Title: Tandem mass spectrometry for the identification and quantitation of tryptolines (tetrahydro-beta-carbolines) in rat brain extracts /
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097415/00001
 Material Information
Title: Tandem mass spectrometry for the identification and quantitation of tryptolines (tetrahydro-beta-carbolines) in rat brain extracts /
Physical Description: ix, 147 leaves : ill. ; 28 cm.
Language: English
Creator: Johnson, Jodie Vincent, 1954-
Publication Date: 1984
Copyright Date: 1984
 Subjects
Subject: Mass spectrometry   ( lcsh )
Brain chemistry   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1984.
Bibliography: Bibliography: leaves 141-146.
Additional Physical Form: Also available on World Wide Web
Statement of Responsibility: by Jodie Vincent Johnson.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097415
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000506237
oclc - 12203427
notis - ACS6554

Downloads

This item has the following downloads:

tandemmassspectr00johnrich ( PDF )


Full Text












TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND QUANTITATION OF
TRYPTOLINES (TETRAHYDRO-BETA-CARBOLINES) IN RAT BRAIN EXTRACTS






















BY

JODIE VINCENT JOHNSON


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PRHLOSOPHY



UNIVERSITY OF FLORIDA


1984

































To my loving wife, Karen,
and our son, Joshua















ACKNOWLEDGEMENTS


I would like to express my sincere gratitude to Dr. Kym F. Faull at

the Stanford University Medical Center for his initiation of this pro-

ject and provision of samples and for his advice, patience, and valued

friendship during this collaborative research. I would like to thank

Dr. Olof Beck, at Stanford University also, who helped plan and prepare

samples for the experiments in Chapter 5, and Jack D. Barchas, M.D., for

securing funding for the work performed at Stanford.

I would like to express my gratitude to Dr. Richard A. Yost for his

acceptance and guidance during this research. His editorial assistance

during the preparation of this dissertation and various papers is very

much appreciated. I will always value his friendship. I thank him also

for offering me a postdoctoral appointment.

I acknowledge the members of my research committee, including Drs.

John G. Dorsey, Gerhard M. Schmid, Martin T. Vala, and Clyde M. Williams

for their suggestions and guidance. I especially thank Dr. Dorsey for

reviving my interest in pursuing a Ph.D. degree.

This research would not have been possible or as much fun without

the support of and discussions with my colleagues and friends in the

MS/MS research group. I would like to give special thanks to Neil

Brotherton and Dean Fetterolf for many helpful discussions. In

addition, I thank Linda Hirschy, Barbara Kirsch, Jonell Kerkhoff, and

Joe Foley for their friendship, help, and lunch-time talks, which helped

to maintain our sanity while taking classes and doing research.

___________________i i_____________---









I gratefully acknowledge the Chemistry Department for the first

year scholarship and Proctor and Gamble for sponsoring my fellowship

during this past year. I also thank Dr. Roy King for accepting me as

his graduate assistant and teaching me the varied aspects of high reso-

lution mass spectrometry. I also acknowledge research funding in part

by grants from the National Science Foundation (CHE-8106533) at the

University of Florida and from the National Institute on Alcohol Abuse

and Alcoholism (AA 0592) and the National Institute of Mental Health

(Program Project grant MH 23861) at Stanford.

I would like to acknowledge the support of my family during all the

years of my education, whether in or out of the classroom. Most of all,

I thank my wife, Karen, for her love, understanding, and patience during

these years in graduate school. She has made it all possible and worth-

while.
















TABLE OF CONTENTS

PAGE


ACKNOWLEDGEMENTS.................. ................................. iii

ABSTRACT ........................................................... viii

CHAPTER

1 INTRODUCTION. ................................................

Organization of Dissertation.............................1
Tryptolines (Tetrahydro-8-Carbolines)...................3
Tandem Mass Spectrometry (MS/MS).........................7
Instrument Description, Operation, and Application...... 10

2 TANDEM MASS SPECTROMETRY FOR TRACE ANALYSIS ................14

Introduction............................................ 14
Requirements for Trace Analysis.........................14
Instrumental Methods to Improve the Limit of
Detection ........................................... 16
Fluorometry. ......................................... 17
Gas Chromatography/Mass Spectrometry (GC/MS)......... 17
Tandem Mass Spectrometry................................ 18
Principles of MS/MS.................................. 18
Applications of MS/MS for Trace Analysis.............19
MS/MS trace analysis with MIKES instruments....... 19
MS/MS trace analysis with triple
quadrupole instruments.........................20
Conclusion.............................................. 23

3 TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND
QUANTITATION OF UNDERIVATIZED TRYPTOLINES...............24

Introduction.... ........................................24
Artefactual Tryptoline Formation.....................24
MS/MS for Mixture Analysis........................... 25
MS/MS for Structure Elucidation......................26
Experimental............................................27
Materials and Reagents...............................27
Instrumentation.....................................28
Procedures........................................... 29
Mass spectra of standards..........................29
Selection of positive or negative
chemical ionization.............................29









PCI-CAD collision energy and collision
gas pressure studies ...........................30
Quantitative studies..............................30
Results and Discussion..................................31
Nomenclature and Structure...........................31
El and EI-CAD Mass Spectra of Standards..............32
PCI and PCI-CAD Mass Spectra of Standards............38
NCI and NCI-CAD Mass Spectra of Standards ............43
Optimization Studies .................................51
Mode of ionization................................ 51
PCI-CAD collision energy..........................52
PCI-CAD collision gas pressure....................53
Quantitative studies.................................56
Conclusion.................................................. 63

4 TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND
QUANTITATION OF TRYPTOLINE-HEPTAFLUOROBUTYRYL
DERIVATIVES.............................................65

Tryptoline-HFB Derivatives..............................65
Experimental............................................66
Materials and Reagents...............................66
Preparation of Extracts..............................68
Chemical Derivatization.............................. 68
Instrumentation. ....................................69
Procedures........................................... 70
Mass spectra of standards.........................70
Collision energy and collision gas
pressure studies ...............................70
Selected ion and selected reaction monitoring.....71
Quantitative studies..............................71
Results and Discussion.................................73
Structure of the Tryptoline-HFB Derivatives..........73
Mass Spectral Characteristics........................73
El normal mass spectra.............................73
PCI normal mass spectra...........................74
NCI normal mass spectra............................74
PCI-CAD daughter mass spectra.....................75
NCI-CAD daughter mass spectra.....................79
Optimization of Experimental Parameters..............84
Source temperature...............................84
CAD conditions....... .................... .... ...84
Quantitative Studies.................................85
Standard calibration curves.......................85
Analyses of derivatized crude brain extracts......94
Conclusion.............................................104



5 TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND
QUANTITATION OF TRYPTOLINES IN RAT BRAIN EXTRACTS...... 105

Experimental............................................... 107
Materials and Methods...............................107
Chemicals and reagents...........................107


- I









Synthesis of standards............................107
Preparation of brain extracts....................108
Chemical derivatization........ ........... ...... 108
Gas Chromatography/Tandem Mass Spectrometry
(GC/MS/MS). ....................................... 110
Results and Discussion.................................. 112
Mass Spectral Characteristics....................... 112
Indoleamines...................................... 112
Tryptolines. ...................................... 113
Assay of Derivatized Rat Brain Extracts.............116
Artefactual tryptoline formation.................117
Quantitation of endogenous TLN...................122
Reproducibility of the TLN-HFB quantitation......126
Assay for other tryptolines......................131
Conclusion .................. .......................... 136

6 CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK...............138

BIBLIOGRAPHY..... ................................................... 141

BIOGRAPHICAL SKETCH............. ......... .......................... 147
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND QUANTITATION
OF TRYPTOLINES (TETRAHYDRO-BETA-CARBOLINES) IN RAT BRAIN EXTRACTS

By

Jodie Vincent Johnson

December, 1984

Chairman: Richard A. Yost
Major Department: Chemistry

The natural occurrence of tryptolines in mammalian tissue is the

subject of controversy. This is due to the lack of sensitivity and/or

selectivity in the methods used for identification and the possibility

of artefactual formation of tryptolines during sample preparation. Due

to its excellent sensitivity and inherent selectivity, tandem mass

spectrometry (MS/MS) has been used successfully for the direct analysis

of complex mixtures for trace components with minimal, if any, sample

clean-up.

The triple quadrupole tandem mass spectrometer consists of, in

series, a dual chemical ionization/electron impact ionization source, a

quadrupole mass filter (Q1), a radio-frequency-only quadrupole (Q2), a

second quadrupole mass filter (Q3), and an electron multiplier. In the

analysis of an extract, methane positive or electron capture negative

chemical ionization (PCI or NCI, respectively) of the extract produced

ions characteristic of the components of the extract. The MS/MS quanti-

tation in these studies was performed by selected reaction monitoring


.____________________' mr i _________-i^ ^ ^ ^ -









(SRM) whereby the ion characteristic of a tryptoline is mass selected by

Q1 for fragmentation in Q2 through collisions with neutral gas mole-

cules, and only the most abundant and characteristic daughter ion is

selected by Q3 for monitoring. The MS/MS quantitation was compared to

selected ion monitoring (SIM), whereby Q1 and Q2 pass all ions and only

the characteristic ion of a tryptoline is selected by Q3 for

monitoring. With sample introduction via a capillary column, the limits

of detection of the heptafluorobutyryl (HFB) derivative of tryptoline

were determined to be 0.50, 0.45, 19, and 60 pg of standard injected

onto the column for NCI-SIM, NCI-SRM, PCI-SIM and PCI-SRM,

respectively. The greater selectivity of the NCI-SRM technique made it

the preferred technique for the analysis of crude extracts of rat brain.

With GC/NCI-SRM analysis of HFB-derivatized crude extracts of rat

brains, it was demonstrated that artefactual formation of tryptoline

during the sample preparation used in these studies was negligible.

Tryptoline (19.2 3.6 ng/g wet tissue, n-7), methtryptoline, 5-hydroxy-

tryptoline, and 5-hydroxymethtryptoline, and their presumed precursor

indoleamines, tryptamine and 5-hydroxytryptamine, were identified in rat

brain extracts.
















CHAPTER 1
INTRODUCTION


This dissertation describes the use of tandem mass spectrometry for

the qualitative and quantitative characterization of tryptolines (tetra-

hydro-B-carbolines) in crude extracts of rat brains. The in vivo pres-

ence of tryptolines in mammalian systems has been an area of much con-

troversy since the first report describing their presence in 1961 (1).

This has largely been due to the inability of the analytical methods

used to consistently identify and quantitate these compounds at the

trace (sub-parts-per-billion) levels reported. This dissertation demon-

strates that, due to its inherent selectivity and sensitivity, tandem

mass spectrometry is able to detect reliably and consistently sub-parts-

per-billion levels of derivatized tryptoline in standards. Furthermore,

it describes the use of tandem mass spectrometry to identify four tryp-

tolines (two of which have not been previously reported) and to quanti-

tate one of these tryptolines in heptafluorobutyryl-derivatized crude

extracts of rat brain.




Organization of Dissertation

This dissertation is divided into six chapters. This introductory

chapter provides background information necessary for understanding the

significance of the research presented in later chapters. An overview

of the tryptolines is presented with regard to their physiological

significance and their in vivo presence in mammalian systems. A brief

historical review of tandem mass spectrometry is followed by a descrip-








tion of the operational modes and applications of the triple quadrupole

tandem mass spectrometer used in these studies.

Chapter 2 describes the use of tandem mass spectrometry for trace

organic analysis. This chapter, in combination with some of the

findings of chapters 4 and 5, has been prepared for publication in

Analytical Chemistry and Applied Spectroscopy Reviews. A brief review

of the fundamental analytical requirements and inherent difficulties of

trace analysis in general is presented. A review of the fundamental

aspects of tandem mass spectrometry is followed by descriptions of

several applications of this technique to trace mixture analysis. The

successful determinations of trace organic in complex matrices in these

applications serve as the basis for the research presented in chapter 3.

In chapter 3, the use of tandem mass spectrometry to structurally

characterize and quantitate underivatized tryptoline standards is dis-

cussed. The fragmentation pathways of the ions resulting from electron

impact and positive and negative chemical ionization are elucidated.

Following optimization of the experimental conditions, quantitation of

underivatized tryptoline standards is performed. The significance of

the limits of detection thus obtained is discussed with regard to deter-

mining the tryptolines in mammalian tissues.

The characterization and determination of the heptafluorobutyryl

derivatives are described in chapter 4. Many of the findings reported

in this chapter have been published recently in Analytical Chemistry

(2). Following a description of the mass spectral characterization of

the standards, the effect of experimental conditions on sensitivity is

discussed and optimized. Quantitation of the heptafluorobutyryl-deriva-

tive of the tryptoline standard is performed by normal and tandem mass









spectrometry with positive and negative chemical ionization and with

sample introduction via packed and capillary gas chromatographic

columns. The sensitivity, limit of detection, and speed of analysis of

each of the techniques are discussed. An application of each technique

to the determination of an amount of tryptoline added to a derivatized

crude extract of rat brain is demonstrated and discussed.

The use of tandem mass spectrometry for the identification and

quantitation of tryptolines in heptafluorobutyryl-derivatized extracts

of rat brains is described in chapter 5. Much of this work has been

submitted for publication as a chapter in "Aldehyde Adducts in

Alcoholism" (3). The tandem mass spectrometric methods developed in

chapter 4 are used to investigate the possibility of artefactual trypto-

line formation occurring during the sample work-up procedure, to quanti-

tate tryptoline, and to tentatively identify three other, two previously

unreported, tryptolines in derivatized crude extracts of rat brains.

The final chapter summarizes the results of this work and presents

some ideas for continued research in the area of tryptolines as well as

in the area of tandem mass spectrometry.



Tryptolines (Tetrahydro-B-Carbolines)

The tryptolines (1,2,3,4-tetrahydro-a-carbolines) are a class of

compounds resulting from the Pictet-Spengler condensation of indole-

amines and aldehydes (Figure 1-1) (4). The laboratory synthesis of

tryptolines by this reaction occurs readily under physiological condi-

tions (1,5). This property has been used in the past as the basis for

the histochemical detection of indoleamines by fluorometry, whereby the

indoleamines are converted to tetrahydro-B-carbolines via condensation









with formaldehyde gas and then oxidized to the highly fluorescent 6-

carbolines (6). In vitro formation of tryptolines also results from the

incubation of methyl tetrahydrofolate and indoleamines with various

tissue extracts having certain enzymatic activities (7-9). This reac-

tion involves the enzymatic formation of formaldehyde from methyl tetra-

hydrofolate followed by non-enzymatic condensation of formaldehyde with

indoleamines (10-12). These results, in conjunction with the fact that

indoleamines (13-16) and formaldehyde (17) are common constituents in

mammalian tissues, have led to the speculation that the tryptolines

could occur naturally in mammalian tissues.

Interest in the physiological significance of the tryptolines stems

largely from three sources. Firstly, they are known to elevate the

levels of serotonin (5-hydroxytryptamine) in the brain. Administration

of tryptolines inhibits the action of monamine oxidase A (18-20), in-

hibits the re-uptake of serotonin by the synaptosomal cells (20-24), and

facilitates the release of serotonin (25). Serotonin has been postu-

lated to be a putative neurotransmitter (26), and imbalance in the brain

serotonin levels has been associated with various mental illnesses

(27). Thus, the in vivo presence of tryptolines may have important

influences upon neurotransmission and mental illness (28,29).

Secondly, tryptolines have been associated with the alcohol abuse

syndrome (30). The level of acetaldehyde, the major metabolite of

ethanol in mammalian systems, increases in the plasma and brain fol-

lowing intake of ethanol (31,32), thus increasing the likelihood of

acetaldehyde-indoleamine condensation reactions. This hypothesis has

been supported by the determination of such tryptolines in the urine of

normal subjects after, but not before, intake of ethanol (33,34), and by

























Z
H

0

> ~


2


2

z


0


















-4
,)
Cn










3)


0
0












o
0
.,
U)









0











(o







o





















Ir-.
!,
0
a












-4
(0
































-,
0












r,.
0,




0












1-4









the increased excretion of such tryptolines in the urine of alcoholics

(35,36). In addition, after administration of tryptolines, rats pre-

ferred drinking alcohol over water (37,38). Thus, the in vivo produc-

tion of tryptolines may have important implications with respect to

alcoholism.

Thirdly, the identification of receptor sites in mammalian brain

specific for the benzodiazepines, a class of mild tranquilizers, led to

the search for possible endogeneous ligands (39-42). A 8-carboline

derivative was isolated from human urine which showed a very high af-

finity for the benzodiazepine receptors (43). Although this compound

had been chemically altered during the extraction procedure, it did lead

to the speculation that a member of the 8-carbolines or tetrahydro-8-

carbolines could be an endogeneous ligand for the benzodiazepine recep-

tor. This speculation was strengthened when it was shown that several

tryptolines and 8-carbolines which have been found in mammalian systems

were powerful displacers of tritiated flunitrazepam from brain tissue

homogenates (44). Related to this aspect, certain tryptolines have

recently been shown to exert LSD-like effects and to be associated with

the opiate receptors in the brain (45-48).

On the basis of combined thin-layer chromatography (TLC) and fluor-

ometry (49,50), liquid chromatography with electrochemical detection

(51), TLC with scintillation counting (52), and gas chromatography/mass

spectrometry (GC/MS) (33,53-57), the in vivo presence of tryptolines

has been described in various extracts of mammalian tissues and

fluids. Despite these numerous reports, widespread support for the

natural occurrence of these compounds in mammalian tissues has not been

forthcoming. This has been largely due to the lack of the sensitivity








and/or the selectivity of the above methods necessary for the consistent

detection of the tryptolines at the trace levels reported. In addition,

the possibility of artefactual formation of tryptolines during the

sample work-up procedure has further confused this issue (58,59).



Tandem Mass Spectrometry (MS/MS)

The demand for more sensitive, selective, rapid, and cost-effective

techniques for mixture analysis has been the spur for the continuing

improvements in mass spectrometry over the years. The need to quanti-

tatively analyze petroleum fractions was the impetus for the commercial

development of single-focusing magnetic mass spectrometers in the

1940s. The desire for higher selectivity and more information resulted

in the advent of double-focusing instruments which combined the momentum

analyzing property of a magnetic sector and the kinetic energy analyzing

property of an electric sector in order to obtain higher mass spectral

resolution. However, these instruments were fairly expensive to pur-

chase, operate and maintain. This led to the commercial development of

quadrupole mass spectrometers in the 1960s. It was quite apparent early

on that in order to obtain qualitative information about the components

of a mixture, the individual components had to be introduced into the

mass spectrometer in a relatively "pure" form. This was the incentive

for the development of gas chromatography/mass spectrometry (GC/MS) in

the late 1950s (60). The ability to separate and identify hundreds of

components in a mixture at trace levels has made GC/MS one of the most

powerful and widely-used techniques for organic mixture analysis (61).

The need to analyze thermally labile and/or nonvolatile compounds not

amenable to GC prompted the coupling of liquid chromatography to mass









spectrometry (LC/MS) in the late 1970s (62). Both of these chromato-

graphic systems provide, it is hoped, not only the introduction of

relatively pure substances into the mass spectrometer, but also the

additional information contained in the elution time of a component

after its injection onto the column, referred to as its retention

time. The combination of having a characteristic retention time and

having a characteristic mass spectrum of a component in a mixture is now

the accepted criterion used for confirming the identity of a compound.

One of the major drawbacks with the use of chromatography/MS for mixture

analysis is the relatively long analysis time. This is not only due to

the time required for the separation and elution of the component of

interest, but also due to the extensive sample clean-up and derivatiza-

tion procedures often necessary prior to the chromatographic step (62).

The mass separation of ions for subsequent fragmentation and mass

analysis, tandem mass spectrometry (MS/MS), has long been used in the

study of ion-molecule reactions (63) and, more recently, in the study of

ion photodissociation (64,65). With the development of mass-analyzed

ion kinetic energy spectrometry (MIKES) (66) and an instrument specifi-

cally designed for MIKES (67), the ability to separate and study ions

mass spectrometrically was dramatically extended. A MIKES instrument is

a double-focusing mass spectrometer having the magnetic sector prior to

the electric sector (reversed geometry). In MIKES, an ion is mass-

selected by the magnetic sector and allowed to undergo metastable (uni-

molecular) or collisionally activated dissociation in the following

field-free region to produce various daughter ions and neutral frag-

ments. The kinetic energy of the daughter ions is then analyzed with

the electric sector. The resulting MIKE spectrum provides information








about both the daughter ions' mass-to-charge ratios (m/z) and the

kinetic energy released upon fragmentation (68). Such information

enables the elucidation of many of the fragmentation pathways, and thus

the structure, of a molecule following its ionization and fragmentation

under electron impact.

As MIKES can involve the separation of ions resulting from the

ionization and fragmentation of a single compound, the next logical

progression was the exploitation of this MS/MS capability of MIKES for

mixture analysis. Subsequent research in the labs of Cooks (69,70) and

McLafferty (71) demonstrated that the inherent selectivity and sensi-

tivity of MIKES-type tandem mass spectrometry (MS/MS) permitted the

rapid analysis of complex mixtures for trace components with minimal, if

any, sample clean-up and preparation. The development in the late 1970s

by Yost and Enke (72) of a triple quadrupole tandem mass spectrometer,

with an RF-only quadrupole serving as an efficient collision cell and

powerful focusing device, led to the commercial development of practical

MS/MS instruments. Since then, many types of tandem mass spectrometers

have been built utilizing magnetic sectors, electric sectors, quadru-

poles, time-of-flight, ion cyclotron resonance cells and various com-

binations of these, each with their own specific characteristics (73).

As in the early days of mass spectrometry, it was soon realized that

more than just tandem mass spectrometry may often be required for com-

plex mixture analysis. Thus, the combination of various chromatographic

systems with MS/MS has further increased the selectivity of the tech-

nique (2,3,74,75).








Instrument Description, Operation, and Application

The triple quadrupole tandem mass spectrometer used in these

studies consists of, in series, a dual chemical ionization/electron

impact (CI/EI) ionization source, a quadrupole mass filter (Q1), a

radio-frequency-only quadrupole (Q2), a second quadrupole mass filter

(Q3), and an electron multiplier. While Q1 and Q3 are operated as mass

filters in the MS/MS modes, Q2 acts as a collision chamber and focusing

device, allowing all ions to be efficiently transmitted. In mixture

analysis, "soft" ionization, e.g. chemical ionization, of a mixture is

utilized in the ion source to produce ions characteristic of the compo-

nents of the mixture. The separation and analysis of the component of

interest is performed by the mass selection of its characteristic ion by

Q1 for fragmentation in Q2 through collisions with neutral gas mole-

cules, and the mass analysis of the resulting daughter ions by Q3.

The four most common MS/MS scan or operating modes of the triple

quadrupole instrument are daughter scan, parent scan, neutral loss scan,

and selected reaction monitoring (Figure 1-2). The specific MS/MS

operational mode chosen for a particular analysis will depend upon the

information desired. A daughter scan (Figure 1-2a) consists of se-

lecting a single parent ion characteristic of the analyte by Q1, frag-

menting it by collisionally activated dissociation (CAD) in Q2, and then

scanning Q3 to obtain a daughter mass spectrum. Analogous to a normal

mass spectrum, the daughter mass spectrum can be used for identification

of an analyte by standard mass spectral interpretation or by matching

the spectrum to that of an authentic sample. In the parent scan mode

(Figure 1-2b), Q1 is scanned over a specific mass range, allowing parent

ions of different m/z to sequentially enter and undergo fragmentation in









Q2 to produce various daughter ions. Then Q3, instead of scanning,

selects only an ion of a particular m/z to be transmitted to the

detector. The resulting parent mass spectrum contains all the ions

which fragment to yield a specific daughter ion, and can be used to

screen for a class of compounds which fragment to yield a common sub-

structure. Fragmentation of the positive CI (M+H)+ ions of most of the

phthalates yields the characteristic daughter ion 149+. Therefore, a

parent scan of 149 would be a method for screening mixtures for phthal-

ates. In a neutral loss scan (Figure 1-2c), both Q1 and Q3 are scanned

with a specific mass difference between them. The resulting neutral

loss spectrum contains the daughter ions which result from the loss of a

specific neutral fragment from the parent ions, and is useful for

screening for a class of compounds characterized by the loss of a spe-

cific fragment. The molecular ions of chlorinated organic often lose

Cl or HC1 during CAD, and thus a neutral loss scan of 35 or 36 (and/or

37 and 38) would provide a rapid screening procedure for chlorinated

organic in a mixture. Although the three operational modes just

described are very selective, as in full scan normal mass spectrometry,

they may not have the sensitivity necessary for the determination of

trace components. Thus, for trace analysis, selected reaction moni-

toring (SRM) (Figure i-2d) is normally employed, wherein only one

characteristic daughter ion, typically the most abundant, resulting from

the fragmentation of the analyte's characteristic parent ion, is

selected by Q3. Thus, an enhancement in the sensitivity is obtained,

albeit at the expense of some selectivity. In addition to these MS/MS

modes, the tandem mass spectrometer can also be operated as a normal

mass spectrometer by allowing all ions to pass through Q1 and Q2, in the







DAUGHTER SPECTRUM


02


- P---T S--E R-
-b-~

,I,


b) PARENT SPECTRUM


- -- --I
4'-


Q3 FULL SPECTRUM


-I -,- 0-----

I8~


NEUTRAL LOSS SPECTRUM


QI Q2 Q3







figure 1-2. MS/MS operational modes.


.------*--------*

__________ *_ -- p_
-~ 0 - 0 -t-- -


S I M


02


03


-Ii


03


SRM


02


03








absence of a collision gas. The second mass analyzer, Q3, can then be

scanned to produce a normal mass spectrum (Figure 1-2e) or it can select

only ions of a specific m/z for selected ion monitoring (SIM) (Figure 1-

2f). This allows direct comparisons to be made between MS and MS/MS

techniques on the same MS/MS instrument.

In light of the difficulties associated with the determination of

tryptolines in tissue extracts, a study was initiated to assess the

applicability of triple quadrupole tandem mass spectrometry for deter-

mining the trace levels of tryptolines reported in rat brain extracts.

With its inherent selectivity and excellent sensitivity, MS/MS should be

able to more reliably determine these compounds than the more conven-

tional techniques presently in use.














CHAPTER 2
TANDEM MASS SPECTROMETRY FOR TRACE ANALYSIS


Introduction

Tandem mass spectrometry (MS/MS) has gained rapid acceptance with

the analytical community since its development in the 1970's. Although

it has been applied successfully for structure elucidation of unknowns

(76), its rapid acceptance has largely been due to its ability to

rapidly provide sensitive and selective analysis of complex mixtures,

often with minimal, if any, sample clean-up (69-72,77,78). A recent

book (79) and several recent reviews (73,80,81) contain extended expla-

nations of the theory, instrumentation, and applications of tandem mass

spectrometry. Here, we will deal with the application of tandem mass

spectrometric techniques to the determination of trace organic compo-

nents in complex matrices. The fundamentals and difficulties associated

with trace organic analyses and the basic MS/MS operating modes will be

reviewed. Various examples taken from research in our labs, as well as

from the general literature, will be used to illustrate how tandem mass

spectrometry meets the needs and overcomes some of these difficulties

associated with trace analyses.




Requirements for Trace Analysis

In order to perform trace analyses successfully, it is necessary to

think in terms of what McLafferty has referred to as the 4 S's of analy-

sis (82): sensitivity, selectivity, speed, and $. The figure of merit









often used to describe an analytical technique is the first "S", sensi-

tivity, defined by the slope of a calibration curve as the change in the

signal obtained from a change in amount of analyte (Equation 2-1).

Sensitivity may be a useful figure of merit for "pure" analytes, but may

become meaningless for the determination of the analyte in a complex, or

even a simple, matrix. This is due to the possibility that




d Signal
(2-1) Sensitivity = slope of the calibration curve = gna
d Amount


other chemical constituents of the matrix or background may give a

response at or interfere with the signal of the analyte. These types of

effects can be referred to as chemical noise. The factor which then

determines the smallest amount of analyte which can be determined by a

technique may not be its sensitivity but rather its ability to discern

the signal of the analyte from the chemical noise. This is the second

"S", selectivity. Thus, a more descriptive figure of merit for an

analytical technique is the limit of detection, defined as the amount of

signal which results in a signal to noise ratio adequate to provide the

desired confidence (typically S/N = 3). With the rearrangement of

equation 2-1, and substitution of the resulting definition of the signal

into the definition of the limit of detection, equation 2-2 is

obtained.. This equation shows that the limit of detection takes into

account both the sensitivity and the selectivity (measured by the level

of chemical noise) of an analytical technique.



noise
(2-2) Limit of detection = (amount for S/N =3) = 3 no
sensitivity









In order to achieve the limits of detection necessary for trace

analyses, the selectivity is normally improved through the use of exten-

sive clean-up, separation, and often derivatization procedures in order

to physically separate or enhance, respectively, the analyte's signal

with respect to the chemical noise. In trace analyses, such sample

manipulations can increase the possibilities for sample contamination

and sample loss through adsorption onto glassware, oxidation, etc. In

addition, if the methods necessary to increase the selectivity become

too time-consuming and/or expensive, then the analytical method may

become too impractical for routine work. Thus, the final two "S's",

speed of analysis and cost effectiveness, are also important figures of

merits.




Instrumental Methods to Improve the Limit of Detection

An alternative method to increase the selectivity is through the

use of two or more types of analytical techniques in conjunction. Cooks

and Busch have shown that, as the number of analytical techniques used

simultaneously for the analysis of a sample increases, the absolute

levels of the signal and noise decrease (83). However, because of the

selectivity that each of the techniques has for the signal over the

noise, the noise level decreases much more rapidly than does the signal,

and an overall improvement in the S/N ratios is obtained. Since the

limit of detection is determined by the S/N ratio, as long as there is a

detectable signal, an increase in the number of analytical techniques

used simultaneously for an analysis will result in improved limits of

detection. Two commonly used analytical methods which utilize this

principle are fluorometry (84) and gas chromatography/mass spectrometry

(GC/MS) (61).








Fluorometry

Fluorometry can be considered to be a combination of absorption and

emission spectroscopy. In the absorption experiment, a specific wave-

length of light is selected by the first monochromator for irradiation

of the sample. Absorption of this radiation by the sample results in

its excitation and possible emission of light. The second monochromator

then selects a specific wavelength of the emitted light for detection as

the analytical signal. As the quantum efficiency for the conversion of

the absorbed energy into emitted light is not one, and only a specific

wavelength is selected for detection, there is a loss in the absolute

signal detected, and therefore, in the sensitivity. However, in order

for a compound to be detected, it must not only absorb energy at a

specific excitation wavelength, but it must also emit radiation at a

specific wavelength. This increase in selectivity reduces the spectral

interference or chemical noise relative to the signal, so that often

lower limits of detection are possible.




Gas Chromatography/Mass Spectrometry (GC/MS)

In GC/MS the selectivity is improved by the actual physical separa-

tion of the components of a mixture by chromatography prior to their

mass analysis. In order to be detected an analyte must elute from the

chromatographic column at a specific retention time and be ionized to

produce ions of specific m/z. The sensitivity is reduced due largely to

the dilution of the analyte during the chromatographic separation. In

addition to reduced sensitivity, another major trade-off for the

increased selectivity is an increased analysis time.









Tandem Mass Spectrometry

Instead of using two different methods of analysis, tandem mass

spectrometry, as its name implies, uses one technique, mass spectro-

metry, twice in tandem. A tandem mass spectrometer consists of an ion

source, two mass analyzers separated by a fragmentation region, and an

ion detection device. The mass analyzers which have been used in these

instruments include quadrupoles, magnetic sectors, electric sectors,

time of flight, ion cyclotron resonance cells, and combinations of these

(73 and references therein). Although each of these has its own spe-

cific characteristics, they are based upon the same MS/MS principles.



Principles of MS/MS

The principles of MS/MS are straightforward, and can be compared to

conventional GC/MS as described above. A mixture is introduced into the

ion source of the tandem mass spectrometer, where "soft" ionization

methods can be used to produce ions characteristic of the mixture com-

ponents. The separation of the analyte from the other mixture compo-

nents (the chromatographic step of GC/MS) is then achieved by the mass

selection of the characteristic ion of the analyte by the first mass

analyzer. The parent ion, thus selected, undergoes collisionally acti-

vated dissociation (CAD) through collisions with neutral gas molecules

in the fragmentation region to yield various fragment or daughter ions,

analogous to the fragmentation occurring during the ionization step of

GC/MS. As in GC/MS, subsequent mass analysis of the daughter ions by

the second mass analyzer results in the analytical signal. When the

second mass analyzer is scanned, a daughter spectrum is obtained. As

with a normal mass spectrum, the daughter mass spectrum can be used for









identification of the parent ion (and thus, the analyte) through conven-

tional mass spectral interpretation or by comparison with an authentic

sample. In order to increase the sensitivity for trace analysis, only a

single characteristic daughter ion, usually the most abundant, may be

selected by the second mass analyzer for monitoring. This selected

reaction monitoring (SRM) is analogous to the selected ion monitoring

(SIM) used for maximal sensitivity in conventional GC/MS. Thus, in

order for an analyte to be detected, it must be ionized to a character-

istic ion and this parent ion must produce a daughter ion of specific

m/z. This again results in decreased sensitivity with respect to normal

MS due to the inefficiencies of the conversions of parent ions to

daughter ions and the subsequent mass analysis of these daughter ions.

However, the significant reduction in chemical noise often results in

increased S/N ratios and improved limits of detection.



Applications of MS/MS for Trace Analysis

MS/MS trace analysis with MIKES instruments. The increase in

selectivity which results from the use of two mass analyzers in tandem,

in conjunction with the excellent sensitivity of the electron multiplier

for ion detection, has enabled the direct analysis of complex mixtures

for trace components with little or no sample preparation. From a

historical viewpoint, the early MS/MS applications were conducted with

MIKES instruments (see Chapter 1) modified to allow pressurization of

the second field-free region (between the magnetic and electric

sectors). One of the first applications of MS/MS for trace analysis of

mixtures was the determination of cocaine in coca leaf samples and urine

(80,86). Solids probe isobutane PCI-SRM analysis of cocaine standards










resulted in a calibration curve extending over 2 or 3 orders of magni-

tude with a limit of detection of approximately 1 ng where the error was

estimated as 30%. This technique permitted the determination of 4 ng of

cocaine in a 1 ug sample of coca leaf diluted in 1 mg of chalk dust, and

of 1.7 ng of cocaine in a 1 1L urine sample. This work was extended

with multiple reaction monitoring (MRM) to the simultaneous mapping of

cocaine and cinnamonylcocaine in 1 mg samples of coca plant tissue. The

only sample preparation used in these samples was grinding of the coca

leaf samples in liquid nitrogen.

The speed of analysis with MS/MS techniques can be increased by

minimizing the sample preparation, as above, and by eliminating any

chromatographic separation of the mixture components prior to mass

spectrometric analysis. Instead, mixtures can be introduced directly

into the ion source by heatable solids probe. The component separation

then occurs by mass selection of ions characteristic of individual

components for fragmentation and subsequent mass analysis of the

daughter ions. The rapidity of sample analysis possible with a solids

probe is illustrated by the PCI-SRM determination of 20 ng of urea in 1

pL samples of diluted blood serum at a rate of 15 samples per hour

(77). Due to the increased speed of analysis, a more reliable deter-

mination of the amount of analyte and an estimation of the precision of

the analysis can be obtained by performing replicate analyses of the

sample. Thus, the precision of the peak heights in the solids probe

PCI-SRM of urea in blood serum was 15 % relative standard deviation.

MS/MS trace analysis with triple quadrupole instruments. The

research with MIKES instruments demonstrated that several trace compo-

nents could be rapidly determined in complex mixtures with little or no









sample preparation. However, MIKES instruments have several disadvan-

tages. Besides having less than unit mass resolution in the daughter

spectra, the scan laws for the parent and neutral loss scans are compli-

cated and the magnetic sector can not be quickly and accurately "jumped"

between parent ions. With the development of a triple quadrupole

instrument, these disadvantages were overcome (72). The quadrupoles,

having a linear scan function and lacking the hysteresis effects of a

magnet, can be quickly and accurately jumped between many different

parent ions. These same characteristics allows all the various MS/MS

scan modes to be placed under computer control. With a center quadru-

pole as a collision cell and focusing device, very efficient CAD of

parent ions and collection of daughter ions for mass analysis by the

third quadrupole are realized.

The application of triple quadrupole MS/MS to trace mixture analy-

sis is illustrated by the direct determination of illicit drugs in the

urine and blood serum of racing animals by solids probe PCI-MS/MS tech-

niques (78). Presently, screening for illicit drugs is performed by

thin layer chromatography, with confirmation performed on any positives

by GC/MS. These methods entail extensive sample workup prior to their

analysis and therefore only the top three to four animals of each race

typically are tested. With the introduction of 1 yL of blood serum via

a heated solids probe, the MS/MS detection limits for most of the il-

licit drugs studied were in the low parts-per-million (ppm) (ng/UL)

range with PCI-SRM. With a simple solvent extraction of the blood

serum, the detection limits were reduced to the low part-per-billion

(pg/UL) range. The selectivity of MS/MS is more dramatically illus-

trated by the fact that three isobaric (same nominal mass) drugs could









be independently quantitated due to their unique daughter ions. Con-

firmation of the drugs at the parts-per-million (ng/uL) level in the

blood serum was possible by comparison of the complete daughter spectra

from the simple extract to those of authentic standards. With this

procedure, it was possible to screen for as many as 50 drugs and metabo-

lites in a single sample in less than 5 minutes. The advantages gained

by the simplicity and time-saving of the MS/MS procedure over that

currently in use are quite apparent.

The selectivity, sensitivity, and speed of analysis possible with

MS/MS is dramatically illustrated in the determination of hexachloro-

benzene (HCB) and 2,4,5-trichlorophenol (TCP) in human blood serum and

urine by GC/triple quadrupole MS/MS (85). Rapid sample introduction was

possible using a 50 cm long, 0.75 mm i.d., packed GC column operated

isothermally to give retention times of 10 and 20 s for HCB and TCP,

respectively. With this rapid means of sample introduction and a simple

1:1 solvent:sample extraction, it was possible to perform triplicate

determinations of TCP (spiked levels ranging from 0.25 to 100 ppb, 1 PL

sample size) in six serum samples, six urine samples, six standards, and

associated blanks in approximately 36 minutes, which corresponds to ca.

100 injections/hr. This speed of analysis did not compromise the sen-

sitivity and selectivity, as the absolute limits of detection for HCB

and TCP were 50 and 250 femtograms, respectively, injected onto the

column. The rapid analysis also made it possible to perform replicate

analyses of each sample, which permitted an estimation of the precision

of quantitation (consistently 10 percent relative standard

deviation). Often with normal GC/MS, the length of time required for

the chromatographic step makes such replicate analyses impractical. A









comparison with the analyses performed with capillary column GC/MS

showed the limits of detection for HCB and TCP with the short packed

column GC/MS/MS to be 4 and 80 times lower, respectively, than those

with GC/MS. In addition, GC/MS/MS was able to perform the same set of

analyses in approximately 1/6 the time of that of the capillary GC/MS

method. This was largely attributed to the time of the actual sample

analyses, although significant savings were also apparent in the sample

and instrument preparation.




Conclusion

The MS/MS examples above demonstrate that MS/MS meets the require-

ments necessary for trace analysis: sensitivity, selectivity, speed,

and low cost per sample. Due to the high sensitivity and increased

selectivity of MS/MS, rapid determinations of picogram and femtogram

quantities of analytes have been demonstrated in small (mg and yL)

quantities of complex mixtures with only minimal, if any, sample clean-

up. The reduced sample clean-up not only increases the speed of analy-

sis but also reduces the possibility of contamination of the sample or

loss of the analyte. In addition, the ability to analyze small quan-

tities becomes a major advantage in biological investigations where the

amount of sample is often limited. Although the initial cost of MS/MS

instruments is in the $105 range, because of the rapidity of sample

analysis, MS/MS becomes a very cost-effective technique for performing

trace analyses.














CHAPTER 3
TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND
QUANTITATION OF UNDERIVATIZED TRYPTOLINES




Introduction

In Chapter 2, the ability of MS/MS to successfully analyze complex

mixtures directly, with minimal or no sample preparation, was illus-

trated. With a reduction in the sample clean-up prior to analysis, the

possibilities of both loss of the trace analytes and contamination of

the sample can be reduced. In particular, the chances of contamination

of the brain extracts with aldehydes, which may lead to artefactual

formation of tryptolines, can be reduced. Thus, this chapter investi-

gates the use of tandem mass spectrometry for the identification and

quantitation of underivatized tryptolines in crude extracts of rat

brains.




Artefactual Tryptoline Formation

The Pictet-Spengler condensation reaction of indoleamines and

aldehydes to produce the corresponding tryptolines occurs readily in the

laboratory under conditions of physiological pH and temperature (1,5).

This fact has led to the possibility of artefactual formation of trypto-

lines occurring during the extensive sample clean-up procedures neces-

sary prior to determination of tryptolines by TLC-fluorescence and/or

GC/MS. It was demonstrated that a major portion of the tryptolines

determined in several instances has largely been due to artefactual









formation, with the source of the problem being traced to the presence

of formaldehyde in the solvents used in the sample clean-up procedures

(8,58). This has necessitated the use of aldehyde-trapping reagents in

the solvents and redistillation of solvents just prior to sample clean-

up. These steps have helped to reduce the level of the problem. In

addition, deuterium-labelled indoleamines are now routinely added in the

first step of sample clean-up as internal checks upon the level of

artefactual formation. The formation of deuterium-labelled tryptolines

would be a measure of the amount of artefactually-formed tryptolines

relative to endogeneous levels. Faull and others were able to demon-

strate that, by reducing the length and complexity of the clean-up

procedure, the possibility of artefactual tryptoline formation during

the sample clean-up could be minimized, if not eliminated (59).

However, in order to obtain high selectivity and sensitivity, it was

still necessary to derivatize the sample and perform gas chromatographic

separation prior to mass spectral analysis.




MS/MS for Mixture Analysis

The use of two mass analyzers in tandem gives MS/MS a high degree

of selectivity. Because of this, it has been possible to successfully

determine targeted compounds or classes of compounds in very complex

matrices by MS/MS techniques with minimal, if any, sample clean-up.

Examples include the mapping of cocaine and cinnamonoylcocaine in coca

leaves (86), the determination of urea in human blood serum (77), and

the determination of illicit drugs in racing animals' serum (78). Not

only is there a time-saving by having minimal sample clean-up, but

further time-saving results from the ability to do mass separation of









components (in a few ms) instead of chromatographic separation (in tens

of min) prior to further mass analysis.

The use of highly selective MS/MS techniques should enable further

reduction in the sample clean-up of brain homogenates. This in turn

should reduce the possibility of contamination with, loss of, and arte-

factual formation of tryptolines during the sample clean-up. In light

of this, tandem mass spectrometry was assessed with regard to its

ability to directly determine trace levels of underivatized tryptolines

without prior chromatographic separation.




MS/MS for Structure Elucidation

In addition to mixture analysis, MS/MS has been used for structure

elucidation studies of organic molecules and ions (72,76,87,88). In

normal El mass spectrometry, many fragment ions are often produced

following the ionization of a compound. Without tandem mass spectro-

metry, the fragmentation pathways (and thus, the structure of the com-

pound) resulting in these fragment ions can be elucidated with the use

of high resolution mass spectrometry for determination of the elemental

compositions of the ions, isotopic labelling studies, and/or with refer-

ence to compilations of logical fragmentation mechanisms which have been

elucidated by similar methods (89,90). With the use of tandem mass

spectrometric techniques, however, the fragmentation pathways of an

ionized compound can be directly determined by systematically frag-

menting and obtaining daughter spectra of all of the ions in the El mass

spectrum of the compound (72). This results in obtaining a "genetic

tree" of an ionized molecule, indicating all the interrelationships and

fragmentation pathways between various daughter ions. Interpretation of










the mass spectral information in order to elucidate the structure of a

compound can then be performed more quickly and reliably.

An additional advantage with MS/MS is realized with the "soft"

ionization techniques generally used in the trace analysis by mass

spectrometry. These "soft" ionization techniques (e.g. chemical ioniza-

tion) increase the selectivity and often the sensitivity of mass spec-

trometric analyses by reducing the fragmentation of the characteristic

ions of the mixture components. However, with normal mass spectrometry,

this results in a loss of structural information about the ions pro-

duced. MS/MS analysis of the ions produced by CI of a mixture is often

able to provide the structural information necessary to identify a

component without the sample separation necessary in conventional mass

spectrometry. Therefore, a second objective of the work presented in

this chapter is to investigate the ability of MS/MS to structurally

characterize the tryptolines in relatively "pure" forms with EI-CAD, and

in mixtures with CI-CAD techniques.




Experimental




Materials and Reagents

All chemicals and reagents were of the highest purity available.

The tryptoline standards were kindly supplied as their HC1 salts by Kym

Faull, Ph.D., and Jack Barchas, M.D. (Department of Psychiatry and

Behavioral Sciences, Stanford Medical Center, Stanford, CA). Ultrahigh

purity methane (Matheson, Morrow, GA) and zero grade nitrogen (Airco

Industrial Gases, Researach Triangle Park, NC) were used as a CI

reagent/GC carrier and collision gas, respectively.









Instrumentation

All data were collected with a Finnigan MAT (San Jose, CA) triple

stage quadrupole GC/MS/MS (91) equipped with a 4500 series ion source,

pulsed positive and negative chemical ionization and INCOS data

system. The Finnigan 9610 gas chromatograph was equipped with packed

and Grob-type capillary injectors. A packed GC column and a heated

direct insertion solids probe were utilized for sample introduction in

these studies.

A short packed glass column was used to introduce samples into the

ion source during the optimization studies. The column was constructed

of a 38 cm length of 6 mm o.d. and 0.75 mm i.d. U-shaped glass tubing.

A 6.4 cm piece of 0.64 cm o.d. and 4 mm i.d. glass tubing was joined to

each end of the 38 cm length to allow room for the injection needle and

for connection to the GC/MS interface. The tubing was rinsed three

times each with the following solvents, in order: distilled water,

acetone, methanol, and methylene chloride. Upon drying, the tubing was

filled with a 10% solution of dimethylchlorosilane in toluene for ca.

1.5 hours, after which it was rinsed three times with toluene and dried

in a GC oven. The tubing was hand-packed with 3% OV-101 on 80/100 mesh

Chromosorb 750. The resulting column was conditioned overnight at 250

"C with 20 ml/min He flow prior to use (note: during conditioning the

column was not connected to the mass spectrometer). The GC/MS interface

consisted of a glass-lined stainless steel tube direct inlet fitted with

a micro-needle valve. During the optimization studies, the GC column

was kept isothermal at 225 C with a carrier gas flow rate of ca. 18

ml/min CH4. The injection port and the GC/MS interface temperatures

were 220 C and 250 "C, respectively. These conditions resulted in a

retention time for tryptoline-HC1 of ca. 10 s.









Procedures

Mass spectra of standards. Standards were introduced into the ion

source by vaporization from a solids probe heated from ca. 50 C to 400

C at varying rates under data system control. Electron impact (El, 70

eV electron energy) and positive and negative chemical ionization (PCI

and NCI, respectively, 100 eV electron energy, 1.0 torr CH4 source

pressure) mass spectra were obtained with a source temperature of 100 C

in the Q3 normal MS mode. Daughter spectra were acquired for the char-

acteristic ions in the El and CI mass spectra of each of the tryptolines

at collision gas pressures of 2.0 and 2.9 mtorr N2, respectively, and at

a collision energy of 24 eV. The NCI-CAD daughter spectra of the

(MHC1-H) (M denotes the tryptoline, while MHC1 denotes the tryptoline-

HC1 salt) ions were obtained at a collision gas pressure of 2.0 mtorr N2

and a collision energy of 26 eV. The El-CAD daughter spectra used to

generate the "genetic tree" of each tryptoline were obtained at a source

temperature of 130 C with a collision gas pressure of 1.3 mtorr N2 and

a collision energy of 20 eV. The mass spectra acquired during the

highest level of the analyte's ion current were averaged and background-

subtracted, if necessary, to yield a representative mass spectrum of

each standard.

Selection of positive or negative chemical ionization. Samples of

tryptoline-HCl were introduced via the heated solids probe while per-

forming Q3 selected ion monitoring (SIM) of the (M+H) ion, m/z 173, and

the (M-H) ion, m/z 171, with the pulsed positive and negative chemical

ionization feature of the tandem mass spectrometer. The integrated ion

currents of the two ions were compared.









PCI-CAD collision energy and collision gas pressure studies. For

each CAD parameter, studies were conducted with two different techniques

at an ion source temperature of 120 "C. In the first study, tryptoline-

HC1 was introduced via the heated solids probe and PCI-CAD daughter

spectra from m/z 140 to m/z 176 were acquired of the (M+H)+ ion, m/z

173, at a collision gas pressure of 1.2 mtorr N2 and at varying colli-

sion energies. This procedure was repeated for a second sample. The

ion intensity ratios of m/z 144 to m/z 173 obtained at each collision

energy were averaged and plotted versus the collision energy (Q2

offset). A similar procedure was utilized for the collision gas pressure

study with the exception of having the collision energy set at 26 eV and

analyzing a single sample for each collision gas pressure.

In a second study, for each combination of collision energy and

collision gas pressure, duplicate or triplicate 1.0 yL injections of a

standard solution of tryptoline-HCl were made onto the packed GC column

(225 C isothermal). PCI-selected reaction monitoring (SRM) of m/z 173

to m/z 144 was performed, and the areas of the resulting GC peaks were

plotted against the desired parameter. The collision energy study was

conducted at 1.2 mtorr N2. The collision gas pressure study was con-

ducted at 24 eV.

Quantitative studies. Serial dilutions were prepared of trypto-

line-HCl and methtryptoline-HC1 to give a series of solutions ranging in

concentration from the parts-per-trillion (pptr) to the parts-per-

million (ppm) level. Triplicate 1.0 uL samples of each solution were

placed in separate 5 pL glass vials, with care to ensure that no large

air bubbles were present. These were allowed to air dry for ca. 1 hr.

The samples were then introduced into the ion source via a solids probe,










whereupon vaporization of the sample occurred by ballistically heating

from 50 C to 400 C in ca. 1 min. PCI-SRM (100 eV, 0.3 mA, 1.0 torr

CH4 source pressure, 140 C ion source-26 eV, 2.0 mtorr N2, 1 nominal

mass unit, u, wide scan at 10 Hz) of the 173+ to 144+ and 187+ to 144+

CAD reactions was performed for tryptoline-HCl and methtryptoline-HC1,

respectively, at an electron multiplier voltage of 2200 V and a preamp

sensitivity of 10-8 A/V. Quantitation was obtained by integrating the

ion current over the scans during which tryptoline-HC1 and methtryp-

toline-HC1 were vaporized from the solids probe. Peak areas are reported

in data system counts. It was estimated that 1 count corresponds to the

detection of one ion.




Results and Discussion



Nomenclature and Structure

The compounds of interest in these studies are tryptoline (TLN),

methtryptoline (MTLN), 5-methoxytryptoline (CH30-TLN), 5-hydroxytrypto-

line (HTLN), and 5-hydroxymethtryptoline (HMTLN) (Table 3-1). These

compounds are substituted 1,2,3,4-tetrahydro-B-carbolines. Holman et.

al. (30) have suggested a change in the numbering system from that of

the 8-carbolines (parenthesized numbers) to one reflecting the numbering

system of the presumed precursor indoleamines (unparenthesized

numbers). They have also suggested the class of compounds be referred

to as tryptolines, which again reflects the presumed indoleamine pre-

cursor tryptamine. Thus, 5-hydroxytryptoline would be the reaction

product resulting from the condensation of 5-hydroxytryptamine with

formaldehye. The tryptoline numbering system and nomenclature will be









used throughout the remainder of the text. When referring to specific

tryptolines, the abbreviations will be used while the term tryptolines

will be applied to the entire class of compounds.




El and EI-CAD Mass Spectra of Standards

The normal El mass spectra of the tryptolines, M, (as their HC1

salts, MHC1) are characterized by relatively intense molecular ions and

numerous fragment ions (Table 3-2). The (M-29)+ fragment ions, pre-

sumably arising from the loss of CH2=NH from the piperidine ring, are

the most abundant ions in the mass spectra of the tryptolines which lack

a substituent at the 9-position (R"=H). For the tryptolines having a 9-

methyl substituent (R"-CH3), i.e. MTLN and 5-HMTLN, cleavage of the

methyl radical from the molecular ion results in the (M-15) fragment

ions being the most abundant, with the (M-29) fragment ions being the

next most abundant ions. The loss of the methyl radical, presumably

from the methoxy group of the (M-29)+ fragment ion of 5-CH30-TLN,

results in the second most abundant fragment ion in its mass spectrum.

The fragmentation of the molecular ions of the tryptolines under

CAD conditions resulted in good yields of several abundant daughter

ions, with the intensities of the parent molecular ions ranging from 5

to 16% of the most abundant daughter ion (Table 3-3). The most abundant

daughter ions resulted from the same processes as seen in the normal El

mass spectra, and serve to support the presumed fragmentation pathways,

i.e. the (M-15) and (M-29) daughter ions from the tryptolines with and

without the 9-methyl substituent, respectively. The (M-29) daughter

ions were also relatively abundant in the daughter spectra of MTLN and

5-HMTLN. A relatively abundant daughter ion at m/z 158, presumably due









Table 3-1. Characteristics of the tryptolines of interest.


Tryptoline

tryptoline

methtryptoline

5-methoxytryptoline

5-hydroxytryptoline

5-hydroxymethtryptoline


Abbr.

TLN

MTLN

CH30-TLN

HTLN

HMTLN


R'

H

H

CH30

HO

HO


R"

H

CH3

H

H

CH3


Molecular Weight

M MHC1

172 208

186 222

202 238

188 224

202 238


5 4 3
5
(5) (4)



(6 N
(7) ( H
(8) (9) (1)
7 8 9R









Table 3-2. El
salts.





Ionsa

M+

(M-CH3) +

(M-R "NH) +

(M-H20)

(M-R"NH-H2)+

(M-CH2=NH)+

[M-(CH2=NH)-CH 3

[M-(CH2=NH)-HCN]+


mass spectral characteristics of the tryptoline-HC1


TLN

172(42b)

nd

156(<2)




154(3)

143(100)


MTLN

186(60)

171(100)

156(35)




154(20)

157(35)


CH3O-TLN

202(70)

nd

186(4)




184(3)

173(100)

158(78)

-


HTLN

188(32)

nd

172(<2)

170(2)

170(2)

159(100)


HMTLN

202(62)

187(100)

172(39)

184(3)

170(16)

173(46)




146(15)


The mechanisms shown for the formation of the ions is supported by the
work of Coutts et. al. (92) and the EI-CAD studies here. R" is the
substituent at the 9-position, as defined in Table 3-1.

percent abundance relative to the most abundant ion.









Table 3-3. EI-CAD (24 eV, 2.0 mtorr N2) daughter mass spectral charac-
teristics of the M ions of the tryptoline-HCl salts.


Ions a




(P-H)+

(P-CH3)+

(P-CH2=NH)+

(P-CH3-HCN)+

EP-(CH2=NH)-CH31 +


TLN

172(5b)

171(2)

nd

143(100)


MTLN

186(16)

185(24)

171(100)

157(37)

144(1)


CH O-TLN

202(10)

201(2)

187(1)

173(100)




158(16)


HTLN

188(6)

187(2)

nd

159(100)


HMTLN

202(15)

201(25)

187(100)

173(44)

160(6)


aThe mechanisms for daughter ion formations is supported by the work of
Coutts et. al. (92). P is the parent ion, i.e. M

Percent abundance relative to the most abundant ion.









to [(M-CH2-NH)-CH3]+ also resulted from the CAD of the M+ ion of 5-CH30-

TLN.

The characteristic loss of CH2-NH from the El molecular ions of the

tryptolines under CAD conditions could be used in the neutral loss MS/MS

mode to screen for other possible compounds having a piperidine ring.

Complete characterization of such an unknown, once isolated, could be

accomplished by systematically obtaining daughter spectra of all the

ions in its normal El mass spectrum. The information thus obtained

would give all the genetic relationships between all the substructures

of the molecule, i.e. from what ions a substructure is produced and to

what ions a substructure fragments, and would allow for easier and more

reliable interpretation of the mass spectral fragmentation pathways.

Such information is illustrated for MTLN in Figure 3-1. Coutts et al.

(92) have previously determined some of the fragmentation pathways of

the 9-alkyl substituted tryptolines using metastable ions and high

resolution mass spectrometry. However, as relatively few metastable

transitions are observed for these compounds, only a few of the fragmen-

tation pathways could be directly confirmed (indicated by asterisks in

Figure 3-1). However, with an EI-CAD generated "genetic tree" of a

molecule, direct confirmation and analysis were possible of all the

fragmentation pathways. The completeness of this information should

lend itself well to computerized structural analysis. Research is

proceeding towards this goal (93).

The numerousness of fragment ions generated under El conditions

makes this the preferred technique for structure elucidation of unknown

compounds by MS/MS. However, this characteristic becomes a liability in

quantitative analyses. In the direct determination of trace components









186 --
185
172
171
169
168
167
158
157 -
156
155
154
145 -
144 *
N 143
E 142 -
130 -
129
128- --
127
118 -
117 -
116 -
115 -
103 -
101 -
91-
90-
89-
77-
54 -
51 -
I I I I r I I 1 T I
0 10 20 30 40 50 60 70 80 90 100

Percent Relative Abundance

Figure 3-1. "Genetic tree" of MTLN-HC1. Horizontal bars are proportional to percent
relative abundance. denotes pathways confirmed by MIKES.








in complex matrices without prior separation, ideally each component

should be ionized to yield a single characteristic ion. This is advan-

tageous for two reasons. Firstly, by reducing the fragmentation of a

molecule after ionization, one can achieve an increase in sensitivity

when monitoring just the characteristic molecular ion. Secondly, by

reducing the number of fragment ions from other compounds in the matrix

having higher molecular weights than the compound of interest, the

possibility of spectral interference at the m/z of interest is reduced,

with a subsequent increase in the selectivity of the technique.




PCI and PCI-CAD Mass Spectra of Standards

Chemical ionization has been shown to be one method of achieving

the above goals (94,95). With methane as a reagent gas, gas-phase

chemical reactions occur in the ion source which result in the ioniza-

tion of the sample with little transfer of energy to the sample mole-

cules. This results in ions of low internal energy, and therefore

little fragmentation of the original ions occurs.

The positive chemical ionization (PCI) mass spectra of the trypto-

line-HC1 salts are dominated by the protonated parent tryptoline mole-

cules, (M+H) and major fragment ions presumably due to loss of CH2=NH

from the (M+H)+ ions to yield (M+H-29)+ (Table 3-4). The 5-substituted

tryptolines have several additional fragment ions, which are explained

below. In addition to the fragment ions, all the tryptolines yield the
+ +
adduct ions (M+29) and (M+41) characteristic of methane PCI. The

presence of these adduct ions serves to confirm the molecular weight of

the compounds.









Table 3-4. Methane PCI mass spectral characteristics of the tryptoline-
HC1 salts.


Ions

(M+41)+

(M+29)+

(M+H)+

(M+H-15)+

(M+H-18)+

(M+H-29)+

(M+H-43)


TLN

213(2a)

201(6)

173(100)

158(1)

nd

144(50)

nd


MTLN

227(2)

215(7)

187(100)

172(4)

nd

158(11)

144(9)


CH30-TLN

243(2)

231(10)

203(100)

188(3)

nd

174(19)

nd


HTLN

229(2)

217(8)

189(100)

174(1)

171(5)

160(20)

nd


HMTLN

243(2)

231(9)

203(100)

188(3)

185(5)

174(16)

160(3)


percent abundance relative to the most abundant peak.









The (M+H)+ ions of the tryptolines fragment efficiently under the

CAD conditions used (2.0 mtorr N2, 26 eV) to yield several abundant

daughter ions (Table 3-5). For the non-methylated tryptolines, the most

abundant daughter ion, (M+H-29) is presumably due to the loss of

CH2=NH from the (M+H) ions. This serves to confirm the fragmentation

pathway of the (M+H)+ ions in the PCI spectra. Although the corre-

sponding daughter ions are seen in the daughter spectra of the (M+H)+

ions of the 9-methylated tryptolines, the most abundant daughter ions

correspond to (M+H-43) ions, presumably from the loss of CH3CH-NH from

the piperidine ring. The (M+H-43) ions could also be explained by a

rearrangement involving migration of the methyl group from the 9-

position to the piperidine nitrogen, and the subsequent loss of

CH2=NCH3. Coutts et. al. have shown some evidence for such a rear-

rangement occurring under El conditions (92). The (M+H-15) ions do

fragment to yield some (M+H-43)+ ions (ca. 40-50 % of the most abundant

(M+H-29)+ daughter ions) by loss of 28 u. Thus, a third possible expla-

nation of the formation of the (M+H-43)+ ion may be loss of the methyl

group followed by loss of H and HCN.

In addition to the formation of the (M+H-43)+ ions, the (M+H) ions

of the methylated tryptolines also yield relatively intense (M+H-17)+

daughter ions. These could result from the loss of the CH3 group,

followed by loss of H2. However, the daughter spectra of the (M+H-15)+

fragment ion has little if any ion corresponding to a loss of H2.

Perhaps a more reasonable explanation is a migration of a hydrogen from

the methyl group to the presumably protonated piperidine NH and subse-

quent loss of NH3. High resolution mass spectrometry and isotopic-

labelling would help to confirm the two possible mechanisms.









Table 3-5. Methane PCI-CAD (24 eV, 2.9 mtorr N2) daughter mass spectral
characteristics of the (M+H)+ ions of the tryptoline-HC1 salts.


Ionsa TLN

P+ 173(2b)

(P-15)+ 158(<0.6)

(P-17)+ 156(<0.3)

(P-18)+ nd

(P-29)+ 144(100)

(P-43) nd

(P-47)1 nd

(P-32) nd

(P-33)+ nd

(P-44)+ nd

(P-61)+ nd

(P-29-R"CN)+ 117(1)

(P-29-R"CN-R'OH)+ nd

(R"CH=N=CH2)+ 42(1)


MTLN

187(1)

172(1)

170(7)

nd

158(11)

144(100)

nd

nd

nd

nd

nd

117(2)

nd

56(0.5)


CH O-TLN

203(2)

188(5)

186(0.4)

nd

174(100)

nd

nd

171(2)

nd

159(79)

142(9)

nd

nd

42(1)


HTLN

189(2)

nd

172(0.3)

171(7)

160(100)

nd

142(57)

nd

nd

nd

nd

nd

115(2)

42(1)


HMTLN

203(2)

188(2)

186(16)

185(7)

174(24)

160(100)

156(13)

nd

170(37)

nd

nd

nd

nd

56(0.5)


a +
a is the parent ion, i.e. (M+H).

percent abundance relative to the


most abundant ion.









The (M+H)+ ions of the 5-hydroxytryptolines yield characteristic

daughter ions (M+H-18) and (M+H-47) The former daughter ions are

presumably due to the protonation of the 5-hydroxy group and subsequent

loss of H20. The latter are presumably due to the combined losses of

H20 and CH2=NH from the (M+H)+ ions. This is confirmed by the formation

of the daughter ion corresponding to (M+H-47)+ from the CAD of the (M+H-

18)+ and (M+H-29)+ PCI fragment ions. In addition, the CAD of the

(M+H) ion of 5-HMTLN yields a daughter ion characteristic of both of

its substituents, i.e. (M+H-33) This could presumably occur by the

combined losses of H20 and CH3 from the (M+H) ion. This is confirmed

by the appropriate losses occurring from the (M+H-15) and (M+H-18) PCI

fragment ions under CAD to yield the daughter ion corresponding to

(M+H-33).

The CAD of the (M+H) of the 5-CH30-TLN also produces unique daugh-

ter ions corresponding to (M+H-32) and (M+H-61) These ions could

possibly result from mechanisms analogous to the formation of (M+H-47)+

ions from the 5-hydroxytryptolines' (M+H) ions. Presumably, protona-

tion of the CH30 group occurs, and its loss as methanol results in the

(M+H-32)+ ion. The subsequent loss of CH2=NH from this ion would result

in the (M+H-61) daughter ion. Little or no evidence was apparent for

loss of methanol from the (M+H-29) fragment ion, and the CAD spectrum

was not obtained of the (M+H-32) fragment ion. However, examination of

the analogous pathways in the 5-hydroxytryptolines reveals that the

major portion of the (M+H-47) ions comes from the loss of CH2=NH from

the (M+H-18)+ fragment ions as opposed to loss of H20 from the (M+H-29)+

fragment ions. This supports the above proposed mechanism.









Thus, PCI of the tryptolines results in the residing of most of the

ion current in the (M+H)+ ions and several fragment ions. The PCI-CAD

of the (M+H) ion of each tryptoline yields a unique daughter spectrum,

reflecting the substituents at the 5- and 9-positions of the parent

tryptoline structure. The daughter ions observed can be readily ex-

plained by the loss of CH2=NH, R", CH2=N-R', and R'H, and combinations

of these losses from the (M+H)+ ions.



NCI and NCI-CAD Mass Spectra of Standards

The electron-capture NCI mass spectra of the tryptoline-HC1 salts

are dominated by the ions resulting from loss of H from both the HC1

salt (MHC1) to yield (MHC1-H) and the parent tryptoline molecule (M) to

yield (M-H) ions (Table 3-6). The (M-H) ions may also arise from the

loss of HC1 from the (MHC1-H) ions. This is supported by the NCI-CAD

of the (MHC1-H) ions which fragment to yield the (M-H) as daughter

ions (Table 3-7). Also possible is the attachment of Cl to the neutral

tryptoline molecules in the gas phase to produce the ion at (M+35),

equivalent to (MHC1-H) Each of the tryptolines shows a major fragment

ion corresponding to (M-29) which is again presumably due to loss of

CH2=NH from the molecular ion and supported by the NCI-CAD of the M-

ions (Table 3-8). The 9-methyl tryptolines have a cluster of three

fragment ions, (M-15) (M-16) and (M-17) presumably corresponding

to loss of CH3 from the M-, (M-H)-, and (M-H2)-, respectively. The 5-

hydroxytryptolines have a major fragment ion (M-17) apparently due to

loss of OH from the M-. Such a loss is substantiated by the daughter

ions corresponding to a loss of OH from the (M-29) fragment ions and a

loss of CH2=NH from the (M-17)- fragment ions. The 5-methoxytryptoline,









in addition, shows a fragment ion at m/z 159 which could correspond to

the loss of CH2=NH from the m/z 188 fragment ion or by loss of CH3 from

the M followed by loss of CO.

All the tryptolines have ions in their NCI mass spectra at higher

m/z than those of their M and (MHC1)- ions. The prominent (M+12)- and

(MHC1+12)- ions have been explained in the CH4 NCI of organic nitriles

by the attachment of C2H5, present in the methane CI plasma, to the

neutral molecule followed by loss of NH3 prior to ionization (96). An

alternative, but less likely, explanation could be the formation of the

reactant ion C which could then attach to and ionize the molecule

(97). The mass analyzer was not scanned to low enough mass to see if

this ion was indeed present. The (M+28) and (MHC1+28) ions are very

prominent in the spectra of the HC1 salts of HTLN and HMTLN, but of only

low relative abundance in the other tryptolines' spectra. These ions

may also be adduct ions due to the addition either of CO to the molecule

prior to ionization or of CO to cause ionization. The CO could pos-

sibly be formed from the reaction of CH4 with 02 (from air leaks) at the

hot filament surface. Another possible source of CO could be the sample

itself, as phenols have been shown to lose CO under El conditions

(89). Subsequent loss of oxygen from the CO-adducts could also be an

explanation of the (M+12) and (MHC1+12) ions. In addition to these

possible adduct ions, the mass spectra of the tryptolines contain ions

corresponding to (M+46) and 46. These ions are of relatively low

abundance for all of the tryptolines, with the exception of TLN-HC1, for

which they represent the two most abundant ions in the NCI mass spec-

trum. The (M+46)- ions could be due to attachment of NO2 to the neutral

molecule prior to ionization, or, due to the attachment of NO2- to the









Table 3-6. Methane electron-capture NCI mass spectral characteristics
of the tryptoline-HCl salts.


Ions

(MHC1+28)

(MHC1+12)

(MHC1-H)

(M+46)

(M+28)

(M+12)

M

(M-H)

(M-H2)-

(M-17)

(M-29)

(M-43)

m/z 46

m/z 35


TLN

236(<2a)

220(4)

207(100)

218(148)

200(<2)

184(20)

172(14)

171(60)

170(25)

nd

143(23)

nd

(381)

(47)


MTLN

250(<0.5)

234(11)

221(100)

232(6)

214(1)

198(27)

186(16)

185(46)

184(25)

169(9)

157(5)

nd

(6)

(17)


CH3O-TLN

266(<0.5)

250(7)

237(100)

248(2)

230(1)

214(15)

202(12)

201(46)

200(18)

nd

173(14)

159(12)

(<0.5)

(18)


percent abundance relative to the
trum, with the exception of TLN's
to the (MHC1-H) ion.


most abundant ion in the mass spec-
spectrum, where ions were normalized


HTLN

252(1)

236(4)

223(56)

234(1)

216(8)

200(15)

188(54)

187(100)

186(49)

171(18)

159(53)

nd

(<0.5)

(18)


HMTLN

266(2)

250(5)

237(100)

248(5)

230(7)

215(14)

202(56)

201(93)

200(44)

185(25)

173(18)

nd

(3)

(18)










Table 3-7. Methane electron-capture NCI-CAD (26 eV, 2.0 mtorr N2)
daughter mass spectral characteristics of the (MHC1-H)" ions of the
tryptoline-HC1 salts.


Ionsa

p-
P

(P-36)-

(P-38)

(P-65)

m/z 35


TLN

207(20b)

171(5)

nd

nd

(100)


MTLN

221(21)

185(32)

nd

nd

(100)


CH O-TLN

237(12)

201(6)

199(7)

nd

(100)


HTLN

223(14)

187(58)

185(13)

158(5)

(100)


HMTLN

237(21)

201(77)

199(40)

172(5)

(100)


ap is the parent ion, i.e. (MHC1-H)-.

Percent abundance relative to the most abundant ion.










Table 3-8. Methane electron-capture NCI-CAD (24 eV, 2.9 mtorr N2)
daughter mass spectral characteristics of the M- ions of the tryptoline-
HC1 salts.


Ionsa

P

(P-H)

(P-15)

(P-16)-

(P-29)

(P-30)

(P-43)

(P-44)

(P-45)

(P-58)


TLN

172(100b)

171(5)

157(1)

nd

143(76)

142(7)

nd

nd

nd

nd


MTLN

186(100)

185(43)

nd

170(2)

157(85)

156(17)

nd

nd

nd

nd


CH30-TLN

202(7)

201(1)

187(100)

186(3)

171(1)

nd

nd

158(14)

157(2)

nd


HTLN

188(100)

187(21)

173(5)

172(2)

159(68)

158(57)

145(16)

144(37)

nd

nd


HMTLN

202(100)

201(25)

187(5)

186(3)

173(58)

172(22)

nd

158(19)

157(4)

144(7)


P is the parent ion, i.e. M.

percent abundance relative to the most abundant ion.









Table 3-9. Methane electron-capture NCI-CAD (24 eV, 2.9 mtorr N2)
daughter mass spectral characteristics of the (M-H)- ions of the trypto-
line-HCl salts.


Ionsa

P

(P-H2)-

(P-15)-

(P-16)-

(P-29)

(P-44)-


TLN

171(95b)

169(1)

nd

nd

142(100)

nd


MTLN

185(99)

183(2)

170(<0.3

169(2)

156(100)

nd


CH O-TLN

201(13)

199(0.4)

) 186(100)

nd

nd

157(18)


HTLN

187(100)

185(2)

172(7)

171(0.4)

158(81)

nd

145(3)


HMTLN

201(100)

199(1)

186(1)

185(2)

172(70)

157(9)

145(2)


aP is the parent ion, i.e. (M-H).

percent abundance relative to the most abundant ion.









neutral molecule. The latter is supported by the presence of 46- ions

in the background. That NO2 may have added to the molecule is further

supported by the NCI-CAD daughter spectrum of the m/z 218 ion in the

TLN-HC1 spectrum. Fragmentation of this (M+46)- ion results in daughter

ions at m/z 171 (loss of 47) and m/z 142 (loss of 76), both of rela-

tively low abundance, and at m/z 46, the most abundant ion. The first

two daughter ions are indicative of the TLN molecule. The 46- ion could

be NO2- as its CAD spectrum contains only a single daughter ion at m/z

16, O presumably due to loss of NO. The NO2- ion could result from

the reaction of N2 and 02 (from an air leak) in the CH4 plasma. NO2 may

also originate from the decomposition of the sample in the reagent

plasma as the m/z 46 ion is seen to increase with sample pressure.

(M+NO2) adduct ions have been reported in the PCI and NCI spectra of

other nitrogen-compounds, but with no explanation of their formation

(98). However, as the compounds studied in this case contained nitro

groups, they could serve as the source of the NO2. The possibility that

all of the adduct ions described above are not adduct ions, but are

instead due to impurities, seems less likely, as the corresponding ions

are not present for such components in the El and PCI spectra.

The source of the air in the ion source and CI plasma is most

likely due to the steady leak of air at the 0-ring fittings of the

solids probe. The reproducibility cf the NCI mass spectra obtained on

different days was much lower than that for the PCI mass spectra. This

is probably directly attributable to the varying amount of air which can

enter through the direct inlet during operation of the solids probe,

depending upon the condition and tightness of the 0-ring seals and the

operator's technique. This susceptibility of CH4 electron-capture NCI









spectra to air as an impurity has been previously noted (97,99,100).

The formation and nonreproducibility of these adduct ions would result

in an unpredictable decrease in the sensitivity when monitoring only a

single ion, and could introduce large errors into the technique.

However, these adduct ions could also serve to help identify the

molecular weight of an unknown compound, just as the (M+29)+ and (M+41)+

adduct ions do in the PCI spectra. A more thorough study of the NCI

conditions and CAD spectra of all the NCI ions would be helpful in order

to obtain more conclusive evidence concerning the above. Due to the

consistently higher sensitivity and reproducibility obtained with PCI of

these compounds, this area was not pursued.

The abundant NCI (M-H) fragment ions do not fragment as effi-

ciently as the (M+H) ions under the same set of CAD conditions (Tables

3-5 and 3-9). However, the (M-H) ions do fragment characteristically,

forming largely daughter ions resulting from the loss of 29 u, CH2=NH,

from the parent ion. The (M-H)- ion of 5-methoxytryptoline is an excep-

tion in that two major daughter ions are formed. The most abundant,

186 is presumably due to loss of the methyl group from the 5-methoxy

group and then subsequent loss of CH2=NH from this ion to produce the

next most abundant daughter ion, 157.

Thus, electron-capture NCI of the tryptolines results in the forma-

tion of abundant (M-H) and (MHC1-H) ions, as well as several fragment

and adduct ions. The NCI mass spectra varied considerably when obtained

on different days, largely with respect to the relative abundance of the

adduct ions and also the relative yields of the (M-H)- and (MHC1-H)

ions. The NCI-CAD of the abundant (M-H)- ions of each of the trypto-

lines yields a simple but unique daughter mass spectrum, dominated by









the daughter ion due to loss of CH2=NH. The NCI-CAD of the (M-H)- ion

of CH30-TLN was an exception to this, in that its most abundant daughter

ion was due to loss of CH3 followed by loss of CH2=NH to yield the next

most abundant daughter ion.



Optimization Studies

Mode of ionization. To determine a mixture component reliably at

trace levels, the highest possible sensitivity and selectivity is de-

sired. Therefore, the optimum mode of ionization and optimum conditions

must be obtained. A comparison was made between positive and negative

CI-SIM of the m/z 173, (M+H)+, and m/z 171, (M-H)-, ions, respectively,

of TLN-HC1. The tandem mass spectrometer is able to sample alternately

positive and negative ions from the ion source very rapidly. Therefore,

it was possible to do this comparison on individual samples, without

taking into account the actual sample amount, but ensuring the signals

for both modes were not near their limits of detection. In doing this

comparison, it was determined that the integrated area ratios,

(M+H) /(M-H)-, were very dependent upon the tuning of the ion optics,

varying from 0.4 to 1000 for seven different tunings and measurements.

In general, when the optics were tuned for the "best" sensitivity of

each of the ionization modes for the ions of perfluoro-tri-N-butylamine,

a mass calibration compound, PCI was more sensitive than NCI for TLN-

HC1, with (M+H) /(M-H)- ratios averaging approximately 60. In addition,

in order to have good sensitivity with SRM, it is necessary to have a

good yield of the characteristic daughter ion from the characteristic

parent ion. From Tables 3-5 and 3-9, it can be seen that the (M+H)+

ions of all the tryptoline-HC1 salts fragment to give higher yields of









their characteristic daughter ions than do the corresponding (M-H) ions

under the same, relatively harsh, CAD conditions. Therefore, PCI was

chosen as the preferred ionization technique due to its greater sensi-

tivity. However, NCI could offer an advantage with regard to selecti-

vity. In analyzing complex mixtures directly, many components which

lack electrophilic atoms are transparent to NCI, and thus chemical

interference at the m/z of interest may be reduced. With methane's

nearly universal protonating ability, however, almost all mixtures

components are ionized. Due to the quantitative results below, however,

NCI was not further investigated.

PCI-CAD collision energy. It has already been mentioned, that in

order to obtain high sensitivity with SRM, it is necessary to have a

good yield of a characteristic daughter ion from the characteristic

parent ion of the analyte. The yield of daughter ions is very much

dependent upon the collision energy and the collision gas pressure. The

former parameter determines the energy which will be transferred to the

parent ion during a collision, and thus the internal energy available

for bond cleavage and daughter ion formation. The latter parameter is a

measure of the number of collisions that a parent ion (and/or its

daughter ions) will undergo in the transit of the collision region.

Thus, it is important to optimize these parameters for the compounds of

interest. The optimization of the CAD conditions was performed in two

different manners. In the first study, the collision energy was varied

during the vaporization and ionization of TLN-HC1 from the solids probe,

while obtaining daughter spectra over a limited mass range (m/z 140 to

m/z 176). The variation with collision energy in the ratio of the

intensities of the major daughter ion at m/z 144 and the parent ion at










m/z 173 revealed an optimum collision energy of 28 eV for the (M+H)+ to

(M+H-29)+ reaction (Figure 3-2). However, the variation in the ion

intensity ratios does not reveal any scattering losses of the parent and

daughter ions. The presence of such losses would result in loss of

sensitivity. In order to evaluate this aspect, a second collision

energy study was performed with gas chromatographic introduction of TLN-

HC1. The variation with collision energy of the GC peak areas of TLN-

HC1 resulting from the SRM of (M+H)+ to (M+H-29)+ was then obtained

(also in Figure 3-2). From this study, a collision energy of ca. 22.5

eV was determined to be optimal for this reaction. The decrease in peak

area following the maximum in both graphs may be due to focusing, scat-

ter, or an increase in the yield of other daughter ions at the expense

of the daughter ion of interest. As a compromise between the two

methods, 26 eV was chosen as the collision energy to be used in the PCI-

SRM studies.

PCI-CAD collision gas pressure. Similar studies were performed for

the optimization of collision gas pressure for maximum CAD sensiti-

vity. With the use of the solids probe for sample introduction, a large

increase is seen in the 144 /173 ratio in going from 0.7 to 2.1 mtorr

N2 (Figure 3-3). With the chromatographic sample introduction, the peak

area increases to a maximum as the collision gas pressure is increased,

and then begins declining (Figure 3-3). This decline in peak area is

most likely due to scattering losses of the parent ions. Although 4.0

mtorr N2 was apparently the optimum pressure for the highest sensi-

tivity, it was felt that operating the instrument at such high collision

gas pressures for long periods of time might be detrimental to the

vacuum system. Therefore, a collision gas pressure of 2.0 mtorr N2 was

selected for PCI-SRM assays for the tryptolines.



























100 -



80 -



60



x 40

S144/ I
173
20 A PCI-SRM GC Peak Area
(173 144)


0 II I I i i iI
8 10 12 14 16 18 20 22 24 26 28 30

Collision Energy (eV, Q2 offset)
Figure 3-2. Optimization of the PCI-CAD collision
energy for TLN-HC1.



























100



80


a 60



x 40 144/1
S/ A PCI-SRM GC Peak Area
0 (173 144)
20


0
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Collision Gas Pressure (mtorr N2)

Figure 3-3. Optimization of the PCI-CAD collision gas pressure for
TLN-HC1.










Quantitative Studies

With the short, packed GC column operated at the conditions used in

the optimization studies, it was possible to get retention times for

TLN- and MTLN-HC1 of ca. 10 s. This allowed very rapid analyses of many

samples much more conveniently than could be done by the solids probe

(ca. 5 min per sample, typically). However, the GC peaks of the trypto-

lines exhibited extreme tailing, most probably due to solute interac-

tions with the stationary phase and, especially, with the active sites

on exposed hot metal in the GC/MS interface. This extreme tailing was

not a large factor in the studies above, as the large levels of trypto-

line used gave very high signal-to-noise ratios (S/N). However, the

adsorption problem might become the limiting factor for trace

analysis. Therefore, quantitative studies were performed by vapori-

zation of the samples from a glass vial inserted in a heated solids

probe.

In the PCI-SRM analyses, the most abundant and characteristic PCI

ion of an analyte is selected as a parent ion by Q1, fragmented in Q2,

and only the most abundant daughter ion is monitored by Q3. Thus, for

PCI-SRM quantitation of TLN-HC1 and MTLN-HC1, the reaction of their

(M+H) ions, m/z 173 and m/z 187, respectively, yielding the m/z 144

daughter ions, was monitored. The quantitation signal was obtained by

integrating the ion current over the time during the vaporization of the

analyte. The limit of detection (LOD) was defined as the amount of

sample necessary to give a (S/N) of 3. The responses obtained from

three solvent blanks were averaged and used as the chemical noise level.

The results of the assay of a series of standard TLN-HC1 and MTLN-

HC1 solutions (Figures 3-4 and 3-5, and Tables 3-10 and 3-11) were used









to construct calibration curves (Figure 3-6). These calibration curves

of the two tryptolines are very similar to each other, as would be

expected from their chemical similarity. From all three figures, it can

be observed that no discernable signal representative of the tryptolines

occurs until the nanogram region is reached, whereupon the signal rises

rapidly with increasing amounts of tryptolines. As more concentrated

solutions were not analyzed, it was not possible to assess the linear

dynamic range of the technique. The lack of reproducibility of the

solids probe quantitation is reflected by the relative standard devia-

tions of the integrated signals varying from +20 to +52 % for the trip-

licate analyses having S/N ratios greater than 3. Most of the lack of

precision is probably due to the process of injecting and drying the 1.0

OL samples in the 5 yL sample vials. During this process, it was impor-

tant to ensure that no air bubbles were enclosed in the sample solution,

so that uniform drying would occur. In order to do this, it was neces-

sary to "jiggle" the syringe needle in the sample solution until the air

bubbles were excluded. However, this could have resulted in the drying

of some of the analyte on the needle and its subsequent removal. With

this amount of variability, the need for incorporating an internal

standard for reliable quantitation is apparent.

In Figures 3-4 and 3-5, some of the analyte peaks are split or have

shoulders on their leading edges. Based upon mass spectrometric evi-

dence, this is not due to a separate component in the sample. Rather,

it is most likely due to the uneven drying of the sample on different

portions of the sample vials. As the vial is not heated evenly, this

may lead to differential vaporization of the sample. An additional

source of this peak splitting may be the interaction of the sample with







27232


I Methanol I 0.0162I


0.162 I 1.62

Amount on Probe (nq)


S 16. 2


I 16.2


Figure 3-4. Solids probe PCI-SRM (173 -+ 144) quantitation of TLN-IICI.


I 162


- I----------1----------


I M






111760


2224


LP
.,I











Methanol 0.546 1 5.46 54.6

Amount on probe (ng)





S Methanol | 0.546 I 5.46 54.6 546

Amount on probe (ng)


Figure 3-5. Solids probe PCI-SRM (187 -+ 144) quartitation of MTLN-IIC1.









107







106 0 'I'TLN-fCl

U A MTLN-IlC1


4-
0




S10
U)







10 4


103





10-3 -2 -1 0 1 2 3
10 10 10- 10 10 10 10

Amount of Sample (ng)

Figure 3-6. Calibration curves for the solids probe PCI-SRM quantitation of TLN-HC1 and MTLN-IIC1.
B denotes the response of the methanol blank.








Table 3-10. PCI-SRM (173+ --> 144 ) solids probe quantitation of TLN-
HC1.


Amount on probe


Blank

1.62 fg

16.2 fg

162 fg

1.62 pg

16.2 pg

162 pg

1 .62 ng

16.2 ng

162 ng


Ia (144+)


7.8

5.8

5.9

6.8

6.2

5.9

5.9

6.4

10.3

1105.9


6.1

6.5

6.2

6.0

5.9

6.2

6.6

6.6

20.5

1099.8


Mean

ia (144+)


6.8

6.6

6.7

6.0

5.9




6.8

6.4

16.6

336.4


6.9

6.3

6.3

6.3

6.0

6.1

6.4

6.5

15.8

847.4


+ %RSDb




12

7

6

7

+3

4

+7

+2

+32

52


S/NC






0.9

0.9

0.9

0.9

0.9

0.9

1.0

2.3

124


aI is the integrated ion current (103 counts) for the m/z 144 daughter
ion.

b%RSD is % relative standard deviation.

CS/N is the signal-to-noise ratio with the mean of the 3 blanks repre-
senting the noise level.










Table 3-11. PCI-SRM (187+ --> 144+) solids probe quantitation of MTLN-
HC1.


Amount on probe

Blank

5.46 pg

54.6 pg

546 pg

5.46 ng

54.6 ng

546 ng 1


5.9

4.9

5.1

5.0

9.8

88.8

0338


(144

5.9

5.2

6.0

5.5

7.1

54.8

14516


5.5

5.6

5.6

5.6

7.6

69.0

15499


Mean

Ia (144+)

5.8

5.2

5.5

5.3

8.2

70.9

13451


%RSDb

4

7

8

6

+18

+24

+20


S/Nc




0.9

1.0

0.9

1.4

12.3

2326


I is the integrated ion current (103 counts) for the m/z 144 daughter
ion.

b%RSD is the % relative standard deviation.

cS/N is the signal-to-noise ratio with the mean of the 3 blanks repre-
senting the noise level.








the glass wall. The sample not adhering intimately to the wall may be

vaporized first, with hotter temperatures being necessary to vaporize

the layer of sample which is interacting with the "active sites" of the

glass wall.

From the definition of the limit of detection above, the limits of

detection were calculated to be approximately 17 ng for tryptoline-HC1

and 13 ng for methtryptoline-HCl on the probe. Tryptoline has been

reported to be present in whole rat brain at levels ranging from 400

pg/g to 40 ng/g of tissue (51,53,56,59), while methtryptoline has not,

as of yet, been reported to be present in brain tissue. As a typical

rat brain weighs ca. 2 g, these LOD's are too high to provide reliable

identification and quantitation of the tryptolines which may be present

in individual rat brains. It should be possible, however, to determine

the tryptolines in the brain homogenates pooled from several rats.

However, much of the biological significance as to individual variation

would be missing in this case.




Conclusion

In conclusion, it has been shown that the LOD's of PCI-SRM with

solids probe sample introduction are not low enough for the determina-

tion of underivatized tryptolines in individual rat brain extracts. It

should be possible, however, to use this technique with brain homo-

genates pooled from several rats. With pooled brain homogenates, this

technique would eliminate to a large degree the extensive sample clean-

up procedures which have been necessary for other methods of analysis

and thus, reduce the possibility of artefactual tryptoline formation.

In addition, the structural characterization possible with EI-CAD, PCI-




64




CAD, and NCI-CAD would allow much easier mass spectral structure eluci-

dation of any unknown tryptolines, once isolated.















CHAPTER 4
TANDEM MASS SPECTROMETRY FOR THE IDENTIFICATION AND
QUANTITATION OF TRYPTOLINE-HEPTAFLUOROBUTYRYL DERIVATIVES




Tryptoline-HFB Derivatives

Tryptoline has been reported to be present at the ng/g and pg/g of

tissue level in brain extracts (51,53,54,56,59), while the presence of

the other tryptolines studied in Chapter 3 has not, as yet, been

reported in rat brain. Thus, in order to be able to investigate the

tryptoline levels in individual rat brains, it is necessary to have a

very sensitive and selective detection method. In the previous chapter

it was determined that the limits of detection of solids probe PCI-SRM

of the underivatized tryptolines were not low enough to accomplish this

goal. Derivatization of the tryptolines is an efficient and straight-

forward method to enhance their chemical ionization sensitivity. Hepta-

fluorobutyryl (HFB) derivatization of the tryptolines has been shown to

give greatly enhanced sensitivity and lower limits of detection than

reported for the underivatized tryptolines in Chapter 3 (33,53-

56,59,101). In addition to increasing the sensitivity, HFB-derivati-

zation improves the chromatographic properties of the tryptolines.

Thus, with chromatographic introduction, an increase in the rate of

sample introduction into the ion source per unit time might be realized,

leading to a narrower, taller peak, a greater S/N ratio, and a still

lower LOD. In addition, the speed of analysis may be substantially

increased through the use of very short packed columns operated to allow









very short retention times of the analyte with minimal chromatographic

separation (85). Rather than use long chromatographic separations,

component separation and subsequent quantitation and identification of

the tryptolines could be performed by MS/MS.

Thus, in this chapter the HFB-derivatives of the tryptolines (Table

4-1) will be studied by GC/MS and GC/MS/MS. The effects of source

temperature and CAD conditions and their optimization will be discussed

with regard to the sensitivity of the methods. Quantitative comparisons

are made between methane positive and electron-capture negative CI and

between selected ion and selected reaction monitoring with a 0.4 m

packed column and a 18 m bonded phase fused silica capillary column. A

practical application is shown for the determination of tryptoline in a

HFB-derivatized crude extract of a rat brain.




Experimental




Materials and Reagents

All chemicals and reagents were of the highest purity available.

The standards of the tryptoline-HFB derivatives were kindly supplied by

Kym Faull, Ph.D., and Jack Barchas, M.D. (Department of Psychiatry and

Behavioral Sciences, Stanford Medical Center, Stanford, CA). Their

synthesis is detailed elsewhere (59). The HFB-derivatized crude

extracts of rat brains were also supplied by Kym Faull and Jack Barchas

and their preparation is described below. Aqueous solutions were pre-

pared in doubly distilled deionized water. Ultrahigh purity methane

(Matheson, Morrow, GA), helium, and zero grade nitrogen (Airco

Industrial Gases, Research Triangle Park, NC) were used as a CI reagent,

GC carrier, and collision gas, respectively.



















l^ N NN HFB
H RR






Table 4-1. Characteristics of the HFB-derivatives of the
tryptolines of interest.a


Compound

TLN-HFB

MTLN-HFB

CH30-TLN-HFB

HTLN-HFB

HMTLN-HFB

HTLN-(HFB)2

HMTLN-(HFB)2


R'

H

H

CH 30

HO

HO

HFB-O

HFB-O


R"

H

CH3

H

H

CH3

H

CH3


Molecular weight

Parent HFB-deriv.

172 368

186 382

202 398

188 384

202 398

188 580

202 594


aTryptoline (TLN), methtryptoline (MTLN), CH30-TLN
(5-methoxytryptoline), HTLN (5-hydroxytryptoline), HMTLN
(5-hydroxymethtryptoline), and HFB (heptafluorobutyryl or
C3F7CO).










Preparation of Extracts

Male rats (Sprague-Dawley, Simonson Labs, CA) were stunned with a

blow to the head and quickly decapitated. The brains were rapidly

removed, weighed, and homogenized in ice-cold perchloric acid (0.4 M, 5

mL/g of tissue). After centrifugation (15000 g, 20 min) the supernatant

was adjusted to pH 3 with 1 N NaOH and passed through a C-18 reverse

phase Sep-PAK cartridge (Waters Associates, Milford, MA) which had been

previously washed with acetonitrile (2 x 5 mL) and water (2 x 5 mL).

The cartridge was then washed with water (2 x 500 UL) and eluted with

acetonitrile (3 x 500 yL). The solvent was removed from the eluate in a

stream of nitrogen and the residue was washed to the bottom of the

collection tubes by the addition of acetonitrile (100 IL) which was also

removed in a stream of nitrogen. The samples were then ready for chem-

ical derivatization. The Sep-PAK cleanup procedure gives recoveries of

added tryptolines of between 65 and 97% (59).



Chemical Derivatization (53)

The authentic compounds and dried extracts were treated with hepta-

fluorobutyrylimidazole (Regis Chemical Co., Morton Grove, IL; 100 yL,

80 C, 60 min) after which methylene chloride was added (2 mL). The

solution was then extracted with water (four times with 2 mL each time)

using centrifugation to separate the phases. The aqueous layers were

discarded and the methylene chloride was evaporated in a stream of

nitrogen. The dried residue (equivalent to ca. 500 mg of tissue for the

brain extracts) was redissolved in ethyl acetate or methanol prior to

injection onto the GC column.









Instrumentation

All data were collected with a Finnigan MAT (San Jose, CA) triple

stage quadrupole GC/MS/MS equipped with a 4500 series ion source, pulsed

positive and negative chemical ionization, and INCOS data system. The

Finnigan 9610 gas chromatograph was equipped with packed and Grob-type

capillary column injectors.

Samples were introduced into the ion source by either an 18-m DB-5

bonded-phase fused silica capillary column (compliments of J & W

Scientific, Rancho Cordova, CA) inserted directly into the ion source or

a short (38 cm x 0.75 mm i.d.) packed glass column, described more fully

in Chapter 3. The capillary column was operated with a Grob-type split-

less injector using a carrier gas split of 40 mL/min and a septum sweep

of 9 mL/min; both were closed for 1 min following injection. The GC

oven was held at 100 C for 1 min following injection and, thereafter,

increased linearly at a rate of 15 OC/min to a maximum of 275 "C. These

conditions resulted in retention times for the tryptoline-HFB deriva-

tives in the 9-13 min range. The short glass column was used for rapid

sample introduction and provided minimal chromatographic resolution

(retention time for TLN-HFB of ca. 55 s). The glass column was sila-

nized with a 10 % solution of dimethylchlorosilane in toluene prior to

manual packing with 3% OV-101 on 80/100 mesh Chromosorb 750. The GC/MS

interface consisted of a glass-lined stainless steel tube direct inlet

fitted with a microneedle valve. The interface and injection ports were

maintained at 250 C and 275 C for packed and capillary column work,

respectively. In the El mode helium was used as the carrier gas. In

the CI mode methane was used as a combined carrier/reagent gas. With

the packed column CI analyses, the carrier gas flow was set to 10 mL/min








CH4. The capillary column CI analyses were conducted with an average

linear velocity of ca. 50 cm/s CH4 (at 275 0C). Additional methane was

added in both cases as makeup gas to give the required ion source pres-

sure.




Procedures

Mass spectra of standards. Standards were introduced on the capil-

lary column and the CI (100 eV electron energy, 1.0 torr CH4 source

pressure) and El (70 eV electron energy) mass spectra were obtained with

an ion source temperature of 100 C in the Q3 normal MS mode. The

effect of the ion source temperature on the fragmentation pattern of the

heptafluorobutyryl derivatives of the tryptolines was studied with

temperatures between 80 and 190 "C. Daughter spectra of M and (M-HF)

and (M+H) ions were obtained at collision gas pressures of 1.0 and 2.8

mtorr N2 and collision energies of 20 and 26 eV (Q2 offset) for negative

and positive CI, respectively. The mass spectra acquired during the

elution of each GC peak were averaged and background subtracted to yield

a representative mass spectrum of each standard.

Collision energy and collision gas pressure studies. For each

combination of collision energy and collision gas pressure, duplicate

1.0 IL injections of a standard solution of TLN-HFB were made onto the

packed GC column (200 *C isothermal, source temperature at 190 C). The

GC peak areas resulting from the PCI-SRM of 369+ to 156+ and the NCI-SRM

of 348 to 179 were determined with the data system. The collision

energy studies were conducted at collision gas pressures of 1.5 and 1.7

mtorr N2 for positive and negative CI-SRM, respectively. The collision

gas pressure studies were conducted at collision energies of 25 and 18

eV for positive and negative CI-SRM, respectively.









Selected ion and selected reaction monitoring. All SIM experiments

were conducted in the Q3 normal MS mode. The SRM experiments were

conducted in the daughter scan mode with collision gas pressures of 1.0

and 2.8 mtorr N2 and collision energies of 20 and 26 eV for negative and

positive CI, respectively. The SRM daughter and the SIM ions were

scanned over a 1 u wide window at 10 Hz. The ions and reactions moni-

tored are listed in Table 4-2.

Quantitative studies. Serial dilutions were made of TLN-HFB with

methanol to give solutions with concentrations ranging from 2.0 pptr to

200 ppm. In two separate studies, triplicate or single 1.0-)L injec-

tions were made of each standard onto the packed GC column (175 C

isothermal, giving a retention time of ca. 55 s for TLN-HFB). Quantita-

tion was performed under the CI-SIM and CI-SRM conditions above with a

source temperature of 100 "C, and an electron multiplier voltage of 1800

V with a preamp sensitivity of 10- A/V. The areas of the TLN-HFB GC

peaks were determined by use of the current data system software and are

reported in units of data system counts. Similar analyses were also

done for quantitation of the crude brain extracts containing added

amounts of TLN-HFB. With the capillary column, a single 1 .0-L split-

less injection of each sample was made and quantitation was accomplished

as for the packed column. The reproducibility of the capillary column

was studied by measuring the peak areas of 6 replicate 1.0 pL injections

of 2 different TLN-HFB standards with an electron multiplier voltage of

1300 V and a preamp sensitivity of 10-8 A/V.
1300 V and a preamp sensitivity of 10 A/V.









Table 4-2. The ions and reactions monitored for the techniques used in the deter-
mination of the tryptoline-HFB derivatives in crude rat brain extract.


Compound

TLN-HFB

MTLN-HFB

CH30-TLN-HFB

HTLN-HFB

HMTLN-IHFB

HTLN-(HFB)2

HMTLN-(HFB)2


Ions for CI-SIM

(M+H)+ (M-HF)-

369 348

383 362

399 378

385 384

399 378

581 560

595 574


Reactions for CI-SRM

(M+H)+ -> (M+H-213)+ (M-HF)" -> 179-

369 --> 156 348 --> 179 (159b)

383 --> 170 362 --> 179 (159)

399 --> 186 378 --> 179 (159)

385 --> 172 364 --> 179 (184,159)

399 --> 186 378 --> 179 (198,159)

_a 560 --> 179 (381,379)

_a 574 -> 179 (395,393)


aWere not determined.


bAlternate daughter ions for multiple reaction monitoring.










Results and Discussion




Structure of the Tryptoline-HFB Derivatives

The derivatization of the tryptolines by heptafluorbutyryl imida-

zole under the experimental conditions used here resulted in the forma-

tion of a single mono-HFB derivative of each of the nonhydroxytrypto-

lines and a single di-HFB derivative and a single mono-HFB derivative of

each of the 5-hydroxytryptolines as determined by capillary GC/MS

(Tables 4-1, 4-3, 4-4, and 4-5). The relative amounts of the di-HFB and

mono-HFB derivatives of the hydroxytryptolines varied, presumably de-

pending upon the experimental conditions. Based upon the mass spectral

evidence below and that of others (33,53,55,59,101), the mono-HFB deri-

vatives of the tryptolines have the HFB group replacing the hydrogen on

the piperidine nitrogen. For the 5-hydroxytryptolines di-HFB deriva-

tives, the HFB groups have replaced the hydrogens on the piperidine

nitrogen and the phenolic oxygen. No evidence was observed for the

replacement of the indolic nitrogen's hydrogen.




Mass Spectral Characteristics

El normal mass spectra. The El mass spectra of the tryptoline-HFB

derivatives show prominent molecular ions and relatively abundant (M-

225) fragment ions (Table 4-3). These fragment ions presumably arise

by loss of a CH2NCOC3F7 group from the piperidine ring. In addition,

the 9-methyl-tryptoline-HFB derivatives show an abundant fragment ion

(M-15)+, presumably due to the loss of the methyl side chain group. The

mass spectra of the di-HFB derivatives of HTLN and HMTLN, although

having prominent molecular ions, are dominated by the uncharacteristic








fragment ion at m/z 69, presumably (CF3). As in the underivatized

tryptolines, there is again extensive fragmentation of the molecular

ions. As it was desired to analyze samples with minimal, if any chroma-

tograpic separation, it was necessary to use the softer chemical ioni-

zation to reduce the fragmentation of all components of the extracts.

Thus, there would be less chemical noise at the m/z of the ions of

interest.

PCI normal mass spectra. The methane PCI mass spectra of all the

HFB derivatives are dominated by the protonated molecules (M+H) and

the methane adduct ions, (M+29) and (M+41) (Table 4-4). The only

significant fragment ions seen in the spectra of the mono-HFB deriva-

tives correspond to (M+H-HF)+. The di-HFB derivatives of HTLN and HMTLN

undergo much more fragmentation than do the mono-HFB derivatives. The

fragment ions at (M+H-198) are most likely due to the protonation of

the phenolic-HFB group and subsequent loss of C3F7CHO from the (M+H).

The other fragment ions are characteristic of the HFB groups, and there-

fore not specific for the tryptolines: Ions at m/z 215, m/z 199, m/z

179, m/z 161, and m/z 141 presumably correspond to (C3F7C(OH) 2)

(C3F7CHOH)+, (C3 FCOH)+, (C3 CHOH)+, and (CF3 4CHO)+, respectively.

NCI normal mass spectra. In contrast to PCI, the methane electron-

capture NCI mass spectra (at a source temperature of 100 "C) of the

derivatives yield molecular ions in relatively low abundance and show

extensive fragmentation (Table 4-5). The most abundant ion in the NCI

mass spectrum of each mono-HFB derivative and HTLN-(HFB)2 is due to the

loss of HF from the molecular ion, with less abundant ions being due to

successive losses of F and HF. Although the (M-HF)- ion is abundant in

the NCI mass spectrum of HMTLN-(HFB)2, the most abundant ion corresponds









to (M-198)-, which is also a major fragment ion of HTLN-(HFB)2. This

loss of 198 u is most probably due to a hydrogen rearrangement and

subsequent loss of C3F7CHO from the phenolic portion of the molecular

ion. In addition, all of the tryptoline-HFB derivatives yield rela-

tively abundant ions at m/z 179, (C3F6HCO)-, and m/z 178, (C3F6CO)-,

attributable to the heptafluorobutyryl portion of the molecule. For the

NCI-SIM experiments, the most abundant and characteristic ion in the CI

mass spectrum of each of the tryptoline derivatives is selected by Q3

for monitoring. Although the (M-198)- ion is the most abundant ion in

the NCI of HMTLN-(HFB)2, the more characteristic (M-HF)- ion was chosen

for SIM. This would allow direct comparison to the SRM experiments

below.

PCI-CAD daughter mass spectra. The PCI-CAD daughter spectra of

HTLN-(HFB)2 and HMTLN-(HFB)2 were inadvertently not acquired. The PCI-

CAD daughter mass spectrum of the (M+H)+ ion of each of the derivatives

is dominated by the daughter ion resulting from the loss of 213 u

(either NH2COC3F7 or HN=COHC3 F 7) from the protonated molecule; as such,

this ion is characteristic of the parent tryptoline structure (Table 4-

6). This characteristic loss from (M+H)+ could be utilized in the MS/MS

neutral loss mode to screen for other possible tryptolines and compounds

with a similar derivatized amine structure. In addition, other, less

abundant daughter ions were present and were useful structurally. Thus,

a loss of the entire HFB group from the (M+H) ions of the mono-HFB

derivatives resulted in daughter ions of low abundance corresponding to

the parent tryptoline structures at (P-197)+. Daughter ions reflecting

the R' and R" substituents also resulted with HTLN-HFB and HMTLN-HFB

losing H20 from their (M+H)+ ions, while the mono-HFB derivatives of









Table 4-3. El mass spectral characteristics of the tryptoline-HFB derivatives.


TLN-HFB

MTLN-HFB

CH30-TLN-HFB

HTLN-HFB

HMTLN-HFB

HTLN-(HFB)2

HMTLN-(HFB)2


M+

368(89)

382(62)

398(100)

384(69)

398(62)

580(58)

594(25)


(M-15)+ (M-169)+

199(17)

367(100) 213(11)

229(17)

215(15)

383(100) 229(12)

411(19)

574(28) 425(5)


(M-197)+

171(18)

185(8)

201(10)

187(11)

201(9)

383(25)

397(6)


(M-199)+

169(16)

183(6)

199(10)

185(11)

199(13)

381(6)

395(2)


(M-212)+

156(13)

170(22)

186(33)

172(25)

186(24)

368(24)

382(6)


(M-213)+

155(23)

169(34)

185(34)

171(30)

185(34)

367(27)

381(7)


(M-214)+

154(25)

168(20)

184(21)

170(23)

184(21)

366(11)

380(3)


(M-225)+

143(100)

157(16)

173(92)

159(100)

173(23)

355(51)

369(4)


129(9)

156(20)

170(18)

158(19)

172(25)

158(84)

169(61)


115(22)

154(33)

158(62)

130(13)

170(32)

69(100)

69(100)








PCI normal mass spectral characteristics of the


TLN-HFB

MTLN-HFB

CH30-HFB

HTLN-HFB

HMTLN-HFB


(M+41)+

409(3)

423(2)

439(1)

425(2)

439(3)


(M+29)+

397(12)

411(8)

427(8)

413(12)

427(10)


(M+H)+

369(100)

383(100)

399(100)

385(100)

399(100)


(M+H-HF)+

349(8)

363(7)

379(10)

365(9)

379(8)


(M+H-198)+

171(0.6)

185(2)

201(0.6)

187(0.6)

201(1)


HTLN-(HFB)2

HMTLN-(HFB)2


(M+41)+

621(2)

635(2)


(M+29)+

609(9)

623(11)


(M+H)+

581(100)

595(100)


(M+H-HF)+ (M+H-198)+

561(19) 383(34)

575(13) 397(450


227+ 215+ 199+ 179+ 161+ 141+

(10) (41) (62) (85) (38) (36)

(13) (32) (77) (92) (36) (38)


(M+H-199)+

170(0.6)

184(2)

200(2)

186(0.2)

200(1)


161+c

(0.4)

(4)

(2)

(2)

(5)


141+d

(0.9)

(5)

(3)

(3)

(4)


tryptoline-HFB derivatives.


Table 4-4.









Table 4-5. NCI normal mass spectral characteristics of the tryptoline-HFB derivatives.


TLN-HFB

MTLN-HFB

CH3OTLN-IIFB

HTLN-HFB

HMTLN-HFB


M~

368(3)

382(14)

398(2)

384(1)

398(3)


(M-HF)-

348(100)

362(100)

378(100)

364(100)

378(100)


(M-2HF)-

328(14)

342(13)

358(12)

344(14)

358(14)


(M-3HF)- 225-

308(13) (0.3)

322(6) (2)

338(14) (0.2)

324(14) (0.1)

338(5) (0.4)


MS

HTLN-(HFB)2 580(2)

HMTLN-(HFB)2 594(0.4)


(M-HF)-

560(100)

574(24)


(M-2HF)-

540(7)

554(0.3)


(M-3HF)- (M-178)- (M-198)- (M-217)- (M-237)-

520(0.6) 402(0.3) 382(85) 363(2) 343(1)

534(<.1) 416(4) 396(100) 377(2) 357(1)


197- 178~ 160-

(2) (11) (3)

(2) (14) (3)


(M-200)-

168(0.2)

182(0.7)

198(0.3)

184(5)

198(14)


179-

(4)

(24)

(6)

(2)

(5)


178-

(2)

(6)

(1)

(1)

(1)


160-

(1)

(5)

(1)

(2)

(3)









MTLN, CH30-TLN, and HMTLN all show loss of CH3 from their (M+H)+ ions

and their [(M+H)-213]+ ions. The (M+H)+ ions of the nonhydroxy- and 5-

hydroxytryptoline derivatives fragment by loss of of R"CH=NCOC3F7 to

yield daughter ions at m/z 144 and m/z 160, respectively. The daughter

ion at m/z 173 in the CH30-TLN-HFB spectrum may also correspond to such

a loss.

NCI-CAD daughter mass spectra. The NCI-CAD daughter spectrum of

each M- ion is dominated by the daughter ion at m/z 225, (CH2NCOC F7)

(Table 4-7). This ion corresponds to the prominent neutral fragment

loss from the molecular ions in the El mass spectra to yield the ions at

(M-225)+ (Table 4-3). This NCI-CAD daughter ion could be used in the

MS/MS parent scan mode to screen for other compounds in the brain ex-

tract having a derivatized CH2-NH substructure. The mono-HFB deriva-

tives of HTLN and HMTLN yield, in addition to the m/z 225 ion, prominent

(P-200) daughter ions which could result from the combined loss of HF

and (C3F6 H2CO) from the M parent ion. This is supported by the abun-

dant (P-180) daughter ions in the NCI-CAD of their (M-HF) fragment

ions (Table 4-8). The NCI-CAD daughter ion mass spectra of the (M-HF)-

ions of the non-hydroxytryptoline-HFB derivatives are dominated by the

ion at m/z 179 corresponding to the (C3F6OH) fragment of the deriva-

tizing group (Table 4-8). Even though this ion is of low mass and low

diagnostic value per se, selectivity is maintained in the MS/MS-SRM

experiments due to the genetic relationship to the (M-HF)- ions.

Although the (M-HF) ions of HTLN-HFB and HMTLN-HFB yield abundant 179

daughter ions, the most abundant daughter ions, (P-180) are presumably

due to the loss of (C3F6H2CO). The (M-HF) ions of HTLN-(HFB)2 and

HMTLN-(HFB)2 yield, in addition to the abundant 179- ion, two prominent









Table 4-6. PCI-CAD daughter mass spectral characteristics of the (M+H)+ ions of the HFB deriva-
tives of the tryptolines.a


TLN-HFB

MTLN-HFB

CH30-TLN-HFB

HTLN-HFB

HMTLN-HFB


P+ (P-15)+ (P-18)+

369(32C)

383(40) 368(1) -

399(46) 384(6) -

385(64) 367(15)

399(59) 384(1) 381(12)


(P-197)+




186(0.6)

202(0.4)

188(0.9)

202(0.3)


(P-213)+

172(0.9)

170(100)

186(100)

172(100)

186(100)


156(100)

144(9)

173(3)

160(1)

160(7)


144(2) 129(7)


155(4)

171(2)

145(2)

171(2)


158(2) 155(3)

154(1)

366(3)


ap is the parent ion, (M+H)+.

b(P-[HR,"C=NCOC3F7] +


CPercent abundance relative to the most abundant ion.








Table 4-7. NCI-CAD daughter mass spectral characteristics of the M- ions of the
HFB-derivatives of the tryptolines.


TLN-HFB

MTLN-HFB

CH30-TLN-HFB

HTLN-HFB

HMTLN-HFB


p

368(42a)

382(22)

398(23)

384(100)

398(100)


(P-HF)-

348(33)

362(1)

378(9)

364(16)

378(18)


(P-200)-


(100)

(100)

(100)

184(6) (25)

198(12) (38)


225- 179- 178- 159-


(38) (9)

(2) (4)

(12) (2)

(14) (4)

(11) (5)


HTLN-(HFB)2

HMTLN-(HFB)2


P

580(61)

594(100)


(P-HF)-

560(100)

574(76)


(P-H)-

579(22)

593(77)


(P-H,F)- (P-197)-

559(30) 383(2)

573(57) 397(8)


(P-198)- 225- 179-

382(2) (11) (3)

396(12) (14) (8)


percent abundance relative to the most abundant ion.










Table 4-8. NCI-CAD daughter mass spectral characteristics of the (M-HF)- ions of the HFB deri-
vatives of the tryptolines.


TLN-HFB

MTLN-HFB

CH30-TLN-HFB

HTLN-HFB

HMTLN-HFB


HTLN-(IIFB)2

HMTLN-(HFB)2


p- a (P-HF)-

348(21b) 328(1)

362(16) 342(7)

378(19) 358(0.3)

364(41) 344(1)

378(27) 358(15)


560(100)

574(99)


540(1)

554(11)


(P-2HF)"

308(2)

322(0.3)

338(0.5)

324(1)

338(1)


520(0.4)

553(3)c


(P-179)-

169(5)

183(6)

199(3)

185(3)

199(7)


381(23)

395(71)


(P-180)"

168(5)

182(1)

198(7)

184(100)

198(100)


(P-181)-

167(4)

181(0.4)

197(2)

183(1)


372(22)

393(27)


parent ion (P)

percent abundance relative to the most abundant ion.

c(P-HF-F)-.


179-

(100)

(100)

(100)

(84)

(91)


(51)

(100)


159-

(29)

(15)

(23)

(25)

(17)


(2)

(2)









daughter ions, (P-179)- and (P-181)-, presumably due to the loss of

C3F 6HCO and C3F6HCHOH, respectively. The HFB-derivatives of HTLN and

HMTLN are apparently more stable than the nonhydroxytryptolines as

evidenced by the lower yield of daughter ions under the same CAD condi-

tions. This can be attributed to the ability of the phenolic oxygen to

stabilize the negative charge. This stabilizing effect also explains

the large abundance of the (P-180) daughter and the (P-179) and (P-

181) daughter ions of their mono- and di-HFB derivatives, respectively.

For the determination of tryptolines in biological extracts by

selected reaction monitoring, maximum sensitivity is obtained by moni-

toring the most abundant daughter ion from the most abundant and charac-

teristic CI parent ion (Table 4-2). Although the 179 ions are not the

most abundant daughter ions for all of the HFB derivatives of the

hydroxytryptolines, it was chosen for convenience purposes. This would

allow the possibility of using the parent scan of 179 instead of

several different selected reactions, for screening purposes. The yield

of several daughter ions of high abundances can be exploited by multiple

reaction monitoring (MRM). MRM would improve the selectivity, and thus,

the reliability, of determining the tryptolines without much loss in

sensitivity. In the NCI of HMTLN-(HFB)2, the abundance of the (M-HF)-

ion is only about one-fourth that of the most abundant (M-198)- ion.

However, under NCI-CAD, the most abundant daughter ion of the (M-198)-

ion represents only ca. 2 percent of the parent ion's abundance, while

the 179- daughter ion represents ca. 30 percent of the parent (M-HF)-

ion. Thus, for increased sensitivity, the more characteristic, and more

selective (M-HF)- ion fragmenting to the 179- ion was chosen for SRM of

HMTLN-(HFB)2. Because the derivatives behave in a similar fashion under








CI-CAD conditions, and because of our particular interest in the natural

occurrence of tryptoline in mammalian tissues, efforts were concentrated

on this compound.



Optimization of Experimental Parameters

Source temperature. The relative intensities of the NCI molecular

and fragment ions of TLN-HFB were greatly influenced by the ion source

temperature (Figure 4-1). At a source temperature of 80 C, the NCI

spectrum was dominated by an intense (M-HF) ion with relatively small

contribution from the other ions. As the source temperature was

increased, the amount of fragmentation in the NCI mass spectrum

increased dramatically. In particular, the (M-HF) ion decreased from

approximately 67 percent of the reconstructed ion current at 80 C to

only 9 percent at 190 C. This would represent a significant loss in

sensitivity in the SIM and SRM techniques (selecting (M-HF) ). This

susceptibility of the fragmentation of these types of derivatives to

experimental conditions in the ion source has been noted previously

(102). In contrast, the PCI mass spectra showed little variability with

source temperature (Figure 4-1). Because it was difficult to repro-

ducibly maintain the ion source temperature below 100 C and because

contamination of the ion source occurs more rapidly at such low tem-

peratures, subsequent analyses were run with an ion source temperature

of 100 OC.

CAD conditions. In the collisionally activated dissociation pro-

cess, the yield of the daughter ions is very much dependent upon the

collision energy and the pressure of the collision gas, as was demon-

strated in Chapter 3. With NCI-SRM (at a collision gas pressure of 1.7









mtorr N2) the GC peak area from injections of TLN-HFB showed a maximum

at a collision energy of ca. 20 eV (Figure 4-2a). At a collision energy

of 18 eV, a maximum GC peak area was observed at a collision gas pres-

sure of ca. 1.0 mtorr N2 (Figure 4-2b). The decrease in signal after

the maxima is due largely to an increase in the yield of other daughter

ions. With PCI-SRM, the GC peak area showed a general increase in

intensity with higher collision energy (Figure 4-2a) and with increased

collision gas pressure (Figure 4-2b). These results, in conjunction

with those of the ion source temperature study and the characteristics

of the CI spectra, suggest that the PCI ions are much more stable than

are the NCI ions. From Figure 4-2, collision energies of 20 eV and 26

eV and collision gas pressures of 1.0 mtorr and 2.8 mtorr N2 were chosen

as optimum for the negative and positive CI-SRM techniques, respec-

tively.




Quantitative Studies

Standard calibration curves. To compare the sensitivities and

limits of detection of the different techniques, calibration curves were

obtained from analysis of standard TLN-HFB solutions (Figure 4-3). The

calibration curves for all the techniques are summarized in Table 4-9.

The limits of detection were calculated from the calibration data and

corresponded to the amounts of TLN-HFB which would give GC peak areas

three times greater than those obtained with solvent blank. The blank

gave a visible response, presumably due to adsorption on the column and

septum and subsequent desorption by the next injection (59). By fre-

quent replacement of the septum and several injections of solvent

between concentration series, this problem was reduced to a manageable










369'


60-






308-
0 40







20-
179-

348-



0 368-
0 ------------- 611
80 90 100 110 120 130 140 150 160 170 180 190

SOURCE TEMPERATURE (C)

Figure 4-1. Effect of source temperature upon the fragmentation of tryptoline-IlFB
under CI conditions.





















NCI-SRM /


PCI-SRM


20.0


30.0


COLLISION ENERGY (eV)


NCI-SRM





PCI-SRM


0,0 1.0 2.0 30


COLLISION GAS PRESSURE (mTorr Nz)


Figure 4-2.


Effect of collision energy (a) and collision gas
pressure (b) upon the tryptoline-HFB GC peak areas.


100-









degree. The limits of detection thus defined are in the linear portion

of the calibration curves (Figure 4-3). Figure 4-4 illustrates the

method utilized with the short, packed GC column for determination of

the noise and signal levels for the assay of the TLN-HFB standards.

After monitoring the background for approximately 50 s, the filament was

turned off and, simultaneously, 1.0 1L of a standard was injected.

Approximately 10 s later, following elution of most of the solvent, the

filament was turned on again, and monitoring continued. The signal or

GC peak area of the TLN-HFB was determined by integrating the ion cur-

rent above an estimated baseline threshold for a given number of scans,

as shown in Figure 4-4. The noise was determined by a similar integra-

tion over the same number of scans (or a number of scans multiplied by

an appropriate factor to equal the same number of scans) in the back-

ground-monitoring region of the chromatogram prior to injection of TLN-

HFB (Figure 4-4). With the capillary column, the electron multiplier

and filament remained off until approximately 90 s prior to elution of

TLN-HFB. The signal or GC peak area of TLN-HFB and the noise level were

determined as above in the regions of the chromatograms indicated in

Figure 4-5. The signal-to-noise ratio (S/N) of all the limits of detec-

tion ranged from 2 to 6 with the exception of the packed column PCI-SRM

limit of detection which had a S/N of 11 due to a high solvent blank

response. All the calibration curves showed good linearity above their

respective limits of detection for the range of concentrations

studied. The linear portion of the log-log calibration curves of all

the techniques had a mean slope of 1.0 5%.

NCI is observed to give significantly better sensitivity and limits

of detection than does PCI for these heptafluorobutyryl derivatives









(Table 4-9). This is readily explained by the highly electrophilic

nature of the perfluorinated portion of the derivative and the under-

lying process of electron-capture NCI (94,95). The limits of detection

obtainable with the NCI-SIM and NCI-SRM techniques are similar, despite

the significantly greater sensitivity of the SIM technique. The lower

sensitivity of the SRM technique is expected because of the ineffi-

ciencies of the CAD conversion of the parent ion to the daughter ion of

interest and the daughter ion's subsequent mass analysis. However, the

selectivity gained by this parent-daughter reaction reduces the chemical

noise to a greater extent than the analytical signal and compensates to

some extent for the lost sensitivity. This reduction in chemical noise

is apparent in the relative heights of the chromatogram baselines in the

NCI-SIM and NCI-SRM capillary column chromatograms (Figure 4-5). With

the packed GC column, the reduction in chemical noise with PCI-SRM

resulted in it having a limit of detection 13 times lower than that of

the corresponding PCI-SIM (Figure 4-4). The difference in the limits of

detection between the packed column GC techniques and those of the

capillary column techniques may be due to ion optics tuning, conditions

of the injection port and GC columns, and cleanliness of the ion source,

lenses, and quadrupole rods. It was found that the LOD's of the tech-

niques could vary by as much as a factor of 10 when obtained on diffe-

rent days. The above should represent what can be expected after

cleaning the instrument and optimal tuning.

The reproducibility of the GC peak areas obtained with the packed

column NCI-SRM for triplicate injections varied from 31 % relative

standard deviation (RSD) near the limit of detection to +4.5 % RSD at

levels well above the LOD. In a limited study with capillary column












108


7 e NCI-SIM
10
L NCI-SRM A/
PCI-SIM
A PCI-SRMI




/ I
o 10







0^3~ ----^---
l04




103
10-2 10-1 100 01 1 02 10 10
Picograms TLN-HFB injected

Figure 4-3. Calibration curves for the capillary GC techniques for the determination of
TLN-HFB standards. The GC peak areas have been noise corrected. The signal obtained with
the solvent blank is indicated by B. The arrows indicate the calculated limit of detection
for each technique.










Signal
34368 -

a) PCI-SIM


Noise A B
F --------- ^ r /''^
-H





16352 -

b) PCI-SRM




-H



1/2 Noise AB


Scan 0 200 400 600 800 1000 1200 1400
Time 0:0 0:21 0:42 1:03 1:24 1:45 2:06 2:27

Figure 4-4. Comparison of packed column (a) GC/PCI-SIM (369) and (b) GC/PCI-SRM (369 156) of
20.3 ng TLN-HFB. The filament was turned off and the sample simultaneously injected at A. After
elution of most of the solvent, the filament was turned on at B.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs