m- v-n.

m, ... .
«Amy wry

a...
. a;

--I

f.»

.3333" :.

“.44

«vi-13

-'~.

0- a

« .m—r...

m

<om,

M...

iv

- r
aim-"9;
-» a...
‘5- -_. ,.

, mm '
All-"Aw:-
.. ”V"
:3" J3:
, -.....§E_.“~z.3-..
A P

u...

“3R

.‘v' . . ‘ A i

. y .I 44'1‘ at
. ";
anti-min?

" “1:3? ' ~

\.1

 

p;

THESiS

a 1 IHIIHHIUHHIll”UNI!!!”Willi?”(IllUlllllflll

3O 1 555 2403

This is to certify that the
dissertation entitled

Electrospray Mass Spectrometry of Metal
Ion Complexes with Chelating and

Macrocyclic Ligands
presented by

CALIN GEORGE ZNAMIROVSCHI

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Chemistry

v 7
Major professor

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

 

 

 

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES Mom on or before date duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU In An Affltmotlvo Action/Ema! Oppottunlty Intuition
Wan-9.1

ELECTROSPRAY MASS SPECTROMETRY OF METAL ION COMPLEXES
WITH CHELATING AND MACROCYCLIC LIGANDS

By

Calin George Znamirovschi

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1996

ABSTRACT

ELECTROSPRAY MASS SPECTROMETRY OF METAL ION COMPLEXES
WITH CHELATING AND MACROCYCLIC LIGANDS

By

Calin George Znamirovschi

The earlier methods of mass spectrometry, especially electron impact
ionization, were limited in their application to inorganic systems, due to the
low volatility of most ionic species. More recent methods of “soft” ionization,
included but not limited to electrospray, have had a major impact on
inorganic mass spectrometric analysis.

A preliminary part of this work focuses on the behavior of simple
models, like metal ions in solution, mainly to assess the extent to which the
mass spectrometric analytical response is consistent with their speciation in
solution. The appearance of metal - solvent adducts of various structures in
the mass spectra allowed us to assess their stability, as a function of the
solution matrix composition and the operational parameters of electrospray.

The inherent processes that accompany the transfer of ionic and
molecular associations from solution into the gas phase are double-edge

swords and they determine how closely a mass spectrum resembles the

solution chemistry of the analyte. The effect of several instrumental
parameters is presented, with respect to their consequences on ion adduct
formation and their rate of “survival” during the transfer through the
electrospray interface.

We also applied electrospray as a means of sample introduction in
mass spectrometry, taking advantage of its ability to generate gaseous ions
from ionic, non-volatile solution species. We have obtained important results
regarding the gas phase stability of several metal complexes with macrocyclic
ligands, by using collisionally activated dissociation with tandem mass
spectrometry.

The application of electrospray mass spectrometry as a means of
analysis which is complementary to solid state and solution studies in
synthetic coordination chemistry is also a part of our study. Our results on
the complexation of transition metals by a novel calixarene macrocyclic
ligand contribute to a more complete image of their solution chemistry as

well as their structural stability.

Copyright© by
CALIN GEORGE ZNAMIROVSCHI
1996

ACKNOWLEDGEMENTS

The research leading to this dissertation was only possible with the
help, inspiration, guidance, and friendship of many people. People who have
had a tremendous impact on all aspects of my life and career, both in and
outside a classroom or a laboratory. A comprehensive list of friends, teachers
and coworkers who have helped me throughout twenty-two years of formal
education and who deserve recognition cannot be compiled, simply because it
is continuously growing.

First and foremost I would like to convey my deepest gratitude to my
research advisor, Dr. Christie Enke. I have been very fortunate, honored, and
privileged to have such a prestigious scientist, a genuinely talented educator,
and a great friend as my mentor. Words are not enough to describe how
thankful I am to him, for his guidance, support and patience during my
tenure as a graduate student. Far from my family and home, I have often
turned to Dr. Enke seeking a word of advice, at happy or difficult times.
Needless to say, it has always been readily and graciously offered.

Special thanks and words of recognition are immensely deserved by
my first chemistry teacher, Ms. Monica Stanca, who supported and

encouraged me while I was a student at “Nicolae Balcescu” High School in

vi

Cluj-Napoca, Romania. I would also like to thank Dr. Sorin Mager, my
research advisor during my undergraduate studies at the University of Cluj-
Napoca.

The financial support I have received during my graduate studies,
from Michigan State University and the University of New Mexico was
crucially important and it is gratefully acknowledged. I also consider myself
privileged to have been part of a wonderful research group, a true and
supportive family. Eric Hemenway was always willing to help, to share his
knowledge and experience, and I am very thankful to him. Paul Miyares,
Douglas Beussman, Paul Vlasak, Qinching Ji, Fei Liu, Michael Davenport,
Tina Erickson, Terri Constantopoulos, Gregor Overney, Kristi Hughes and
Gail Bradley, all have been, at one point or another, part of our team, and I
enjoyed their friendship and support very much throughout these years.

Lastly but most importantly, I would like to acknowledge my family. I
thank my parents and grandparents for everything they have done for me,
for their priceless contribution to my formation as a citizen and as a scientist.
I have permanently felt their love, thoughts and support, across the ocean
and across the miles. To my wife Karin and our son Patrick, just a few words,
my thanks and my love, along with the pledge that I will always make them

proud.

vii

TABLE OF CONTENTS

Chapterl Introduction 1

Introduction .......................................... 1

Electrospray Mass Spectrometry of Metal-Ligand Systems . . . . 3

Chapter 2 Metal Ions in the Electrospray
Process.........................15

Introduction .......................................... 1 5
Interaction of Metal Ions in Solution ....................... 1 7
Metal Ion - Solvent Adduct Formation ...................... 17
Salvation Effects ........................................ 23

Chapter 3 “Source CAD” - An Advantage or a
Hindrance?.................... 32

Introduction .......................................... 32

Structural Transformations within the Electrospray Source . . . . 34

Metal Ion - Solvent Adducts ............................... 36

viii

Chapter 4 Interaction of Metal Ions with
Macrocyclic Ligands . . . . . . . . . . . . . . 54

Introduction ........................................... 54
Crown Ethers and Aza-Analogs ........................... 56
Interactions in Solution .................................. 58
Interactions in the Gas Phase .............................. 62
Summary .............................................. 79

Chapter 5 Stability Studies in the Gas Phase by
Tandem Mass Spectrometry . . . . . . . 84

Introduction .......................................... 84

Collisionally Activated Dissociation in a Triple Quadrupole

Mass Spectrometer .................................... 86
Effect of Collision Energy ................................ 89
Effect of Collision Gas Pressure ............................ 93

Comparison of Relative Stability of Alkali-Metal Complexes . . . . 104

Summary ............................................. 1 08

Chapter 6 Metal Complexes of a Novel
Macrocyclic Ligand . . . . . . . . . . . . . 111

Introduction ......................................... 1 1 1

Complexes with Scandium(lll), Yttrium (Ill), and Lanthanum(lll) 113

Table 3.1

Table 3.2

Table 4.1

Table 4.2

Table 4.3

Table 5.1

Table 5.2

Table 5.3

Table 5.4

LIST OF TABLES

Rationalized peak intensities for the predominant ions in the

ES-MS spectrum of NaCl in ethanol, as a function of the
transfer capillary temperature

42

Rationalized peak intensities for the predominant ions in the

ES-MS spectrum of KCl in ethanol, as a function of the
transfer capillary temperature

Comparison of log Ks values for the complexation of L, L’,
and L” with K“ and Ag’ in methanol

Absolute and relative peak intensities of the predominant
ions in the ES-MS spectrum of AgL”

Absolute peak intensities of the precursor and product ions
for the CAD of (Ag ”)+, (Ag ”-2H)+, and (Ag ”-4I-I)+

Effect of collision energy (E0011) on the relative intensity of
product ions of (Cst)+ (m /z 661); Collision gas pressure =
0.40 mTorr

Effect of collision gas pressure (P0011) on the relative
intensities of product ions of (CsL)+ (m /z 397);
Collision energy = 20 eV

Effect of collision gas pressure (P0011) on the relative
intensities of product ions of (CsL2)+ (m/z 66 1);
Collision energy = 20 eV

Relative intensities of the product ions in the CAD of
complexes of alkali-metals with 18-crown-6 (L);
Collision energy (Ecm) = 4 eV

43

57

7O

75

91

97

100

107

Figure 1.1

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

LIST OF FIGURES

Application ranges of modern interfacing techniques in
mass spectrometry

ES-MS spectrum of NaCl in methanol. Capillary
temperature = 1 10 C

ES-MS spectrum of KCl in methanol. Capillary
temperature = 1 10 C

Generic structures of primary salvation shells around
an alkali-metal ion. Ligand: (a) water, (b) methanol,
and (c) dimethylether

ES-MS spectrum of KCl in l-propanol. Capillary
temperature = 120 C

ES-MS spectrum of KCl in l-propanol. Capillary
temperature = 160 C

The schematic of an electrospray - mass spectrometer
interface

ES-MS spectrum of NaCl in ethanol. Capillary
temperature = 110 C

ES-MS spectrum of NaCl in ethanol. Capillary
temperature = 160 C

ES-MS spectrum of NaCl in ethanol. Capillary
temperature = 220 C

ES-MS spectrum of KCl in ethanol. Capillary
temperature = l 10 C

ES-MS spectrum of KCl in ethanol. Capillary
temperature = 160 C

xi

20

22

25

26

27

35

38

39

41

45

46

Figure 3.7 ES-MS spectrum of KCl in ethanol. Capillary

temperature = 220 C 47
Figure 3.8 MS/CAD/MS product ion spectrum of m/z 195 48
Figure 3.9 MS/CAD/MS product ion spectrum of m/z l 17 50
Figure 3.10 MS/CAD/MS precursor ion spectrum of m/z 39 51

Figure 4.1 The structures of the 18-member macrocyclic ligands
considered in the present complexation studies 56

Figure 4.2 ES-MS spectra of: (a) Potassium complex of 18-crown-6
(KL)+, (b) Potassium complex of aza-18-crown-6 (KL’)+, and
(c) Potassium complex of diaza-18-crown-6 (KL”)+ 63

Figure 4.3 ES-MS spectra of: (a) Silver complex of 18-crown-6 (AgL)+
(b) Silver complex of aza-18-crown-6 (AgL’)+, and
(0) Silver complex of diaza-l8-crown-6 (AgL”)+ 65

Figure 4.4 Proposed scheme for the dehydrogenation of the silver
complex of diaza-18-crown-6 (AgL”)+ during the
electrospray process 69

Figure 4.5 MS/CAD/MS spectra of: (a)Potassium complex of 18-crown-6
(KL)+, (b) Potassium complex of aza-18-crown-6 (KL’)+, and
(c) Potassium complex of diaza-lS-crown-G (KL”)+ 71

Figure 4.6 MS/CAD/MS spectra of: (a) Silver complex of l8-crown-6
(AgL)+, (b) Silver complex of aza-18-crown-6 (AgL’)+, and
(c) Silver complex of diaza-18-crown-6 (AgL”)+ 73

Figure 4.7 Proposed mechanism for dehydrogenation and loss of silver
hydride (AgH) during CAD of ions of: (a) m/z 371,
(b) m/z 369, (c) m/z 367, and (d) m/z 365 77
Figure 5.1 The block diagram of the TSQ-7000 mass spectrometer 88

Figure 5.2 ES-MS spectrum of a CsCl : L (1:2) solution in methanol.
Capillary temperature = 110 C 90

xii

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6

Figure 5.7

Figure 5.8

Figure 5.9

Figure 5.10

Figure 6.1

Figure 6.2

Figure 6.3

Figure 6.4

Figure 6.5

MS/CAD/MS product ion spectra of (CsL2)+ (m /z 661).
Collision gas pressure = 0.40 mTorr; Collision energy:
(a) 4 eV, (b) 46 eV 92

Rationalized peak intensities of product ion 08*, (CsL)+,
and (Cst)* in the CAD of (Cst)+ (m/z 661) as a function
of collision energy. Collision gas pressure = 0.40 mTorr 94

MS/CAD/MS product ion spectra of (CsL)+ (m /z 397).
Collision energy = 20 eV; Collision gas pressure:
(a) 0.10 mTorr, (b) 2.20 mTorr 96

Rationalized peak intensity of product ion Cs+ (m /z 133) in
the CAD of (CsL)+ (m/z 397) as a function of collision gas
pressure. Collision energy = 20 eV 99

MS/CAD/MS product ion spectra of (CsL2)+ (m/z 661).
Collision energy = 60 eV; Collision gas pressure:

(a) 0.35 mTorr, (b) 0.75 mTorr 101
Rationalized peak intensities of product ions Cs”, (CsL)*,
and (CsL2)+ in the CAD of (CsL2)+ (m/z 661) as a function
of collision energy. Collision gas pressure = 0.40 mTorr 102

Rationalized peak intensity of product ion 03* (m /z 133) in the
CAD of (Cst)+ (m /z 661) as a function of collision gas pressure.

Collision energy = 60 eV 103
ES-MS spectrum of an equiformular mixture of (N aL)+,

(KL)+, (RbL)+, and (CsL)+. Capillary temperature = 200 C 106
The structure of hexahomotriazacalix[3]arene (HaL) l 12
ES-MS spectrum of a Sc3+:H3L (1: 1) solution in

acetonitrile. Capillary temperature = 200 C 1 14
ES-MS spectrum of a Sc3+zH3L (1:1) solution in methanol.
Capillary temperature = 200 C 1 16
MS/CAD/MS product ion spectrum of m/z 309 117
ES-MS spectrum of a Y3+:H3L ( 1:1) solution in methanol.
Capillary temperature = 200 C 119

xiii

Figure 6.6

Figure 6.7

Figure 6.8

Figure 6.9

Figure 6.10

Structures of representative fragmentation products that
result in the ES-MS processes involving H3L and its
metal(III) complexes

MS/CAD/MS product ion spectrum of m/z 472

Product ion spectra of CAD of m/z 707. Collision gas
pressure = 0.6 mTorr. Collision energy: (a) 20 eV,
and (b) 30 eV

ES-MS spectrum of a La3+zH3L (1:1) solution in methanol.

Capillary temperature = 200 C
Product ion spectrum of CAD of m/z 400.

Collision gas pressure = 0.6 mTorr; Collision
energy = 35 eV

xiv

120

122

123

124

125

Chapter 1

Introduction

Introduction

In recent decades the advances in high-resolution mass analyzers and the
discovery of new methods of sample introduction and ionization have made
mass spectrometry a widespread and absolutely necessary tool in chemical
research. Although organic and biochemical areas previously accounted for
the vast majority of its uses and applications, mass spectrometry has lately
emerged as a very powerful means of investigation in inorganic, coordination,
and organometallic chemistry [1-3].

For many years the main obstacle to the study of mass spectra of
inorganic and organometallic compounds was the process of generating ions
in the gas phase. The mass spectrometry of inorganic compounds and metal
complexes usually required the volatilization of species prior to ion

formation. Even though many desorption techniques were gradually

developed [4,5], most of the studies were limited to non-ionic, volatile
compounds, and were mainly based on electron impact (EI) ionization [6,7].
Although used as reagents in various mass spectrometric techniques since
the early 19703 [8], metal ions, along with other related chemical species,
such as metal adducts and complexes, were not suitable for E1 ionization
mass spectrometric studies because of their low volatility, high thermal
lability, and tendency to undergo irreversible secondary reactions during the
ionization process. Although derivatization techniques were generally
available, the methodology was time consuming and required large amounts
of sample.

The development of soft ionization techniques, such as fast atom
bombardment (FAB) [9,10], matrix-assisted LASER desorption ionization
(MALDD [11], and chemical ionization (CI) [12,13] lifted many of the earlier
limitations. This trend was later continued by the development of desorption
chemical ionization (DOD [14], 2520f plasma desorption (PD) [15], and the
whole family of field-assisted desorption methods, which includes
electrospray (ES), thermospray (TS), and ionspray (IS) ionization [16].

In a field-assisted desorption process, a strong electrostatic field is
used to extract ions from a liquid substrate. Two generic methods have
emerged, one in which ions desorb directly into the vacuum, named
electrohydrodynamic (EH) ionization [17 ,18], and one in which the analyte is
dispersed as a mist of droplets, from which the solvent gradually evaporates

and the ions desorb into a bath gas, under atmospheric pressure. The latter

is generically called atmospheric pressure (AP) ionization. The ionization
either occurs in the gas phase, as in the case of atmospheric pressure
chemical ionization (APCD, or ions are preformed in the analyte solution and
subsequently nebulized into a chamber, also held at atmospheric pressure, in
the process of electrospray (ES). Electrospray has lately become one of the
most attractive and widely employed variant of the atmospheric pressure
ionization methods. Also, electrospray enjoys today the reputation of being
the most versatile interfacing technique for coupling liquid chromatography
with mass spectrometry (Figure 1.1) [19] due to its principal features, namely
the capability to produce multiply charged ions, even from thermally labile
and extremely fragile molecules, and its gentleness, which allows such
species to be transferred to the gas phase while undergoing little or no
fragmentation. The possibility of generating intact gaseous ionic species has
triggered a great deal of research aimed at obtaining new and improved

structure information in the gas phase.

Electrospray Mass Spectrometry of Metal-Ligand Systems

Even though the study of electrospray phenomena extends back early this
century, to the work of Bose [20] and Zeleny [21], and later to the studies of
Dole et al. [22,23], electrospray ionization combined with mass spectrometry

was reported, essentially simultaneously, much later, in 1984, by both

a a .8055QO mums s assets c.5853. 58E 6 SEE “.3833. 2 939;

CH

Eats 33832

OH

OH OH

m N

.2

 

 

    

mwuamofioflm

88:
30am

:ofimflnom 183530
9330.3 oflmaawofiuw
98 3388543.

 

   
   

mzdw

 

 

AEOQGOZ

KamIOd awreuv

oEoH

Yamashita and Fenn [24] and Aleksandrov et al. [25,26]. Since its inception,
in 1988, first as a tool for the analysis of large biomolecules [27], the practice
of ES-MS has grown impressively [28-31]. Recently, electrospray has shown
itself capable of producing intact ions, either positive or negative, multiply
charged, from remarkably large, complex, and fragile analyte molecules
[32,33].

Electrospray mass spectrometry (ES-MS) provides an unrivaled
method for the transfer of preexisting ions and ionic associations from
solution into the gas phase. Although the goal of most researchers has been
on the utilization of electrospray as a sample introduction device for the mass
analysis of organic and biochemical species [34,35], Kebarle and co-workers
[36-39] have focused most of their efforts on the gas phase chemistry of metal
ions and their reactions. The potential of electrospray to evolve into a generic
ion source for elemental mass spectrometry has been reported by different
groups [40,41] who consider it may become competitive with inductively
coupled plasma (ICP) mass spectrometry.

During the development of ES-MS, mass spectra of alkali metals were
generally observed [24], while spectra of alkaline earth and transition metals
have been reported only in the last few years [37-39]. The reasons may be
that other sensitive and selective analytical techniques existed for metal
ions, or that it had been believed for a long time that desorption of multiply

charged metal ions from solution would be diflicult or impossible [42].

6

The determination of gas phase equilibria involving singly charged
ions and solvent molecules was initiated more than two decades ago [43] and
results providing extensive data an ion-solvent and ion-ligand interactions
[44] were reported. However, the transfer of singly and multiply [37-39]
charged ions from the solution to the gas phase involves processes that often
alter their form (charge, degree of salvation, coordination pattern).

As fiequently occurs with a rapidly advancing technique, the
understanding of the physica-chemical processes behind electrospray
ionization has not kept pace with the tremendously rapid expansion in
applications. The mechanism of electrospray, that is, the mechanism of
formation and desolvation of ions during their transfer from solution to gas
phase, does not have yet an unanimously accepted scenario [22,45-48].

Electrospray mass spectrometric studies of aqueous solutions of metal
salts [41] revealed that an unambiguous correlation between the peak
intensities in the ES-MS spectra and the ionic concentrations in solution may
not be readily observable. Although the solution chemistry dictates the types
of ions that are present in the electrospray spectrum, other processes, mainly
desorption-related, may modify the relative abundances of these ions [49-51].
Most of these processes occur in the electrospray source, being therefore
affected by various instrumental and operational parameters. In specifically
designed experiments, the salvation and desorption-related contributions

could be minimized or even canceled, which would allow quantitative

correlations between the equilibrium concentrations of the ionic species in
solution and the electrospray mass spectrometric response.

One of the major problems faced in the analytical application of ES-
MS to metal analysis is the' formation of metal-solvent adduct ions. These
adducts complicate the spectra and spread the analyte signal over several
different ionic species, thereby raising the detection limits. This problem is
less severe for singly charged ions, but it becomes quite complicated for
multiply charged species [38,39]. Another problem encountered in the
analysis of metal ions and their adducts is the loss of charge state
information, that is, the original charge state of the ionic species in solution
is not always preserved in the gas phase.

The first of the problems, and possibly the second, can be largely
overcome by the use of particular macrocyclic compounds, such as crown
ethers, as complexation agents for metal ions [52]. The ability to form host-
guest type complexes with metal ions is one of the most important properties
of crown ethers and related compounds [53]. Macrocyclic ligands were found
to prevent the formation of various metal-solvent adducts and to help
preserve the charge state of the metal ions [54].

The involvement of mass spectrometry in this area of coordination
chemistry started with the need to characterize these complex systems with
progressively higher masses and lower volatility. Mass spectrometric studies
of various aspects of the gas phase chemistry of metal coordination

compounds had been performed even before ES ionization proved its

suitability for such applications. Hence, investigations to elucidate the
mechanisms of ion formation and transport and the stability of the gas phase
complexes were carried out by means of fast atom bombardment [55] and
plasma desorption mass spectrometry [15]. Significant results on ion-
molecule reactions involving particularly transition metals and polyethers
were obtained by Huang and Allison [56] by using ion cyclotron resonance
(ICR) spectrometry.

The application of ES-MS in studying the interaction of metals with
chelating and macrocyclic ligands has lately been found to be very useful in
understanding the behavior of metal ions in complex organic matrices [57].
Metal chelates play an important role in environmental chemistry [58]
especially by regulating the bioavailability of certain metals in natural
waters. The study of such complexes as they exist in solution requires in situ
methods of analysis. ES-MS of solvated metal chelates provides the
structural information which complements data obtained by other
techniques.

Electrospray mass spectrometry, due to its ability to transfer intact
ionic species from solution to the gas phase is expected to enable more
insightful investigations on a wider array of metal ions and their interactions
with various chelating and macrocyclic ligands. The use of an electrospray
ion source in conjunction with tandem mass spectrometry (MS/MS) [59] can

generate new and more complex information about fundamental ion

formation processes, structure and reactivity, as well as the gas phase
chemical equilibria between metal ions and the species with which they
associate.

Our study has a high degree of complexity, being structured in five
distinct parts. The general behavior of solvated metal ions under electrospray
conditions is the subject of Chapter 2. Its emphasis is on qualitative
correlations that exist between the solution concentrations and the
distribution of ions in the gas phase. The appearance of ES-MS spectra help
rationalize the relative affinity of various solvent molecules for certain metal
ions. The ligand-metal binding strength within the clusters translates into
the ease with which the ionic associations preformed in solution “survive”
desolvation and the transport processes within the electrospray source and
through the interface.

Although the degree of structural degradation of solvated ionic
associations is greatly minimized in electrospray when compared to other
“soft” ionization techniques, dissociation occurs to a certain extent, mainly as
a function of experimental conditions. Hence, the so-called “source CAD” or
“up-front CAD” is exploited in order to achieve an “orchestrated survival” of
the species of interest, subsequently used as precursor ions in tandem mass
spectrometry. Chapter 3 deals with the study of the processes involved in
“source CAD”, and the optimization of instrumental parameters according to

our experimental goals.

10

A comparative study of the coordination chemistry of two metal ions
known for their opposite behavior both in solution and in the gas phase,
potassium and silver, is the main objective of Chapter 4. The character and
strength of the metal-ligand interactions are the key factors that determine
the stability of the complexes. Our goal is to establish a relationship, beyond
qualitative considerations, between the appearance of the electrospray mass
spectra and the composition of the analyte solution.

Chapter 5 consists of a more systematic study of collisionally activated
dissociation in conjunction with tandem mass spectrometry. The effect of
particular experimental factors, such as the collision gas pressure and the
collision energy is rigorously studied on specific, well-defined examples. The
study of CAD processes provides a great deal of information about
mechanistic aspects of reactions, as well as the stability of various ions in the
gas phase.

The last chapter of the dissertation concentrates on the applicability of
electrospray mass spectrometry as a method of analysis complementary to
the “classic” methods widely employed in analytical and synthetic
coordination chemistry. A novel macrocyclic ligand and its complexes with
three transition metal ions gave us the opportunity to perform experiments
leading to results unattainable through most other techniques, regarding the
structure and reactivity of the complexes and related structures in solution

and in the gas phase.

11

References

1. Gregor, I. K.; Guilhaus, M. Mass Spectrom. Rev. 1984, 3, 39.

2. Allison, J. Prog. Inorg. Chem. 1986, 34, 627.

3. Bruce, M. I.; Liddell, M. J. Appl. Organometallic Chem. 1987, 1, 191.

4. van Breemen, R. B.; Martin, L. B.; Schreiner, A. F. Anal. Chem. 1988, 60,
1314.

5. Better, L. D.; Conner, K. A.; Rothwell, A. P.; Walton, R. A. Inorg. Chimica
Acta 1988, 143, 109.

6. Distefano, G.; Innorta, G.; Pignataro, S.; Foffani, A. J. Organomet. Chem.
1968, 14, 165.

7. Distefano, G.; Innorta, G.; Pignataro, S.; Foffani, A. J. Organomet. Chem.
1974, 65, 209.

8. Teesch, L. M.; Adams, J. Org. Mass Spectrom. 1992, 27, 931.

9. Cochran, R. L. Appl. Spectrosc. Rev. 1986,22, 137.

10.Miller, J. M. Mass Spectrom. Rev. 1989, .9, 319.

11.Karas, M.; Killenkamp, F. Anal. Chem. 1988, 60, 2299.

12.Allison, J.; Ridge, P. D. J. Am. Chem. Soc. 1979, 101, 4998.

13.Bombick, D.; Pinkston, J. D.; Allison, J. Anal. Chem. 1984, 56, 396.

14.Bryant, J. A.; Blanda, M. T.; Vincenti, M.; Cram, D. J. J. Am. Chem. Soc.

1991, 113, 2167.

15.Malhotra, N.; Roepstorfi, P.; Hansen, T. K.; Becher, J. J. Am. Chem. Soc.

1990, 112, 3709.

16.Li, Y.-T.; Hsieh, Y.-T.; Henion, J. D.; Ocain, T. D.; Schiehser, G. A.;

Ganem, B. J. Am. Chem. Soc. 1994, 116, 7487.

12

17.Man, V. F.; Dale Lin, J.; Cook, K. D. J. Am. Chem. Soc. 1985, 107, 4635.
18.Cook, K. S. Mass Spectrom. Rev. 1986, 5, 467.
19.Volmer, D. A.; Vollmer, D. L. LC—GC 1996, 14, 236.

20.Bose, G. M. In Recherches sur la cause et sur la véritable the’orie de
l’électricité, Wittenberg, 1745.

21.Zeleny, J. J. Phys. Rev. 1917, 10, 1.

22.Dole, M.; Mach, L. L.; Hines, R. L.; Mobley, R. C.; Ferguson, L. D.; Alice,
M. B. J. Chem. Phys. 1968, 49, 2240.

23.Mach, L. L.; Kralik, P.; Rheude, A.; Dole, M. J. Chem. Phys. 1970, 52,
4977.

24.Yamashita, M.; Fenn, J. B. J. Phys. Chem. 1984, 88, 4451.

25.Aleksandrov, M. L.; Gall, L. N.; Krasnov, V. N.; Nikolaev, V. I.; Pavlenko,
V. A.; Skurov, V. A. Dokl. Akad. Nauk. SSSR 1984, 277, 379.

26.Aleksandrov, M. L.; Gall, L. N.; Krasnov, V. N .; Nikolaev, V. I.; Pavlenko,
V. A.; Skurov, V. A.; Baram, G. I.; Gracher, M. A.; Knorre, V. D.; Kusner,
Y. S. Bioorg. Kim. 1984, 10, 712.

27.Fenn, J. B; Mann, M.; Meng, C. K.; Wang, S. F.; Whitehouse, C. M.
Science, 1989, 246, 64.

28.Katta, V.; Chowdhury, S. K.; Chait, B. T. J. Am. Chem. Soc. 1990, 112,
5348.

29.Colton, R.; Traeger, J. C. Inorg. Chimica Acta 1992, 201, 153.

30.Maleknia, S.; Brodbelt, J. S. Rapid Commun. Mass Spectrom. 1992, 6,
376.

31.Colton R.; Klaui, W. Inorg. Chimica Acta 1993, 211, 235.
32.Cody, R. B.; Tamura, J.; Mussehnan, B. Anal. Chem. 1992, 64, 1561.

33.Nohmi, T.; Fenn, J. B. J. Am. Chem. Soc. 1992, 114, 3421.

13

34.Smith, R. D.; Loo, J. A.; Edmonds, C. G.; Barinaga, C. J.; Udseth, H. R.;
Anal. Chem. 1990, 62, 882.

35.Smith, R. D.; Loo, J. A.; Ogorzalek Loo, R. R.; Busman, M.; Udseth, H. R.
Mass Spectrom. Rev. 1991, 10, 359.

36.Blades, A. T.; Jayaweera, P.; Ikonomou, M. G.; Kebarle, P. J. Chem. Phys.
1990, 92, 5900.

37.Blades, A. T.; Jayaweera, P.; Ikonomou, M. G.; Kebarle, P. Int. J. Mass
Spectrom. Ion Processes 1990, 101, 325.

38.Blades, A. T.; Jayaweera, P.; Ikonomou, M. G.; Kebarle, P. Int. J. Mass
Spectrom. Ion Processes 1990, 102, 251.

39.Jayaweera, P.; Blades, A. T.; Ikonomou, M. G.; Kebarle, P. J. Am. Chem.
Soc. 1990, 112, 2452.

40.Agnes, G. R.; Horlick, G. Appl. Spectrosc. 1992, 46, 401.

41.Cheng, Z. L.; Siu, K. W. M.; Guevremont, R.; Berman, S. S. J. Am. Soc.
Mass Spectrom. 1992, 3, 281.

42.1ribarne, J. V.; Dziedzic, P. J.; Thomson, B. A. Int. J. Mass Spectrom. Ion
Processes 1983, 50, 331.

43.Dzidic, 1.; Kebarle, P. J. Phys. Chem. 1970, 74, 1466.
44.Keesee, R. G.; Castleman, A. W. J. Phys. Chem. Ref. Data 1986, 15, 1011.
45.Iribarne, J. V.; Thomson, B. A. J. Chem. Phys. 1976, 64, 2287.

46.Schmelzeisen-Redeker, G.; Butfering, L.; Rollgen, F. W. Int. J. Mass
Spectrom. Ion Processes 1989, 90, 139.

47 .Luttgens, U.; Rollgen, F. W. In Proceedings of the 39th ASMS Conference
on Mass Spectrometry and Allied Topics, Nashville, TN, May 19-24, 1991.

48.Fenn, J. B.; Rosell, J. In Proceedings of the 43rd ASMS Conference on
Mass Spectrometry and Allied Topics, Atlanta, GA, May 21-26, 1995.

49.Tang, L.; Kebarle, P. Anal. Chem. 1991, 63, 2709.

50.Fenn, J. B. J. Am. Soc. Mass Spectrom. 1993, 4, 524.

14

51.Dupont, A.; Leize, E.; Marquis-Rigault, A.; Funeriu, D.; Lehn, J.-M.; Van
Doesselaer, A. In Proceedings of the 43rd ASMS Conference on Mass
Spectrometry and Allied Topics, Atlanta, GA, May 21-26, 1995.

52.Zagorevskii, D. V.; Nekrasov, Y. S. Org. Mass Spectrom. 1993, 28, 1562.

53.Weber, E.; Toner, J. L.; Goldberg, I.; Vtigtle, F.; Laidler, D. A.; Stoddart, J.
F.; Bartsch, R. A.; Liotta, C. L. In Crown Ethers and Analogs, S. Patai and
Z. Rappoport, Eds.; John Wiley & Sons: Chichester-New York-Brisbane-
Toronto-Singapore, 1989, Chapter 4.

54.Van Berkel, G. J .; McLuckey, S. A; Glish, G. L. In Proceedings of the 39th
ASMS Conference on Mass Spectrometry and Allied Topics, Nashville, TN,
May 19-24, 1991.

55.Bonas, G.; Basso, C.; Vignon, M. R. Rapid Commun. Mass Spectrom.
1988, 2, 88.

56.Huang, S. K.; Allison, J. Organometallics 1983, 2, 883.

57.Wahl, K. L.; Stromatt, R. W.; Blanchard, D. L. In Proceedings of the 43rd
ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta, GA,
May 21-26, 1995.

58.Ross, A. R. S.; Ikonomou, M. G.; Orians, K. J. In Proceedings of the 43rd
ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta, GA,
May 21-26, 1995.

59.Gatlin, C. L.; Turecek, F.; Vaisar, T. In Proceedings of the 43rd ASMS
Conference on Mass Spectrometry and Allied Topics, Atlanta, GA, May 21-
26, 1995.

Chapter 2

Metal Ions in the Electrospray Process

Introduction

The increased involvement of metal ions in analytical mass spectrometry has
lately been the immediate result of a continuously emerging need for new
and improved information for structure analysis. Adducts and complexes of
metal ions can be used to determine relative molecular masses of unknown
species, to assist in mixture analysis, and to elucidate structures. Metal ions
used to be primarily used as analytical reagents mainly in conjunction with
desorption ionization techniques such as field desorption (FD) [1,2],
secondary ion mass spectrometry (SIMS) [3,4], plasma desorption (PD) [5],
matrix-assisted LASER desorption ionization (MALDD [6], and fast atom

bombardment (FAB) [7 ,8]. Bath Hodges and Beauchamp [9] and Allison and

15

16

ca-workers [10,11] proposed and demonstrated the use of alkali-metal ions as
chemical ionization (CI) reagents for relative molecular mass determinations.

The main drawback of the above methods consisted of the very limited
amount of structural information that could be obtained. The most direct
information concerning the nature and strength of the interaction of an ion
with a given solvent is obtained when the interaction takes place in a dilute
gaseous phase. The interaction of the ion with one or more solvent molecules
to form so called “salvation clusters” can be studied under such conditions by
applying appropriate experimental conditions [12]. It has been only during
the last two decades that the gas phase chemistry of such compounds has
been investigated to a greater extent for comprehensive structural
information [13,14].

The development of electrospray mass spectrometry (ES-MS), which
exploits the formation of molecular associations involving metal ions in
solution, has recently emerged as the method of choice for the investigation of
metal adducts and complexes, due to its versatility and relative simplicity. To
date. it has been shown that ES-MS, operated under specific experimental
conditions, can be used to quantitatively determine trace levels of both
elemental and molecular inorganic cations and anions with ppb and sub-ppb
detection limits [15]. Also, ES-MS can provide, in some cases, valence state
information [16] and can be employed to establish the presence of ligands
and study complexation [17 ,18]. The sharp increase in the volume of recent

research concerning electrospray in connection with the chemistry of metal

l7

complexes [19-21] has been triggered by its potential to generate abundant

information.

Interaction of Metal Ions in Solution

An ideal analysis technique for solution samples would provide both
qualitative and quantitative information on all solution components. Solution
components include, in addition to the solvent matrix, anions, cations, and
all the forms they may be found in, that is, elemental, molecular, and
complexed. Because solvated ions are sampled directly from the liquid phase,
ES-MS can yield two types of complementary information regarding ion
salvation: the distribution of adduct ions reflects the patterns and degrees of
association in solution, whereas their stability in the gas phase is a measure

of ion-ligand binding in the absence of the liquid phase.

Metal Ion - Solvent Adduct Formation

The potential of ES-MS to evolve into a generic method for elemental
analysis was reported by Agnes and Horlick [15]. One of the limitations most
widely encountered was that a large variety of peaks due to metal-solvent
clusters may be present in the mass spectrum, clusters that further

complicate the spectra and spread the analytical signal of metal-containing

18

ions over a large range of masses. The distribution of these peaks can
however be altered as a function of the experimental conditions [15,22].
Studying simple inorganic metal ions as analytes in the electrospray
process has the advantage that such species have no solvophobic moieties
and they provide good models for ion salvation-desolvation phenomena.
Solutions of metal salts are suitable analytes for our study because first,
metal ions produce readily interpretable electrospray mass spectra, and
second, the solution chemistry of metal salts is generally well characterized,
reliable stability constants of a large number of metal salt species being

available [23].

Experimental. Solutions of 0.1 M NaCl and KC] respectively
were made by dissolving the appropriate amount of salt (ACS-grade)
in minimum required amount of water, followed by dilution with
HPLC-grade methanol (Merck) to the desired concentration. The water
content of the solutions was kept around 0.3%. Electrospray mass
spectra (positive ion made) were obtained using a Finnigan MAT—
TSQO 7000 triple-stage quadrupole mass spectrometer system
equipped with an electrospray source (Finnigan MAT, San Jose, CA).
The solutions were introduced into the electrospray source using a
syringe pump (Harvard Apparatus, South Natick, MA) at a flow rate of

5 ,uL/min. Electrospray voltage was 5 kV and the temperature of the

19

transfer capillary was maintained at 110 C. For the study of the
composition and distribution of metal-solvent clusters, “mild”
conditions (i.e., low voltage settings and temperature on the interface

components) were employed.

In the ES-MS spectrum of the NaCl solution (Figure 2.1) it is evident
that the quantitation of sodium is practically impossible. Sodium is contained
in at least nine ionic species whose peaks have outstanding intensities and
vary significantly with any change in solution matrix or experimental
conditions. Besides, the salvation energies and the instrumental response
factors for various species in the spectrum are unknown, a reliable
relationship between peak intensities and the concentrations in solution
being unfeasible at this point.

Even though the ES mass spectrum does not reflect the relative
concentrations of various ions in solution, it qualitatively consists of species
that either preexist in solution or are formed during the solvent evaporation
and interface processes. Metal ion adducts containing methanol in their
salvation shell are generally present in higher abundances than the ones
containing water molecules. Although the very low water concentration in
the analyte solution warrants a partial reason for this situation, the
enhancement in the stability of adduct cations can be attributed to a more
favorable structure of the primary salvation shell toward coordination of the

metal ion, due to the presence of methanol molecules. Clusters with

20

com

o o 2 u 238353 basque 8:385 s as: do 238% msam 5 9:9“.

 

 

 

 

 

 

 

 

 

 

 

 

es
omm com one cow on o
— p p P _ p + . . F
_ 2g 3:11—qu A! . I O
\ :35 we
a: as gates. M
Egg-59%.; M.
5: as M
.EIooéioiaz. %
/ / . a.
Ana #5 \M
/ .5 low
a: as
ESE-z. Gauss
.1 .2835.
335 A: as
._.=o.§.aoiwz_ .zzoixoaaaa
/ roe
E #5

Hie-aux.

21

coordination numbers of 1 (m/z 55), 2 (m/z 87), and 3 (m/z 119),
corresponding to three of the predominant ions in the spectrum illustrate not
only that they or at least their precursors originate in solution, but also their
ability to “survive”, as intact structures, the desolvation and ion transfer
processes. Species containing water, such as the (1:1) sodium-water adduct
(m/z 41) and the (1:1:1) sodium-water-methanol ion (m/z 73) are present in
the spectrum, but their peaks have relatively low intensity.

Furthermore, another solvent-derived ligand takes part in the metal
adduct formation. We assume that in situ intermolecular dehydration of
methanol (Reaction 2.1) is responsible for the existence of dimethylether as a
ligand in some varieties of partially desolvated clusters. Joint coordination of
sodium by methanol and dimethylether is characteristic of the ions at m/z
101, m/z 133, and m/z 179. It can be noticed that coordination with
dimethylether allows for a coordination number of 4 (i.e., m/z 179), compared
to a maximum of 3, encountered in sodium clusters that contain methanol

only, or methanol and water.

 

CH30H + CH30H > CH30CH3 + H20 (2. l)

The ES-MS spectrum of a potassium chloride solution (Figure 2.2) is
similar to that of sodium chloride, in terms of the generic structures present.
However, the relative ionic abundances of clusters with higher coordination

numbers (3 and 4) are greater than those of analogous sodium clusters. Due

22

com

o o: u 85.23.23 Emsqmo .chSoE 5 Ex do E383» msam «N 2:9...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ne
0mm cow om? cow om o
p p _ F P b — r P . p p
i, .4; 1.“ 434..— l. 10
E as .
.zoaxi
Mu
B:
B
. fl..
A
4/ 6
/ an 5.5 w.
c. e
A: as . w
\ .255: . «M.
to/zo
we as as as (\
.zzoégioia .=:o.axo~:Z
a: s5 \
.Eiioia \ /
as as r cow
as as his
.3689:

E. as
.zzooéioii

23

to its larger ionic radius and electron affinity [24], potassium can more easily
accomodate more solvent and solvent-derived ligands in its primary

coordination sphere.

Salvation Effects

The term “selective salvation” is applied to the case where the composition of
the solvent components in the neighborhood of the ions, that is, their
salvation shell, is difl’erent from the composition of the bulk solution. The
selective salvation of ions in binary mixtures depends on the free energy of
salvation of these ions in the two pure solvents. In mixed solvents it must be
recognized that different types of solvent molecules may interact individually
and to different extents with the metal ions present in solution.

It was commonly assumed [25] that ions in binary solvents are
predominantly surrounded by molecules of the more polar constituent,
namely, water in partially aqueous organic media, such as the methanol-
water solutions employed in our study. Grunwald and co-workers [26]
reported that simple inorganic ions are, on the contrary, appreciably solvated
by the organic solvent molecules in mixed aqueous organic solutions.

The ion cluster formation and binding in solution is part of the reason
for their behavior during the electrospray and transport processes. The
complexity of ion-solvent interactions is well illustrated by the evidence

concerning the relative “basicity” of, for instance, methanol and water. The

24

coordination of methanol molecules around an alkali-metal ion is enhanced
compared to the coordination of water (Figure 2.3.a and 2.3.b), due to the
electron-releasing (+I) effect of the methyl functional groups involved. This
electronic effect increases the electron density on the oxygen atom, therefore
increasing its donor character, that is, the coordination affinity for the metal
ion. In the case of dimethylether, the effect is even more significant, two
methyl groups being now responsible for the increased binding ability of the
oxygen atom in the ligand (Figure 2.3.c).

A similar experiment, conducted with a solution of 0.1 M potassium
chloride in l-propanol, is useful for the partial assessment of the salvation
trend within a homologous series of alcohols. The temperature of the transfer
capillary was in this case 120 C. Unlike the methanol solutions, according to
the appearance of the ES-MS spectrum (Figure 2.4), there is no indication of
intermolecular condensation of the solvent molecules involved in
coordination. Propanol and propanol-water adducts are present with variable
coordination numbers between 1 and 4. The peak intensity of the
uncomplexed metal ion (IQ) is significantly lower compared to its intensity in
methanol solution. This is due to a more favored complexation of K+ by the
alcohol molecules, facilitated by an increased donor character of 1-propanol.

By increasing the temperature of the transfer capillary to 160 C, the
appearance of the ES-MS spectrum changes (Figure 2.5). Smaller relative

peak intensities for the adducts containing water, such as m/z 177 and m/z

2S

 

Figure 2.3 Generic structures of primary salvation shells around an
alkali-metal ion. Ligand: (a) water, (b) methanol, and
(c) dimethylether

26

0 on“ u EEEmQEoL >5:qu 6:395; 5 Ex mo :53QO m2-mm vN 2:9“.

 

 

 

 

 

 

 

~\E
com com com o? o3 om o
- b - P P I P b - b - L
in} 3 67.4. .4 ..
f \ / an as
as as .x
c: as .zzoéx.
.zouafoéa
Mu
B.
was
1 M.
a
M
9
u
S
as as / rm"
.25.“: am. as . \o/M
.1295. (\
c8 s5
.zoaaafoii /
\ E— 65
hogafoéx.
Gad—5 I 00—.

 

.225.—

 

27

o 8 a M 958353 55me 3:323 E Bx do SEES» w:-mm flu 2:9...

 

 

es

com 0mm com 09. oo_. o

. _ . . . _ t

25 j J— / I o
\ / an as

as as \ \ as as .5.

5:95.. EN ea. c : as axes:
._.o£§o£x_ .zoazxxoéx.

an. e5
.1395.

 

\

c: as
._.o£§o£x_

 

am. as
.2333:

 

(%) insuetuu GAQBIGH

 

:2

28

237, denote a lower “survival” rate of such species during their transfer
through the electrospray interface. The relative intensities of all-propanol
adducts, such as m/z 159 and m/z 219 increases, these ions being probably
products of in-source, heat-induced dehydration of the species at m/z 17 7
and m/z 237 , respectively. No quantitative correlation between the peak
intensities and the extent of water loss from the coordination shells can be
found, due to the fact that the salvation energies of the adducts involved are
unknown. However, the qualitative conclusion regarding the binding
strength between the metal ion and the two ligands is consistent with the
difference in the donor character of water and 1-propanol.

[K(PrOH)(HzO)]* (m/z 117) is an apparent exception to the above
rationale, its ionic abundance slightly increasing with an increase in the
capillary temperature, while the peak intensity of its immediate dehydration
product [K(PrOH)]+ (m/z 99) follows the trend. This is consistent with the
assumption that, from a mechanistic point of view, [K(PrOI-D(H20)]+ is an
intermediate in a multiple-step fragmentation of heavier propanol-water
adducts, a process that ultimately leads to the formation of [K(PrOH)]+ or
even Kt.

Due to its versatility and relative simplicity, ES-MS is perfectly suited
for studying the strength of such metal-ligand interactions in a variety of
solvents. By modifying the operational parameters of electrospray and/or the

interface voltage or temperature settings, various degrees of desolvation can

29

be achieved. Chapter 3 consists of a more detailed discussion about specific

processes that occur during ion formation, desolvation, and transport.

30

References

1. Schulten, H. M. Int. J. Mass Spectrom. Ion Phys. 1979, 32, 97.

2. Budzikiewitz, H.; Linscheid, M. Biomed. Mass Spectrom. 1977, 4, 103.
3. Detter, L. D.; Walton, R. A. Polyhedron, 1986, 5, 1321.

4. Day, R. J.; Unger, S. E.; Cooks, R. G. Anal. Chem. 1980, 52, 557A.
5. Sundqvist, B.; McFarlane, R. D. Mass Spectrom. Rev. 1985, 4, 421.
6. Karas, M.; Killenkamp, F. Anal. Chem. 1988, 60, 2299.

7. Cochran, R. L. Appl. Spectrosc. Rev. 1986, 22, 137.

8. Miller, J. M. Mass Spectrom. Rev. 1989, 9, 319.

9. Hodges, R. V.; Beauchamp, J. L. Anal. Chem. 1976, 48, 825.
10.Allison, J.; Ridge, P. D. J. Am. Chem. Soc. 1979, 101, 4998.
11.Bombick, D.; Pinkston, J. D.; Allison, J. Anal. Chem. 1984, 56, 396.

12.Marcus, Y. Ion Salvation, John Wiley & Sons Ltd.: Chichester-New York-
Brisbane-Toronto-Singapore, 1985, p 12.

13.Kebarle, P.; Davidson, W. R.; French, M.; Cumming, J. B.; McMahon, T.
B. Disc. Faraday Soc. 1977, 64, 220.

14.Kebarle, P. Ann. Rev. Phys. Chem. 1977, 28, 445.
l5.Agnes, G. R.; Horlick, G. Appl. Spectrosc. 1994, 48, 655.

l6.Cheng, Z. L.; Siu, K. W. M.; Guevremont, R.; Berman, S. S. J. Am. Soc.
Mass Spectrom. 1992, 3, 281.

17.Jayaweera, P.; Blades, A. T.; Ikonomou, M. G.; Kebarle, P. J. Am. Chem.
Soc. 1990, 112, 2452.

31

1800an R.; Klaui, W. Inorg. Chimica Acta 1993, 211, 235.

19.Wahl, K. L.; Stromatt, R. W.; Blanchard, D. L. In Proceedings of the 43rd
ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta, GA,
May 21-26, 1995.

20.Xu, Y.; Zhang, X.; Yergey, A. L. In Proceedings of the 43rd ASMS
Conference on Mass Spectrometry and Allied Topics, Atlanta, GA, May 21-
26, 1995.

2 l.Wang, J .; Ke, F.; Guevremont, R.; Siu, K. W. M. In Proceedings of the 43rd
ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta, GA,
May 21-26, 1995.

22.Van Berkel, G. J .; McLuckey, S. A.; Glish, G. L. In Proceedings of the 39th
ASMS Conference on Mass Spectrometry and Allied Topics, Nashville, TN,
May 19-24, 1991.

23.Smith, R. M.; Martel], A. E. Critical Stability Constants, vol. 4: Inorganic
Complexes; Plenum: New York, 1981, p 135.

24.Handbook of Chemistry and Physics, 65'Eh ed.; CRC Press, Inc.: Boca
Raton, Florida, 1984.

25.Schneider, H. In Solute-Solvent Interactions; Coetzee, J. F. and Ritchie, C.
D., Eds; Marcel Dekker: New York and London, 1969.

26.Grunwald, E.; Baughman, G.; Kohnstam, G. J. Am. Chem. Soc. 1960, 82,
5201.

Chapter 3

“Source CAD” - An Advantage or a

Hindrance?

Introduction

As discussed in Chapter 2, in order to take advantage of the main feature of
electrospray “ionization”, namely, achieving minimum fragmentation of
preexisting ionic species during their transfer from solution into the gas
phase, all operational parameters must be carefully chosen and closely
monitored. Electrospray voltage, solution flow rate, sheath and auxiliary gas
flow rates, voltage and temperature on the interface components, the entire
array of ion optics settings, and so on, all affect, in different ways and to
various extents, the “deviation” of electrospray from perfection. In other

words, they determine the closeness of the appearance of an ES-MS spectrum

32

33

to being an ideal reproduction of the ionic composition of a given analyte
solution.

Depending on the purpose of a certain experiment, one can determine
the most favorable set of parameters, usually at the expense of trading off
other potential collateral experimental findings. Three positive-ion modes are
common, a metal-ion made, an ion cluster mode, and an intermediate mode
[1], named for the type of information eventually available from the mass
spectrum. For instance, for the elemental analysis of metal ions in a liquid
matrix one would have to employ a procedure that strips metal adducts of
their coordination shells, regardless of their structure and binding strength,
to the maximum possible extent. In the ion cluster mode, which this chapter
focuses on, the electrospray source is run under so-called “mild” conditions,
and the mass spectra are quite complex since the ion-solvent cluster
distribution generated during the ES process is preserved relatively intact.
With this mode it is possible to determine the solution valence states of the
cations, and the data obtained in this mode are generally more
representative of solution speciation. The third mode consists of an
intermediate approach in which partial declustering occurs, some of the ion
clusters may contain counterions, and various species with reduced charge

may be observed.

34

Structural Transformations within the Electrospray Source

A schematic of the dual-stage Finnigan MAT electrospray source is shown in
Figure 3.1. Immediately after the liquid droplets emerge from the tip of the
electrospray needle, the solvent begins to evaporate. Because of the pressure
difference between the chambers at both ends of the transfer capillary,
solvent evaporation, ion desolvation, and ion transport all occur during the
relatively short time required for the transfer to the first pumping stage. At
the end of the capillary, a “steering” voltage is applied, the ions being
therefore accelerated. Various degrees of ion fragmentation occur in this
stage, during the free jet expansion and ion acceleration, the process being
analogous to some extent to CAD that takes place in the collision chamber of
an MS/MS experimental arrangement. It was reported that the product ions
generated during “source CAD” are more efficiently transferred into the
quadrupole mass analyzer than the CAD products in a triple quadrupole
tandem system [2].

Thus, the in-source fragmentation can be regarded as a hindrance
reducing the ability of ES-MS to provide comprehensive solution chemistry
elucidation, or it can be exploited as a tremendous advantage. Many ionic
species that preexist in solution or are formed during ion desolvation and

transport are very fragile, some are unobtainable through any other gas

35

meshes 5.65s; 82: - $38383 :m :0 £528 2: 3 9:9“.

3%
9.553
28%

 

U...

/
5652.83 mums

28285
_

39m.

9.353

 

\

 

Hem

32%

4E

5:58
.929»

822 3985on

23 3:5
4

L c.8520 :25an .285

 

 

Ill Til
29:8 2.5:

 

 

2:8: Eamecom

 

 

 

 

36

phase processes, and only a careful and systematic investigation taking into
account all processes and phenomena that occur in the electrospray source
can be of benefit. This chapter deals with specific aspects and examples of a
variety of structural transformations that occur within an electrospray

source, including structure-reactivity relationships.

Metal Ion - Solvent Adducts

The ion adduct formation involving certain alkali-metal ions in solutions of
two aliphatic alcohols was partially covered in Chapter 2, as a part of the
general behavior of metal ions in the electrospray process. The reason one of
the most common alcohols, ethanol, though not common in electrospray
studies, was not employed in the general study is that it offered us a wealth

of fascinating results, worth a separate, more detailed subchapter.

Experimental. Solutions of NaCl and KCl (0.1 M) respectively were
made according to the general procedure, by dissolving the salt (ACS
grade) in the minimum required amount of distilled water, followed by
dilution with dehydrated absolute ethanol (McCormick Distilling Co.).
The sample solutions were introduced continuously into the ES source
with the aid of a syringe pump (Harvard Apparatus) at a flow rate of 5
uL/min. The electrospray voltage was maintained at 5 kV. The

temperature of the transfer capillary was varied between 110 C and

37

220 C. Argon was used as the collision gas in the CAD experiments.
Both ES-MS and MS-MS spectra were acquired using a signal

averaging procedure, 60 spectra being averaged throughout the study.

Concerning their general appearance, both ES-MS spectra of NaCl and
KCl in ethanol consist of much fewer outstanding peaks than the spectra of
methanol solutions of both metals recorded under similar conditions (Figure
2.1 and 2.2). Most of the adducts of both metal ions have structurally
identical ligand arrangements around the cation, significant differences
being noticeable regarding the intensity of their peaks.

At a capillary temperature of 110 C the ES-MS spectrum of NaCl
(Figure 3.2) is dominated by the peaks at m/z 147 and m/z 179. Other
significant peaks present include Na+ (m/z 23), [Na(EtOH)]+ (m/z 69), and
[Na(EtOH)2]+ (m/z 115). At this point it is not clear, based only on the
appearance of ES-MS spectra whether the peak at m/z 179 represents a
singly charged or a multiply charged ionic cluster. With an increase in the
capillary temperature, significant changes in the relative intensities of these
and other peaks take place, amid a slight decrease in the total ion current. At
160 C the peak at m/z 147 becomes the base peak (Figure 3.3), mainly due to
the dramatic decrease in the relative intensity of m/z 179. Also, as the
temperature of the transfer capillary is raised, the relative intensity of an

apparently unexpected peak (m/z 287) increases steadily, becoming the base

38

o o: u 258353 >5:me 6:26 5 Bmz a: 8:863 msam Nm 93mm

 

 

 

 

 

NE
8: com cow 2: o
. _ p _ p p . P .
.. o
i. \ .
am :5
.s .
so as
axons; . w
n m.
.. m
:2 as m
. m...
a: as W.
.Ezomsz. . W.
a: as .
s u
auzofzxomxa A 2. \E. r 8?

39

o 2: ... $2352 bmsqmo 3:23 5 amz 3 Eggs $8 on 23:

 

 

 

 

 

 

NE
‘ 8w com com 2: o
— . r P _ _ r
‘14:“..1 i— . lo
5 NE
.2 .
60 £5
:3 $5 .zxomwz. . H
m.
. m
a: £5 . w
fizoinfomwa :2 p35 m.
:3 E. 52

40

peak at a temperature of 220 C (Figure 3.4). Again, for the determination of
its structure a tandem mass analysis with precursor ion selection is
necessary. The peak intensities corresponding to the predominant species in
the spectra were rationalized to the total ion current for the entire scan range
(m/z 10 to m/z 450) and are summarized in Table 3.1.

Overall, several conclusions can be drawn regarding the ES-MS

spectra of sodium chloride in ethanol:

0 The ion abundance of clusters containing water decreases with an
increase in the capillary temperature relative to the anhydrous
adducts. This is due to the fact that water is less strongly
coordinated to the metal cation (see Chapter 2) compared to, for
instance, an aliphatic alcohol;

0 Adduct stripping is enhanced by a higher temperature, one of the
efl‘ects being an increase in the peak intensity of Na+ (m /z 23);

° The peak intensity of m/z 287 increases, absolutely and relatively,
the ion being probably a product of an intermolecular collisionally

activated reaction involving solvent molecules.

Potassium chloride solutions behave similarly under identical ES-MS
experimental conditions (Table 3.2). There are, however, a few discrepancies
between the spectra of potassium solutions versus sodium. The ion of m/z

195 (structurally analogous to m/z 179 in the ES-MS spectra of sodium

41

0 8m u. eaquEmn $3me .chso s 6.2 B 25%QO m2.mm in 9:3

 

 

 

 

 

 

 

 

NE
8... 08 SN 2: o
_ r F . . . _ . .
n _ I o
_ A . _
SNWEV . H.
.«z m.
. m
m.
:2 NE . m...
a: £5 . lax/W
/ :3 NE .
:3 ”E f
2:

42

:3 N}: 3 3 N}: mummy ~88. 25 .8.“ 32.26 :3 H33 on... 3 zuwmqmpfi Mama 2% wgacfimu mega—SQ ..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Em 3 m6 3 3 Na «in w.m can
fix 3 m4 mm mm «xc fin wd cam
flu fin km S 3 ad QN mé cw“
ad m.m 3 «v ms : QN arm 2;
m4 9m 3 av as 2 «mm m; 2:
m6 was ow mm Wm c6 m4 m4 3;

c HA .3 mm as Nu m4 , «A c:

adv a}: pwu a}: a: a): h: a): a: $5 2: aid .5 a}: an a): 6v 235.—wagon.
3% «56:3:— Adon— 60313393“ 3.35.30

 

 

mgzuoLmQES bSNES 5‘35: 2%. .8 ~82ng c me .3233
E 832 s senses mzwm 2: 5 Se “assessors 2: é assesses «8a sfingsssm s .m 2.3.

 

43

c9. N}: 3 3 ~\E mung 58m as... .Sm 39:3 :3 ~33 one 8 36¢an Mama 9% wflfifinofiwu mononom «

 

 

 

 

 

 

 

 

ma 8 c m... o 3 2 2 8m
3 E a «A c 3 3 2.. 8m
.3 an no 3 3. ms 3 mm a:
Z S S. 3 3. S 3 mm o:
3 . E 3” 3. 3 . mm 3 E 2:
3 B 3 3 3 mm 3” S a:
3 3 mm 3 3. an we 3 o:
as ea 2” ea 5: ea 3; £8 1.: ea s2 s8 mm s8 an s8 6v 232383

 

 

 

 

 

 

 

 

 

 

 

15 {36:02: Adam wonzdnomadm thud—:35
BEGSQES $83.38 .5326: 3: \o :02ng d me .3553
5 5M .3 siege mgmm 2: s 33 “ceaseless 2: é meanness seq Easiest a.” sea.

 

44

solutions) is depleting at a much faster rate as the temperature increases.
This is due to the fact that the potassium adduct is probably less stable in the
gas phase compared to its sodium counterpart. The surprising appearance of
a possible product of an intermolecular collisionally activated reaction at m/z
287 in the case of sodium solutions is confirmed once again when potassium
is involved. Furthermore, the peak intensity of the analogous ionic species at
m/z 303 increases more steeply with the temperature of the transfer
capillary (Figure 3.5, 3.6, and 3.7). Thus the hypothesis that an
intermolecular condensationldehydrogenation process involving probably six
ethanol molecules may lead to the in situ formation of 18-crown-6 ether,
known for its very high affinity for alkali-metal ions.

A tandem mass spectrometric experiment was performed in order to
elucidate the charge states and structures of some unassigned peaks present
in the ES-MS spectra. First, CAD of ions of m/z 195 was performed aiming at
the determination of their fragmentation path and charge state. The product
ion spectrum is shown in Figure 3.8. A very “clean” dissociation of the
precursor ion generates a symmetrical fragmentation pattern, somewhat
characteristic of doubly charged precursor ions (see Chapter 6). In this case
however, the product ions are not symmetrically distributed on both sides of
the precursor on the m/z scale, but they are both the result of an identical

neutral loss, the difference between the two reaction pathways being the

45

o o: n 25285 5:8 3:25 5 6x3 Essa» Sam 3.. 2%

 

 

 

NE
oov com com co? 0
p P _ p b p
/ A -.
an as
8% NE / .x
a» £5 . who
axons: m...
38 NE . m
. m
as as / .
a~_.o~:zzoéx_ t: u}:

 

 

roo—

46

o a: N 232353 $36 3.5% 5 6x .3 Essa” 23m 3 2%:

 

 

 

E
00% com com 2:. o
— p p p P b p
I o
as. $5 /
.8 s5 . m
.zzomi . W.
/ . m.
\ an S: . m,
68 Q5 sass .x . M
a~_.o~:E:omZ 10..
c: E.

 

 

roo—

47

0 8w u assesses 338 3:28 5 6x .8 eggs aim 2 9:3

NE
cos com owm

P p p p

o? o
_

 

:2 s5 .

 

g. as . /

y
(%Hl!SU91UI eAueIea

/ .

as as. -02

 

 

48

me ~an assume. :2 Essa 355%: mm 93m

NE

com 9.: om: om . o
_ _

 

44 1 :1:V44 ‘ 1 ‘ IO

32 NE
gazoflzizeag

(%) Kusuelun amenaa

 

\

E. NE
signifier. 52

 

 

49

occurrence of charge reduction. The proposed mechanism is given in

Equations 3.1 and 3.2.

[K(EtOH)3(HzO)]22+ —-> [K2(EtOH)3(HzO)]2+ + 3 EtOH + H20 (3.1)
(m/z 195) (m/z 117)

[K(E1:C)H‘)3(Hz())]22+ ——-> 2 K“ + 6 EtOH + 2 H20 (3.2)
(m/z 195) (m/z 39)

Thus, the peaks at m/z 117 and m/z 195 in the ES-MS spectra of KCl
solutions in ethanol represent doubly charged ionic species that originate in
solution. Additional proof is provided in the product ion spectrum that results
in CAD of the ions of m/z 117 (Figure 3.9). The collision energy of the
precursor ions was 20 eV, and the collision gas pressure 0.3 mTorr. The sole

fragmentation product is the bare metal ion K+ (m/z 39), in agreement to

Equation 3.3.
[K2(EtOH)3(I-I2O)]2+ —> 2 K+ + 3 EtOH + H20 (3.3)
(m/z 117) (m/z 39)

Apparently, according to our preliminary ion assignments, most of the
adducts of interest contain the metal ions. Tandem mass spectrometry with
CAD also provides a confirmation of this fact, by means of a “reverse”
analysis approach, namely a product ion selection in conjunction to the mass

analysis of its precursors. Therefore K+ (m /z 39) was selected as the product

50

N : ~an 5:33” 8.. cusses $5,..an 2” 23:

 

 

(%)M!sue1UI eAneIea

 

1///// r
c: E.
azouxizommx. /

a» £5 ,

 

 

[cow

51

2.. #5 go .550on :2 8&8ch «555%: ofim 9:9“.

 

 

 

 

 

NE
com 03 8F om o
. P . _ . — p p .
no
:5: i: j... _ _ __ _‘/_ .
an as .
.x
semi / .
.2132 6» NE .
/ . aroma: T
sass
auxouzizomvg .

/ .

t: gs
azofzzomg r 2:

 

 

(%) Kusuexuu eAuelea

52

ion of interest, and the CAD spectrum containing the precursor ions which
were fragmented into K+ is presented in Figure 3.10. The collision energy
employed was 30 eV, and the collision gas pressure 0.4 mTorr. As it can be
seen, all the adducts of m/z 85, m/z 117, m/z 131, m/z 163, and m/z 195
include potassium in their structures. The reason the peak at m/z 303 is not
present in this precursor ion spectrum is not that it does not contain
potassium, but rather that the dissociation did not occur under the specific
conditions applied, the collisions induced being energetically insufficient for
the fragmentation of a complex with a very high stability, exhibited both in
solution and in the gas phase. A similar experiment for sodium results in the
ruling that the ions of m/z 101 and m/z 179, respectively, are doubly
charged as well, and structurally analogous to the potassium-containing ions

of m/z 117 and m/z 195, respectively.

53

References

1. Agnes, G. R.; Horlick, G. Appl. Spectrosc. 1994, 48, 655.

2. Voyksner, R. D.; Pack, T. Rapid Commun. Mass Spectrom. 1991, 5, 263.

Chapter 4

Interaction of Metal Ions with Macrocyclic

Ligands

Introduction

Although various aspects of metal ion-ligand interactions had been studied
before 1974 [1,2], it was then that Cram [3] introduced the term “host-guest
chemistry” to describe the developing field of synthetic complexation
chemistry, exemplified by cryptands, crown ethers, and related structures.
Host-guest chemistry is primarily concerned with elucidating the “rules of
non-covalency” [4], involved in the recognition and binding of a guest by a
specific receptor (host).

A host is usually a large and geometrically concave organic molecule
that can non-covalently interact with and bind a guest. Hosts may be acyclic,
macrocyclic, or oligomeric, and possess cavities or clefts into which the guest

54

55

fits. The host’s recognition site or sites for the guest may be inherent in its
normal structure or may be organized during the process of interaction. The
binding sites may interact with guests by combinations of various non-
covalent means, such as hydrogen-bonding, ion-ion, ion-dipole, a, van der
Waals, electron donor-acceptor, and hydrophobic interactions. A certain
degree of covalency may sometimes exist, along with non-covalent
interactions, depending on the nature and structure of the guest and the
particular coordination sites involved.

Guests are simpler organic or inorganic molecules or ions, whose
epitopes present divergent binding sites complementary in charge and steric
requirements to the host. The term “epitope” defines the part of the guest
that actually interacts with the host, an extension of the use of the term in
immunology, where the epitope is the portion of the antigen actually
recognized by an antibody [4]. Typical guests include metal ions, ammonium
ions, polar neutral species, hydrogen-bonding compounds, aromatic
substrates, diazonium salts, halides, and many others. The interaction of a
host with a guest produces a complex. One of the most significant properties
of macrocyclic ligands, such as crown ethers and analogous structures is the

formation of “host-guest”-type complexes with a wide variety of metal ions.

56

Crown Ethers and Aza-Analogs

The macrocyclic polyethers, termed “crown ethers” from their structural
resemblance to crowns, were first synthesized by Pedersen [5] and since then
cyclic polyethers containing up to twenty oxygen atoms have been prepared
[6,7]. Crown ethers were among the first synthetic reagents found to bind
strongly to alkali metal cations and they have been widely used to transfer
alkali metal salts into non-aqueous phases, and in solvent extraction studies
[4].

The structural features of polyether macrocyclic ligands are
represented by systems containing the unsubstituted 1,4,7,10,13,16-

hexaoxacyclooctadecane (L), commonly named 18-crown-6 (Figure 4.1).

[($01] Fufi Fifi]
Loj [Levi] Lad

L L, L”
18-crown-6 aza-18-crown-6 diaza-lB-crown-G

Figure 4.1 The structures of the 18-member macrocyclic ligands
considered in the present complexation studies

57

Substituting one or more oxygen atoms with other heteroatoms, such
as nitrogen or sulfur dramatically changes the complexation ability of the
ligand and the properties of the resulting complexes, both in solution and in
the gas phase [8]. Our comparative study focuses on the behavior of
potassium and silver complexes with a series of macrocyclic systems having
the 18-crown-6 hexaether skeleton, namely 18-crown-6 (L), aza-18-crown-6
(1,4,7,10,13-pentaoxa-16-azacyclooctadecane) (L'), and diaza-18-crown-6
(1,4,10,13-tetraoxa-7,16-diazacyclooctadecane) (L") (Figure 4.1).

While the steric compatibility between the cations and the cavity of the
ligand is recognized as a key factor in complexation [9] the electronic
structure of the metal ions also affects the selectivity of the macrocyclic
ligands towards cations [10]. The stability constants for the interaction of
potassium and silver with 18-crown-6 hexaether and its aza and diaza
derivatives in solution (Table 4.1) show that the ligand affinity for K+
decreases while that for Ag+ increases in going from the oxa (L), to the aza

(L’), and then diaza (L”) ligand.

Table 4.1 Comparison of log Ks values for the complexation
of L, L’, and L” with K" and Ag“ in methanol [1 1]

 

 

 

 

 

 

 

 

Cation Ligand
L L! L”
K‘ 6.29 4.18 1.80
Ag‘ 4.05 6.03 9.99

 

 

58

As it will be further elaborated in this chapter, the relative peak
intensities of the potassium and silver complexes, as well as their
uncomplexed ligands in the ES-MS spectra follow, qualitatively, the trends
expected from Table 4.1. Tandem mass spectrometry with collisionally
activated dissociation (CAD) provides additional information regarding the
contrasting behavior of the complexes of the two metals in the gas phase. The
different dissociation patterns observed are direct consequences of the
binding strength between the metal and the coordination sites, as well as the
nature and distribution of the potential protonation centers within the
complex structure. Also, the relative gas phase stabilities of the complexes
considered can be qualitatively assessed based on the various degrees of

induced fragmentation observed.

Interactions in Solution

Many solution studies have been performed on the relative stability of
complexes of different metal cations with various crown ethers and aza-
analogs with particular emphasis on discrimination between the metals
[9,12]. However, determination of the stoichiometry of complexes in solution
is significantly more difficult because the systems are very labile and
especially because most of the techniques used give only average

compositions. Christensen and co-workers [13] suggested that caution be

59

exercised in assessing the structures in solution from the solid state
structures of crown ether complexes with metals.

The stability constants of complexes of macrocyclic ligands with a large
number of metal cations were determined by using a wide variety of physico-
chemical methods [14]. As a measure of the complexing strength between
metal ions and ligands in solution, the stability constants are a function of
ligand structure, cation size and type, and solvent.

In methanol solutions, the stability constants of the complexes studied
follow opposite trends for potassium and silver, respectively [11]. The
stability of potassium complexes in solution decreases with an increase in the
number of nitrogen atoms in the ligands, whereas in the case of silver
complexes, their stability increases with the number of nitrogen donors in the
ligand. Potassium is an A-type, “hard” acceptor [15], and so it interacts most
readily with A-type, “hard” donor atoms, like oxygen. Consequently,
potassium complexes derived from ligands containing only oxygen atoms
have high stability constants (K) values (Table 4.1). The introduction of B-
type, “soft” donors such as nitrogen atoms in the ligand structure gives a
destabilizing influence to the coordination process of “hard” cations. As
shown in Table 4.1, the stability of the nitrogen-containing complexes of
potassium, (KL’)+ and (KL”)+ is much lower than the stability of their oxa-
crown analog (KL)+, while increased Ks values characterize the complexation

of silver, a B-type, “soft” cation, with ligands containing nitrogen atoms.

60

Complexation of potassium is weakened appreciably by increasing the
number of nitrogen atoms in the host. This is just as expected: as the
negative charge on the binding site drops, the electrostatic attraction
between it and the metal cation is diminished. The efi‘ects of nitrogen atoms
on silver complexation are exactly the opposite: binding strength is greatly
increased at these sites, although it is not electrostatic interaction that
matters here, but rather an increased degree of covalent bonding, which is
somewhat characteristic of silver complexation with amine functional groups.

The degree of covalency involved in the complexation of silver cannot
be determined from solution chemistry alone. Neither can the contribution of
electrostatic interactions to the stability of complexes be assessed. On the
other hand, though it is highly unlikely that covalency plays any role in the
complexation of potassium, solution chemistry studies do not rule that out
either.

The technique of electrospray mass spectrometry (ES-MS) is well
suited to complement existing methods for studying solutions of metal
complexes with macrocyclic ligands. Unlike other forms of ionization,
electrospray allows the preservation of most ionic structures that preexist in
solution, and most of their characteristics during the transfer from solution

into the gas phase.

Experimental. All solutions were made by dissolving KCl and AgNOa

(ACS-grade) in minimum amount of water followed by dilution with

61

HPLC-grade methanol (Merck). The total metal ion concentrations
were 1 mM, while the water content of each solution was kept around
0.3%. The ligands were used as supplied (Aldrich) and added to each
metal salt solution to give a ligand concentration of 1 mM.

Positive ion electrospray mass spectra were obtained using a
Finnigan MAT-—TSQ‘D 7000 triple-stage quadrupole mass spectrometer
system equipped with an electrospray source (Finnigan MAT, San
Jose, CA). The solutions, prepared as described, were directly
introduced into the electrospray source using a syringe pump (Harvard
Apparatus) at a flow rate of 5 ,uL/min. The electrospray voltage was 5
kV and the temperature of the transfer capillary was maintained at
200 C.

For the MS/MS study, argon was used as collision gas, its
collision cell pressure being 1 mTorr. The laboratory collision energy
(ELAB) of the precursor ions was adjusted according to their mass to
provide similar center-of-mass collision energy (Ecm), for consistency.
Hence, ELAB was set and maintained at 22 eV for potassium complexes
and 26 eV for silver complexes, ECM being, in both cases, approximately
equal to 2.5 eV. The ion selection stage, that is, the first quadrupole
analyzer, was tuned to unit mass resolution to allow for unambiguous
precursor selection. Typically, 60 signal-averaged spectra were

acquired for both full and product spectra throughout the study.

62

Interactions in the Gas Phase

M

In the full ES-MS spectra of the potassium complexes (Figure 4.2) the
predominant ions are K+ (m/z 39), and the singly charged intact (1:1)
complexes, KL+ (m/z 303), (KL')+ (m/z 302), and (KL")+ (m/z 301)
respectively. In the case of the oxa-crown potassium complex (KL)+ (Figure
4.2.a) no protonated uncomplexed ligand ions are observed. The peak
intensity of protonated nitrogen-containing uncomplexed ligands, (L'H)+ (m /z
264) and (L"H)+ (m /z 263) was found to increase with the number of nitrogen
atoms involved in coordination. Furthermore, in the case of the potassium
complex of L", the doubly protonated uncomplexed ligand (L"+2H)2+ (m/z
132) is present in the spectrum (Figure 4.2.c). The peaks at m/z 346 (Figure
4.2.b) and m/z 347 (Figure 4.2.c) are due to the complexation of potassium by
the aza and diaza analogs of 21-crown-7 ether, present as impurities in aza-
18-crown-6 and diaza-18-crown-6, respectively.

The nitrogen heteroatoms act, therefore, as more favorable protonation
sites than the oxygen atoms. According to the variation in the peak
intensities of species derived from uncomplexed ligands, the amount of
uncomplexed ligand available for protonation was found to increase with an

increase in the number of nitrogen atoms in the ligand. Considering the

63

 

 

 

 

 

 

 

 

 

 

 

 

. (KLY‘
100i (M 303)

.‘ a /

E"; :

:3 ¢

:2 -.

g :

E d K. No protonated

% _‘ (W139) uncomplexed ligand Ions
‘1 : \

01-..! ...........................

0 100 [2320 300 400
A . b (m/z 302)

E; 1 /
'2 ‘ (Nari
E j K+ (m/z 286)
. +
5% 3 W1 39) (L ”) [mam-2107))+
g 1 / (m/z 264) (m 346)
0i 1 \ - -i/

o ' 160 ' [2,260 300 400
1003 C (KL")+
g (m 301)
2* 4
'g 1 K, (NaL")*

g : (m/z 285)
0 ‘ m/ u
a : (.3 392L..+2H)2+ It; g3 [K(diaza-21C7)]*
& ‘ l (m/2132) ( Z \) (”V2 345)
0.: \L l 1 I
0 100 [2.920 300 400

Figure 4.2 ES-MS spectra of: a) Potassium complex of 18—crown-6 (KL)+
b) Potassium complex of aza-18-crown-6 (KL') "
c) Potassium complex of diaza-18crown-6 (KL") +

64

solution equilibria and stability of (KL)+, (KL')+, and (KL")+ respectively
[11,15], the concentration of free ligand in solution is consistent with the
appearance and peak intensities of protonated ligands in the ES-MS spectra.

Unlike the potassium complexes discussed above, the ES-MS spectra
of silver complexes (Figure 4.3), recorded under identical experimental
conditions, do not contain the bare metal ion Ag‘. The ions representing the
intact (1:1) complexes, (AgL)+ (m/z 371), (AgL')+ (m/z 370), and (Ag ")+ (m/z
369) are still predominant, but more fragments derived from uncomplexed
ligands are present. Protonation of uncomplexed oxa-crown ligand (L) is not
noticeable in the ES-MS spectrum of AgL. In addition to the protonated aza
and diaza ligands (L'H)+ (m/z 264), (L"H)+ (m/z 263), and (L"+2H)2+ (m/z
132), when protonation is accompanied by hydrogen loss, ions of m/z 262
(Figure 4.3.b) and m/z 261 (Figure 4.3.c) are formed and present in the ES-
MS spectra of (AgL')+ and (AgL”)+, respectively.

The absence of Ag+ is expected considering its high ionization potential
(7.576 eV) [16] and its low charge affinity. Hence, it is more likely that other
species with higher proton affinity (i.e., uncomplexed ligand) would
preferentially win the competition for the excess charge generated during the
ion formation process and therefore appear in the spectrum. The ES-MS
response factors for various ionic species are also strongly dependent on their
relative solvation energies [17]. In other words, the formation of bare K+ ions

is favored compared to the formation of bare Ag‘ ions during the electrospray

65

+33» 0-5.998 do 3358 52.6 ms do 8:53» m2.mm Q3 9:9“.

 

 

 

 

 

 

 

 

 

 

 

 

 

N\E
oov com com 9.: o
— p p p p b p p . b P p — b p p p
41 . 1 - 1 to
80m N\EV // .
has 2.3 can. 96 9:85. .
c/mm £5 .flafihhoarz

+3mzv . W
n04.
. m.
M
. a
m.
«M.
omni--~mn---£n_-t§ r \w
I]: 1 o lo”

7 r m T

32.2.: /. a m .

AmnmeEv 2.5ng C .3 (8.. :5.
+Al_m<mcwv +Al_u<h°rv r COP

66

k. .38 0.232928% do 6358 82..» ms do 25.86% msam 93. 9:2“.

 

 

 

(%) htsueiw eAueIea

 

 

 

 

 

N\E
cow own emu 2.: o
d a 4 1 i 4 I o
\ / .
3% NE 38 NE
+3qu MOT/adv .
in: o .. o .
z/ .
fi/OJ _/\+/\_
Chm/L .
.Egz . H .
/ +9.35%. C 2: w m
“mum NxEv SB Ev gang—rat #5.: fig: .
.3282 +3233 h 2:.

 

 

 

67

+«...B$ 323292-83“. go 3350 62.6 ms do 238% m2-mm 0.3 9:9...

 

 

 

   

 

 

 

NE
08 an 2.x 2: o
\ - A t 4 - J ./. t - t 0
$35 / A V .
hazy 53$ 3 NE
32) 3a) .
80m N\EV O O O O a
32) m U F. L . a
O O O O n W...
m U (C CC . a
O N O m
CC . w
s. m...
\\ in . . . fin . - - own - - - «8 fl l/w
a (at
E” as . m .
2222; H .
\ +£29.35...H W. ,.
. w
80m MES L:~.....u<8.v C +2.. 6.342.; ( .
£3053 .r.3<=..\. 25.32.... 8.

 

 

 

 

 

too—

68

process. The appearance of (NaL)+ (m/z 287) and (KL)+ (m/z 303) in the
spectrum (Figure 4.3.a), is not surprising, being consistent with the very high
aflinity of the oxa macrocycle for alkali-metal ions [18] which exist as
impurities in the solution matrix. However, the intensity of the peaks
corresponding to complexes of L, L', and L" with alkali-metal impurities
decreases with the increasing number of nitrogen atoms in the ligand, the
trend following, once again, the stability sequence of potassium complexes in
solution [11]. The complexation of silver with the nitrogen-containing ligands
L' and L" generates characteristic ion patterns in the ( 1: 1) intact complex ion
region (Figure 4.3.b and 4.3.c). These patterns are the result of
superimposing the characteristic isotope patterns of silver, and the patterns
due to the successive dehydrogenation (i.e., 2 m/z units for each loss of two
hydrogen atoms) (Figure 4.4).

The deconvolution of the ions represented by these patterns is
necessary in order to avoid ambiguity in subsequent MS/MS experiments
involving especially the ions of m/z 367 and m/z 369 as selected precursors
in CAD processes. Based on the natural abundances of silver isotopes (1°7Ag =
51.84% and 109Ag = 48.16%) and the assumption that all reaction
mechanisms and yields are identical for the two isotopes, the deconvolution
of the predominant peaks in the ES-MS spectrum of AgL” resulted in the

data shown in Table 4.2.

69

882a xmcqmofimam as 8.2% 39¢ 9:388.» 35% 6
53:8 325 as B Samcmaosémn 2: Q 328 888$ v. v 9:9“.

Em N\E I N - mom N\E I N .. Em N\E
.fv 3...?st Allll “53.428; AIII .A..._m<s.v

mom N\E 5m N\E mom N\E
“I? I:I—m<hcwv +AINI:I_U<NO—.v + A:I_D<NOFV

DQU _\/I_..m\/_Iw I\/.
m</ IAINIH flMMflw/L/ Illml. flow/Mg

\ I
\z/V_OU +_i)r\.u\z/U_OU
.. _| . TC

70

Table 4.2 Absolute and relative peak intensities of the
predominant ions in the ES-MS spectrum of Ag ”

 

 

 

 

 

 

 

 

 

 

Ion Absolute Intensity Relative Intensity
m/z Species (counts 1' 107) (% of base peak)
371 (1°9AgL”)* 8.5 60
(‘°9AgL” - 2H)* 9,2
369 (1°7AgL”)+ 5.0
Total 14 100
(‘°°AgL” - 4H)+ 5.4
367 (1°7AgL”-2H)+ 1.6
Total 6.9 49
365 (‘°7AgL”-4H)* 1.5 , 11

 

 

ES-MS/MS

Unlike ES-MS experiments, which generate information about the solution
chemistry and processes that occur at the interface between solution and gas
phase, CAD is a much “cleaner” gas phase experiment. The MS/CAD/MS
product spectra also exhibit substantial differences between the complexes of
the two cations. The dissociation of (KL)+, (KL')+, and (KL")+ follows a one-
step pathway, the process resulting in a unique product ion, K+ (m/z 39)
(Figure 4.5). Therefore, only two ionic species, the precursor complex ion and

K+ as the sole product are present in the spectra.

 

71

 

 

 

 

 

 

 

 

 

 

 

100-. (KL)+
: a (an303)
E ‘2
a a
5 :
a : (In/239)
a . /
0 100 "1,2200 300
100-, (KL')+
: (m/z302)
5.2: 3 b /
a: :
.6 : K“
t; 1 (M39)
'3 j /
‘6 ..
o: .
°J """"" i'o' """" 260' ‘‘‘‘‘ 360"”
0 0 m/z
q (KI-«)4-
1001 (m/z301)
A 1 c /
2% 4 +
a 1 K
§ : (WZ39)
% 4 /
s. : l
a: ‘.
10 20 30
O 0 m/z 0 0
Figure 4.5 MS/CAD/MS spectra of: a) Potassium complex of 18erown-6 (KL)+

b) Potassium complex of aza-18-crown-6 (KL') +
0) Potassium complex of diaza-18erown-6 (KL") +

72

The ratios of peak intensities of the precursor ion and K+ are
qualitatively indicative of the relative afinities of the three ligands, L, L’,
and L” respectively, for the potassium ions. This conclusion is in agreement
with the observations of Brodbelt et al. [19] regarding the gas phase
stabilities of alkali-metal complexes with a variety of macrocyclic ligands.
The trend depicted in Figure 4.5 suggests a similar stability sequence to the
stability of potassium complexes in solution (Table 4.1), although no
quantitative correlation between the relative peak intensities and the
stability constants in solution can be found based solely on this experiment.

The relative peak intensity of these ions is also a function of the
collision energy and the collision gas pressure. An increase in either of these
two parameters leads to a greater fragmentation, that is, a larger relative
peak intensity of the bare metal ion. Chapter 5 deals extensively with CAD of
alkali-metal complexes and the effect of various operational parameters on
their gas phase stability. Under identical experimental conditions the
complexes containing a larger number of “soft” nitrogen donor sites involved
in complexation, that is, complexes of K+ with aza and diaza ligands generate
larger abundances of the bare metal ion.

The silver complex precursor ions undergo different dissociation
pathways. The CAD spectrum of (AgL)+ contains the fragment ions Ag* (m /z
107), ligand fragments (e.g., m/z 45), and silver-containing complex

fragments (e.g., m/z 182) (Figure 4.6.a). Following collisions with target gas

73

 

 

 

 

"’07 a (Ang*
3 (m/z 371)
a -. /
E ‘. (021-150)” A +
5 (m 45) 9 +
g 1 (W210?) (C,Hp,Ag)
o 1 / (m/z 181)
m ..
0.: J J .4 L l J ‘ L ‘ J
O I 160 1 1279/2 ' 360 ' 460
’00? b (AgL')

(AgL'- 2H)’ (m/z 370)

(cHi’Ho fl ("V235‘\l/

 

l l 14 ‘ l ‘ 1

Relative Intensity (%)
§~ ~32
afb

‘IAA..

o
l .
1 k
).
).____
i
t
l
*—

 

 

 

vvv'vvvv‘vvvi'vvvv'vi vvvvvvvvvvvvvvvvvv

 

 

 

0 100 209 300 400
(AgLu. 4H)+ u ‘I'
"’07 (m/z 365 (ML -2H)
E ' .
‘3‘ ‘ + m (AgL")+
g (C a") (an 369)
3 - ( Z 44) H /
,5}; j / (m/z 261)
03 MALLAL AA - A
0 ' 160 T 269 ' 360 ' 460

Figure 4'6 MS/CAD/MS sPectra of: a) Silver complex of 18—crown-6 (AgL)*
9) Silver complex of aza-18-crown-6 (A guy
0) Silver complex of diaza-18crown-6 (AgL")+

74

molecules, the nitrogen-containing precursor ions (AgL')+ and (Ag ")+ do not
form Ag'. Instead, these complex ions release silver in the form of a neutral
product, AgH, a process characteristic of the precursor ions (AgL')+ (m/z 370)
(Figure 4.6.b), and (Ag ")* (m/z 369) (Figure 4.6.c). Except for the ions
formed by dehydrogenation, no other product ions containing silver are
present in the MS/MS spectrum.

Hydrogen loss involving the nitrogen atom sites is a significant process
characteristic of CAD of (AgL')+ and (Ag ")+. This process can occur during
either “in-source fragmentation” during electrospray ion formation (Figure
4.3.b and 4.3.c) or it can be induced by CAD of selected precursor complex
ions. For instance, CAD of ions of m/z 371, m/z 369, m/z 367, and m/z 365,
ions that exist in the ES-MS spectrum of AgL” (Figure 4.3.c) provides crucial
information about silver binding within the complex as a function of
structural changes in the ligand constitution. The selection of ions of m/z 365
and m/z 371 as precursors presents no problem, the peaks representing
unique ionic species, (1°7AgL”-4H)+ and (1°9AgL”)+ respectively. Tuning the
first mass analyzer to unit mass resolution allows therefore for unambiguous
precursor ion selection. The analysis of the product ions resulting in CAD of
species of m/z 367 and m/z 369, which consist of overlapped peaks, is
however more complicated, involving the deconvolution of both the precursor
and product ions. Based on the peak intensities of the product ions of

interest, which are summarized in Table 4.3, we conclude that the second

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- - - - we .AIV-.§§V 8...
Io - 2 S .28
- S - mo 2 .AIN-.§§V 5m
- - - - 3 25-223;
3 mo 2 t 3 .92
Io E e 3 9 .5sz 8m
- S - m: S .AIN - 22...;
3 S 2 3 a .25.; E
#272... I27: IT: IT: I 38% as
E: . £58. .235 :2... 828.2 :2 38395

 

 

 

+~mVuamev¢ cha ”flame: NMVVV «+5. \vaVy \Q $0
3: Lox 25.» 338.5 vac somgzoogq 2% \c 8.33335 «8Q 3383.4 m6 533.

 

76

dehydrogenation step (k2) is favored compared to the first step (In) (Figure
4.7).

The loss of AgH occurs when intact complex ions (e.g., (AgL”)+) collide
with target gas molecules (Figure 4.7.a). The loss of AgH from (Ag ”)+ during
CAD is dependent upon the degree of previous dehydrogenation, as shown in
Table 4.3. The release of AgH from ions that have previously lost two
hydrogen atoms (e.g., (AgL”-2H)+) is significantly hindered with respect to
that from (AgL”)+ (Figure 4.7.b and 4.7.c), and silver hydride is not released
at all from ions that have already lost four hydrogen atoms (e.g., (AgL”-4H)+)
(Figure 4.7.d).

The formation of double bonds following dehydrogenation has a
tremendous effect on the coordination of silver. The degree of covalency
involved in silver binding is dramatically increased by the contribution of
carbon-nitrogen double bonds formed (Figure 4.7.a). Therefore, the
interaction of silver with the ligand is enhanced, making the neutral loss of
AgH more difficult when associated with an increased degree of

dehydrogenation.

77

an 093 to 835303
8m NE «a new $9.. NE 3 .%m E a» .5 NE E as 23.. .6 96 5%
«In S was? $36 .8 m3 Em :oumcmmougcmb é EmEmfimE bmmoqoi NV 2:9”.

5:82 3.55:: 5:32 3:895
mmm #2. EN ~\E
. a
2.. do a)

.. new new

Ia<o3 . =U<2= . 20(8. .
an E 8” E E E
.. a: .22.... .. .3 522.. made.
I
2 >3

o< AIN- \a<\OH_

 

78

(”Au L")
b -( M369 2 H
"Ml
m/z261
Preferred reaction
and
c
and

a

Figure 4.7 (continued from page 77)

(107MLI. 2H,. ('01A'L'JH “rt
M367 2 H m/2365
_ 1°7A°Hl '107A0Hl
m/z 259
Unfavored reaction No reaction
(‘°°AgL"- 2H)+ TH" (1%ng 4H)+
m/z 369 ' m/z 367
.1” “Hi _109A°Hl
m/z 259
Unfavored reaction No reaction

(107AgLu_ 2H)+ _’ (1°1AgL". 4H)+
m/z 367 '2 H m/z365
.101 Ang _101A°Hl
m/z 259
Unfavored reaction No reaction
(WA/352?“? ——>_ 2 H No reaction
m
.109 A 9H1
No reaction
(WA/3:254“) W No reaction
m
_107A°Hl
No reaction

79

Summary

The stability of potassium and silver complexes in solution strictly follows
the theoretical considerations regarding preferential “hard-hard” and “soft-
soft” interactions between the metal ions and the coordination sites in the
ligand, respectively. Hence, the interaction strength of potassium with L, L’,
and L” is weakened with an increase in the number of nitrogen atoms, trend
which is qualitatively evident in both ES-MS and CAD spectra of the
complexes chosen.

In the electrospray spectra, similar trends in the stability of the
complexes were noticed, although a quantitative correlation, regarding the
extent to which the appearance of the ES-MS spectra of the compounds
considered resembles their equilibrium concentrations in solution was not
readily observable. The complex processes behind the electrospray ion
formation, desolvation, and transport prevent making an immediate
correlation between relative peak intensities and the relative concentrations
of those ionic species in solution. Solvent evaporation and ion desolvation
occur difi'erently for different ionic species, so that the ES-MS spectra are not
quantitatively representative of their solution chemistry. A quantitative
study would only be possible if the solvation energies of all species involved

were identical or known. However, the qualitative observations are on more

80

solid ground; at least some of the ionic species present in the ES-MS spectra
do exist in solution.

The ES-MS spectra of silver and potassium complexes with the
macrocyclic ligands chosen illustrate their contrasting behavior. Silver does
not generally have a preference for charge retention, therefore Ag+ is not
present in the ES-MS spectra, especially when nitrogen-containing ligands
are involved. The nitrogen atoms in the ligands are more favored protonation
sites than the oxygen atoms, so that protonation of L’ and L” may occur at
every nitrogen atom available (Figure 4.3.b and 4.3.c). Potassium, however,
is always present as K+ (m/z 39) in the ES-MS spectra, regardless the nature
and the number of donor sites in the ligands. Very limited protonation of
nitrogen-containing ligands was observed in the presence of potassium,
clearly indicating the greater ability of potassium to retain the charge in
competition with various coexisting species. This is further illustrated in the
CAD spectra of potassium complexes, where a single dissociation product, K“,
prevails over any other fragment that might be formed during the collision
processes in the gas phase.

Electrospray combined with tandem mass spectrometry could be
further employed in studying complexes of transition metals with macrocyclic
and chelating ligands. Not only does electrospray provide qualitative data
about the ionic species that exist in solution, but CAD with tandem mass

spectrometry complements this information with observations regarding

81

their gas phase chemistry. The assessment of preferential metal-ligand
binding, and especially the interaction strength between metal ions and
various coordination sites are research objectives that may be reached by

using ES-MS/MS.

82

References

8.

9.

Harrigan, E. T.; Hirota, N. Chem. Phys. Lett. 1967, 1, 281.

Andretti, G. D.; Cavalca, L.; Sgarabotto, P. Gazz. Chim. Ital. 1970, 100,
697.

Cram, D. J.; Cram, J. M. Science 1974, 183, 803.

Weber, E.; Toner, J. L.; Goldberg, I.; Vtigtle, F.; Laidler, D. A.; Stoddart, J.
F.; Bartsch, R. A.; Liotta, C. L. In Crown Ethers and Analogs, S. Patai and
Z. Rappoport, Eds; John Wiley & Sons: Chichester-New York-Brisbane-
Toronto-Singapore, 1989.

. Pedersen, C. J. J. Am. Chem. Soc. 1967, 89, 7017.

Izatt, R. M.; Christensen, J. J. In Synthetic Multidentate Macrocyclic
Compounds; Academic Press: New York, 1978.

Gokel, G. W.; Durst, H. D. Synthesis 1976, 3, 168.
Frensdorfi‘, H. K. J. Am. Chem. Soc. 1971, .93, 600.

Man, V. F.; Dale Lin, J.; Cook, K. D. J. Am. Chem. Soc. 1985, 107, 4635.

10.Gokel, G. W.; Goli, D. M.; Minganti, C.; Echegoyen, L. J. Am. Chem. Soc.

1983, 105, 6786.

11.Stroka, J.; Ossowski, T.; Cox, B. G.; Thaler, A.; Schneider, H. Inorg.

Chimica Acta 1994, 21.9, 31.

12.Boss, R. D.; Popov, A. I. Inorg. Chem. 1985, 24, 3660.

13.Christensen, J. J.; Eatough, D. J.; Izatt, R. M. Chem. Rev. 1974, 74, 351.

14.Izatt, R. M.; Pawlak, K.; Bradshaw, J. S. Chem. Rev. 1991, 91, 1721.

15.Pearson, R. G.; J. Am. Chem. Soc. 1963, 85, 3533.

83

16.Handbook of Chemistry and Physics, 65th ed.; CRC Press, Inc.: Boca
Raton, FL, 1984.

17.Dupont, A.; Leize, E.; Marquis-Rigault, A.; Funeriu, D.; Lehn, J.-M.; Van
Dorsselaer, A. In Proceedings of the 43'“ ASMS Conference on Mass
Spectrometry and Allied Topics, Atlanta, GA, 1995; p 250.

18.Coxon, A. C.; Laidler, D. A.; Pettman, R. B.; Stoddart, J. F. J. Am. Chem.
Soc. 1978, 100, 8260.

19.Maleknia, S.; Brodbelt, J. J. Am. Chem. Soc. 1992, 114, 4295.

Chapter 5

Stability Studies in the Gas Phase by Tandem

Mass Spectrometry

Introduction

Tandem mass spectrometry is entering its third decade of application in the
area of complex mixture analysis and its fourth decade of use in basic studies
of ion chemistry in the gas phase [1]. One of the most exciting and
rewarding areas in mass spectrometry is selected ion fragmentation, in which
an ion is mass-selected, fragmented, and the resulting fragment ions are
mass-analyzed. The most widely used process to obtain fragmentation of ions,
which can be initially produced in a large variety of ionization processes is
collisionally activated dissociation (CAD), also known as collision-induced

dissociation (CID) [2].

84

85

CAD has grown dramatically in the past decade as an important
method for ion structure determination and complex mixture analysis [3,4],
especially with the development of new ionization techniques, such as fast
atom bombardment and electrospray, both capable of producing ions from
large molecules. Current applications of CAD include polypeptide sequencing
[5], characterization of oligosaccharides [6], drug metabolites [7], inorganic
coordination compounds [8], etc. The CAD process involves activation of ions
following collisions with neutral gas molecules, or sometimes a surface, the
latter being called surface-induced dissociation (SID) [9]. Subsequent
fi'agmentation of activated ions to various product ions depends upon the
amount of energy transferred into the ion.

The experimental methods used to study CAD have lately evolved
significantly. Tandem arrays of momentum and energy analyzers and
differentially pumped collision chambers provide a convenient means of
selecting ions of interest, colliding these ions with neutral gas particles in
cells located in field-free regions and conducting mass/energy analyses of the
product ions. Sectors [2,10], quadrupoles [11], ion cyclotron resonance [12],
time-of-flight [13], and ion traps [14] have all been used alone or in various

combinations to suit individual requirements.

86

Collisionally Activated Dissociation in a Triple Quadrupole Mass

Spectrometer

CAD of polyatomic ions occurs according to a two-step mechanism. Collisional
activation (Equation 5.1), where a fraction of the kinetic energy of the target
ion is transferred into internal energy, is followed by the dissociation of the

internally excited ion (Equation 5.2).

M1+ + M2 > M1“ + M2 (5. l)

 

M1” _"'—> M3+ + M4 (5.2)

The essence of this model is that the collision time is short compared to
the dissociation time, thus separating the activation and dissociation in time.
During this time delay, internal energy is redistributed among various
internal degrees of fieedom of the ion. The mode of energy deposition in
collision and the extent of internal energy redistribution are essential
mechanistic issues in CAD of polyatomic ions.

Most commonly, a target mass (precursor ion) is selected in a primary
mass analyzer (first quadrupole) and accelerated into the collision chamber
at modestly elevated pressure. The collision chamber is commonly named

“second quadrupole”, even though it is built as an octapole in most of the

87

modern commercial instruments (Figure 5.1). The fragment ions formed in
the collision chamber are subsequently mass discriminated and analyzed in
the second analyzer (third quadrupole). These CAD fragments are known as
“product ions”. Although routinely referred to as MS/MS, tandem mass
spectrometry involving collisionally activated dissociation is in fact
MS/CAD/MS.

Unlike reactions induced in the condensed phase, including but not
limited to electrospray ionization, CAD allows for “cleaner” gas phase studies,
conducted in a solvent-free environment. The comparative study presented in
Chapter 4 deals specifically with the complexes of two metal ions, K+ and Ag“
with three particular ligands, and their solution and gas phase chemistry.
This chapter focuses to a greater extent on aspects of the gas phase stability
of several alkali-metal ion complexes with 18-crown-6 (L) under CAD

conditions. The effect of experimental CAD parameters is also discussed.

Experimental. The analyte, a solution of 0.1 M CsCl (Aldrich) in
methanol was prepared according to the general procedure. The ligand
used was 18-crown-6 (L), and it was added to the solution such as its
concentration was 0.2 M. The electrospray voltage was 5 kV, and the
temperature of the transfer capillary was 110 C. Argon was used as
the collision gas in the CAD experiments. Both ES-MS and MS/MS
spectra were acquired using a signal averaging procedure, 60 spectra

being averaged throughout the experiments.

88

ACQUQ—Qm mmwe

E3268: cum .8 .3358»
3.828% was 888 E 05 s 53% «83 a: B can

 

 

$6 30

 

 

 

uSoBmQ :2 835.5
_ _ _ _ _
$29258 ”G
sebum—m _
_
)9
Di
_
9.35 82%

:o_m..m>:o0

_
«a

_
Spawn”.

53:8

 

 

 

 

 

 

 

 

 

 

the . é mm

_
.28928 mme

:2 83:85

 

mmwam :ofimficom

macaw

89

The ES-MS spectrum of the species that result in the complexation of
CS“ with 18-crown-6 is shown in Figure 5.2. The experimental conditions
were chosen such as the peak intensity of the “sandwich” (CsL2)+ complex
was at a maximum. Even though a lower temperature of the transfer
capillary would have probably yielded a larger amount of (Cst)+ (see
Chapter 3), its operation at at least 110 C was necessary to ensure the
complete solvent evaporation from the emerging analyte solution droplets.

For the MS/MS operation (CsL2)* (m/z 661) was selected as precursor
in the first mass analyzer. Regardless the values of collision energy (E0011) and
collision gas pressure (P0011), the general fi'agmentation pathway led to the
exclusive formation of (CsL)+ (m/z 397) and 03* (m/z 133). The extent of the
fragmentation was a function of the collision parameters. The specific

influence of these variables is presented in detail in the present chapter.

Effect of Collision Energy

With the collision gas pressure set and maintained at approximately 0.4
mTorr, the voltage offset of Q2 was varied from -2 to -50 V and the peak
intensities for the product ions that result in the CAD of (CsL2)+ monitored
(Table 5.1). Typical MS/MS spectra, for two sets of experimental parameters
are shown in Figure 5.3. For consistency, the relative peak intensities in

spectra recorded under different sets of experimental parameters were

o o: u eamcqsg amino .3852: E 8.58 an: n .. 68 m do .5588 msam Wm mama

own cow

a:
own ems com

emu cow

b h

 

 

/

:8 NE
gassed

9O

/

 

has

Aammnssv

 

/

82 NE
.8

(%) Mtsuetw eAnelea

 

roo—

91

Table 5.1 Effect of collision energy (Ecoll) on the relative intensity
of product ions of (Cst)+ (m/z 661),-
Collision gas pressure = 0.40 mTorr

Collision Relative Intensity (%)

energy Cs+ (CsL)+ (CsL2)*
e m/z 133 m/z 397 m/z 661

 

92

 

 

 

 

 

 

 

 

10c (m2).
‘ (m/z 661)
A a
i
.5
E (0er
g. (m/2 397)
a?
0‘ 1 I f . .
0 200 400 600
m/z
100- (CsLli
‘ (m/z 397)
b
g
g
is
§ 08* (63L;
(In/2133) (m/z 661)
0+ w l U u 1 u 1 i u
0 200 400 600
m/z

Figure 5.3 MS/CAD/MS product ion spectra of (CsL2)+(m/z 661).
Collision gas pressure = 0.40 m Torr; Collision energy:
(a) 4 eV, (b) 46 eV

93

rationalized to the summed intensity of all the peaks present in the spectrum
(Figure 5.4). As the collision energy increases from 2 eV to approximately 20
eV, a sharp decrease in the relative intensity of Cst+ peak is evident, while a
mirror-like increase is noticed in the relative intensity of (CsL)+. The
appearance energy of (CsL)+ is therefore very low, according to the trend
depicted in Figure 5.4. The appearance energy of Cs+ (m/z 133) is
approximately 20 eV, and its peak becomes more intense as the collision
energy further increases. Even though the increase in 05* peak intensity is
qualitatively parallel with a decrease in the intensity of the precursor ion,
the fragmentation mechanism (i.e., whether single or multiple-collision
dissociations are involved) can not be studied based solely on the assessment

of the collision energy effect.

Efl'ect of Collision Gas Pressure

The conclusions drawn from the appearance of ES-MS spectra and from
studying the effect of collision energy on fragmentation do not provide
answers to crucial issues regarding various dissociation mechanisms and
reaction pathways. There are several possible fragmentation mechanisms
that govern CAD of complex ions. The reaction pathways, as well as the

distribution of the product ions depend on the structure of the precursor ion,

94

SEE 36 n 9883 man 8260 6388 8.28 .6 885‘ m 8 :8 NE» .393 3
98 as s 3.38» pa . argue . 3 .8 8.. assess saws: «ma E383 E. 9am

33 385 86:60

 

 

 

 

om ow om cm or o
o o 0'slolsll... o c c it--. a a a o o n n w ”-ouo fio
4
Q
4
4

. m
m.
U
W

:8 as... :8 as... 32 s5.
:8 as. . mm
B
:8 as? :8 £5135: x.
can as. n. W
:8 as... :8 #513 as. m.
.8. as. .M

 

 

 

 

95

the range of collision energies applied, and a series of other factors among
which the collision gas pressure is particularly important.

The fragmentation of (Cst)+ (m/z 661) can occur according to the

 

 

following mechanisms:
a) (Cst)+ + G > Cs+ + 2L + G (5.3)
(m/z 661) (m/z 133)
b) (CsL2)+ + G ; (CsL)+ + L + G (5.4.1)
(m/z 661) (m/z 397)
(CsL)+ + G ———> Cs+ + L + G (5.4.2)
(m/z 397) (m/z 133)

The general equation describing the formation of Cs+ (m/z 133) in

either of the two instances above is:

d[Cs+]

dt = k[(Cst)*1“P€ou (5.5)

where Pcoll is the collision gas (G) pressure in mTorr, (a) is the reaction order

in Cst, and (b) is the reaction order in G.

The mechanism described by Equation 5.3 implies fragmentation
occuring as a result of a single collision between the precursor ion and a
target gas molecule, while Equation 5.4 describes a multiple-collision
dissociation. According to the Equation 5.5, the rate of formation of Cs+ from
precursors depends on the collision gas pressure.

CAD of (CsL)+ (m/z 397) is a good example of a single-collision
fragmentation. The relative peak intensities of the only two ionic species

present in the spectra (Figure 5.5) are presented in Table 5.2. The

96

 

 

 

 

 

 

 

 

 

100' (CsLY
* i (W2 397)
\
3 a
z. 4
'2
E ‘ Cs’
("V1\33)
0. 1 A I I I
0 100 200 300 400
m/z
100' Cs‘
‘ (/m/z 133)
E b
g;
E * (081-?
g y (m/2397)
§ i
0‘ I T I I I
0 100 200 300 400
m/z

Figure 5.5 MS/CA D/MS product ion spectra of (CsL) +(m/z 397).
Collision energy = 20 eV; Collision gas pressure:
(a) 0.10 mTorr, (b) 2.20 mTorr

97

Table 5.2 Effect of collision gas pressure (Pcoll) on the relative
intensities of product ions of (CsL)+ (m /z 397);
Collision energy = 20 eV

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Collision Relative Intensity (%)
pressure Cs+ (CsL)+
(mTorr) m/z 133 m/z 397
0.10 1.7 100
0.21 5.4 100
0.28 8.1 100
0.35 1 l 100
0.42 16 100
0.46 20 100
0.50 2 1 100
0.61 30 100
0.65 31 100
0.71 34 100
0.80 35 100
0.86 45 100
0.94 58 100
1.0 62 100
1.1 67 100
1.2 83 100
1.3 92 100
1.5 100 86
1.6 100 75
1.8 100 66
1.9 100 54
2.1 100 40
2.2 100 35

 

 

98

rationalized peak intensity for the bare cesium ion (m /z 133) varies linearly
with the collision gas pressure (Figure 5.6). Therefore, the fragmentation of
(CsL)+ follows a mechanism that is first order in G (Equation 5.4.2), which
denotes a single-collision dissociation. Deviations from linearity were
observed for collision pressures larger than 2 mTorr, amid an overall
decrease in the total intensity of the peaks in the spectra.

Unlike the CAD of the (1:1) complex ion, the collision-induced
fragmentation of the “sannd ” complex (CsL2)+ (m/z 661) yields three
predominant product ions. Their relative peak intensities are summarized in
Table 5.3. The appearance of the MS/MS spectra varies as a function of
collision gas pressure, two examples being presented in Figure 5.7. By
rationalizing the peak intensities of Cs+ (m/z 133), (CsL)+ (m/z 397), and
(Cst)+ (m/z 661) to the total peak intensities in the spectra, it can be noted
that the abundance of the precursor ion follows a decreasing trend, while the
abundance of the bare metal ion increases with an increase in collision gas
pressure (Figure 5.8). The rationalized peak intensity of the (1: 1) complex ion
reaches a maximum (at Pcoll = 0.35 mTorr), and its variation suggests the role
of (CsL)+ as an intermediate in a multiple-step fragmentation mechanism
(Equations 5.4.1 and 5.4.2).

Indeed, the rationalized peak intensity of 03* as a function of collision

gas pressure follows a power trendline (Figure 5.9). Curve fitting of

Rafionalized Peak Intensity (Cs+)

99

 

0.8

 

0.7 a y = 0.349x
R2 = 0.992

 

 

 

0.6 r

0.5 r

0.4 -

0.3 —

 

0.2 r

0.1 4 °

 

 

 

0 ° r r r
0 0.5 1 1.5 2 2.5
Collision Gas Pressure (mTorr)

_4

Figure 5.6 Rationalized peak intensity of product ion Cs+(m/z 133) in the
CAD of (CsL) i (m/z 39 7) as a function of collision gas pressure.
Collision energy = 20 eV

100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 5.3 Effect of collision gas pressure (Pcoll) on the relative intensities
of product ions of (Cst)* (m/z 661). Collision energy = 60 eV.
Collision Relative Intensity (%)
pressure 03* (CsL)+ (CsLa)+
(mTorr) m/z 133 m/z 397 m/z 661

0.12 0.60 100 75
0.12 0.54 100 74
0.14 0.71 100 71
0.17 1.1 100 45
0.20 1.5 100 32
0.2 l 1.3 100 29
0.24 2.0 100 2 1
0.26 2.9 100 18
0.30 4.8 100 13
0.31 4.4 100 13
0.32 4.6 100 13
0.35 5.1 100 7.9
0.41 7.1 100 4.0
0.41 6.9 100 4.1
0.49 13 100 2.2
0.54 18 100 1.1
0.55 17 100 1.1
0.62 29 100 0.99
0.64 31 100 0.91
0.70 46 100 0.77
0.7 5 61 100 0.65
0.81 83 100 0.58

 

 

 

 

 

 

lOl

 

 

 

 

 

 

 

 

 

31°"? .ng
7' a
g .
s i
% .
E?
J C8. ((3
(m/2133) (m 661)
o 260 460 600
m/z
100- (CsL)’
, b (m/z 397)
33 .
3. 8
g (M133)
3
r‘i'ééir
Z
0- \
o 250 460 600
m/z

Figure 5.7 MS/CA D/MS product ion spectra of (0st )* (m/z 661).
Collision energy = 60 eV; Collision gas pressure:
(a) 0. 35 mTorr, (b) 0. 75 mTorr

102

ESE 9.6 u 28on 8m 8338 x986 8.538 .6 Sauce m mm :8 NE» +38» no

 

 

 

  
   

 

 

 

 

 

96 as s a. a. .u so» as . 3. see . Emu 80.. “seas meanness «e8 nonsense 3 23:
rats cease «.8 82.8
2. no so as o
Iii I .III I e w . o orollolrpro r O
- mo
0
/ ud
o :8 as. :8 as. A8. 95. /4 a o W.
... + C r
o :8 ss. . W
:8 £5.38 #518. as: d
:8 as: D m
:83: + cassifiéez , ed M
89 E. u m
m.
a 2
F

Rationalized Peak Intensity (05+)

103

 

 

 

 

 

 

0.5
0

y = 0749x 2.617 .

0.4« R2: 0995
0
99 °
0.29
/
0.1 ~ /
/
o , ,

 

 

0 0.2 0.4 0.6 0.8.

Collision Gas Pressure (mTorr)

Figure 5.9 Rationalized peak intensity of product ion Cs +(m/z 133) in the
CAD of (Cng )+ (m/z 661) as a function of collision gas pressure.
Collision energy = 60 eV

104

experimental points results in a power equation with an order of 2.6.
According to Equation 5.5, the rate of formation of Cs+ in CAD of (CsL2)+ has
an apparent order of reaction of b = 2.6, a consequence of multiple collisions
with argon molecules in the collision cell. Even though theoretically the
reaction order in G should be 2, as a result of two collisions occurring per
reaction, the higher apparent order is due to a deviation from ideality,
namely, a certain amount of energy lost by the incident precursor ion

following the first collision.

Comparison of Relative Stability of Alkali-Metal Complexes

In order to evaluate the intrinsic binding interactions involved in host-guest
complexation of macrocyclic ligands with alkali-metal ions, 3 solvent-flee
environment appears to be a necessary experimental condition. Gas phase
selectivities of crown ethers for various metal ions were previously

determined by application of the kinetic method [15].

Experimental. An equiformular solution of NaCl, KCl, RbCl, 0501
(0.1 M) and 18-crown-6 ether in methanol was used as the analyte for
the ES-MS investigation. All reagents were used as supplied. A
syringe pump (Harvard Apparatus) was used for sample introduction,
the solution flow rate being 5 uL/min. The electrospray voltage was 5

kV and the temperature of the transfer capillary 200 C.

105

The ES-MS spectrum of the (1:1) complexes of Na“, K+, Rb“, and Cs+
with 18-crown-6 (L) is shown in Figure 5.10. The ES-MS intensities for the
peaks corresponding to the uncomplexed metals are in agreement with their
solvation energies [16]. As mentioned in Chapter 4, no quantitative
correlation can be made between the peak intensities of the complex and
metal ions and their equilibrium concentrations in solution, even though the
latter are generally known. A comprehensive quantitative interpretation of
the ES-MS spectra can be done only if the solvation energies of all the ionic
species involved are similar or known.

Tandem mass spectrometry is once again useful in this relative
analysis within a series of complexes, allowing for the evaluation of the
relative gas phase stabilities of the complexes of the four alkali-metal ions
considered. The (1:1) complexes (ML+), present in relatively large abundance
in the ES-MS spectrum (Figure 5.10) were successively selected as precursor
ions. The collision gas (Ar) pressure was set at 0.40 mTorr, and the collision
energy (ELAB) adjusted such as the center-of-mass collision energy (Ecm) was
identical for each precursor ion selected, for consistency. The product ions
present in each MS/MS spectrum were the bare metal ion, and the intact
complex ion, the fiagmentation being, as determined earlier, a single-
collision process. The CAD collison energies and the peak intensities in the
MS/MS spectra are summarized in Table 5.4.

Based on the relative intensities of the bare metal ion peaks with

respect to their precursors, a qualitative conclusion about the gas phase

106

o 88 u 25.8qu 538 SQ Be 58: a: 5% no 2:23 assess. es s 5.58% 83m o F .m 9:9...

as
ems 98 gm
_

‘ {l ‘ ‘ lo
U

2: o
P

 

 

 

 

 

:8 as / .az .
.._.o :8 as
:8 as :8 85 .x .
Lam baz Sn #5
.nm

 

/ firms. .

:8 as

 

 

107

 

 

 

 

 

 

Table 5.4 Relative intensities of the product ions in the CAD
of complexes of alkali-metals with 18-crown-6' (L)
Collision energy (ECM = 4 eV)
M etal Ion Collision Energy Relative Intensity (%)

(ELAB) (9V) M‘ (ML)*

Na+ 33 3.5 100

K+ 34 23 100

Rb+ 39 39 100

Cs+ 44 52 100

 

 

 

 

 

while the least stable, CsL+, produces the largest abundance of its metal ion

“core”. Therefore, the stability sequence in the gas phase is:

Unlike in solution, where the stability sequence is, according to the
complex equilibria, (KL)+ > (N aL)+ > (RbL)+ > (CsL)+, and is governed by the
“best fit” concept between the cation size and the ligand cavity [17,18] the

concept of “maximum contact point” [15] best describes the gas phase

chemistry of these complexes.

(N aL)+ > (KL)+ > (RbL)* > (CsL)*

 

108

Summary

The gas phase chemistry of coordination compounds can be studied, at
least to a certain extent, by performing MS/CAD/MS experiments in a triple
quadrupole mass spectrometer. The introduction of intact non-volatile
inorganic species into the mass spectrometer had been a problem until the
implementation of electrospray. The method has a double duty as a “launch
pad” for tandem mass spectrometric analyses. ES does not only serve as a
sample introduction means, ensuring the transfer into the gas phase of
virtually any solvated ionic coordination compound, regardless its
(non)volatility, but this transfer occurs with little or no fragmentation, even
for unstable and thermally fiagile ionic species.

Collisionally activated dissociation processes are strongly dependent
on the reaction conditions, such as collision gas pressure and collision energy.
Careful monitoring of instrumental parameters allows for mechanistic and

relative structural studies, involving selected ions, in the gas phase.

109

References

. Busch, K. L.; Glish, G. L.; McLuckey, S. A Mass Spectrometry/Mass
Spectrometry: Techniques and Applications of Tandem Mass Spectrometry;
VCH Publishers: New York, 1988.

. Jennings, K.R. Int. J. Mass Spectrom. Ion Phys. 1968, 1, 227.

. Cooks, R. G. In Collision Spectroscopy, R. G. Cooks, Ed.; Plenum: New
York, 1978.

. Tandem Mass Spectrometry; F. W. McLafl'erty, Ed.; John Wiley & Sons:
New York, 1983.

. Carr, S. A.; Roberts, G. D.; Hermier, M. E. In Mass Spectrometry of
Biological Materials; C. McEwen and B. Larsen, Eds; Marcel Dekker:
New York, 1990.

. Domon, B.; Costello, C. E.; Biochemistry, 1988, 27, 1534.

. Haroldsen, P. E.; Reilly, M. H.; Huges, H.; Gaskell, S. J.; Porter, C. J.
Biomed. Environ. Mass Spectrom. 1988, 15, 615.

. Znamirovschi, C. G.; Enke, C. G. J. Am. Soc. Mass Spectrom. (in press).

. Mabud, Md. A.; Dekrey, M. J.; Cooks, R. G. Int. J. Mass Spectrom. Ion
Processes 1985, 67, 285.

10.Hadden, W. F.; McLafl'erty, F. W. J. Am. Chem. Soc. 1968, .90, 4745.

11.Yost, R. A.; Enke, C. G. J. Am. Chem. Soc. 1978, 100, 2274.

12.Cody, R. B.; Freiser, B. S. Int. J. Mass Spectrom. Ion Phys. 1982, 41, 199.

13.Schey, K.; Cooks, R. G.; Grix, R.; Wollnik, H. Int. J. Mass Spectrom. Ion

Processes 1987, 77, 49.

14.Louris, J. N.; Cooks, R. G.; Syka, J. E. P.; Kelley, P. E.; Reynolds, W. E.;

Todd, J. F. J. Anal. Chem. 1987, 5.9, 1677.

110

15.Maleknia, S.; Brodbelt, J. J. Am. Chem. Soc. 1992, 114, 4295.

16.Dupont, A; Leize, E.; Marquis-Rigault, A.; Funeriu, D.; Lehn, J .-M.; Van
Doesselaer, A. In Proceedings of the 43rd ASMS Conference on Mass
Spectrometry and Allied Topics, Atlanta, GA, May 21-26, 1995.

17.Gokel, G. W.; Doli, D. M.; Minganti, C.; Echegoyen, L. J. Am. Chem. Soc.
1983, 105, 6786.

18.Zhang, H.; Chu, I.-H.; Leming, S.; Dearden, D. V. J. Am. Chem. Soc. 1991,
113, 7415.

Chapter 6

Metal Complexes of a Novel Macrocyclic

Ligand

Introduction

Electrospray mass spectrometry applied to coordination and organometallic
systems has recently become a preferred tool for inorganic analytical and
synthetic chemists. ES-MS operation in either positive or negative mode,
coupled or not with tandem mass analysis, allows for structural and
mechanistic investigations not approachable through most other means.
Colton and co-workers have successfully applied the technique of
electrospray to a wide variety of organometallic systems [1-3] as well as
coordination complexes [4].

In the present study we applied electrospray mass spectrometry to the

study of a novel polydentate, macrocylic ligand, hexahomotriazacalix[3]arene

111

112

(HsL) (Figure 6.1), synthesized by Hampton and co-workers [5]. Calixarenes
and related macrocycles have recently received considerable attention for
their host-guest chemistry and for their coordination chemistry with metals
[6,7]. The hexahomotriazacalix[3]arene which this study focuses on is the
first structurally characterized compound of its kind. Following
comprehensive structural studies of its complexation in solution [8], we
attempted to acquire additional information about its host-guest coordination
chemistry involving transition metal ions such as scandium (III), yttrium

(III), and lanthanum (III) by means of electrospray mass spectrometry.

was
cc ad-

0
OCH3

Figure 6.1 The structure of hexahomotriazacalix[3]arene (HsL)

113

Complexes with Scandium (Ill), Yttrium (Ill), and Lanthanum (Ill)

Egperimental. Solutions of ScCla, YCls, and LaCla (ACS-grade) were
prepared in accordance with the procedure previously described.
Equivalent amounts of H3L were added to each solution such as the
concentrations of both the metal ion and the ligand were 0.1 M. The
solvents used were HPLC-grade methanol or acetonitrile (Merck). The
solutions were introduced in the electrospray source using a syringe
pump, at a flow rate of 5 uL Imin. The electrospray voltage was set at 5
kV, and the temperature of the transfer capillary was maintained at
200 C. For the CAD study, argon was used as collision gas. Data

acquisition was done by means of a signal averaging procedure.

The ES-MS spectrum of the scandium complex in acetonitrile is shown
in Figure 6.2. The protonated uncomplexed ligand (HuL)+ accounts for the
base peak (m/z 664), while various ligand fi‘agments are also present in the
spectrum, in large abundances. The structure of the macrocyclic ligand
consists of three monomeric entities (L’), each of them having one pendant
arm (P) attached. The protonated trimer (m/z 664) does not lose any of the
three pendant arms during the electrospray process, fact that denotes its

greater stability compared to the protonated monomer (m/z 222) and the

114

0 saw u 958358 :3:me 83:9QO 5 8.38 9: q 81.. .88 m .8 .3588 «3.8 No 8:9“.

N\E
so.“~ 8N o
.

h b b F I b b

431.144. 3.1L -.

\ \ / H

:8: as mmflué 88 as
.318. a. 3 £8. :9..sz

2mm owe

 

 

(%) lusueim aAueIea

 

\ -
:3 as

.3211.
V85 / 4V8; -

:8 as
s
.3 z. +E.~A.._.n_.__ .SNE u oo—

 

 

 

 

115

protonated dimer (m /z 443), respectively. The loss of a pendant arm accounts
for the neutral loss of 89 m/z units, for singly charged ionic species, the
appearance of peaks at m/z 133 and m/z 354 being direct consequences of
such in-source fiagmentation processes. Scandium ions are complexed only to
a little extent, according to the low intensities of the peaks at m/z 706 and
m/z 457, respectively.

However, the peak at m/z 457 suggests a structure that was not
determined as a part of the complexation studies in solution. The appearance
of this triply charged (1:2) “sandwich” complex is surprising, especially
taking into account the widely recognized dificulty of generating multiply
charged ions in the gas phase [9,10].

When complexation occurred in a different solvent (methanol), the
appearance of the ES-MS spectra changed significantly (Figure 6.3). More
scandium-containing ions are present, leading to the conclusion that
methanol is a more favorable medium for this particular complexation
reaction in solution. The singly charged (1:1) complex ion (ScHL)+ (m/z 706)
is now the base peak in the spectrum, while the triply charged ion of m/z 457
is more abundant than in the acetonitrile solution. For the assignment of the
peak at m/z 309, CAD was the experimental method of choice. The ion of
m/z 309 was selected as a precursor, fragmentation being induced by
collisions with argon in the collision chamber of a triple quadrupole

experimental arrangement. The product ion spectrum is shown in Figure 6.4.

116

a 88 u a asses s 838 at a m: .. saw as 5888 83m 3 2:9...

com com
— —

P b h

3...

 

 

/

:8 as

 

 

:8 as

:2 as
.318.

P

:3 as

 

 

+3.5 Lesion... 83 as

Him:

:8 as
Lamina

P b b

 

 

 

 

:8 95
L842”;

:8 85

com

b

/

:8 as

.5“;

/

:8 as
an";

 

roe

(%) lusuetun situates

117

\

:8 as

 

com
—

 

:8 as 8 ages 8.. saga 385:: E 2::

 

\

:3 as

com
_

:5

co?
_

/

:8 as
3...“;

(%) Altsuetm eAneIeu

 

roo—

118

Based only on this experiment the only information obtained about the ion of
m/z 309 is that a likely structure would be consistent with the coordination
of Sc(III) by a dimeric ligand substructure.

The structural similarity between most of the complexes involving
Sc(III) and Y(III) is beneficial in the context of our study. The ES-MS
spectrum of yttrium complex in methanol (Figure 6.5) provides some answers
to questions raised during the analysis of scandium complexation. First, the
same protonated uncomplexed ligand and ligand fragments are present in
the spectrum, independent of the nature of the metal, confirming the initial
assignments. The structures of some ligand fiagments are shown in Figure
6.6. The base peak in the spectrum is at m/z 331 and, like in the case of
scandium, its assignment is difficult without taking advantage of the above
mentioned parallelism between scandium and yttrium complexation
patterns. Indeed, if we assume that the ions at m/z 309 (Figure 6.3) and m/z
331 (Figure 6.5) have the same charge and ligand structure, the difference
between their m/z values (22 m/z units) suggests that both ions may be
doubly charged and each contain one Sc (A=45) ion and Y ( =89) ion,
respectively. However, the evidence concerning this aspect is not yet
sufficient.

Like the spectrum of the scandium complex, the spectrum of yttrium
complex contains the triply charged “sandwich” complex as well (m/z 472).

Tandem mass spectrometry helps once again solve the problem regarding its

119

0 8: u 858358 88.38 .3888: s 8.38 an: .8 81.. a; m :0 E283: 88mm 00 8:9...

 

 

 

 

 

 

 

 

 

 

 

8:
com 8o 8.: com o
P . . p _ p . a F p p _ . .
:8 NE
\ \ 8....sz .
:8 NE -
:8 es 313...: \ W
.31: :3 NE - m.
.333: / w
\ :8 £5 a.
:2 NE .3sz . M
958385: :8 NE . \A/ro
/ than; Mr
:8 NE .
bf.
\
:8 85 .2:

120

0
H300
NH2 Pendant arm (M=89)
+H2
C °
“30 CH2 m/z133
o
H3CO +NH2
H3C 0
CH2 mm

Figure 6.6 Structures of representative fragmentation products
that result in the ES-MS processes involving HsL
and its metal(lll) complexes

121

structural assignment. The MS/MS product spectrum resulting fiom CAD of
m/z 472 confirms the multiple charge carried by the precursor ion, since one
of the products, (YHL)‘r (m /z 750) is the result of a favored structure splitting
combined with a charge reduction process (Figure 6.7).

The structural confirmation regarding the ion at m/z 707 also emerges
from an MS/MS experiment. By using a collision energy of 20 eV, the CAD
product spectrum in Figure 6.8.a is obtained. A more advanced
fiagmentation can be achieved, as expected, by using a larger collision
energy (30 eV) (Figure 6.8.b). The products of the dissociation process
unambiguously confirm the charge and the structure of the ion of m/z 707.
This conclusion can now be extended to the ES-MS spectrum of the scandium
complex (Figure 6.3) as well, confirming the assignment of the peak at m/z
685 to [Sc(HaL)2-H]2+.

The third metal considered in this comparative study, lanthanum (III),
behaves, however, differently. The complexation between the metal ion and
HaL occurs to a very limited extent, even in methanol, the (1:1) complex ion
(LaHL)+ (m/z 800) being virtually inexistent in the ES-MS spectrum (Figure
6.9). The usual protonated ligand-derived ions are present as predominant
peaks, as a consequence of the availability Of a large fiaction of uncomplexed
ligand in the solution matrix.

An ion that may be relevant with respect to the possible complexation

of lanthanum is present at m/z 400. While its mass-to-charge ratio is

122

8: 8e 8 5:888 8.1888 888:: 8 8::

 

 

 

8:
com com 8: com o
— b b . P . p p _ p p p b . .
4 1 «+14 r - tit- i -8111: « rt 1 .1; i i. 1 no
:8 8s / -
.31: :8 85 8
.385 . W
m
\ . u
3
:8 8s . m.
8
:8A .
n
83.. IE I cow

 

 

123

 

 

 

 

 

 

 

 

 

 

100 - "mm”
(M707)
3 a
8
s
s
E Mi uHLi
% (m/z 664) (m/z 750)
m J
0‘ r J ‘ “ ’ % “ :J . l
0 200 400 600 800
m/z
lYll'lngrHl”

100 q . (m/z 707)

J (0:ch

In 664) HL ”
b \ /”'“’
8
g [Hull’lz-Pl .
-- (m/z 354) “(L311
.% (m/z 443)
7, (HM
a: (tn/2?
0 _‘ 4.1.1: .11: u .11.
0 200 400 600 800
m/z

Figure 6.8 Product ion spectra of CAD of m/z 707. Collision gas
pressure = 0. 6 mTorr. Collision energy: (a) 20 eV, and (b) 30 eV

124

0 8m u Eambqsp. ‘5“..me .‘ocmsmE E 8.38 at g m: .. 53 m Co Sagan 23m 3 2:9“.

 

 

 

 

S:
2.5 2.5 Se 08 o
p p . . . . r _ L b p _ . . .
i +1 a i 4 i . : 4: 31244141 T O
88 NE /
.Eii
:8 NE . mm
\\ .3;an. M
:8 £5 r m
3. 38 NE . m.
/ .
aw gs

 

 

.3"; r 2:

125

>0 mm n 38% 86.58
.53.: o d n 8383 8m 5358 62. NE B :33QO :2 8:85 o F .o 9%:

 

 

 

NE
own coo Gov CON
i d 1 1 4 1 . =. .1154 _ _«/1
89. $5 32 NE
33 £5 .33“;
+59 / hllgll’,
\ $33 illaaos /
+82. an». E.

 

(%) Misuewl annelaa

 

Se

126

consistent with a structure in which the (LaHL)+ (m/z 800) would undergo a
charge increase reaction, such a structure is unlikely. CAD of the ion at m/z
400 (Figure 6.10) does not help much, even though the dissociation pattern

denotes a multiple-step fragmentation mechanism, not elucidated at this

point.

MICHIGAN STATE UNIV. LIBRARIES
llllllllllllllllllllllllllllllllllllllllllll
31293015552403