Associative ion-molecule reactions and ion trapping with a triple quadrupole mass spectrometer

MISSING IMAGE

Material Information

Title:
Associative ion-molecule reactions and ion trapping with a triple quadrupole mass spectrometer
Physical Description:
ix, 219 leaves : ill. ; 29 cm.
Language:
English
Creator:
Annacchino, Anthony Peter, 1965-
Publication Date:

Subjects

Subjects / Keywords:
Quadrupoles   ( lcsh )
Mass spectrometry   ( lcsh )
DNA   ( lcsh )
RNA   ( lcsh )
Chemistry thesis Ph. D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 212-218).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Anthony Peter Annacchino, Jr.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001950911
oclc - 31200844
notis - AKC7453
System ID:
AA00003234:00001

Full Text








ASSOCIATIVE ION-MOLECULE REACTIONS AND ION TRAPPING
WITH A TRIPLE QUADRUPOLE MASS SPECTROMETER
















By

ANTHONY PETER ANNACCHINO, JR.


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


UNIVERSITY OF FLORIDA































To my family with love.












ACKNOWLEDGEMENTS


would


to extend my


sincere gratitude


to my


research


advisor,


Richard A.


Yost.


Had it not been for his support, guidance and patience (and a


couple of well-timed lunches at Joe's


Deli), I am certain that this work would not


have


been


realized.


would


to thank


members


graduate


committee, Drs. James D.


Winefordner, David H. Powell, John R. Eyler, and Charles


L. Beatty.


would


to acknowledge


U.S.


Environmental Protection Agency-


Environmental


Research


Laboratory/Duluth


(U.S.


EPA-ERL/D)


providing


financial support for much of this work.


Additionally, I would like to thank Dr.


Douglas W. Kuehl of the U.S. EPA-ERL/D for helpful discussions.


I would like to thank all the members of the


Yost group, past and present,


with whom I have had the pleasure to work during my time here in Gainesville.


particular, I would like to acknowledge two of my office mates and close friends, Uli


Bernier and Rafael Vargas.


In their company, I was able to enjoy this period of my


life and experience such wonders as the "foul towel" and the Bayou Plaza.


memories, indeed.


Fond


I would also like to thank all the friends that I have gained in


this community through music.


More than once, these friends, and the music that







I shared with them, reinvigorated me when I needed it the most.


Also, I was always


reminded that there is more to life than graduate school.


I would like to thank my family, Mom,


Dad, and Ken,


who supported me


throughout my studies and always had confidence in me,


even during the rough


times.


Finally, I would like to express my sincere thanks and love to my fiance,


Jennifer Buckingham.


Her companionship and understanding allowed me to devote


more time to school than was fair to her.


But, then again, she said I would pay later.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS


ABSTRACT


CHAPTERS


INTRODUCTION


9 9 9 9 a a . a 4 9 a p a a a a a a 1


Methods for
Ion-Molecule
Triple
Trapping Ion
Organization


the Determination of Mutagenic
Reactions in the Collision Cell
Quadrupole Mass Spectrometer
s with Linear RF-only Multipole
of Dissertation


Compounds
of a

Devices .


* a a a a a a a a a
* a p p S P


REACTIONS OF DNA/RNA BASE AND
NUCLEOSIDE IONS WITH ALLYL HALIDES


Introduction . . . ....
Ionization of DNA/RNA Bases and Nucleosides
Technique of Performing Ion-Molecule Reaction


in Q2


a a p a a *
p 9 4 P


Experimental
Chem icals . . . . .
Instrumentation . . . .
Procedures ........
Reactions of DNA/RNA Base and Nucleoside I
with Allyl Halides
Selection of Molecular Ions or Protonate
Molecules for Reactions .
Relative Reactivities of DNA/RNA Base
and Nucleoside Ions . .


d


4 9 9 9 p p p 9 P P p a a

* 9 4 9 9 a a p 9 P P P 9 a a a


>ns
* P P 9 9 4 9 9 4 9 9* 9 9 *


* P P P 9 P P P a a a p a 9 p *

* a a a a a a a a a a 4 S S
. p a a a a a a a a P P a


Correlation of Reactivities to Mutagenicities .
Identification and Correlation of Side Reaction Products
Reactions of Guanine Ions with Allyl Halides .. ..


a a a a p a 9 P .
* a 9 9 9 9 a a p
* a a p 4 a a a p a


E








Comparison of Reactivities of DNA/RNA Base and Nucleoside
Ions Towards Allyl Halides to Those of Pyridine Ions ..
C conclusions . . . . . . . . .


PRACTICAL ASPECTS OF REACTIONS OF DNA/RNA BASE
AND NUCLEOSIDE IONS WITH ALLYL HALIDES ......


Introduction . . . . . .
Pressure Effects of Allyl Halides on Formation of Product Ions


upon
Expe
Reac
Calibration
with
Expe


Reaction with DNA/RNA Base Ions


rimental ..
:tions of Uracil Ions v
Studies of Reactions
Allyl Halides . .
rimental


vith Allyl Halides
of DNA/RNA Base Ions


* a a a a a a a a


* a -
* a a 4


* a .* a a 4 4 S


Evaluation
Effect of U
Calibration
Conclusions


of Sensitivity Measurements . . .
racil N Ion Intensity on Product Ion Formation
Studies of Allyl Halides ..


DEVELOPMENT AND CHARACTERIZATION OF A SYSTEM
FOR TRAPPING IONS IN THE COLLISION CELL OF
A TRIPLE QUADRUPOLE MASS SPECTROMETER ........


Introduction . . . .
The Finnigan MAT TSQ70
Mass Spectrometer
Trapping Ions in the Collisi


Triple Quadrupole a a a a a a a a a a
Triple Quadrupole


Cell of the TSQ70


* a a a a a a a a a
* a a a a a a a


Experimental . a .
Experimental Conditions .... ......
Determination of Ion Trapping Efficiencies


TSQ70 Timing Studies
Methods for Trapping Ions in Q2
Precursory Techniques .....
External Control of the Exit Le
Enhanced Ion Detection
Helium Buffer Gas Effects ..... .
Pressure Studies
Ion Trapping Efficiencies....
Optimization of Ion Optics for Q2 Tra
Entrance Lens. L23 . .


* a a a a a a
* a a a a a a


ns (L31) for


* a a a a a a a a *.
* a a a a a a a a .
* a a a a a a a a a a a .
* a a a a a a a a a a a .
* a a a a a a S S 9 S .
* a a a .a a a a a.


a a a a a a a a a a a a a a' a a 4 4 4
* a a a a a a a a a a a a a a a
* a a a a a a a a a a a a a a a 4 4 4 4


S*ppi
pping


* a a a a a a a a 4 a a a a a a a 4 a
r . . . . .


Collision (Q2) Offset Potential.
T ,.1 +. ,-,-, 1'.. .i n Ir ta LV) T an*> n n cv








INVESTIGATIONS OF ION TRAPPING: FUNDAMENTAL
STUDIES AND APPLICATIONS FOR ENHANCEMENT
OF ASSOCIATIVE ION-MOLECULE REACTIONS


Introduction


Expi
Opt
Ion
Fill
Traj
Full
Rea


ire mental


-A *.-..-n frt.. a a a a a a a a .
imization of Mass Scanning Param
Accumulation in the Collision Cel
Time Studies . . . .
pping Time Studies
Scan Data while Trapping Ions in
actions of DNA/RNA Base Ions wi
under Ion Trapping Conditions


eters
1 .


*a* a a a a a a a a a a a a a a a
* a a a a a a a a at a a a a a a a a a
a a a a a a a a a a a a a a a a a a
. a a a a a a a a a a a a a a a a


i . .
th Allyl Chloride


Conclusions


CONCLUSIONS AND FUTURE WORK


Conclusions . . . .
Suggestions for Future Work .
Reactions of Other Nucleop
Comparison of Quadrupole
Radial Detection of Stored
Replacement of Q3 with a (
(Q-Q-ITMS) . .
Resonance Excitation in 02


bhile Ions .
and Octopo
Ions via RF
Quadrupole


le Collision Cells .
Ramping
Ion Trap


* a a a a a a a a a a a a a a a a a a a a a a a a
* a a a a a a a a a a a a a a a a a a a a a a a a


APPENDICES


TUNING GUIDELINES FOR LOW-ENERGY
ION-MOLECULE REACTIONS AND ION TRAPPING
IN THE COLLISION CELL OF THE TSQ70........


INSTRUMENT CONTROL LANGUAGE (ICL) PROCEDURES
USED TO FACILITATE TUNING FOR ION TRAPPING
IN THE COLLISION CELL OF THE TSQ70 ................

INSTRUMENT CONTROL LANGUAGE (ICL) PROCEDURES
USED TO PERFORM ION TRAPPING IN THE COLLISION
CELL OF THE TSQ70 . . . . . . . .


LITERATURE CITED .. ..

TT/1 r A TIT TT/ AT (TT-Fnp/-T T


~~-----













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

ASSOCIATIVE ION-MOLECULE REACTIONS AND ION TRAPPING


WITH A


TRIPLE QUADRUPOLE MASS SPECTROMETER


By

Anthony Peter Annacchino, Jr.


December 1993


Chairperson: Dr. Richard A.


Yost


Major Department: Chemistry


Gas-phase


reactions


ionized


DNA/RNA


bases


nucleosides


with


electrophilic


halides


collision


a triple


quadrupole


mass


spectrometer have been demonstrated.


Reactions of this type are important because


of their correspondence with the reactions of mutagenic electrophiles with biological


macromolecules in vivo.


Relative reactivities of the DNA/RNA base and nucleoside


ions toward the allyl halides were


determined based on


the production


of allyl-


nucleophile


adduct


ions,


Side


reaction


products


to charge


exchange


processes were also observed for many of the reagent combinations.


Significant


intensities of side reaction products were typically observed only when allyl iodide

was used as the electrophile; in many cases the abundance of side reaction products


1 *-A ..- -----l-- -- -_--. 1 .----_ .t A1_ -1 r.*


r







reactions of each DNA/RNA base ion with allyl halides were generally not linear;


limits of detection for the most reactive DNA/RNA base ion, uracil,


were on the


order of 10-100 femtomoles of allyl halide.

A system for trapping ions in the collision cell of a triple quadrupole mass


spectrometer has been developed and characterized.


Mass-selected ions are injected


collision


trapped


within


a variable


period


time.


Subsequently, the ions are gated out of Q2 and mass-analyzed by Q3.


Ion optical


parameters significant for effective ion trapping have been detailed and optimized.


Moreover,


importance


helium


buffer


to Q2


ion trapping


been


demonstrated via calculations of ion trapping efficiencies.


It has been determined


that buffer gas can enhance ion confinement and detection by over 2000%.

Utilizing the Q2 ion trapping system, applications have been demonstrated


that illustrate the power and utility of the technique.


It has been shown that large


numbers of ions can


stored


the collision


several


hours or


more.


Furthermore, it was demonstrated that Q2 ion trapping can be used to study ion-


molecule reactions.


In particular, the method was utilized to enhance formation of


product ions resulting from reactions of DNA/RNA base ions with allyl chloride.












CHAPTER 1
INTRODUCTION


this dissertation


studies


are presented


demonstrate


utility


associative ion-molecule reactions and ion trapping with a triple quadrupole mass


spectrometer (TQMS).


Gas-phase reactions of DNA/RNA base and nucleoside ions


with electrophilic allyl halides have been studied as an extension of previous work

[1] done in our laboratory involving ion-molecule reactions of various electrophiles


with small model bases such as pyridine and piperidine.


The studies presented here


have shown that, while the reactions are often complex, characteristic product ions


can signal the presence of potential carcinogens.


Furthermore, the method can be


used to roughly estimate the relative reactivities of these potential carcinogens. Also,


a technique for ion trapping in the collision cell of the


and characterized.


TQMS has been developed


The method allows execution of experiments that are normally


impractical, or impossible, to perform on standard tandem quadrupole instruments.

In particular, the technique is useful for the study of kinetically slow ion-molecule

reactions.

This introductory chapter begins with a review of various analytical techniques


that have been utilized for the determination of mutagenic compounds.


It continues


with a short discussion of ion-molecule reactions in the collision cell of the


TQMS.







2

quadrupoles is then presented and the general mechanism of the trapping method


is described.


The chapter concludes with an overview of the dissertation.


Methods for the Determination of Mutaaenic Compounds


The detection of adducts that result when biological macromolecules such as


DNA react with potentially hazardous


of great importance.


chemicals such as mutagenic electrophiles is


These adducts signal modification of DNA molecules and may


ultimately lead to carcinogenesis. Several methods have been utilized for the study

of modified DNA or model DNA molecules. Via determination of the potential that


electrophiles


have to form covalent bonds with small DNA models, various analytical


techniques have been utilized to study the potential mutagenicities of electrophiles


toward DNA [2-6].


Of these, methods involving mass spectrometric characterization


of DNA adducts have been particularly significant.


Fast atom bombardment (FAB)


[7-12], secondary ion mass spectrometry [13] and laser desorption [14] have all been


utilized to study modified DNA.


Continuous flow-fast atom


bombardment (CF-


FAB) mass spectrometry has been used to identify DNA adducts [15] as well as


water-soluble


hydrocarbon conjugates that indicate exposure


to carcinogens [16].


Mass spectrometry


has also been


employed


to detect DNA


adducts


formed via


offline electrochemical oxidation [17-19],


addition,


used to simulate metabolic activation.


gas-phase reactions of pyridine and guanine with polynuclear aromatic


hl-,urru- rhbnnc /PAI--^ hvsP hbiPn rnndrtprd in a maq ntprtrnmPtPr inn tnnre.p [701







3

Although many groups have clearly shown that analysis of modified DNA via

mass spectrometry is practical, these methods still require batch preparation of the


adducts in vitro.


This does not allow for rapid analyses and makes the simultaneous


determination of mutagenicities for several compounds impossible.


To alleviate


these


limitations a method


has been developed,


as is described in


following


section, that allows for the real-time determination of potential carcinogens via ion-


molecule reactions in the collision cell of a


TQMS.


Ion-Molecule Reactions in the Collision Cell of a


Triple Quadrupole Mass Spectrometer


Certainly, the most frequently utilized type of ion-molecule process that takes


place


in the collision


of a


triple


quadrupole


mass spectrometer


(TQMS)


collision-induced dissociation (CID).


The pioneering research of Yost et al. [21]


showed that CID can be accomplished with high efficiency in the collision cell of a


TQMS


spawned


development


several


commercially


available


triple


quadrupole mass spectrometers, including the Finnigan MAT TSQ70.


In turn, CID


has been utilized by many researchers [22-24] in analytical tandem mass spectrometry

(MS/MS) to fragment ions for mixture analysis or structure elucidation.


Although CID in the collision cell of a TQMS is a widely used technique,


study of other types of ion-molecule reactions performed in the collision cell has


become


increasingly common


as researchers


realize


potential


utility


-









quadrupole


(Q1)


reaction


with


neutral


cyclohexene


benzene


molecules


in the interquadrupole region.


Similar studies have been performed with


a triple quadrupole mass spectrometer [34-35] in which various reactant ions were

mass-selected by Q1 for subsequent charge exchange, proton transfer and hydride


abstraction


chemical


ionization


reactions


in the


collision


This


research


demonstrated


that control


of the


energetic of the


charge exchange and


proton


transfer processes could be achieved, facilitating the interpretation of the resulting


mass spectra.


Ion-molecule reactions in the collision cell have also been utilized to


synthesize covalently bonded adduct ions between protonated esters and ammonia


[36].


Reaction


between both the neutral and ionic species of vinylmethylether and


propene have been investigated with a triple quadrupole mass spectrometer [37] in

order to elucidate the mechanism of the reaction under CI conditions in the ion


source.


Furthermore, the dependence of the protonation of neutral ammonia on the


translational energy of protonated reactant ions has been studied in the gas phase


inside the collision cell of a


TQMS [38].


The ion-molecule reaction of the tolyl


cation with dimethyl ether has also been investigated using triple quadrupole mass

spectrometry [39].

In this dissertation, results from investigations of mass-selected reactions in

the collision cell between DNA/RNA base and nucleoside ions and allyl halides are


presented.


These studies allowed for the determination of the relative reactivities


of the nucleophilic base and nucleoside ions toward the allyl halides.


Furthermore,







5

presence frequently caused the measured reactivities of the nucleophiles to be much


lower than expected.


The ability to detect gas-phase ion-molecule reaction products


between


mutagenic/carcinogenic


electrophiles


simple


base


ions


(e.g.,


molecular ion of pyridine) which model DNA has been demonstrated previously in


our laboratory [1,40].


The technique is unique in that electrophilic mutagens may


be separated and characterized on a chromatographic time scale. Thus, the method

allows samples to be rapidly screened for the presence of mutagens. Furthermore,


the results of the investigations compared well to those for other methods used to

estimate mutagenicities, such as the Ames test [41-42].


Trapping Ions with Linear RF-only Multipole Devices


motivation


to study


trapping


collision


a triple


quadrupole mass spectrometer (TQMS) was partly due to the general popularity of


other types of ion


trapping apparatus.


Quadrupole


ion trap


mass spectrometers


(QITMS) and Fourier-transform ion-cyclotron resonance (FT-ICR) instruments are


the two most common types of ion trapping devices.


The quadrupole ion trap mass


spectrometer [43-45] is essentially the three-dimensional analogue of the quadrupole


mass filter [46-48].


Ions over a range of m/z values are trapped by appropriately


selected RF and DC voltages applied to the two endcap electrodes and a doughnut-


shaped ring electrode.


A mass spectrum is produced by ramping the RF potential


such


that ions of successively higher m/z values become


unstable and strike







6
operates on the principle that ions in a magnetic field move in circular orbits at


frequencies characteristic of their m/z


values.


Ion detection is achieved via image


currents that are generated in the receiver plates due to ions excited by a frequency-


swept excitation signal.


Because this signal contains all the excitation frequencies,


ions transmit a


complex


signal that contains


frequency components


characteristic


ions.


resulting


time-dependent


image


current


subjected to Fourier transformation, which produces the data for the mass spectrum.


trapping


collision


(Q2)


a triple


quadrupole


mass


spectrometer was first demonstrated by


Yost [51] in 1982.


It was proposed that the


trapping


ions


in the


center


quadrupole


would


provide


a way


to study


energetic of the CID process, as well as provide enhanced fragmentation patterns


for particularly stable ions.


Since then, only a few research groups have performed


trapping experiments with linear


RF-only devices.


Beaugrand et al.


[52-53]


performed the first definitive work utilizing Q2 ion trapping for studies of collision-


induced dissociation


, decomposition of metastable ions, and ion-molecule reactions.


It was shown that ion storage time can dramatically affect the relative abundances


of the resulting product ions for each of these processes.


Further investigations by


Beaugrand et al.


[54-5


] demonstrated the utility of the technique for the study of


the kinetic and, in some cases,


thermodynamic parameters of gas-phase reactions


resulting from the interaction of ammonium ions with a mixture of piperidine and


pyrrolidine.


Similar instrumentation and procedures required for Q2 ion trapping









been


utilized


by this group


to study


acetone


dimerization


protonation


reactions as well as ion-molecule reactions between protonated acetaldehyde and

methanol [57].


Ions


are trapped


a linear


RF-only


quadrupole


octopole)


combination of two mechanisms.


Ions are confined in two dimensions by the RF


potential applied to the rods of the quadrupole, as defined by the stability diagram


(Figure 1-1).


As is shown in Figure 1-1, ion trajectories within a quadrupole device


can be either stable or unstable depending on the values of the Mathieu parameters


q, and as,


which are related to the applied RF and DC potentials, respectively.


values of the parameters are defined by


4V e
"RF
mr 2
mr -w


(1-1)


DC potential,


e is the charge of the ion, m


is the mass of the ion, and r, is the


inscribed radius of the quadrupole.


Ions that have Mathieu parameters within the


triangular


transmitted


envelope


through


stability


device.


diagram


When


have


operated


stable


as a mass


trajectories


filter,


notentials are annlied to the rods (in addition to a suoulemental DC voltage, the


mr 2o
0


where Vn is the zero-to-peak RF voltage with an angular frequency of w, VDc is the

















0.26

0.24

0.22

0.20

0.18

0.16

0.14


0.12


Scan


Line


0.10


0.08

0.06

0.04

0.02

0.00


Figure


Stability


Region


I

I
I


Low-mass
Cutoff


0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1


Stability diagram for the quadrupole mass filter.









masses in a narrow band are stable at one time.


This is achieved by operating the


device over a scan line that intersects the apex of the stability diagram that defines


the RF/DC ratio.


This ratio, in turn,


determines the range of masses that are stable.


By continually increasing the


and DC potentials linearly with


time (but at a


constant RF/DC ratio), ions of increasing m/z are brought within the window of

stability and a mass scan is obtained.

Operation of the quadrupole mass filter as an RF-only device (such as the


collision cell in a


(other than


TQMS)


is a much simpler proposition since no DC


the offset voltage) are applied to the quadrupole


rods.


]


potentials

Looking at


Figure


, this means that ions are stable all along the x-axis up until a value of


q=0.908 where


low masses become


unstable (see


Equation


1-1).


This point is


termed the low-mass cutoff of the device.


By carefully selecting an appropriate RF


potential,


however,


a wide


range


masses


can


stable


within


an RF-only


quadrupole, making it suitable for applications such as a collision cell.


It should be noted that the collision cell of the


presented here is an RF-only octopole.


TSQ70 used for the studies


Nonetheless, in general the basic operating


principles of the RF-only quadrupole presented above are also applicable to the RF-


only octopole.


Even though a stability diagram cannot be explicitly defined for an


octopole


mass


filter


to nonlinear and


coupled


ion motions [58],


a stability


envelope still exists, as does the low-mass cutoff.


This, in effect, allows the RF-only


octoDole to be operated similarly to an RF-onlv cuadrupole.







10

While the time-varying electric field produced by the RF potential confines


radially


in two


dimensions,


potentials


applied


to the


lenses


immediately before and after the collision cell can be employed to trap ions axially


(in the third dimension).


In this way, Q2 ion trapping can be likened to the trapping


mechanism


a Fourier-transform


ion-cyclotron


resonance


(FT-ICR)


mass


spectrometer [49-50] where ions are confined magnetically in two dimensions while

DC potentials applied to the trapping plates trap ions in the third dimension.


Organization of Dissertation


This dissertation is organized into six individual chapters.


Chapter


1 has


presented


general


overview


topics


to be


discussed


bulk


of the


dissertation and has also provided an overall perspective of the research.


To begin


with,


various methods used


the determination of mutagenic compounds were


described,


focusing


primarily


on mass


spectrometric


techniques.


Additionally,


previous studies involving ion-molecule reactions in the collision cell were discussed.

Furthermore, a discussion of ion trapping in the collision cell of a triple quadrupole

mass spectrometer was presented along with a review of previous ion trapping studies

performed using linear quadrupoles.

Chapter 2 presents research involving reactions of nucleophilic DNA/RNA

base and nucleoside ions with electrophilic allyl halides in the collision cell of the


triple auadrunole mass spectrometer (TOMS).


A brief description of the technique









of the chapter focuses on characterization of the reaction products that are observed


for the electrophile/nucleophile reactions.


In addition, the relative reactivities of the


DNA/RNA base and nucleoside ions are reported and compared.


These relative


reactivities can provide a rough estimate for the mutagenicities of the allyl halides,


even though side reactions


often predominate for these processes and subsequently


render the relative reactivity measurements inaccurate.

Chapter 3 describes several practical aspects that must be considered when

studying collision cell reactions of DNA/RNA base and nucleoside ions with allyl


halides.


In particular, the effects of allyl halide pressure on reaction products are


described and characterized.


Also, calibration studies of reactions


of DNA/RNA


base


ions


with


halides are


presented.


Some


nucleophiles are


much


more


sensitive


(due


to higher reactivities)


towards allyl


halides


than


others and


thus


provide lower limits of detection.

Chapter 4 describes the development and characterization of a system for


trapping ions


in the collision cell of the TSQ70 triple quadrupole mass spectrometer.


The unique ion optical design of the


is discussed.


TSQ70, particularly the octopole collision cell,


The mechanism of the Q2 ion trapping technique is described, and key


ion optical parameters that are important for efficient Q2 ion trapping are noted.

Furthermore, studies are presented that show that helium buffer gas greatly improves


Q2 ion trapping performance.


Also, the various methods that were utilized during


the deve.lonment of the 02 ion


tranoin, system are discussed in detail.


Finally,









trapping experiments and tune the instrument for Q2 ion trapping are described,

with details provided in Appendix B.


Chapter


reports several


investigations that utilized the Q2 ion


trapping


system described in Chapter 4.


The section begins with a discussion of the mass


scanning parameters


(scan


time,


scan


range,


etc.)


required


when


acquiring


trapping dat

presented.


ha.


Several examples of ion accumulation in the collision cell are then


If precautions are not taken, ions can amass in Q2 from cycle to cycle


and cause erroneous results for ion-molecule reaction studies.


Examples are then


shown


indicate


maximum


number


ions


collision


accommodate.


Also, investigations concerning ion


confinement times are discussed.


trapping efficiencies for long


A method to obtain full scan data


under ion


trapping conditions


is demonstrated,


the accompanying ICL procedures are


described.


The chapter concludes by evaluating several ion-molecule reactions using


ion trapping


system.


particular,


enhanced product


ion formation


demonstrated for reactions of DNA/RNA base ions with ally


chloride.


Chapter 6 draws general conclusions about the


research


presented in


dissertation.


In addition, several ideas for future studies are suggested.


can












CHAPTER 2
REACTIONS OF DNA/RNA BASE AND
NUCLEOSIDE IONS WITH ALLYL HALIDES


This chapter describes studies of ion-molecule reactions in the collision cell

of a triple quadrupole mass spectrometer to evaluate the reactivities of a series of


mutagenic allyl halides toward DNA/RNA base and nucleoside ions.


In performing


these studies, the method initially developed with small model nucleophiles


1,40] has


been


substantiated


use of


actual


DNA/RNA


bases


nucleosides.


Additionally, performing these ion-molecule reactions for four DNA/RNA bases and


their


corresponding nucleosides


has allowed


determination


their


relative


reactivities toward the allyl halides. AI

systematically identified and correlated.


[so, predominant side reactions have been

As these reactions were often extensive,


their occurrence frequently caused the measured reactivities of the nucleophiles to


be much lower than expected.


Thus, the ability to understand and predict these side


reactions is important for the results to be meaningfully interpreted.


Introduction


Ionization of DNA/RNA Bases and Nucleosides


Previous studies


1] have demonstrated that, in contrast to the molecular ions









[N+H]+


, generally have little or no reactivity toward these mutagenic allyl halides.


This


reactivity


reflects


protonation


(most)


basic


site of


nucleophile, reducing or eliminating the neutral's


nucleophilicity.


Whereas the DNA


and RNA bases, with the exception of guanine, are thermally stable and volatile

enough to produce N' ions by EI, the nucleosides are thermally labile, producing


primarily


base


to loss of the


ribose


moiety.


Commonly used


ionization techniques such as chemical ionization (CI) [59], fast-atom bombardment


(FAB)


[60-62],


electrospray


(ESI)


[63-65


matrix-assisted


laser


desorption


ionization (MALDI) [66-68] produce abundant protonated molecules, [N+H]+


well as adducts such as [N+Na]+


and [N+K]+


, and are thus inappropriate for these


studies.


Because


problems


with


ionization


methods


described


above,


benzene charge exchange [69-71] ionization was utilized to produce nucleoside N


ions.


Chemical ionization via charge exchange can be a relatively soft ionization


method if an appropriate reagent is selected [34-35,72].


Typically, charge exchange


ionization will occur if the recombination energy (RE) of the reactant ion,


greater than the ionization energy (IE) of the sample molecule, N.


The degree of


fragmentation is determined by the difference between the RE of the reactant ion


and the IE of the sample molecule.


As indicated in


Table


1, when performing


charge exchange using benzene reagent gas very little energy is transferred to the


sample molecules


because of the relatively low recombination energy of C6H6+


ions














Table 2-1:


Relevant properties of compounds used in these studies.


Compound


Ionization Energy,
eVab


Proton Affinity,
kcal/mola


Gas Phase
Basicity,
kcal/mola


Benzene


Allyl chloride
Allyl bromide
Allyl iodide


181.3


9.90
10.06
9.30


220.8


Pyridine
Adenine


213.1
215.7


223.8


Cytosine
Thymine


Uracil


208.8


208.0


216.0


201.0


200.0


Adenosine
Cytidine
Thymidine


Uridine


208.0


208.0


a Values taken from reference 73.
b Recombination energies for the associated molecular ions are approximately equal
in magnitude to the ionization energies of the neutral molecules.
0 NF = no literature reference was found.









IEurjijije


=9.00


eV).


This


minimizes


fragmentation


provides


maximum


intensity of molecular ion (N+*) for subsequent reactions in the collision cell.


Technique of Performing Ion-Molecule Reactions in Q2


technique


performing such


ion-molecule


reactions


second


quadrupole (collision cell, Q2) of a triple quadrupole mass spectrometer has been


described previously [40].


Also, the concepts and instrumentation required for GC


introduction into the collision cell were reported [35].


A brief discussion of the


method


offered


here.


technique


involves


ionization


a nucleophile


(DNA/RNA base or nucleoside) and mass-selection of nucleophile ions (normally


molecular ions, N+*) with the first quadrupole (Q1).


These ions then pass into the


collision


at low


energies


subsequently


react


volatile


electrophiles (the allyl halides) eluting from a gas chromatograph (GC) into the


collision cell (Q2).


The third quadrupole (Q3) then scans over a mass range such


that any product or fragment ions that may result are detected.


The resultant mass


chromatograms


thus


identify


product


with


values


correspond


formation of particular adduct ions.


Figure 2-1 illustrates a typical ion-molecule


reaction


involving


a DNA


base


an allyl


halide.


this case,


radical


molecular ions (N+) of the nucleophile cytosine are mass selected to react in the


collision cell with neutral allyl chloride to yield product ions of m/z 152.


The product


ions are formed by nucleophilic substitution of the allyl chloride and loss of the


with.
















'-N


ZI


ZI









Experimental


Chemicals


Nucleophiles


investigated


included


pyridine


DNA/RNA


bases


(adenine, cytosine, guanine, thymine and uracil) and nucleosides (adenosine, cytidine,

thymidine and uridine), which were purchased from Sigma Chemical Company (St.


Louis, MO).


The structures of these bases and nucleosides are shown in Figures 2-2


and 2-3, respectively.


Allyl bromide,


allyl chloride and allyl iodide were acquired


from Aldrich Chemical Company (Milwaukee,


WI).


HPLC grade pentane solvent


benzene,


utilized


charge


exchange


ionization


nucleosides,


were


obtained from Fischer Scientific (Fairlawn, NJ).


High purity helium carrier gas and


nitrogen collision gas were purchased from Liquid Air Corporation (Walnut Creek,


CA).


Argon was acquired from Matheson Gas Products (East Rutherford,


NJ).


Instrumentation


All experiments were performed on a Finnigan MAT (San Jose, CA) TSQ70

triple quadrupole mass spectrometer modified with an octopole collision cell and a


20 kV dynode.


The mass spectrometer was equipped with a Varian (Walnut Creek,


CA) 3400 gas chromatograph,


interfaced to the collision cell, Q2,


via a resistively


heated transfer line [35].


Experiments were performed at an emission current of 200







NH2


NH2

N\N


Adenine


Cytosine


HN


HN


CH3


H2N


Guanine


Thymine


HN


Uracil


N














HO-CH2

H


NH2
A N.

N


HO-CH2


NH2


Adenosine


Cytidine


H2C


HO--CH2


HO-CH2


OH H


Thymidine


Figure 2-3:


Uridine


Nucleosides used in these studies.







21

mass spectrometer was mass-calibrated using perfluorotributylamine (PFTBA) and

then tuned to maximize product ion formation for low-energy ion-molecule reactions


in the collision cell.


The details involved in these tuning optimizations are presented


in Appendix A.


For all studies, gas chromatography was carried out on a J &


W Scientific


DB-


(17 m long, 0.178 mm i.d.,


mode (split ratio


0.4 pIm film thickness) capillary column in the split


= 100:1) with helium carrier gas at an inlet pressure of


GC oven was temperature programmed from


O0C to


psig.


100C at 20C/min with an


injection port temperature of 200C and a transfer line temperature of 200C.


One


microliter


amounts of


injections


allyl


were


chloride


made


, allyl


in triplicate


bromide


mixtures


iodide


containing


in pentane


equimolar


ranging in


concentrations from 40 pmol/pL to 3 pmol/pL.


Procedures


Bases and nucleosides contained in capped aluminum vials were introduced


into the ion source via a heated solids probe.


The solids probe temperature was


precisely adjusted in each case to give a steady ion current of constant intensity.


DNA/RNA bases were ionized by electron ionization (EI), and the nucleosides were

ionized under CI conditions using benzene charge exchange at an ion source pressure


approximately


mTorr,


indicated


Granville


Phillips


(Boulder,


CO)


Convectron gauge.


Investigations to determine the relative reactivities of guanine


V







22
and constant, ion intensities needed for reaction with the allyl halides as they elute


from the GC.


At the solids probe temperatures required for vaporization of guanine


and guanosine (


> 300oC),


sample decomposition readily occurs, thus the quantities


of molecular ion needed for the experiments were not attainable.


a direct exposure probe (DEP),


However, using


it was possible to produce brief intensities of guanine


molecular


ions


could


then


react


separately


with


each


allyl


halides


introduced into the collision cell via a leak valve.

Reference electron ionization (EI) spectra were taken of all the DNA/RNA


bases and nucleosides used in the study.


Also,


reference spectra using benzene


charge exchange ionization were acquired for the nucleosides.

spectra were obtained for the molecular ions (either N+ or [p


MS/MS daughter ion


N+H]+) of each of the


DNA/RNA bases and nucleosides in order to help decipher the complicated product


ion spectra.


an attempt


to predict what


charge exchange


products


may


observed, daughter ion spectra were taken when allowing mass selected argon ions


(I.E.


= 15.76 eV) to react with the allyl halides in the collision cell.


For all of these


studies, instrumental conditions, including collision energy, collision gas pressure and

reagent gas pressure, were optimized for maximum sensitivity of the ions of interest.


Reactions of DNA/RNA Base and Nucleoside Ions with Allvl Halides


Selection of Molecular Ions or Protonated Molecules for Reactions









been demonstrated [1] that, while the [N+H]+


ion of piperidine is reactive toward


the allyl halides, the protonated molecule of pyridine is not.


except reactions performed with cytidine, N

subsequent reactions with allyl halides in Q2

higher intensity of [N+H]+ than N+ under be


Therefore, in all cases


molecular ions were mass-selected for

. Cytidine, however, produced a much


enzene charge exchange conditions, and


only the [N+H]


ions were reactive toward the allyl halides.


the only nucleoside that produced a reactive [N+H]


species.


Indeed, cytidine was

The reason for the


unusual reactivity of cytidine [N+H] ions is unclear, but it may be that cytidine has

a higher gas-phase basicity (or proton affinity) than any of the other nucleosides.

Values of gas-phase basicities for the nucleosides are not recorded in the literature;

but to the extent that the basicities of the corresponding DNA/RNA bases are an

indication of the basicities of the related nucleosides, then this theory may be correct.


Table


1 shows that cytosine, which also produces a reactive [N+H]


species, has


the highest gas-phase basicity (and proton affinity) of the DNA/RNA bases used in


this study.


This suggests that cytidine would also have a proportionately large gas-


phase basicity compared to the other nucleosides and would explain why cytidine

forms almost exclusively [N+H]+ ions under benzene charge exchange ionization.

Cytidine [N+HJ] ions retain reactivity toward the electrophilic allyl halides


because there are three potentially basic amine sites on the neutral molecule.


with one site protonated to form [N+H]


Even


, the ion evidently still is basic enough to


react with the allyl


halides.


Given that some basicity remains when the cytidine









reactive [N+H]


species as well, since they too each have more than one basic site.


One way to roughly estimate the basicities of the potentially reactive sites on each

DNA/RNA base or nucleoside is to examine their acid dissociation constants (pKIt,


pKa2, and pK3) in aqueous solution, as listed in


Table


Although none of the


DNA/RNA bases or nucleosides are basic enough to exist as diprotonated ions in

solution, the pKi values for the association of one proton can provide some insight


into the relative basicities of the free protonated molecules.


It is clear that cytidine


is the most basic of the nucleosides in solution and that only cytidine and adenosine


exist primarily as protonated molecules at low pH.


Furthermore, of the DNA/RNA


bases, adenine and cytosine are the most basic.


This suggests that the [N+H]


ions


of adenine and adenosine, as well as those of cytosine and cytidine, might be reactive


toward the allyl halides.


Indeed, experiments have shown that adenine [N+H] ions


do react slightly with


iodide


to produce


[N+H+41]


species,


although


protonated molecules of adenosine do not show any reactivity at all.


The lack of


reactivity


of the adenosine [N+H]


species


to the


intrinsically low


reactivity of adenosine,


in general, as is discussed below.


Moreover, the somewhat


lower


value


adenosine


may


indicate


protonated


molecule


adenosine is less basic, and less reactive, than that of cytidine.


It is surprising that,


unlike the [N+H]


species, cytidine


ions are not


reactive toward the allyl halides.


reactive than [N+H]


One would expect N+


+ ions since, presumably, the N+


ions to always be more


species is more basic than the


pI~,



















Table 2-2:


Acid dissociation constants for DNA/RNA bases and nucleosides.


Base or Nucleoside (B)


pKaI


PK 2


pKa3


Adenine


Cytosine
Thymine


> 13.0


Uracil


Adenosine
Cytidine
Thymidine


>13.0


Uridine


a Values taken from reference 74.


b For the reaction BH+


0 For the reaction B [B-H]


For the reaction [B-H]


, [B-2H]2-


e NF = no literature reference was found.


~B+


+H+


+H+










a reflection of cytidine's


very high proton affinity; the N


ion may preferentially add


a hydrogen atom from allyl halide to form [N+H] instead of adding the allyl group.

In other words, the side reaction that forms [N+H] may be so favorable that P+


([N+41]) is not observed. This argument is supported by the fact that cytidine

produces much more [N+H]+ than N+ when ionized under benzene charge exchange


conditions, while the other nucleosides readily form N'


ions.


Relative Reactivities of DNA/RNA Base and Nucleoside Ions


Results of a typical experiment are shown in Figures 2-4 and


In this case,


uracil N


ions (m/z


112) were allowed to react with allyl halides (


.6 nmol each)


eluting


from


a GC


column


Upon


inspection


mass


chromatogram in Figure


it is clear that product


ions,


, corresponding to


[N+41]


ions were easily detected as the three allyl halides eluted from the GC.


Relative reactivities of the individual DNA/RNA base and nucleoside ions toward

the allyl halides were calculated using peak areas from mass chromatograms such as


those


in Figure


Looking


at Figure


negative


peaks


chromatogram correspond to elution of the three allyl halides to form m/z


large negative peak at a retention time of 1.1 minutes coincides with the elution of


n-pentane solvent,


which


causes extensive


ion scattering.


Due


to probable


column overload, some allyl halide also elutes with the solvent and gives rise to a


Ir C1,, t2 s


,,,,,,,,,1:,,,,,1, L1. nu II


~n _____ ~___1_



















m/z 112, Uracil N+. Ion


m/z 153, [N+41]+ Product Ion


Allyl Chloride
\


Allyl Bromide


Allyl Iodide


0:30 1:00 1:30 2:00


retention time (min:s)


Figure


Mass chromatograms for uracil N+


ions (m/z 112) reacting with allyl


halides in the collision cell to form product ions


of m/z















Allyl Chloride






69
I


50 100 150 200


Allyl Bromide






69


50 100 150 200


Allyl Iodide


[E+(E-I)] +


Figure


Product ion spectra (corresponding to the retention times of the three


allul hnaliAc in P ari 39_A\ n nro il 1LT+- ,- aI .-


1 1 \ 1ra,+,n rr t.n +lt







29

on the allyl bromide and allyl iodide peaks can also be attributed to column overload

and poor chromatography.

Figure 2-5 shows the spectra derived from the chromatograms in Figure 2-4.


Each spectrum is an average of several scans over the


tops of


each allyl


halide


chromatographic peak and is not background subtracted.


In all three spectra, it is


clear that the expected product ion,


(m/z


153),


is observed as well


as some


unreacted N ion at m/z 112 and a low abundance of ion at m/z 69 arising from CID


of the N" ion (loss of CONH).


In the case of allyl iodide, a number of products of


other reactions are also observed, as discussed below.

Figures 2-6 and 2-7 show the calculated relative reactivities for the DNA/RNA


base


nucleoside


utilized


these


studies.


Relative


reactivities


were


calculated by first


determining the P+


areas (for at least three individual trials for


each reaction) from the appropriate mass chromatograms.

normalized for the flux of N+ ions passing through Q1


These areas were then


into the collision cell by


dividing by the average (over several scans) N intensities immediately before the

pentane solvent eluted (preceding any allyl halide elution) from the GC into the


collision cell.


The different concentrations of the allyl halide solutions were then


accounted for by dividing by the individual solution concentrations (in mol/L) that


were used for each particular experiment.


For example, for one trial of the reaction


of uracil N ions with allyl bromide, the GC peak area of 1533939 counts for the m/z


product ion is divided by the average incident N


ion intensity of


4789373
















































Adenine


Cytosine


Thymine


Uracil


Figure 2-6:


Relative reactivities of DNA/RNA base ions with allyl halides.


Allyl CI

Allyl Br

Allyl I


I


~ma\d~







































Adenosine Cytidine Thymidine Uridine


Figure 2-7:


Relative reactivities of nucleoside ions with allyl halides.









bromide


solution


injected


(0.558


mol/L)


yields


relative


reactivity


0.574.


Averaging this value with the results calculated for the other trials generates the


relative reactivity shown in Figure 2-6 of 0.601.


No corrections for the amount of


allyl halide introduced on the column were necessary since


pL injections were


always made with the same split ratio of 100:1.


was


previously


stated,


relative


reactivities with


halides were


calculated by measuring the production of one specific product ion, [N+41]+


. In


some instances, however, it may make more sense to include several product ions to

evaluate nucleophile reactivities, particularly those that contain both the nucleophile


and electrophile groups or fragments there of

studies, several other product ions listed in


. Although this was not done in these

Tables 2-3 and 2-4, such as [N+X]+


(X=halide)


[BH+(E-X)]+


Table


could


have


been


included


calculations of relative reactivities.


The goal of this work, however, was to evaluate


the reactivities of these nucleophiles with allyl halides based on production of the


[N+41]I


exclusively.


Thus,


no other product ions were considered for the


determination of reactivities.


Comparing Figures 2-6 and


consistently less reactive (by a factor of


it is evident that the nucleoside ions are


10, with the sole exception of uracil/uridine


with allyl iodide) towards the allyl halides than the corresponding DNA/RNA base


ions.


In large part,


major reason for the


lower apparent reactivities of the


nucleoside N


ions is their lower stability, and therefore their greater tendency to








33




4-1 ,_

5~ 0
S. S I

tii nl~li
o6
3U o
rd 7 O
+ *
rQ a ^




2o M ao

0+0
I IV V

UO
VIV
*









(OO
In Cl en














'0 0 0
5 0 v tim
o + v en

Z
Cl c














'04
O
'Ct
-a 3 a -o -


+ I V h *@
SN re r) r oO O
In an In 0o tir



2 '-4 __ _- ^ ^^ ^ __^^ ^^ ^ ^- ^
ha d -1h h~O
Qv^ -* en A-' ^ -* 'a' A-'<: ^ N
*s 6 0 00- eN "1 3
+ C ON v 0 0 C 0 In 6 6
. '- '. 0 ^ ^ rs. Cl Ci N N e 00 00
CS No N t-im in N \0 '.0 tim in 002
oE \ i-r^,_<,i v, ^ soc(^,- ^ ^
U00
I- LL\O~ rl
rl rl r
^ ^ O fl M '- Ol Ffl O fl O fl


-^^ -^^ 1^^ -^- 11 -' -
0^ '^^ -*******- -4k -^ -. -^ -^ ^ -N^ --- -^ -^^ -H ^^h
s;

(" E3O~ U f s ou c -
r1 -: -1 --l ---





0) *3 4 C


a 9, .r (1 O ?.^
00




ct uIT3 X [ I ) rd
1')

















































-cl
'- .


+

2
6 +


t+Z


U2 c
+~


-- I rrs -c ..- -r























'-CU
0.*.
a) c


O~\o
czlo~


o\ 0o 0(
























































Vt
bi


hrd
t: o
Z 6
O
I~ CM
Q C3


df -m


5;17
S
a
u









(1-2 eV) employed here.


This competing reaction, in effect, reduces the number of


intact molecular ions that are available for reactions and decreases the number of


P+ ions ([N+41]) that are ultimately formed.


As is discussed below, other ion-


molecule reactions involving the nucleoside fragment ions


do occur, but were not


included in the calculation of relative reactivities.


Correlation of Reactivities to Mutagenicities


The relative reactivities of each individual DNA/RNA base or nucleoside ion


can, in some cases, roughly approximate the mutagenicity of each allyl halide.


Table


lists the mutagenicities of the three allyl halides used in this study, as determined


Ames


[41-42],


a bacteriological


assay.


Allyl


chloride


is much


mutagenic than allyl bromide,


allyl iodide.


and allyl bromide is substantially less mutagenic than


Since the mechanism of mutagenicity for electrophiles such as the allyl


halides is direct modification of nucleophilic sites on DNA [7


76], one would expect


to observe increasing relative reactivities, from allyl chloride to allyl iodide, upon

reaction with DNA/RNA base and nucleoside ions, as has been previously observed


upon reaction with pyridine ions [40].


Figures 2-6 and


7 show that, for reactions


of six of the eight DNA/RNA base and nucleoside ions, this is what is observed.

However, for the reactions of cytosine and uracil ions, the quantity of product ion,

P+, formed upon reaction with allyl iodide is less than that which would be expected.

In fart it ic Pven Inwer than the nrndnr.t ion intensity obtained in the reaction with
























Table 2-5:


Mutagenicities of allyl halides as determined by the Ames


Test.


Allyl Halide
Allyl chloride
Allyl bromide


Allyl iodide


a Values were taken from reference 41.


Mutagenicitya


2000


Mutagenicity in rev/pmol without metabolic


activation.







39

This phenomenon is apparent in the relatively small chromatographic peak observed


for allyl iodide in the m/z


153 mass chromatogram in Figure 2-4 from reaction with


uracil N+


ions.


This apparent contradiction is explained by noting (Figure 2-5c) that


extensive side reactions occur when uracil N+


ions react with allyl iodide.


case, among other ions, allyl iodide molecular ions (E+


, m/z


In this


168) and dimer ions


([E2-I]+


, m/z 209 and [E2-I-HI]+


81) are produced along with the expected P+


ions.


Products of such side reactions are often observed in the product ion spectra;


frequently the most intense ion is a product ion other than the expected P+


Identification and Correlation of Side Reaction Products


It is immediately apparent upon inspection of Tables 2-3, 2-4, 2-6 and 2-7 that


number and variety of side


reactions that take place


for these


ion-molecule


reactions is significant.


However, many product ions that may initially appear to be


unrelated can


be grouped into different categories,


with analogs for each of the


reactions for the different DNA/RNA base and nucleoside ions.


A large fraction of


the side reaction products arise directly from the allyl halides (Table 2-6), including


, allyl halide ions, E+


, allyl dimer ions, [E2-HX-X]+


and halide-


bound ally


dimers, [E2-X]+


These ions arise from charge exchange reactions in


competition with the desired associative ion-molecule reactions to form E+


ions and


[E-X]+


(allyl)


ions


(from


dissociative


charge


exchange


collision-induced


dissociation (CID) of E+


or allyl-containing adduct ions).


Further reactions of


allyl ions, [E-X]f










yU e~
1~ I o
El ~3
u ~
P1 E
cl S
Q1 ~I
s! w
O w


f














pOc~o





+~b
I


+~U

















LI












Cu
+~


z

0


sCO
\0'"'
c~i













Ca~r




0'0


"0


c~O
fr0
- (a'
cto


r lI


=1i~
ld~
el FI
otd
Irl II


r


P)
d
o err


r






















-O -



ON


en-



6I
4-





00
eN






'C\




Cl


F-'










CC
00FS
od


'-C


'CIS
'4-


cu ci


N3


C1 'C
-r


Co
Q\r4


-r


U
cu
-1
p
0
_:C3



S'ij
OS

**
SC/3



*M
S-

0 *4
a -


1 2


+W '4-

zQ
^*^


~S


0\


4,


\O
Q\
CC)









The greater intensity and variety of these ions with allyl iodide are due to it'


much


lower ionization energy compared to the other allyl halides, as shown in Table 2-1.

Similarly, the DNA/RNA bases with the highest ionization energies (and thus the


base


ions with


highest recombination energies) result in


most abundant


electrophile ions.


Figure


8, for example, illustrates the charge exchange reaction


that occurs extensively between uracil N


Figure


ions and allyl iodide.


6 shows that, upon reaction with the most reactive of the base and


nucleoside


ions,


uracil


cytosme,


relative


phase


reactivities


bromide and allyl iodide are opposite to those that would be expected.


This may


point to another reason, in addition


to excessive side


reactions, for decreases in


observed


product


ion formation


upon


reaction


with


allyl


iodide.


Simply


reactivity can't increase ad infinitum.


For reactions involving uracil and cytosine ions,


the nucleophile is certainly highly reactive with both allyl bromide and allyl iodide.

However, since it is apparent that side reactions are prominent when allyl iodide is


used as the electrophile,


these additional products may decrease [N+41]+ formation


by using up some of the nucleophile ions that would otherwise be available.

In addition to associative ion-molecule reactions, collision-induced dissociation

(CID) of the DNA/RNA base or nucleoside ions may also take place when the allyl


halides elute from the GC into the collision cell.


That this process occurs is evident


upon inspection of Table 2-7


. In all cases except the reaction of adenosine N


ions


with allvl bromide or iodide,


the base peak is the ion that was mass-selected by the


)




















ZI







47

nucleophile ions, significant CID occurs due to the presence of allyl halide neutrals


CID product ions are


more


prominent in


daughter spectra


of the


nucleoside ions than of the corresponding DNA/RNA base ions.


This difference in


extent of CID is due


to the


intrinsic stability of the base


ions compared


to the


nucleoside ions.

Daughter ions arising from CID of the nucleoside molecular ions include the


aglycone, or base, fragment ions [B+H]+


and [B+2HJ]


These ions are formed by


cleavage of the glycosidic bonds of the nucleoside ions with transfer of one or two


hydrogens and are analogous to the corresponding DNA/RNA base ions, N


[N+H]+


It is not surprising,


then,


to find that many of these


nucleoside CID


product ions go on to react with the allyl halides in the collision cell to produce


either [BH+ (E-X)]+


or [BH2+(E-X)]


ions, as shown in Table 2-4.


In the case of


uridine,


[B+H]+


fragment goes on


to react with


allyl


iodide


to produce


[BH+ (E-X)]+ ion (m/z


153) with a relative abundance 2/3 as large as that of the P+


([N+41]+) ion.


Similarly, the uridine N+


ion loses H20 to form a daughter at m/z


226 which then goes on to react with all three allyl halides via addition of allyl to


form m/z


. Reaction of adenine molecular ions with the allyl halides yields the


largest


number


product


ions


to reactions


CID


product


nucleoside, due to the significant extent of CID of the adenosine N" ion (Table 2-4).

In the case of the reaction with allyl iodide, no less than nine significant product ions


can be attributed to these tvoes of processes.


The most abundant of these, m/z 364,









[N-30]+


at m/z


of adenosine.


The [BH+(E-X)]


ion is also observed for the


reaction of adenosine ions with allyl iodide,


albeit with much less intensity than for


the corresponding uridine reactions.


previous


discussion


implies


formation


ions


such


[BH+(E-X)]J


always occurs stepwise,


via CID of the


nucleoside


ion to form


[B+H] or [B+2H]+ species which subsequently reacts to form the adduct.


It should


be pointed out, however, that it is probably just as likely that a single collision yields


an [N+E] adduct which can then fragment not just by loss of X to form P+


but also


by a coinciding loss of the sugar moiety (minus one or two hydrogens) to form the


[BH+(E-X)]+


or [BH2+(E-X)]


species.


In summary, the lower reactivities of the nucleoside ions compared to those

of the DNA/RNA bases (see Figures 2-6 and 2-7) reflect the greater ease with which


nucleoside


ions can


undergo CID.


Although


the CID products of the


nucleoside N+


ions can react with the allyl halides to form side reaction products


(Table 2-4), the decrease in the number of N+


ions remaining to undergo charge


exchange with the allyl halides reduces the number of charge exchange side products


observed


(Table


typically


most


abundant


CID


product


ions


nucleosides,


[B+2H]+


(corresponding


to the


protonated


DNA/RNA


base)


unlikely to charge exchange with the allyl halides.


Although ionization energies have


not been reported for three of the four nucleosides, they are probably all lower than


those for the corresponding bases, as noted in Table


1 for uridine (with an IE 0.2









eV less than uracil).


This may also account for the fewer charge exchange side


reaction products observed for the nucleosides (Table 2-6).


Reactions of Guanine Ions with Allyl Halides


Because


guanme


very


low volatility,


it is impractical


to achieve


constant flux of molecular ions required for Q2 ion-molecule reactions with allyl


halide introduction via gas chromatography.


Furthermore,


when using the solids


probe, temperatures in excess of 300C are required for vaporization of guanine, and


this leads to sample decomposition and reduced molecular ion intensities.


Also, a


dramatic decrease in sensitivity (over several scans) was noted when solids probe


vaporization of guanine was attempted.


It is unclear as to why this occurred, but it


is speculated that vaporized guanine molecules may have rapidly cooled and coated


the mass analyzer (Q1), thu


impeding its operation due to field aberrations.


these


reasons,


a direct


exposure


probe


(DEP)


been


utilized


vaporize guanine.


DEP vaporization is accomplished by coating a fine wire on the


end of the probe with a small amount of analyte.


flows through the wire,


By increasing the current which


it is rapidly heated, causing the substance to be vaporized


with much less sample decomposition as compared to solids probe vaporization.

Electron ionization (EI) is then used for ion production. Since the temperature of


filament of


DEP


is ramped


(due


to increasing current),


method


harhrrtPri77Pl hb/ vannriT7ti'lfn nf the analvte as a distinct nealk


Thus. reactions in the







50

eluting from a GC are not practical since a constant flux of guanine ions is not

available.


Nonetheless,


it has


been


demonstrated


guanmne


molecular


ions


reactive toward the allyl halides.


This was accomplished by using a leak valve to


provide a constant pressure of each allyl halide separately in the collision cell while


guanine was vaporized and ionized using the DEP


Unfortunately, this method does


not allow relative reactivities (such as those shown in Figure 2-6) to be calculated


since


, unlike GC introduction of the allyl halides, it is impossible to accurately know


how much allyl halide is present in the collision cell for each reaction.


Thus, the


products


resulting


from


these


reactions


have


not been


included


tabulated data presented in


this chapter since


the method used for guanine ion


reactions was quite different from the technique used for the other reactions.

Product ion spectra resulting from reactions of guanine molecular ions (m/z


151) with allyl chloride, allyl bromide and allyl iodide are shown in Figure 2-9.


clear that the [N+41]+ product ion (P+


,m/z


192) is observed for all three reactions.


Furthermore, the lowest P+ intensity is observed in the allyl chloride reaction while


the allyl iodide reaction produces the largest intensity of P+


, indicating that the order


of reactivity follows the relative mutagenicities of the three allyl halides (see


Table


This conclusion, however, may be erroneous since the absolute amounts of


each allyl halide that were allowed to react in the collision cell are not known.


Thus,


:enn n n ~ nr r~l n2nnc n+r Lar nnn *tt~ra4 rn n.n)n4- I-tn', +1. ranf ~ rnnrifl*tn flCrl fot Ohfl h1d















Allyl Chloride


108 124
I t


_,&mA I.J


50 100 150 200


Allyl Bromide


108 124


p+
192 206 229 249
168 I I 233 278
I I I I ll I. 278


50 100 150 200 250


Allyl Iodide


207
P+
192
168 197 233
II


50 100 150 200 250


Figure 2-9:


Product ion spectra of guanine N


ions (m/z


1) reacting with a) ally


I


ri 11 I r. 1_\ fl' _1, ', L1 ,,,tlf-f-, ~~11









reaction products are present in each of the spectra.


In particular, Figure 2-9c shows


that the side reaction products [E2-HX-X] (m/z


81) and E+


(m/z


168) are formed


upon reaction


guanine molecular


ions with


allyl


iodide.


However,


(corresponding to the allyl iodide ion) is observed for the allyl chloride and allyl


bromide reactions as well (Figure 2-9a,b).

contamination by each allyl halide of the


halide introduction into 02


This unusual result can be attributed to

leak valve and plumbing used for allyl


the allyl halides are only slightly volatile and tend to


condense in the conduit leading to the collision cell.


Other side reaction products


that are observed in the spectra such as m/z


206 (or 207), 233,


249 and 278 do not


correlate to any of the side reaction products listed in


Tables 2-3 and 2-6 and are


probably due to contamination of the plumbing as well, since they are common to

all three spectra.


Comparison of Reactivities of DNA/RNA Base and Nucleoside Ions


Towards Allvl Halides to


Those of Pvridine Ions


Figure


compares


relative


reactivities


small


model


base


pyridinee) ion to those of the most reactive DNA/RNA base (uracil) and nucleoside


(uridine) ions toward the allyl halides.


The relative reactivities were calculated as


described previously.


Looking at Figure


10, it may initially appear surprising that both uracil and


uridine ions are, in general, more reactive toward the allyl halides than pyridine ions,







































0.1


0




Figure 2-10:


Pyridine Uracil Uridine


Relative reactivities of pyridine (a model base), uracil (a DNA/RNA
base) and uridine (a nucleoside) ions with allyl halides.


Allyl CI

Allyl Br

Allyl I









previous studies [1] conditions for pyridine reactions with allyl halides were such that

a pyridine pressure of 0.15 mTorr in the ion source resulted in maximum product ion


, m/z 120) formation.


The results reported here, however, were obtained using


enough


pyridine


ion source


to obtain


a sufficient


intensity for


subsequent reaction in Q2.


Thus, it is clear that pyridine ion relative reactivities


appear much greater at higher ion source pressures due to a much larger number of


neutral


pyridine


molecules


collision


which


subsequently


react


with


electrophile charge exchange products such as [E-X]+ and E+


to form the P+ species.


The observation that pyridine ions are less reactive toward the allyl halides


than


uracil


uridine


ions


(except for


reaction


of uridine


ions with


chloride) can be at least partially attributed to side reactions in the collision cell.

Similarly to the results obtained for uracil, pyridine ions produce the least amount


of P+


ions upon reaction with allyl iodide due to excessive side reaction products


arising from charge exchange processes.


In fact, one would expect that pyridine ions


are more


prone


to excessive side


reactions via


charge exchange since pyridine's


ionization energy

(see Table 2-1).


eV) is higher than any of the DNA/RNA bases or nucleosides


Furthermore, pyridine's somewhat high ionization energy may allow


it to charge exchange more readily with allyl bromide and allyl chloride than the


DNA/RNA


base or nucleoside


ions.


This would


account for the overall lower


reactivities of pyridine ions compared to the reactivities of uracil, and in some cases,


J(s









Conclusions


The ability to perform gas-phase ion-molecule reactions of DNA/RNA base


and nucleoside ions with allyl halides has been demonstrated.


Relative reactivities


of the base and nucleoside ions toward the allyl halides were also calculated.


general, the DNA/RNA base ions were much more reactive than the corresponding


nucleosides.


This effect can be attributed to the fact that the nucleoside ions tend


to fragment upon collision with the allyl halides to a larger degree than do the base


ions,


inhibiting


product


formation.


Also,


presumably


to their


higher


recombination energies, the DNA/RNA base ions tended to induce significant side


reactions with


the electrophiles, and, in particular with allyl iodide.


These side


reactions, arising from charge exchange between the base ions and the allyl halides,

often distorted the measured relative reactivities by effectively decreasing the number

of nucleophile ions that were available for reaction.












CHAPTER 3
PRACTICAL ASPECTS OF REACTIONS OF DNA/RNA BASE AND
NUCLEOSIDE IONS WITH ALLYL HALIDES


Introduction


The characteristics of ion-molecule


reactions for analytical determinations


must


considered


differently


from


typical


analyses


performed


using


mass


spectrometric methods (e.g., GC/MS).


Ion-molecule reactions can offer significant


improvements in selectivity (as in being highly selective for reactive electrophiles

here) compared to GC/MS due to the specific nature of the reactions themselves.

Furthermore, the sensitivity of the ion-molecule reaction method depends on more


parameters than does a GC/MS experiment.


Whereas the sensitivity of a GC/MS


analysis depends entirely on the properties of the instrument and the behavior of the


analyte,


an analysis


performed


using


a particular


ion-molecule


reaction


as the


method of determination also depends on the characteristics of the reaction itself.

These aspects, in turn, can lead to unexpected or unwanted results that often hinder

the analysis.

Chapter 2 illustrated that associative ion-molecule reactions in the collision

cell of a triple quadrupole mass spectrometer are useful for studies of reactions of


ionized DNA/RNA bases and nucleosides with neutral allyl halides.


In particular,









varying accuracy) to the mutagenicities of the three allyl halides. Furthermore, it was

demonstrated that product ions due to undesired side reactions often overshadow the

product ions of interest, causing erroneous relative reactivity measurements.

In an attempt to more fully understand the processes that take place for the


ion-molecule reactions described in Chapter


studies have been performed which


address how the reaction products are altered under different circumstances, and


also demonstrate the analytical utility of the method.


The effects of varying allyl


halide pressure in the collision cell on the formation of product ions upon reaction


with ions of the DNA/RNA base uracil have been studied.


Also, calibration curves


have been determined for the reactions of each of the DNA/RNA base ions with the


allyl halides.


Furthermore, the effect of nucleophile ion (uracil) flux on reactions


with allyl halides has been investigated.


Pressure Effects of Allvl Halides on Formation of Product Ions


upon Reaction with DNA/RNA Base Ions


Experimental


Experiments were performed on a Finnigan MAT


(San Jose,


CA) TSQ70


triple quadrupole mass spectrometer modified with an octopole collision cell and a


20 kV dynode.


The mass spectrometer was equipped with a Varian (Walnut Creek,


CA) 3400 gas chromatograph, interfaced to the collision cell,


h1Patd tfrancfer line [ISl


via a resistively


FPneriments were performed at an emission current of 200








100C


, and an ion source temperature of 170C for El (150C for CI).


The mass


spectrometer was mass-calibrated using perfluorotributylamine (PFTBA), and then

tuned to maximize product ion formation for low-energy ion-molecule reactions in


the collision cell, as detailed in Appendix A.


For studies involving the pressure


effects of allyl halides on the reactivity of uracil ions, introduction of the allyl halides


into Q2 was accomplished via a Negretti (Hampshire,

to provide constant and precise pressures. Allyl hal


England) fine-metering valve


ide pressures within Q2 were


indicated by a Granville Phillips (Boulder,


CO) Convectron gauge,


calibrated for


nitrogen.


Reactions of Uracil Ions with Allvl Halides


As an example of the effect of allyl halide pressure in the collision cell on


product ion formation, uracil molecular ions, N+


chloride


, were allowed to react with allyl


, allyl bromide and allyl iodide introduced individually via a variable leak


valve into the collision cell at various pressures.


Figures 3-1 through 3-3 show that


the allyl halide pressure has a profound effect on the relative abundances of the

various product ions that are formed for each reaction (note the logarithmic intensity


scale).


particular,


as the


halide


pressure


increases


each


case,


intensities of all the ions start to decrease at indicated pressures above 1


.0 mTorr.


This is due


to increased ion scattering losses


in the collision cell,


resulting in a


smaller percentage of ions that are ultimately detected.










































eeeeo m/z
se-s m/Z
AAAAA rn/Z
OOO m/z
H-H-O m/z


69
110
112
153


0.01


I I I I I


I II,'


I 1 1


S i-- -f i f ifI If I-fU I II


I I I I I


U I I I I


I I I


I' ~


Ally


Chloride


sure


indicated


(mTorr)


Figure 3-1:


Effect of allyl chloride pressure on product ion distribution for the
reaction of uracil N ions, m/z 112, with allyl chloride.


I


I


































o58Ee)
s-s-l-.-
.4444


m/z
m/z
m/z
m/z
m/z


0.01


0.5 1.0 1.5 2.0 2.5 3.0


Bromide


"essure


indicated


(mTorr)


Figure


Effect of allyl bromide pressure on product ion distribution for the


reaction of uracil N+


ions, m/z 112,


with allyl bromide.

























0.1




0.01


0.001


Allyl


Iodide


Pressure


Indicated


(mTorr)


Figure 3-3:


Effect of


iodide


pressure


on product


distribution


reaction of uracil N+ ions, m/z 112, with allyl iodide.









Upon inspection of Figure 3-1,


it is clear that the reactions that take place


between uracil molecular ions and allyl chloride are relatively simple and few in


number.


As N+


(m/z


112) is depleted, the intensity of the product ion, P+ (m/z


153),


increases until scattering begins to occur.


The P+


intensity then decreases gradually


at higher allyl chloride pressures.


The reaction that occurs between the major CID


daughter ion of the uracil N+


ion, m/z


69, and allyl chloride behaves similarly.


the pressure of allyl chloride in the collision cell increases, m/z


69 initially increases


in intensity due to CID, and then decreases due to reaction with allyl chloride to


form m/z 110,


which is analogous


to m/z 153 (P+).


Indeed, the curves for m/z


and m/z


153 mirror one another, as do the curves for m/z


69 and m/z


112 at higher


pressures.

Examination of Figure 3-2 shows similar behavior for the reaction of uracil


ions and allyl


bromide.


addition


to the


ions depicted


Figure


significant amount of m/z


81 is also formed at indicated pressures above


1 mTorr.


This species, presumed to be the 3-methylcyclopentenyl ion, [E2-HX-X]+


evolves


from reaction of an allyl bromide ion (formed via charge exchange with uracil N+


ions) with an allyl bromide molecule.


It is somewhat surprising that allyl bromide


ions would form via charge exchange with uracil molecular ions since the ionization

energy of allyl bromide (10.06 eV) is higher than the recombination energy of uracil


ions (9.2 eV).


However, small intensities of the allyl bromide ions 120+ and 122+


are nhgerved in the nroduct ion snectra (not shown in Figure 3-2L maximizing at an









laboratory [77] has substantiated that m/z


81 arises via the reaction between allyl


bromide


ions


bromide


neutrals.


surprisingly,


reaches


maximum intensity at about the same allyl bromide pressure as does P+ (m/z


Then, at increasing ally


153).


bromide pressures, all the ion intensities decrease due to


scattering losses.


reaction


of uracil N


ions with allyl


iodide


(Figure 3-3)


produces a


somewhat different reaction profile than the reactions with either allyl chloride or


allyl bromide, with many more side reaction products observed.


In particular, m/z


81 dominates the spectra at pressures of allyl iodide greater than 1 mTorr, eventually


surpassing the intensity of m/z


153 (P+) by almost two orders of magnitude.


Other


product ions are also formed that can be directly related to charge exchange between


uracil N


ions and allyl iodide neutrals.


In addition to m/z


, ions at m/z


and 209 are also produced via the charge exchange process, corresponding to the


, the allyl iodide


ion, E+


, and the iodine-bound allyl dimer ion,


[E2-X]+


, respectively.


The intensities of both m/z


41 and m/z


168 can be observed


to increase at indicated allyl iodide pressures below approximately 0.3 mTorr.


higher pressures, the intensities of m/z


41 and m/z


168 decrease while m/z


81 and 209


production continues to increase due to further reactions of the allyl and/or allyl

iodide ions with neutral allyl iodide.


allyl ion, [E-XI+









Calibration Studies of Reactions of DNA/RNA Base Ions with Allyl Halides


Experimental


Calibration studies of reactions of the DNA/RNA base ions with allyl halides

were conducted under the identical mass spectrometric conditions as those employed


for the experiments described previously.


Introduction of the allyl halide


into Q2,


however, was accomplished via gas chromatography, instead of a fine-metering valve,


so that precise amounts of each allyl


halide could be allowed


to react with


nucleophile


ions.


Gas chromatography was carried


out on


Scientific


(Folsom,


CA)


DB-


(20 m


long, 0.25


mm


thickness) capillary


column in the split mode (split ratio


pressure of 5 psig.


= 100:1) with helium carrier gas at an inlet


The GC oven was temperature programmed from 50C to 100C


at 20C/min, yielding retention times for allyl chloride, allyl bromide, and allyl iodide


of 45

were


and 75


s, respectively.


maintained at 200C.


Injection port and transfer line temperatures


One microliter injections were made


in triplicate of


mixtures containing equimolar amounts of allyl chloride, allyl bromide and allyl


iodide in pentane.


The solutions utilized ranged in concentration from 9.8 fmol/pL


to 980 nmol/iL.


Evaluation of Sensitivity Measurements


Thbrs arp several nrnreseP that must be considered when evaluating the









These


critical


issues


affect the


shapes and


linearities of


calibration


curves and


ultimately determine the apparent limits of detection of three allyl halides.


various


parameters


can


broken


down


following


general


events


categories:

1. The partial pressure of neutral nucleophile molecules available for ionization

in the ion source.

2. The ionization efficiency of the nucleophile which determines the number of

ions in the source.

3. The percentage of ions in the source that are the ion of interest, normally the

molecular ion, N+.

4. Ion transmission from the ion source through the first mass analyzer into the


collision cell; all four of these factors determine the flux of N+


ions available


for reaction.


The gas-phase reactivity of the N+

The variety and extent of competii


well


ion to form the desired product ion, P+.

ig side reactions that may take place as


in the studies to date, most of these side reactions involve the formation


and further reaction of electrophiles ionized by charge exchange,


making the


difference in ionization energies between the nucleophile and electrophile the

most critical parameter.


The pressure of the electrophile,


or other species, in Q2,


which affects the


extent of CID and scattering losses.









transmission


out of


collision


through


second


mass


analyzer to the detector, which determines the number of product ions that

are detected by the electron multiplier.


Effect of Uracil N+


Ion Intensity on Product Ion Formation


The effect of uracil N+


ion intensity (i.e


the amount that is available for


reaction) on product ion (P+


has been studied.


153) formation via reaction with the allyl halides


Figure 3-4 shows the single reaction monitoring (SRM) curves for


the collision cell reactions of uracil N


ions with allyl halides.


Linear log-log curves


are obtained over two orders of magnitude when varying the number of uracil N


that are


allowed


to react


with


9770


pmol


each


halide


introduced


separately into the collision cell via GC.


Furthermore


, the slopes of the three curves


are all approximately equal to one, indicating that the curves (when plotted using


linear axes) are truly linear throughout the range of N+


fluxes utilized.


The results


indicate that, in this

molecule reactions.

to the amount of N+


case,


uracil ion is the "limiting reagent" for each of the ion


The fact that the product ion intensity is directly proportional

available for reaction suggests that there is always a surplus of


allyl halide in the collision cell.


Also, the results point out that the absolute amount


of nucleophile available for reaction should be as large as possible for quantitative


determinations so


that maximum


sensitivity is


realized,


as previously shown


ranntl nfi an: el~ n+ rnnAna nne. Hn


ra a n~ ntr a



































-4-,
C)
I0
oO. 1
L.
01


oo_ oo _o Ally
o ..qAlly
n 3a Ally
A A AIIV


Chloride,
Bromide,
Iodide. s


slope= -0.96
slope= 1.00


ope=


1 10
Counts of m,


100


112


(million


Figure 3-4:


Single


reaction


monitoring


(SRM)


curves


collision


reactions of uracil N'


ions with allyl halides (varying flux of N


enters Q2; 9770 pmol of each allyl halide on column).







68

It is also interesting to note that, for each of these studies, product ions that


indicate


presence


neutral


uracil


collision


(i.e.,


[N2+H]+


[N,+X]+) were never observed.


Other studies [1] of reactions of the more volatile


model


DNA


bases


pyridine


piperidine


with


halides


have


shown


products resulting from reactions of neutral nucleophile molecules in the collision


cell are common.


These observations can be attributed to the very high pressures


of pyridine and piperidine that were utilized in the ion source for ion production.

Invariably, at higher pressures some neutral pyridine or piperidine would migrate


into the collision cell where it could take part in unwanted side


reactions.


partial pressure of the DNA base uracil (a solid at room temperature), however, is


much


lower


than


those


either


pyridine


or piperidine


(both


liquids


room


temperature).


Therefore, for the studies performed here insignificant amounts of


neutral


nucleophile


(uracil)


entered


the collision


and no product


ions


contain two or more nucleophile species were observed.


Calibration Studies of Allvl Halides


Calibration curves have been produced for the reactions of the DNA/RNA


base ions with the allyl halides.


Figures 3-5 through 3-8 show the single reaction


monitoring curves obtained for the collision cell reactions of each DNA/RNA base

ion while varying the amount of each allyl halide introduced into the collision cell


fnr reap.tion


It ia clear that the ranTe of allvl halide concentrations that can be
















o o o o Allyl Chloride
a.__ Allyl Bromide
A A Allyl Iodide, ,







r


, slope=1.57
, slope=1.21


ope= 1.04


0.001


O

o0.0001


Amount


100 1000


Column


10000


(picomo


Figure 3-5:


Single reaction monitoring
cell reactions of adenine ]


(SRM) calibration curves for the
+ ions with allyl halides.


collision


0.01


sl

































0.01


- ____


0.001


oQooo Ally

Ally
aA A A A yI


Chloride,
Bromide.


Iodide


0


slope=0.58
slope=0.53
ilope=0.54


0.0001


I I 1111111 I
4 10-3


I I IIIII 2 I
10-2


Amount


I I I II
10


I I 1111111


I I IIIIII


Column


I I 1111111
102


(picomole


I I il ll I
103
s)


I 1 11 ilt
10*


Figure 3-6:


Single reaction monitoring (SRM) calibration curves for the collision


cell reactions of cytosine N+


ions with allyl halides.


i 313131313131~

















oo9o0 Allyl Chlori(
o D__o. Allyl Bromi
AAAA -Allyl Iodide










r


de, slope=1.51
de, slope= 1.12
, slope=0.61


0.001


0.0001


Ir I1IIIII


I I I IIIII


I I IIIIII


I I 1IlliI1


rt i illl


II I I Il I


i I I II


0.001


0.01


Amount


Column


100
(picomo


1000
)


10000


Figure 3-7:


Single reaction monitoring (SRM) calibration curves for the collision


cell reactions of thymine N


+" ions with allyl halides.


S


0.01




































0.01


4,<






f J







ooaO_ Allyl Chloride, slope= -0.52
a_ sa Allyl Bromid e, slope- 0.51
,,, Allyl Iodide, slope=0.52


0.001


I 11111Ill


I I 111111


I I 11 lll I


I I I 1111


1i t 11lll


I 1 I IIIli


I I 1 11111


0.001


0.01 0.1
Amount


Column


3 100
(picomole


1000


10000


Figure


Single reaction monitoring (SRM) calibration curves for the collision


cell reactions of uracil N+


ions with allyl halides.







73

relative reactivities reported in Figure 2-6 would indicate, a very reactive combination

such as uracil ions with allyl bromide produces detectable intensities of product ions

over a wide range of concentrations (over seven orders of magnitude), as shown in


Figure 3-8.


In contrast, an only slightly reactive combination such as adenine ions


with allyl chloride (see Figure 2-6) produces detectable intensities of product ions

over only two orders of magnitude, as shown in Figure 3-5.

Looking at Figures 3-5 through 3-8, it is obvious that the log-log slopes of the


calibration


curves


vary


substantially


depending


on the


nucleophile/electrophile


combination used for each reaction.


Slopes between 0.


1 for the reaction of uracil


ions with allyl bromide (Figure 3-8) and 1


7 for the reaction of adenine ions with


allyl chloride (Figure


have been


observed.


Log-log slopes for


each


of the


calibration


curves


tabulated


along


with


relative


reactivities


each


DNA/RNA base ion toward the allyl halides (as calculated in Chapter 2) in


Table


. It is clear that there is a correlation between the relative reactivities of each


nucleophile/electrophile pair and the log-log slope of the corresponding calibration


curve.


In general, reaction pairs that are quite unreactive (small relative reactivities)


have slopes larger than one,


maximizing at around 1.6, whereas reaction pairs that


are very


reactive


(large


relative


reactivities)


have


slopes


much


than


one,


minimizing around 0.5.

It should be pointed out that log-log slopes not equal to one indicate curves

that are non-linear when plotted on linear axes; curves with log-log slopes less than















Table 3-1:


Comparison of the relative reactivities of the DNA/RNA base ions and


the log-log slopes


of their calibration curves.


Nucleophile
Ion, N+


Electrophile,


Relative Reactivitya


Log-Log Slope
of Calibration
Curveb


Adenine


Cytosine


Thymine


Uracil


Allyl Cl
Allyl Br
Allyl I
Allyl Cl
Allyl Br
Allyl I
Allyl Cl
Allyl Br
Allyl I
Allyl Cl
Allyl Br


0.001


0.005


0.011


0.293


0.58


0.496
0.285
0.001


0.075


0.313


0.330
0.601


Allyl I


0.157


0.52


a Relative reactivities calculated as outlined in Chapter


b Log-log slopes as indicated in Figures


2 and reported in Figure 2-6.


through 3-8.










slopes greater than one skew upward as


the amount on column increases.


This


means that, for the curve generated by the reactions of uracil ions with allyl bromide,

product ion intensities increase by successively smaller proportions as the amount of


bromide


introduced


collision


is increased.


This behavior


probably due to the high reactivity of uracil ions toward allyl bromide.


Since uracil


ions are highly reactive, they are consumed more quickly (whether it be to form side

reaction products or the desired product ion) when a larger quantity of allyl bromide


is available for reaction.


This same argument can be applied for the other reaction


pairs (i.e.,


cytosine ions with the three allyl halide


s) that have log-log calibration


curve slopes less than one.


Conversely, for the curve generated by a quite unreactive


(i.e.,


adenine


ions


chloride),


product


ion intensities


increase


successively larger proportions as the amount of allyl chloride introduced into the

collision cell is increased, as indicated by a log-log slope greater than one.


Conclusions


Several


important


considerations


when


performing


reactions


ionized


DNA/RNA bases with allyl halides in the collision cell of a triple quadrupole mass


spectrometer have been addressed.


It has been shown that allyl halide pressure can


have a great effect on the relative intensities of the various product ions observed


upon


reaction


with


uracil


ions.


particular,


when allyl


iodide


is used


as the







76

dominate the spectra at higher allyl iodide pressures, overshadowing the desired allyl


adduct ion, P


SFurthermore, it has been observed that at high pressures of allyl


halide, ion losses due to scattering begin to occur, decreasing the sensitivity of the

technique.


The technique of using Q2 ion-molecule reactions


for quantitative analysis is


much more involved than typical mass spectrometric techniques such as GC/MS.

While the sensitivity of a technique such as GC/MS is determined primarily by the

characteristics of the instrument and the properties of the analyte, the sensitivity of

analytical ion-molecule reactions also depends on the attributes of the reaction itself,


as well as those of possible competing reactions.


With this in mind,


calibration


curves for reactions of DNA/RNA base ions with allyl halides have been obtained.

Varying the nucleophile ion intensity for the reaction of uracil ions with allyl halides


resulted in linear log-log plots with slopes of one.


Upon variation of the amount


of allyl halide neutral available for reaction with nucleophile ions, however, log-log


calibration curves were observed to be linear, but with slopes ranging from 0.


7 depending on the particular nucleophile ion.


The individual slopes are related


to the relative reactivities of each specific nucleophile/electrophile reactant couple,


with higher slopes


resulting from


reactive


species,


and more


reactive


pairs


producing curves with lower slopes.













CHAPTER 4


DEVELOPMENT AND CHARACTERIZATION OF


A SYSTEM


FOR TRAPPING IONS IN THE COLLISION CELL OF
A TRIPLE QUADRUPOLE MASS SPECTROMETER


Introduction


As will be demonstrated in this chapter and the next, trapping ions in the


collision cell of a linear quadrupole


instrument can be useful for studies of ion-


molecule


reactions.


technique,


however, is not very common, having been


implemented by only a few research groups [51-57].


This lack of utilization can be


attributed, in part, to the unique designs of the many tandem quadrupole instruments


in use today.


The ion optical layouts, instrumental hardware, instrument control


software and data system configurations are different for each commercial design.


Additionally,


"home-made" instruments that have unique designs of their own are


being used in several laboratories.


Moreover, many groups have found it necessary


to further modify their instruments to successfully implement ion trapping in the


linear


scheme


quadrupole


cannot


devices.


universally


Therefore,


applied


it is clear that a


every


particular


tandem


ion trapping


quadrupole


instrument, and that each


kind of device must be considered individually.


This


points out that each instrument must be characterized before successful ion trapping


9 *


9









of the


TSQ70 triple quadrupole mass spectrometer used in these studies and its


development for use as an ion trapping device.


The Finnizan MAT TSQ70 Triple Quadrupole Mass Spectrometer


Finnigan MAT TSQ70 triple quadrupole


mass spectrometer


(and its


successor,


TSQ700) is certainly one of the most popular tandem quadrupole


instruments in use today.


This, and the fact that it is a highly complex and powerful


computer controlled instrument, make it a good choice for development of a Q2 ion


trapping system.


A system developed for the instrument in our lab could be easily


applied to other TSQ70 instruments worldwide.


Additionally, the instrument is ideal


for methods such as Q2 ion trapping since all instrumental parameters are under


microprocessor control, and can be changed in real time.


Control of the instrument


by the user is accomplished using Instrument Control Language (ICL) procedures.

This unique interface allows for computer control of any number of instrumental


parameters during an experiment.


Furthermore, ICL is structured similarly to many


common


computer


programming


languages,


such


as BASIC


which


allows


instrument to receive incoming data and react accordingly.


When performing ion


trapping


in the


collision


means


many


operations


would


controlled by


hardware


modifications on


other instruments can


regulated by


software programmed on the


TSQ70.


r7ifllrP A-lI1 ehnirxc 2a -ar2rn nf tha macc analrrrrrtr ccrnrnhh, n^f th TnCf7Aft


/I,









79




S




CtC4
YU
clFL.I 5 ,
tn U,

IO
Tc4)








z 0'




Li]

o b







CC1)






Wh* *
t! ~I
nfl
Ul 41









filters


Q3),


an enclosed


bent octopole


collision


(Q2)


several


focusing lenses mounted on an optical rail.


To avoid confusion, only the lenses (L23


and L31) utilized for trapping ions in the collision cell have been labelled.


One


component of the TSQ70 that is quite different from all other triple quadrupole mass


spectrometers is the bent collision cell.


The bent design allows improved signal-to-


noise


ratios


compared


to linear


collision


cells,


without


a compromise


in ion


transmission [78-79].


The benefits are most apparent when using ionization methods


which produce an abundance of excited neutrals, such as fast atom bombardment


(FAB) ionization.


The bent design reduces the number of neutrals that traverse the


length of the instrument and reach the detector, and thus decreases the noise from

the electron multiplier.


The collision cell of the


TSQ70 used in these studies is, in fact, different from


that in most other


TSQ70 mass spectrometers in use.


Typical TSQ70 instruments


have a quadrupole collision cell, but the instrument in our laboratory was upgraded


to the octopole


design


used


in the


newer


TSQ700 mass spectrometers.


Under


normal operating conditions,


been observed.


almost no noticeable changes in spectral quality have


However, many groups that have studied multiple devices with two-


dimensional electric fields [58,80-84] have found that, in some cases, higher-order


multipoles may be advantageous when used as


RF-only collision cells.


Through


theoretical treatments and simulations, Higg and Szabo [58,81-82] determined that


octoooles are not best suited to operating as mass filters in the RF/DC mode.


Thev







81

inherent nonlinearities, give rise to mass resolution too low for the purpose of mass


analysis.


However,


they


state


octopoles


can be


more


effective


than


quadrupoles in the RF-only mode for guiding and transporting low-energy ions, such


as in a collision cell.


In a related study,


Syka and Szabo [83] evaluated the ion


transmission characteristics of quadrupole, hexapole and octopole collision cells in


the RF-only mode on a


TSQ70 mass spectrometer.


They suggested that octopole


collision cells are modestly advantageous in comparison to quadrupole collision cells

due to somewhat extended stability regions, leading to wider mass ranges, and more


uniform transmission qualities.


It was also pointed out that RF-only octopoles would


be useful in studies where correct ion kinetic energies must be known, due to the


flat-bottomed pseudo-potential wells generated by higher-order multiple fields.


representation of this concept, generated using the SIMION ion simulation program


[85] and the exact measurements of the two commercially available


cells, is shown in Figure 4-2.


TSQ70 collision


Davis and Wright [84] have also studied multiple


collision cells, using a computer modelling technique.


In their work, calculations


were made to determine factors affecting transmission of ions through higher-order


multipoles under more extreme entry conditions.


It was determined that,


under


these circumstances,


the trajectories may be


less stable


in an octopole


than in a


quadrupole.


It was also noted, however,


that if a well-ordered beam enters the


octopole,


higher-order


multiple


advantageous


in that


it restricts


transverse


kinetic


enerev


causes


an overall


smearing


e parent ion


I~J


--














100

90

80

70

60

50

40

30

20

10

0-O


Quadrupole


Octopole


-10--r
-3.0


I I I I I
-2.0
Position


I I I i II
-1.0
Relative


0.0
to Center


1.0
of


! I I I I I
2.0
Rods (mm)


I I I.
3.0


Figure 4-2:


Comparison of pseudo-potential wells of the quadrupole
collision cells of the TSQ70. Data generated using the
simulation program with offset potentials of 0 V and RF
+ 100 V applied to opposite rods.


and octopole
SIMION ion
potentials of







83
lenses immediately before the collision cell constrict the ion beam so that it enters


the collision cell close to the central axis.


This allows the octopole to act as an


efficient ion containment device.


Trapping Ions in the Collision Cell of the TSQ70


The concept behind trapping ions in the


relatively simple.


RF-only octopole collision cell is


The technique involves utilizing the lenses immediately before and


after Q2 as electrostatic mirrors, trapping ions within the cell.


The ions are then


held within


the octopole


field for a


variable


time


period, allowing ion-molecule


reactions to occur, and subsequently scanned out via the third quadrupole.


Figure 4-1


shows the configuration of the


TSQ70 triple


quadrupole


mass


spectrometer used for trapping ions in Q2.


Essentially, ions are trapped in Q2 by


varying the


potentials applied


to the entrance (L23)


exit (L31)


lenses in


specific and repetitive manner.


An illustration of the time profiles of the voltages


applied to the entrance and exit lenses is depicted in Figure 4-3.


The lens potentials


are changed according


to Instrument Control


Language


procedures


specifically


designed to facilitate ion trapping in Q2.


Additionally, these procedures allow for


control of variables significant to the trapping process such as ion trapping time and

Q2 fill time, as discussed below.

Inspection of Figure 4-3 shows that there are three distinct stages in a Q2 ion


trapping cycle.


The first stage, called the fill time, allows ions to be gated into the









84






C)

bO



c, ,
a.
Q 0H
En

U.)
V(


rC.
c)l
I-

C.
C.)






.U.)



'~I-








I CI)
'IY Llr -i
C"t





acl 0, C ;
AP 4 enI 2
uo a.)
U '-4









potential while the exit lens, L31,


is closed.


In other words, to fill the collision cell


with positive ions,


lens L23


is held at a negative voltage while lens L31


is held


positive.


Through careful selection of the potentials applied to L23 and L31,


ions


are permitted to enter Q2 but not allowed to exit out the back.


cell is filled with ions.


Thus, the collision


In the second phase of the cycle, the trapping time, lenses


L23 and L31 are both closed so that ions may neither enter nor exit the collision cell.


During


variable


time


period


processes


such


as ion-molecule


reactions


collision-induced dissociation are allowed to take place.


the last stage of the


cycle, the exit lens, L31, is opened allowing the trapped ions to pass through Q3 and


be detected my the electron multiplier.


During this time,


the entrance lens, L23,


still closed so that ions continually formed in the ion source do not pass through Q2

and Q3 and are not detected. An ion pulse typical of what is normally observed is


shown in Figure 4-4.


Although the basic Q2 ion trapping scheme is very simple, many parameters

must be carefully adjusted to achieve optimum trapping performance. In particular,


the potential applied to the entrance lens, L23,


ions is of the utmost importance.


when filling the collision cell with


Additionally, it was discovered that the DC offset


voltage applied to the collision cell greatly affects ion trapping efficiency.


Also, it


was


determined


presence


helium


buffer


in the


collision


dramatically


improves


ion trapping ability


system.


These and


other


parameters will be discussed in detail in the chapter.





































42.5 ms


Figure 4-4:


Trace of a typical ion pulse detected by a digital storage oscilloscope
upon opening of the exit lens, L31, during a trap and pulse cycle.









Experimental


Experimental Conditions


All experiments were performed on a Finnigan MAT (San Jose, CA) TSQ70

triple quadrupole mass spectrometer modified with an octopole collision cell and a


conversion


dynode.


Optimization


studies


were


performed


using


perfluorotributylamine (PFTBA) calibration compound. PFTBAwas introduced into

an El ion volume in the ion source via a specially designed probe, constructed in


house, which allowed for precise metering of the PFTBA vapor pressure.


consisted of a


The probe


o.d. hollow stainless steel probe connected to a variable leak


solenoid valve


under computer


control.


PFTBA


pressure within the probe was


indicated by a Granville Phillips (Boulder,


performed under


CO) Convectron gauge.


conditions at an electron energy of 70 eV


Studies were


and an emission


current of 200 pA.


The manifold was maintained at temperatures ranging from 70-


C while the ion source was held at 170C.


For the Q2 ion trapping optimization studies, electron multiplier potentials


ranging from -600 to


-1400


V were utilized.


As is discussed in this and the next


chapter, several


experimental parameters greatly affect the


number of ions that


ultimately strike


the electron


multiplier.


Among these,


the mode


which


second mass analyzer, Q3, is operated, and the presence or absence of helium buffer


eas in the collision cell have the largest effects.


In general, experiments that were









electron multiplier voltages on the order of 700 V


. On the other hand, studies that


were performed without helium buffer gas in Q2 or with Q3 in the RF/DC mode


required


electron


multiplier


voltages


upward


-1000


cases,


preamplifier gain was set at 108 V/A and the conversion dynode was held at


The mass spectrometer was tuned with PFTBA to optimize ion transmission

and to calibrate the mass assignment using tuning procedures written by Hail [71].

Additionally, further tuning operations were performed using either of the two sets

of tuning guidelines recommended in Appendix A.


Experiments


involving Q2


ion trapping were


regulated


several


of the


Instrument Control Language procedures listed in Appendix C.


These programs


were specifically written to perform the atypical operations of dynamically controlling


the various ion optical devices needed to effect ion trapping in the collision cell.


addition to utilizing these specially written procedures, several other parameters were


optimized for Q2 ion trapping.


For all studies, the prescan voltage settling time,


SETTIM,


was set to its minimum value,


to minimize


the time


that the


trapping


cycles


took


to execute.


Also,


offset


potential


applied


to the


octopole


collision


was


optimized


or -2


'respectively,


depending


whether or not helium buffer gas was employed.


For experiments where helium was


utilized


as a


buffer gas,


it was


introduced into Q2


via a


Negretti


(Hampshire,


England) fine-metering valve to an indicated pressure of approximately 6.0 mTorr.


Fi rtherm rep


the mnde


in which 03 was nneratecd was varied.


For exneriments










by turning off the Q3 DC potential using the DDOFF command.


If, however, mass


analysis was to be performed on the pulse of ions as it exited the collision cell, Q3

was operated in the normal RF/DC mass filter mode.


Parameters


defined


ion trapping


process


were


carefully


controlled.


The period of time that the collision cell was allowed to fill with ions,


the fill time, was typically on the order of 0.1 to 1 second.


However, trapping times


ranged from several hundred milliseconds to several minutes or more,


the particular experiment.


depending on


In any case, specific values for fill times and trapping


times are stated in the main text when the situation warrants.


Determination of Ion Trapping Efficiencies


In order to evaluate the effectiveness of the octopole collision cell as an ion


trapping device,


studies were performed to determine the number of ions that are


ultimately detected as a percentage of those ions that are initially injected into Q2.


For the sake of brevity, this quantity is referred to as


(Out/In).


The procedures


required


to determine


%(Out/In)


are not intuitively


obvious.


particular,


calculations of the absolute numbers of ions that enter and exit the collision cell


during


a typical


trapping


cycle


are difficult.


Fortunately,


calculation


percentage of ions that are ultimately detected does not require that the absolute


numbers


known.


reason,


absolute


quantities


were


- S... -


- ~ ~ ~ ~ ~ ~ ~ ~ ~ n a a nt- flf 4%.*V l-A% *f


,S ft f


anl j.4 .


,,., ~


,1, ,,


/K*/







90

There are several instrumental considerations that must be taken into account


when performing %(Out/In) studies.


Ideally,


instrument should have


transmission through Q2 and Q3 for these experiments, to facilitate the calculations.


This


is not the


case


however.


Fortunately,


even


though


there


is not 100%


transmission,


net effects are canceled.


Comparisons of the


number of ions


injected to the number of ions detected are valid because imperfect transmission


characteristics equally affect both processes.


Additionally,


the Q1


and Q3 mass


analyzers must be static (i.e


not scanning) for all the experiments so that a valid


comparison of ion intensities can be made.


Since ions


are gated into the collision


cell under static conditions in a trapping experiment, ion current determinations for

both the number of ions injected and the number of ions detected must be made


under


these


same


conditions.


Normally


TSQ70


does


not allow


data


acquisition when the instrument is not scannix

one analyzer can be set to pass a chosen m/z

converter (DAC) is varied over a small range


However, by using a "DAC scan"


while a lens voltage digital-to-analog

. This, in effect, allows for detection


of a continuous ion signal under static conditions.

DAC scans were used in studies to determine the relative number of ions


injected into the collision cell.


instrument was set to scan over the ion


interest in the daughter mode while Q3 was operated as an RF-only device with a


low-mass cutoff well below the m/z of the particular ion. Thu

ion source entered 01 where a specific m/z was mass selected.


s, ions formed in the

The ions were then









multiplier.


In order to observe the ion currents, DAC scans were executed where


lens IA1 was varied over a small range.


In fact, due to the nature of operation of


the TSQ70 mass spectrometer, lens LA1 was never varied at all since, in the daughter


mode


, the instrument does not permit ion optical components after the collision cell


to be changed.


At constant electron multiplier voltages and preamplifier gains,


potentials representing ion currents were then measured from the SIG1 test point

on the electrometer board of the TSQ70 using a LeCroy (Chestnut Ridge, NY) 9400


Digital Storage Oscilloscope (DSO).


These values were then multiplied by the fill


times used for the trapping experiments to obtain quantities, I, related to the total

number of ions (charge) that were gated into the collision cell.


Determinations of the


number of ions detected after being trapped were


conducted at electron multiplier voltages and preamplifier gains identical to those

used in the corresponding experiments where the relative numbers of ions injected


into 02


were


determined.


This was


done


so that additional


factors were


introduced.


these


experiments,


trapping


procedure


TQ2FILL5K


(Appendix C) was utilized to perform ion trapping, with and without helium buffer


gas, in Q2.


Using the same


fill times as were utilized for the other studies,


TSQ70 parameter TIC read back the total ion currents, A, for the ion pulses as they


exited the collision cell.


These values provided a measure of the relative areas of the


ion pulses.


Additionally, the HEIGHT'


of the peaks were recorded and converted


to TIC'


that would have been expected had the signals been square nulses.


s. R