquuinoline 12 36.7 2
2-Phenyl-2-Methylbutane 12 $2 1
n-Pentylbenzene 12 $2 3
p-Ethylcumene 12 $6 1
2,4,6-Trimethylacetophenone 12 $6 1
3-Phenylpentane 12 37.3 1
Pentamethylbenzene 12 376 1
3,5-Dimethyl-1-i- 12 $6 1
Propylhenzene 12 $3 2
1,4-Dimethyl-2-i- 12 39.8 1
Propylbenzene 12 40.1 1
2-t-Butyl-m-Creeol 12 409 1
1,3-Dimethyl-5-i- 12 432 2
Propylbenzene 12 43.9 1
2-N-Hexylpyridine 12 450 3
1,3-Dimethyl-4-i-

Propylbenzene 12 580 2
l-Methylbutyl Benzene 12 59.0 1
N -Benzylpyrrole

2-Methoxynaphthalene 12 789 1
1,1-Dimethylindan

p-t-Butylaniaole

n-Valerophenone

4-t-Butyl-o-Cresol

2,5-Dimethyl-1-i-

Propylbenzene

2,4-Dimethyl-1-i-

Propylbenzene

2-Phenyl-2-Methylbutane

ar-Ethyl-1,2,4-

Trimethylbenzene

t-Butyltoluene
m-Ethylcu mene

Ethyl Methyletyrene
3-Phenylcyclohexene
1-Phenylcyclohexene
1,2,4-Trimethyl-5-i-
Propylbenzene
Cyclododecetriene
2,3-Dihydro-1,4c
Dimethylnaphthalene
l-Phenyl-Z-Ethylbutane
Acenaphthene
m-Di-i-Pro pylhenzene
1,3-Dimethyl-6«t-
Butylhenzene
3-Phenyl-3-Methylpentane
Li-Propyl-3-i-
Propenylbenzene
p-Di-i-Propylbenzene
4,5,7 -Trimethylindan
1-i-Propyl-4-i-
Propenylbenzene
1,6-Dimethylnaphthalene
2,3-Dimethyl-3-
Phenylbutane
1,4-Di-n-Propylbenzene
1-Phenyl-3-Methylpentane
2A,3,4,5-Tetrahydro-
acenaphthene
p-toButylatyrene

1, 1-Dimethyl Tetralin

0mm!
1-Ethyl-(13.3,4-
Tetrahydronaphthalene)
1,2-Dimethyl-3,4-
Diethylbenzene
l-Ethylnaphthalene
1,3-Dimethyl-4-eec-
Butylbenzene
Hexamethylbenzene
1,1,6-Trimethylindan
Triethylbenzene
1,4,7-Trimethylindan
1,1,5-Trimethylindan
1,1,4-Trimethylindan
1,5,7-Trimethylindan
1,2-Di-i-Propylbenzene
Biphenyl
ar-t-Butyl-ar-Ethylbenzene
1,1,3-Trimethylindan
3-Phenylhexane
4-Phenylcyclohexene
2-Phenylhexane
ar-i-Propenyl-l,2,4-
Trimethylbenzene
2-Phenyl-2-Methylpentane
ar-i~Propyl-1,2,4o
Trimethylbenzene
100.0 2,6-Dimethylnaphthalene

10.0
429
11.1
8.3
100
126
10.0
333
14.3
11.1
126
250
9.1
75.0
33.3
$0