Thesis Title: Gas phase studies of metal complexes, isomeric carbanions and distonic radical anions under soft ionization mass spectral conditions

Gas phase studies of metal complexes, isomeric

carbanions and distonic radical anions under

soft ionization mass spectral conditions

THESIS

SUBMITTED TO

OSMANIA UNIVERSITY

FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN CHEMISTRY

By

M. Kiran Kumar, M.Sc.

NATIONAL CENTRE FOR MASS SPECTROMETRY

INDIAN INSTITUTE OF CHEMICAL TECHNOLOGY

Hyderabad -500 007, INDIA

April, 2007

Dedicated To My Beloved Parents

and wife

DECLARATION The research work presented in this thesis entitled “Gas phase studies

of metal complexes, isomeric carbanions and distonic radical anions

under soft ionization mass spectral conditions” was carried out by me

independently in this institute under the supervision of Dr. M. Vairamani,

Scientist-in-Charge, National Centre for Mass Spectrometry, Indian Institute of

Chemical Technology, Hyderabad. This work is original and has not been

submitted in part or full, for any degree or diploma of this or any other

university.

Dt : (M. Kiran Kumar)

)

National Center for Mass Spectrometry

Indian Institute of Chemical Technology

Hyderabad, AP-500 007.

CERTIFICATE This is to certify that the research work incorporated in this thesis

entitled “Gas phase studies of metal complexes, isomeric carbanions and

distonic radical anions under soft ionization mass spectral conditions”

submitted by Mr. M. Kiran Kumar was carried out by the candidate under my

supervision. This work is original and has not been submitted for any other

research degree or diploma of this or any other university.

Dt : (Dr. M. Vairamani)

Tel : +91-40-27193482 Fax : +91-40-27193156

e-mail : [email protected]

National Centre for Mass Spectrometry Indian Institute of Chemical Technology

Council of Scientific & Industrial Research Hyderabad-500 007, A.P., India

Dr. M. Vairamani Scientist ‘G’, Head Analytical Division

ACKNOWLEDGEMENTS I am very much thankful to my guide and supervisor Dr. M Vairamani, Head Analytical

Chemistry Division, National Centre for Mass Spectrometry (NCMS), for welcoming me into his

research group and providing me enough impetus to carry out my work independently. My heartfelt

gratitude to him for his guidance and valuable suggestions throughout my research

My stepping into the arena of research remains incomplete without mentioning about Prof. G.

L. David Krupadanam and Dr. D. Sitha Ram, whose inspiration and guidance encouraged me to step

into the world of Mass Spectrometry.

Special thanks go to Dr. G Narahari Shastry for introducing me to the theoretical chemistry

and also for his constant encouragement.

I express my heartfelt gratitude to Dr. S. Prabhakar, for his timely help and valuable

suggestions throughout the course of my work. I am very thankful to him as he listened to all my

problems with utmost patience and suggested me solutions in an appropriate manner. I cannot imagine

anybody like him taking better care of me and he has been a great source of inspiration for me.

My sincere thanks to Dr. R. Srinivas, Mr. L. K. Rao, Dr. N. S. Swamy, Mr. R. Narsimha,

Dr. N. P. Raju,, Dr. U. V. R. V. Saradhi, V. V. S. Lakshmi and M. R. V. S. Murty for their

cooperation and encouragement.

My special thanks to Dr. N. P. Raju for reading my thesis with patience.

I would also like to thank my inter and degree college classmates, Ravi,, Nagi Reddy and

Hari, who stood beside me all the way to keep me in the right path. It is my pleasure to thank all my

past and present collegues, Shama, Veni, Srikanth, Jagadeshwar Reddy, Bhaskar, Ramu, Murali,

Shivaleela, Ramesh and Sangeeta, for making my stay at NCMS a pleasant experience. I thank my

colleagues Srinivasa Rao, Sateesh Kumar and Nagaraju.from Molecular Modeling Division.

I am grateful to my entire family for their support and encouragement throughout my studies.

There are many, many people who have helped me along the way. My regrets to those whom I

have forgotten if any, but one can be assured that, his help has been greatly appreciated!

Lastly, I would like to thank CSIR, New Delhi for the financial support in the form of

Research Fellowship (JRF/SRF). I take this opportunity to thank Dr. J. S. Yadav Director, IICT,

and Dr. K. V. Raghavan, former director for providing the facilities to carryout my research work.

-Morishetti Kiran Kumar

Abbrevations

ACN : Acetonitrile

BSSE : Basis Set Super position Error

CID : Collision-Induced Dissociation

∆E : Binding energy difference

EI : Electron Ionization

EPR : Electron Paramagnetic Resonance

ESI : Electrospray Ionization

ESR : Electron Spin Resonance

FA : Flowing Afterglow

FAB : Fast Atom Bombardment

FTICR : Fourier transform ion-cyclotron resonance

FTMS : Fourier transform mass spectrometer

FWHM : Full Width at Half Maximum

HF : Hartree-Fock

LAMMA : Laser Microprobe Mass Analysis

MALDI : Matrix Assisted Laser Desorption Ionization

MO : Molecular Orbital

NMR : Nuclear Magnetic Resonance

Pc/Pd : Precursor/Product ratio

QITMS : Quadrupole ion trap mass spectrometers

SORI : Sustained off-resonance irradiation

Teff : Effective Temperature

Index Contents Page No

CHAPTER 1 Gas phase ion Chemistry of CrIII(Salen) complex under electrospray ionization conditions

1. Prologue 1 1.1. Brief introduction of ESIMS 5 1.2. Metal-salen complexes analysis by ESIMS 7 2. Scope of the work 12 3. Results and discussion 12 3.1. Source experiments 13 3.2. Ligand-pickup experiments 20 3.3. Collision induced dissociation (CID) experiments 25 4. Conclusions 28 5. Experimental 29 6. References 31 CHAPTER 2 Proton and alkali metal ion affinities of bidentate bases: spacer chain length effects

Part 1: Proton and alkali metal ion affinities of α,ω-Diamines: Spacer

chain length effects

1. Prologue 35 1.1. The Kinetic method 40 2. Scope of the work 43 3. Results and discussion 44 3.1. Li+ ion affinity ladder construction 45 3.2. Na+ and K+ ion affinity ladder construction 48 3.3. Proton affinity ladder construction 51 3.4. Relative alkali metal ion binding energy calculations 52

3.5. Comparison between the proton and alkali metal ion affinity

orders 53

3.6. Theoretical studies 54 4. Conclusions 59

Part 2: Proton and alkali metal ion affinities of α,ω-Diols: Spacer chain length effects

1. Prologue 61 2. Scope of the work 62 3. Results and discussion 63 3.1. Proton affinity ladder construction 63 3.2. Li+, Na+ and K+ ion affinity ladder construction 67 4. Conclusions 72 5. Experimental 73 6. References 74 CHAPTER 3 Generation of regiospecific carbanions under electrospray ionization conditions and characterization by ion-molecule reactions with carbon dioxide

Part 1: Generation of regiospecific carbanions from aromatic hydroxy acids and dicarboxylic acids

1. Prologue 79 1.1. The generation of carbanions in the gas phase 80 1.1.1. Proton abstraction method 81 1.1.2. Fluorodesilylation method 82 1.1.3. Collision induced decarboxylation method 83 1.2. Characterization of carbanions 84 1.3. Stability studies of carbanions 86 1.4. Generation and characterization of specific carbanions 88 2. Scope of the work 95 3. Results and discussion 96 3.1. Geometrical isomers 99 3.2. Positional isomers 102 3.2.1. Aromatic dicarboxylic acids 102 3.2.2. Aromatic hydroxy acids 106 3.3. Effect of desolvation temperature 110 3.4. Theoretical calculations 112 4. Conclusions 118 Part 2: Generation of regiospecific carbanions from sulfobenzoic acids

1. Prologue 119 2. Scope of the work 120 3. Results and discussion 121

ℵ Abstract

3.1. Isomeric sulfobenzoic acids 122 3.2. Isomeric benzenedisulfonic acids 127 4. Conclusions 130 5. Experimental 131 6. References 133 Chapter 4 Generation of distonic dehydrophenoxide radical anions under electrospray and atmospheric pressure chemical ionization conditions

Part 1: Generation of distonic dehydrophenoxide radical anions from substituted phenols under Electrospray ionization conditions

1. Prologue 140 1.1. Formation of radical anions in the gas phase 141 1.1.1. Electron attachment 141 1.1.2. Electron transfer 142 1.1.3. Ion-molecule reactions 144 1.2. Distonic radical anions 145 1.3. Characterization of radical anions 148 2. Scope of the work 149 3. Results and discussions 150 3.1. Isomeric nitrobenzoic acids 150 3.2. Isomeric hydroxytoluenes 154 3.3. Isomeric nitrophenols and hydroxy benzaldehydes 157 3.3. Ion-molecule reactions in the collision cell with CO2 160 4. Conclusions 163 Part 2: Generation of distonic dehydrophenoxide radical anions from substituted nitrobenzenes under atmospheric pressure chemical ionization mass spectral conditions

1. Prologue 164 1.1. Atmospheric pressure chemical ionization 165 2. Scope of the work 166 3. Results and discussions 167 3.1. Isomeric nitrobenzaldehydes 168 3.2. Isomeric nitroacetophenones 172 4. Conclusions 176 5. Experimental 177 6. References 179

N N

O OCr

L

L

III

GGAASS PPHHAASSEE IIOONN CCHHEEMMIISSTTRRYY OOFF CCrrIIIIII((SSAALLEENN)) CCOOMMPPLLEEXX UUNNDDEERR

EELLEECCTTRROOSSPPRRAAYY IIOONNIIZZAATTIIOONN CCOONNDDIITTIIOONNSS

Chapter 1 Chemistry of CrIII-Salen complex…

1

CCCHHHAAAPPPTTTEEERRR 111

1. PROLOGUE

t is important to characterize the metal complexes and to identify the crucial

intermediates in metal-mediated reactions in order to understand the nature

and reactivity of metal complexes and their reaction pathways.1-4 Variety of

techniques based on X-ray diffraction, infrared spectra, nuclear magnetic

resonance (NMR), and electron paramagnetic resonance (EPR) have been used

to gather coordination structure information.5 For example, the use of NMR is

limited for characterization of metal complexes that contain a paramagnetic

metal atom; this technique is less applicable if metal complexes are present at

low concentrations or as complex mixtures.5 Consequently, researchers have

chosen the advantage of using mass spectrometry (MS) as a technique of

choice to gather coordination structure information of metal complexes. The

study of metal complex systems using MS (i.e., in the gas phase) is a rapidly

expanding field of research.1,2 As the mass spectrometer is operated in either

of the positive or negative ion mode, metal complexes can readily be isolated

and studied without interferences from counter ions, solvent or additional

complexes those are usually present in solution.1,2 These experimental

I

GGaass pphhaassee iioonn CChheemmiissttrryy ooff CCrrIIIIII((SSaalleenn)) ccoommpplleexx uunnddeerr eelleeccttrroosspprraayy

iioonniizzaattiioonn ccoonnddiittiioonnss


2

conditions are ideally suited for studying the intrinsic properties and reactivity

of various chemical entities may be clearly unrevealed.

Knowledge of the gas-phase structures of metal complexes is important

for analytical applications, as evidenced by several reviews.1,2,6,7 Recent mass

spectrometric experiments have drawn direct correlations to metal complex

mediated catalytic processes involved in various reactions.2,8-13 Mass

spectrometric investigations benefit from the ability to evaluate catalytically

active species in the gas phase that are too transient to study in solution. The

species with short lifetimes in solution do not pose a problem in the high-

vacuum environment of a mass spectrometer.

Complexes of N,N-bis(salicylidene)ethylenediamine, commonly known

as H2Salen (Figure 1), belong to a fundamental class of compounds in

coordination chemistry, known since 1933.15 This compound also belongs to the

class of Schiff base ligands, because of the preparation of this compound is by

the condensation of salicylaldehyde and ethylene diamine. Schiff base ligands

are able to coordinate metals through imine nitrogen and also the through the

hydroxyl oxygen in the case of Salen complexes. In fact, Schiff bases are able

to stabilize many different metals in various oxidation states, controlling the

performance of metals in a large variety of useful catalytic transformations. The

Salen type complexes have been extensively studied and more than 2500

complexes have been synthesized.16 Interest in Salen type complexes

intensified in 1990 when the groups of Jacobsen17 and Katsuki18 discovered the


3

enantioselective epoxidation of unfunctionalised alkenes using chiral

MnIII(Salen) complexes as catalysts (Scheme 1).

NN

OH HO

Figure 1: H2Salen.

+

O

+ PhIPhIO[MIII(Salen)]+

Scheme 1: Epoxidation of olefins with Metal(Salen) complex.

Since that time, an extremely wide variety of reactions catalyzed by

Salen complexes have been investigated. These include oxidation of

hydrocarbons,19 aziridination of alkenes,20 Diels Alder reaction,21 hydrolytic

kinetic resolution of epoxides,22 alkylation of aldehydes23 and oxidation of

sulfides to sulfoxides.24 Different mass spectrometry techniques have been

used to characterize the Metal-Salen complexes in the gas phase.8,25-28 In

general, application of the traditional method of ionization i.e. electron

ionization (EI), the earlier ionization method of MS, was limited to some metal

complexes, because most of the metal complexes are non-volatile and

thermally labile.25,27 However, there are few reports on the EI studies on a few

metal Salen complexes.25 The reported complexes include Co, Ni and Cu Salen

complexes. The EI spectra of these complexes showed abundant molecular


4

ions and fragment ions. Rohly et al.27 compared the EI mass spectra of metal

Salen complexes with the laser microprobe mass analysis (LAMMA) spectra,

wherein they report positive ion LAMMA spectra failed to provide the

information that is obtained in EI and negative ion LAMMA spectra were

dominated with only the carbon cluster ions.

The soft ionization methods such as chemical ionization, field desorption,

plasma desorption, secondary ion and fast atom bombardment (FAB) have

been developed to overcome the drawbacks encountered in EIMS towards the

analysis of metal-complexes. The FAB ionization technique has been extended

to new areas of inorganic and organometallic chemistry.28-30 Zhao at al.28

analyzed metal Salen complexes by using positive ion FAB technique. They

found that among many solvents tried to dissolve the complexes, the use of

trifluoro acetic acid (TFA) was crucial for producing good FAB spectra. Though,

much of the researchers used the FAB technique to analyze various metal

complexes, still there are some problems, like complications arising from

recombination of fragments, or interactions with the matrix used.2

Recent developments of ionization methods like matrix assisted laser

desorption ionization (MALDI)31-34 and electrospray ionization (ESI),1,7-14,35-37

have also been applied for the characterization of the metal complexes. MALDI

technique, though often used for characterization of high molecular weight

compounds, is relatively not explored much in characterization of metal

complexes. The selection of a suitable matrix is crucial in MALDI experiments,

and the ligand exchange reactions by MALDI matrix are known to complicate


5

the spectra. Relatively more studies are available for the analysis of metal

complexes using the ESI technique. The major impact of ESIMS to date is that it

can be used in identification of metal complexes, because it has allowed

observation of mass spectra for low as well as high molecular weight

compounds of ionic and nonvolatile, such as salts. In addition to the ionization

of the analytes, ESI process also transfers pre-existing ions in solution, if any, to

the gas phase, and hence is ideal for inorganic and organometallic compounds.

Further, ESI has proven to be a soft ionization method that keeps intact any

weakly bound ligands in a complex ion.36,37 Moreover, the ionization

techniques like ESI need very small amounts of sample to generate reasonably

good spectra. Use of low level quantities of samples for ESI enables the

technique for the analysis of environmental or biological samples, where the

samples are precious. With these advantages, ESI has become increasingly

popular as an analytical tool in inorganic/organometallic chemistry. This

technique, in combination with tandem mass spectrometry (MS/MS), has been

employed to study mechanistic pathways of reactions.1,7-14,35-37

1.1. BRIEF INTRODUCTION OF ESIMS

ESI technique involves spraying of a solution of the sample through a

electrically charged needle the so-called capillary which is at atmospheric

pressure (Figure 2). The spraying process can be streamlined by using a

nebulizing gas. The charged droplets are produced where the positive or

negative ions are solvated with solvent molecules. Hot gas or a dry gas, usually

called as desolvation gas, is applied to the charged droplets to cause solvent


6

evaporation. The desolvation process decreases the droplet size, leads to the

columbic repulsion between the like charges present in the droplet and further

the droplet fission leads to the formation of individual gas phase analyte ions.

The charged ions are then focused into the mass analyzer.

Figure 2: Schematic diagram of a typical ESIMS source phenomenon

operating in positive mode. Solvent evaporation of the charged

droplet generated in the source can be clearly seen (courtesy

reference Gaskell SJ, J. of Mass Spectrom., 1997)38

Application of an electrostatic field in the region between the capillary

exit and cone causes collisional activation of the solvated analyte ions. The

electrostatic field can be easily varied and provides control over the amount of


7

collisional activation. At low levels of cone voltage, the generated ions can be

sampled without causing any fragmentation. At higher levels of cone voltages,

the generated ions can be induced to undergo dissociation to give structurally

informative fragment ions. Such fragmentation in the source is called ‘cone-

voltage fragmentation’ or ‘source fragmentation’.

1.2. METAL-SALEN COMPLEXES ANALYSIS BY ESIMS

It is well known that Mn- and Cr-Salen complexes catalyze the oxidation

of organic substrates through the formation of a high-valent metal-oxo species,

(Salen)M=O. Kochi et al. used metal-Salen complexes as versatile epoxidation

catalysts in the 1980s.39-41 A typical mechanism is shown in Scheme 2.

NN

O O

MIII

+PhIO-PhI

NN

O O

M

OV

[O=MV(salen)]+ +

O

+ [MIII(salen)]+

[MIII(Salen)]+ [O=MV(salen)]+

Scheme 2: The mechanism for olefin epoxidation with Metal(salen) complex.

Epoxidation of various alkenes was successfully carried out with

iodosylbenzene in the presence of catalytic amounts of CrIII(Salen), and the

epoxidation reaction failed in the absence of the chromium complex. They

successfully isolated the catalytically active oxo-chromium(V) (O=CrV)

complexes in the condensed phase by careful recrystalization and

characterized by X-Ray and ESR studies.39 The successful isolation and


8

characterization of O=CrV(Salen) revealed the basis for oxygen activation in

the O=Cr(V) functionality. Kochi et al.39 further developed the use of

MnIII(Salen) complexes as much more versatile oxidation catalysts. An

enatioselective version of the reaction was developed later by Jacobsen et al.17

and Katsuki et al.18 by using chiral MnIII(Salen) complexes (Schemes 1 and 2).

However, the mechanistic studies on the MnIII(Salen) systems were hampered

by the fact that the catalytically active oxomanganese (O=MnV) species appear

only as short-lived putative intermediates.40 At that time, it was suggested that

the concentration of MnV-oxo complex was regulated by an equilibrium

involving µ-oxo-manganese (V) as depicted in Scheme 2. However, the

reactive species were neither isolated nor characterized in the condensed

phase. Plattner et al. successfully applied ESI technique to give a direct proof

for the epoxidation reactions using MnIII(Salen) complexes.4 They used ESI

method in combination with tandem mass spectrometry to study the

mechanistic pathway for oxygen transfer to organic substrates in the gas phase.

The [MnIII(Salen)]+ salts with iodosylbenzene were electrosprayed and the

resulted ESI spectrum showed two oxidized species, i.e. [(Salen)Mn=O]+ at m/z

337, [PhIO(Salen)Mn–O–Mn(Salen)OIPh]2+ at m/z 549.4,11 The collision-induced

dissociation (CID) of the ion at m/z 549 resulted in the decomposition products

of MnIII and MnV-oxo derivatives [Scheme 3]. These findings represented the

first experimental evidence for the formation (conproportionation) and

decomposition (disproportionation) of a µ-oxo bridged MnIV dimer acting as

reservoir of the catalytically active species involved in the oxidation reaction.12


9

[PhIO(Salen)Mn-O-Mn(Salen)OIPh]2+

m/z 549

[PhIO(Salen)MnIII]++ [PhIO(Salen)MnV=O]+

m/z 541 m/z 557

Scheme 3

The capability of MnV-oxo species to transfer oxygen to suitable organic

substrates in the gas phase was also demonstrated by collision experiments.

When [(Salen)MnV=O]+ ions were mass selected and submitted to CID

experiments with either Ar or Xe as collision gases, no fragmentation could be

obtained, [Scheme 4]. However, if the inert gas was replaced with oxygen

acceptors like sulfides or electron-rich olefins, formation of [(Salen)MnIII]+ ions,

that is, the reduction product of the oxidation reaction, was detected, [Scheme

4].

[(Salen)MnV=O]+ [(Salen)MnV=O]+

m/z 337 m/z 337

[(Salen)MnV=O]+ [(Salen)MnIII]+

m/z 337 m/z 321

Ar/Xe gas

O

+

(1)

(2)

Scheme 4: Collision cell experiments for [(Salen)MnIII]+ ions with (1) Ar/Xe

gas (2) electron rich olefin gas.

Further, Plattner et al.12 evaluated the effects of various substituents in

the 5- and 5'-positions of the Salen and found that the electron-withdrawing

substituents enhance the reactivity of the Mn=O moiety. The importance of the

axial positions in MnIII(Salen) complexes was also demonstrated by the


10

enhancement of epoxidation reaction yields with addition of a donor ligand that

stabilizes the oxometal-Salen complex. Further, the significance of axial ligands

was demonstrated by their studies on the coordination chemistry of MnIII(Salen)

and oxo MnV(Salen) complexes by applying ion-molecule reactions in the

collision cell (ligand-pick up experiments).10 Thus, the role of axial ligation on

geometry and reactivity of the high-valent oxo complex appeared to be quite

drastic.

Study of solvent clusters with ionic species in the gas phase provides

basic insights into the chemical reactivity and dynamics of ions in the

condensed phase. Such studies also provide a wealth of information on

interaction between singly charged metal ions and small ligands such as water,

methanol, acetonitrile etc. Beauchamp et al.37 studied the evaporation kinetics

on hydrated Cr and Mn Salen complexes by using ESI technique in a “soft

sampling” mode. In this study, they observed that the kinetics of water

evaporation from solvated Salen complexes is highly dependent on the central

metal ion. The clusters of CrIII(Salen) ions with two water molecules attached

exhibit special stability, indicated by their prominence in the overall cluster

distribution. These results were in accord with the solution phase chemistry

and with the ligand field theory.

Madusudanan et al.36 studied the axial interactions of CrIII(Salprn), where

Salprn = N,N-bis(salicylidene)propanediamine complexes with nucleotides

and nucleosides using ESI-MS. The nucleosides formed 1:1 and 2:1 adducts

with [CrIII(Salprn)]+ and dinucleotides formed only the 1:1 adducts. The CID of


11

these adducts revealed the attachment of Cr+ ion to the bases in nucleosides

and to both the phosphate and base in nucleotides.

It is well known that the complexes of transition metal ions are known to

undergo redox reactions during the ESI process (Scheme 4).42-47 Within the

transition metal ions, only copper ion is shown as an oxidant in several

examples for peptides and amino acids.42-46 O’Hair et al.47 studied the redox

processes in various metal ions other than copper by taking the advantage of

vacant axial positions of the metal(Salen) complexes (metal = Cr, Mn, Fe and

Co). In this process they have generated singly charged metal(Salen) ternary

complexes with hexapeptides under ESI conditions. The CID experiments on

these ternary complexes produced peptide radical cations (P+.) by redox

process (Scheme 5). The authors suggested that the redox process occur

either by a homolytic cleavage or by a heterolytic process followed by

subsequent electron transfer. In the fragmentation reactions of ternary

complexes, produced P+. were found to be highly dependent on the metal ion

used. The redox pathway was favored with FeIII or MnIII complexes when Salen

ligand contained an electron withdrawing group. The resulting peptide radical

cations are odd electron species of nonvolatile precursors, and are not

typically available under ESI or MALDI processes.

[MIII(Salen)(P)]+

[MII(Salen)] + P+.

[MIII(Salen)]+ + P

Redoxprocess

Scheme 5: Redox process of [MIII(Salen)(P)]+ in the CID experiment.


12

2. SCOPE OF THE WORK

Since all of the metal-Salen complexes generally are used as a catalyst in

the reactions, which specifically give enantioselective products, here, in this

study we have selected [CrIII(Salen)]PF6 complex for gas phase chiral

discrimination of enantiomeric compounds. It is well known that, CrIII(Salen)

complex have two free axial positions, hence, we started to make use of these

positions towards chiral recognition by using the kinetic method.

Unfortunately, we are unsuccessful in achieving the chiral discrimination with

R- and S- phenyl ethyl amines and napthyl ethyl amines in the gas phase. In this

experiment, we found that the affinity of these amines towards the axial

positions of metal-Salen complexes is fair, hence, we attempted to check the

interaction among the mono and bidentate ligands. To the best of our

knowledge, detailed studies on the behavior of [CrIII(Salen)] complexes at its

axial positions and coordination chemistry in the gas phase are not available in

the literature. The use of CoIII-Schiff base complexes with two amines in the

axial positions as antimicrobial agents was reported earlier. Therefore we

employed in our work the ESI method in combination with tandem mass

spectrometry to study the coordination chemistry of axial positions on the

unsubstituted [CrIII(Salen)] complex with amines and diamines.

3. RESULTS AND DISCUSSION

In the process of analysis of [CrIII(Salen)]PF6 complex in acetonitrile

(ACN) under ESI conditions with mono and diamines; we applied different


13

types of experiments, i.e. i) Source, ii) Ligand pick-up, and iii) CID

experiments.

3.1. SOURCE EXPERIMENTS

The positive ion ESI mass spectrum of [CrIII(Salen)]+ complex in ACN

shows major ions at m/z 318, 359 and 400, corresponding to [CrIII(Salen)]+,

[CrIII(Salen)(ACN)]+ and [CrIII(Salen)(ACN)2]+ ions, respectively (Figure 3).

Such attachment of solvent molecules to a central metal atom in the ESI process

is well documented.5,48 However, it was found that the relative abundances of

these ions are very much dependent on experimental conditions, especially

the cone voltage. Hence, we recorded the spectra at cone voltages of 10, 20

and 30 V (Figure 3(a-c)) to help understand the effect of solvent co-ordination

in the gas-phase. The spectrum recorded at a cone voltage of 10 V showed

mainly two peaks at m/z 359 and 400, corresponding [CrIII(Salen)(ACN)]+ and

[CrIII(Salen)(ACN)2]+, respectively, in addition to a minor peak at m/z 318 that

corresponds to [CrIII(Salen)]+ ion. The [CrIII(Salen)(ACN)2]+ ion was found to be

the base peak in the spectrum under these conditions. It is interesting to note

that the [CrIII(Salen)]+ did not pick up more than two acetonitrile molecules. If

the ions at m/z 359 and 400 had resulted from simple solvation of the

[CrIII(Salen)]+ complex ion by acetonitrile, one would have expected a series of

ions corresponding to [CrIII(Salen)(ACN)n]+ where n = 1,2,3 etc. The absence of

such higher adducts (n > 2) indicates that the [CrIII(Salen)]+ complex is able to

accept only two acetonitrile molecules, and that the central metal ion can adopt

a maximum of six as co-ordination state in the gas-phase.


14

H2(Salen) is a tetra-dentate ligand that occupy four coordination sites of

the central metal ion through the N, N', O, O' atoms. The ions observed at m/z

359 and 400 represent the occupation of the axial positions of the CrIII(Salen)

complex by one and two acetonitrile molecules, respectively, proving the

capability of the [CrIII(Salen)]+ complex to form five- or six-coordinated species

in the gas-phase. Similar behavior has been reported for the [CrIII(Salprn)]+

complex under ESI conditions.36 It is also in good agreement with reported

crystallographic studies in which [MnIII(Salen)] complexes were shown to bind

with one or two solvent molecules such as acetone, ethanol etc., that were used

for recrystallization, usually in axial positions.49

The ESI spectrum recorded at a cone voltage of 20 V showed the ion at

m/z 318 as the base peak with the acetonitrile adducts present at reasonable

abundance. However, the spectrum obtained at cone voltage of 30 V contains

mainly the ions at m/z 318 (base peak) and 359, and the ion at m/z 400 is

absent. This demonstrates that, at higher energies (i.e. at high cone voltages, ≥

30V), the solvent molecules are dissociated from the complex to leave the

[CrIII(Salen)]+ ion. This result prompted us to study the coordination chemistry

of the [CrIII(Salen)]+ complex with mono- and bi-dentate ligands (amines and

diamines) in detail.


15

Figure 3: The ESI mass spectra of CrIII(Salen) at different cone voltages. a)

10, b) 20, c) 30V.

The ESI mass spectrum of the [CrIII(Salen)]+ complex in the presence of

propylamine (Pr-NH2) clearly demonstrates that the displacement of solvent

molecules present in the axial positions by the stronger ligand. At low cone

voltage (10V) the [CrIII(Salen)(Pr-NH2)2]+ ion at m/z 436 is dominant, that of the

[CrIII(Salen)(Pr-NH2)(ACN)]+ ion at m/z 418 is of significant abundance, and the

[CrIII(Salen)]+ ion is of negligible importance (Figure 4a).


16

Figure 4: The positive ion ESI mass spectra of [CrIII(Salen)]PF6 complex in the

presence of propylamine (Pr-NH2) at the cone voltage of a) 10 eV b) 20 eV and

c) 30 eV.

The ion at m/z 418 is dominant over the ion at m/z 436 in the spectrum

recorded at a cone voltage 20V (Figure 4b). It may be due to decomposition of

a fraction of [CrIII(Salen)(Pr-NH2)2]+ to [CrIII(Salen)(Pr-NH2)]+ that immediately

picks up one molecule of acetonitrile, the surrounding solvent molecule, to

result in a stable six-coordinated [CrIII(Salen)(Pr-NH2)(ACN)]+ ion (m/z 418).

The ion at m/z 418 is stable even at high cone voltage (30V) but abundant

[CrIII(Salen)(Pr-NH2)(ACN)2]+


17

[CrIII(Salen)]+ ions and its acetonitrile adduct ions at m/z 359 and 400 also are

observed in the spectrum as the result of the fragmentation of the

[CrIII(Salen)(Pr-NH2)2]+ ions (Figure 4c) (the abundance of the [CrIII(Salen)(Pr-

NH2)2]+ ion is considerably reduced at high cone voltages ≥30 V). Thus, the

relative abundances of the ions at m/z 318, 359 and 400 in this experiment can

be used as a measure of the stability of the [CrIII(Salen)(Pr-NH2)2]+ ion. These

experiments reveal the ability of Pr-NH2 to occupy both the axial positions of

the complex to form six-coordinate complex ions that survive at low cone

voltage. This prompted us to study the coordination of diamines which are

bidentate in nature and can occupy both axial positions of the [CrIII(Salen]+

complex. We carried out three types of experiments, i.e. source experiments,

ligand pick-up experiments and CID experiments, for this purpose.

A series of primary α,ω-diamines (DA) were selected to study not only

the effect of the bidentate nature of diamines but also the effect of chain length

of the ligand on the occupation of the axial positions of [CrIII(Salen)]+. We used

1,2-diaminoethane (1), 1,3-diaminopropane (2), 1,4-diaminobutane (3), 1,5-

diaminopentane (4), 1,6-diaminohexane (5), 1,7-diaminoheptane (6) and 1,8-

diaminooctane (7) as bidentate ligands. The ESI mass spectra of an equimolar

(100 µM) mixture of [CrIII(Salen)]+ and diamines were recorded at cone

voltages of 10 and 30 V. All the spectra recorded at 10V show mainly the

[CrIII(Salen)(DA)]+ ions, and other ions are negligible. However, the spectra

recorded at 30V still showed the [CrIII(Salen)(DA)]+ ion as the base peak but, in

addition, [CrIII(Salen)]+, [CrIII(Salen)(ACN)]+ and [CrIII(Salen)(ACN)2]+ ions are


18

also present in significant abundances. This indicates that the

[CrIII(Salen)(DA)]+ ion is stable at high cone voltage (30V), unlike

[CrIII(Salen)(ACN)2]+ and [CrIII(Salen)(Pr-NH2)2]+. Though the

[CrIII(Salen)(DA)]+ ion is dominant, the appearance of the other low mass ions

in the spectra recorded at cone voltage 30 V but not at 10 V implies partial

decomposition of [CrIII(Salen)(DA)]+ ion to give unbound [CrIII(Salen)]+ ion

(m/z 318) that picks up surrounding solvent molecules in the API interface

region to yield the ions at m/z 359 and 400 (Figure 4).

NN

O OCrIII

PF6

+ H2N-(CH2)n-NH2NN

O OCrIII

H2N

NH2

(CH2)n

ESI

+

Scheme 6: Reaction of [CrIII(Salen)]+ with a diamines (n=2-8, 1-7) to form a

bidentate complex.

The relative abundances of these ions were found to vary depending on

the size of the ligand used. Hence, we consider the spectra recorded at a cone

voltage of 30V for further discussion, as the relative abundances of the ions at

m/z 318, 359 and 400 are found to reflect the stability of the diamine co-

ordination complexes formed with [CrIII(Salen)]+. The ESI mass spectra of

[CrIII(Salen)] in the presence of the chosen diamines are listed in Table 1. It is

clear that ligands 1 and 2 are able to form the [CrIII(Salen)(DA)]+ ion with ease,

and the other ions due to loss of DA at m/z 318, 359 and 400 are less abundant

(<8%). [CrIII(Salen)(DA)2]+ and [CrIII(Salen)(DA)(ACN)]+ ions are absent,


19

consistent with the effective occupation of the two empty axial positions of

[CrIII(Salen)]+ by the two amino groups in the case of ligands 1 and 2 (Scheme

6).

Relative abundance (%) Ion

1 2 3 4 5 6 7

[CrIII(Salen)]+

m/z 318 2.9 4.3 10 19 44 34 40

[CrIII(Salen)(ACN)]+

m/z 359 5.8 7.9 20 37 65 61 66

[CrIII(Salen)(ACN)2]+

m/z 400 4.4 7.1 18 35 56 54 48

[CrIII(Salen)(DA)]+ 100 100 100 100 100 100 100

[CrIII(Salen)(DA)(ACN)]+ - - 2.9 5.1 5.1 0.7 0.3

[CrIII(Salen)(DA)2]+ - - 1.5 4.4 5.1 8.7 13

Table 1: Positive ion ESI mass spectra (cone voltage 30 V) of mixtures of

[CrIII(Salen)]+ (as the PF6- salt) with diamines (DA) ligands (1-7) in

acetonitrile (ACN) solvent.

In the case of higher diamine homologues (3-6), the [CrIII(Salen)(DA)2]+

and [CrIII(Salen)(DA)(ACN)]+ ions are also present, reflecting the decreasing

ability of the higher homologues to yield stable bidentate complexes. It is

interesting to note that the relative abundances of the ions m/z 318, 359 and 400

gradually increase from ligand 1 to 5. In the case of 6 and 7, the relative

abundances of these ions are lower than those found for ligand 5. The gradual

increase of the abundances of the fragment ions at m/z 318, 359 and 400 from

ligand 1 to 5 reflects a gradual decrease in the stability of [CrIII(Salen)(DA)]+

ion from 1 to 5 under the ESI conditions used here. The stability of


20

[CrIII(Salen)(DA)]+ ion for ligands 6 and 7 seems to be higher than that of ligand

5 and lower than those from 1-4. There is a marginal but consistent increase in

the relative abundances of [CrIII(Salen)(DA)(ACN)]+ and [CrIII(Salen)(DA)2]+

ions from 3 to 5, reflecting that the ability of the diamine to occupy two axial

positions decreases from 1 to 5. In the case of 6 and 7, the abundance of the

[CrIII(Salen)(DA)(ACN)]+ ion is negligible and that of the [CrIII(Salen)(DA)2]+ ion

is higher than for the other ligands used. From these observations it can be

inferred that the bidentate nature decreases as the chain length increases from

1 to 5, while that of ligands 6 and 7 shows mixed behavior. In order to

understand the stability of these ionic species in the gas phase, we performed

ligand-pickup experiments in the collision cell.

3.2. LIGAND-PICKUP EXPERIMENTS

In these experiments the ions of interest were selected using MS1 and

allowed to undergo ion-molecule reactions in the collision cell with the ligand

of interest introduced into the collision cell. The resultant product ions were

then analyzed by MS2 (Figure 5). Experiments done on [CrIII(Salen)]+ as the

mass-selected precursor ion using acetonitrile as the collision gas showed the

pickup of one and two acetonitrile molecules by [CrIII(Salen)]+ to form five- and

six-coordinated species (ions at m/z 359 and 400, respectively, Figure 6). The

same behavior was observed when [CrIII(Salen)(ACN)]+ and [CrIII(Salen)

(ACN)2]+ are selected as precursor ions; the former ion showed addition of one

acetonitrile and the latter ion did not undergo any further addition of

acetonitrile (Figure 6).


21

Figure 5: Schematic diagram of a triple quadrupole instrument.

Figure 6: The spectra obtained from ligand-pickup experiments using

acetonitrile as the collision gas for precursor ions:

a) [CrIII(Salen)]+ ion, m/z 318

b) [CrIII(Salen)(ACN)]+ ion, m/z 359 and

c) [CrIII(Salen) (ACN)2]+ ion, m/z 400.


22

The displacement of weaker ligands (ACN) in the axial positions of

[CrIII(Salen)]+ by relatively stronger ligands (amines) was also observed in the

collision cell experiments using propylamine as collision gas. The propylamine

formed abundant [CrIII(Salen)(Pr-NH2)]+ and [CrIII(Salen)(Pr-NH2)2]+ ions with

[CrIII(Salen)]+ as the precursor ion; the same product ions are also observed

when [CrIII(Salen)(ACN)]+ and [CrIII(Salen)(ACN)2]+ are selected as precursors

(Figure 7). These experiments clearly indicate that the empty axial positions of

unsubstituted [CrIII(Salen)]+ ion are easily occupied by any ligand, and that the

displacement of weaker ligands by relatively stronger ligands occurs when a

complex with weaker ligands is selected as precursor ion. Similar ligand-

pickup experiments were reported previously by Plattner et al.10 for MnIII

(Salen) species; they showed that MnIII is able to form only five-coordinated

species unless there is an electron-deficient substituent on Salen. However, in

the present case, [CrIII (Salen)]+ is able to form six-coordinate complexes

easily, possibly due to differences in the electronic configurations of MnIII and

CrIII ions. Note that it is difficult to achieve equilibrium between the collision

(reactant) vapor and the selected ion species in the collision cell under the

mass spectral conditions used, as the experimental time window for the ion in

the collision cell is about 10-100 milliseconds. However, the results indicate

that it is possible to study the relative efficiencies of ligand exchange in the

collision cell.


23


propylamine as the collision gas for precursor ions:

a) [CrIII(Salen)]+ ion, m/z 318

b) [CrIII(Salen)(ACN)]+ ion, m/z 359 and

c) [CrIII(Salen) (ACN)2]+ ion, m/z 400.

We extended the ligand-pickup experiments to study of the bidentate

nature of diamines and the stability of diamine complexes, using acetonitrile in

the collision cell. From the ESI source mass spectra (Table 1) it is evident that

the five-coordinated complex [CrIII(Salen)(L)]+ (L = acetonitrile or

propylamine) exists in the gas-phase in addition to the stable six-coordinated


24

species. The observation of [CrIII(Salen)(DA)]+ ion in the ESI mass spectrum of

the mixture of [CrIII(Salen)]+ and diamine poses the question whether or not

both axial positions are occupied by the two amino groups of the diamine

ligand. When we selected a complex ion containing a monodentate ligand (Pr-

NH2) for ligand-pickup experiments, for example [CrIII(Salen)(Pr-NH2)]+, it

picked up one acetonitrile molecule in the collision cell to yield [CrIII(Salen)(Pr-

NH2)(ACN)]+ (Figure 8). This observation is as expected because one axial

position in [CrIII(Salen)(Pr-NH2)]+ is free and acetonitrile can readily occupy the

vacant axial position. When we selected [CrIII(Salen)(DA)]+ ion for ligands 1-4

in MS1 for ligand-pickup experiments with acetonitrile in the collision cell, no

addition of acetonitrile to the selected species was observed. The same

experiments for [CrIII(Salen)(DA)]+ ions from 5-7 resulted in

[CrIII(Salen)(DA)(ACN)]+ ions of low abundance (5.1, 1.7 and 0.5% for 5, 6 and

7, respectively); the spectrum for the case of ligand 5 is shown in Figure 8b as

an example. Similar results were obtained from ligand-pickup experiments

between [CrIII(Salen)(DA)]+ ions of 1-7 and propylamine in the collision cell.

From these experiments it can be concluded that [CrIII(Salen)(DA)]+ ions from

ligands 1-4 are stable in the collision cell under the present experimental

conditions. In the case of ligands 5-7, one of the axial coordinate bonds

between the central metal ion and the amino group of the ligand becomes

weaker, so that the DA can be displaced by acetonitrile in the collision cell.

Further, these experiments confirm that the [CrIII(Salen)(DA)]+ species is


25

relatively less stable for ligand 5 than for 6 and 7. These observations are

consistent with the results obtained from the ESI mass spectra (source).


acetonitrile as the collision gas for precursor ions:

a) [CrIII(Salen)(Pr-NH2)]+ ion, m/z 377 and

b) [CrIII(Salen)(hexd)]+ ion, m/z 434.

3.3. COLLISION INDUCED DISSOCIATION (CID) EXPERIMENTS

With a view to study the stability of diamine complexes with

[CrIII(Salen)]+, we also performed CID experiments on [CrIII(Salen)(DA)]+ ions

using argon as the collision gas at different collision energies (10, 12 and 14


26

eV). All the spectra resulted in only one product ion corresponding to

[CrIII(Salen)]+, and the relative abundance of this ion was found to depend on

the nature of the diamine used. It is demonstrated in the literature that the

precursor/product (Pc/Pd) abundance ratio can be used to measure the

relative stability of adduct ions, since a stable precursor ion undergoes less

decomposition.50 The Pc/Pd ratios, i.e. relative abundance ratio

[CrIII(Salen)(DA)]+/[CrIII(Salen)]+, obtained from the CID spectra of

[CrIII(Salen)(DA)]+ ions from 1-7 at different collision energies, are presented in

graphical form in Figure 9.

Figure 9: The plot of Pc/ Pd ratios ([Cr(III)(Salen)(DA)]+/ [Cr(III)(Salen)]+)

obtained at collision energies of 10, 12 and 14 eV from CID of

[Cr(III)(Salen)(DA)]+ ions for ligands (Diamines) 1-7.

The order of stabilities of [CrIII(Salen)(DA)]+ complexes for diamines 1-7

can be given as 2> 1> 3> 4 ≈ 7> 6> 5. This agrees well with the similar stability

order obtained from source experiments (mass spectra) already discussed,


27

except for ligands 1 and 2. The collision cell experiments show that ligand 2

forms a more stable complex with [CrIII(Salen)]+ compared to ligand 1.

In the source experiments the behavior of ligands 1 and 2 is reversed

but the difference is marginal. Hence, it can be concluded from both source

and collision cell experiments that the feasibility of complexation of diamines

with unsubstituted [CrIII(Salen)]+, by occupying the axial positions, decreases

as the chain length increases from ligand 1 to 5. We cannot offer an explanation

from the available experimental data for the marginal increase in the stability

of the complexes with 1 and 2 and similarly with ligands 6 and 7.


28

4. CONCLUSIONS

The positive ion ESI mass spectra for [CrIII(Salen)]+ complex in the

presence of amines as ligands (propylamine and a series of diamines (1-7))

were studied with a view to understand the coordination chemistry of the

complex in the gas phase. The ESI mass spectra of [CrIII(Salen)]+, either in

acetonitrile alone or in the presence of propylamine, showed ions

corresponding to five- and six-coordinated species, respectively. The

[CrIII(Salen)]+ in the presence of bidentate ligands (L = diamines) mainly

resulted in [CrIII(Salen)(L)]+ ions in which the two empty axial positions in

[CrIII(Salen)]+ species are occupied by the two amino groups of the diamine. In

addition to five- and six-coordinated complex ions, other ions corresponding to

[CrIII(Salen)]+ and its solvent adduct ions are also observed in the ESI mass

spectra, and the relative abundances of these ions were found to depend on the

cone voltage. However, the relative abundances of the above ions at constant

cone voltage reflected the stability of the [CrIII(Salen)(L)]+ ions. The

[CrIII(Salen)(L)]+ ion is most stable for 1,2-diaminoethane and 1,3-

diaminopropane ligands. The stability of the complex ion decreased from 1,4-

diaminobutane to 1,6-diaminohexane, and there is a slight increase for 1,7-

diaminoheptane and 1,8-diaminooctane. A similar trend was observed from the

ligand-pickup experiments in the collision cell using acetronitrile or

propylamine as collision gas, and from CID experiments on [CrIII(Salen)(L)]+

ions.


29

5. EXPERIMENTAL

The [CrIII(Salen)]PF6 was synthesized using a known procedure.51 All the

ligands (propylamine and diamines, 1-7) used in the present study were

purchased from Sigma-Aldrich (Steinheim, Germany) and were used without

further purification. The solvents (HPLC-grade) were purchased from Merck

(Mumbai, India). Stock (1mM) solutions of all ligands and of [CrIII(Salen)]PF6

were made in acetonitrile. The stock solutions of the ligand of choice and of

[CrIII(Salen)]PF6 were mixed in appropriate volumes (1:1) and diluted with

acetonitrile to achieve final concentrations of 100µM each.

All the mass spectra were recorded using a Quattro LC triple-

quadrupole mass spectrometer (Micromass, Manchester, UK) coupled with an

HP1100 series liquid chromatograph (Agilent, Palo Alto, USA); the data were

acquired using Masslynx software (version 3.2). The ESI capillary voltage was

maintained between 4 and 4.2 kV, and the cone voltage was kept at 30 V unless

otherwise stated. Nitrogen was used as desolvation and nebulization gas. The

source and desolvation temperatures were kept at 100o C. The ESI mass spectra

were recorded by scanning MS1 and the sample solutions were injected

through the Quattro LC injector with a Valco six-port valve with a 10 µL loop,

using acetonitrile at a flow rate of 100 µL/min using the HPLC pump. The CID

spectra and ligand-pickup experiments were obtained by selecting the

precursor ion of interest with MS1 and scanning MS2. For these experiments,

the sample solutions were introduced into the source of the mass spectrometer


30

using an infusion pump (Harvard Apparatus) at a flow rate of 10 µL/min. Argon

was used as the collision gas for CID experiments and the collision cell

pressure was maintained at 9x10-4 mbar. For ligand-pickup experiments,

acetonitrile or propylamine was used as the collision gas, maintaining the

collision cell pressure at 9x10-4 mbar. All the spectra reported here were

obtained as the averages of 20 scans.


31

6. REFERENCES

1. Cotton R, D’Agostino A, Traeger JC. Mass. Spectrom. Rev., 1995; 14: 79.

2. Henderson W, Nicholson BK, McCaffrey LJ. Polyhedron, 1998; 17: 4291.

3. Plattner DA. Int. J. Mass Spectrom., 2001; 207: 125.

4. Plattner DA, Feichtinger D, El-Bahraoui J, Wiest O. Int. J. Mass Spectrom.,

2000; 195/196: 351.

5. Chipperfield JR, Clayton J, Khan SJ, Woodword S. J. Chem.Soc., Dalton

Trans., 2000; 1087.

6. Di Marco VB, Bombi GG. Mass Spectrom. Rev., 2006; 25: 347.

7. Operti L, Rabezzana R. Mass Spectrom. Rev., 2006; 25: 483.

8. Hinderling C, Plattner DA, Chen P. Angew. Chem., Int. Ed. Engl., 1997; 36:

243.

9. Hinderling C, Feichtinger D, Plattner DA, Chen P. J. Am. Chem. Soc., 1997;

119: 10793.

10. Feichtinger D, Plattner DA. Angew. Chem., Int. Ed. Engl., 1997; 36: 1718.

11. Feichtinger D, Plattner DA. J. Chem. Soc., Perkin. Trans., 2000; 2: 1023.

12. Feichtinger D, Plattner DA. Chem. Eur. J., 2001; 7: 591.

13. El-Bahraoui J, Wiest O, Feichtinger D, Plattner DA. Angew. Chem. Int. Ed.,

2001; 40: 2073.

14. Plattner DA. Top. Curr. Chem., 2003; 225: 153.

15. Pfeiffer P, Breith E, Lübbe E, Tsumaki T. Liebigs Ann., 1933; 503: 84.

16. Dalton CT, Ryan KM, Wall VM, Bousquet C, Gilheany DG, Top. Catal., 1998;

5: 75.


32

17. Zhang W, Loebach JL, Wilson SR, Jacobsen EN. J. Am. Chem. Soc., 1990; 112:

2801.

18. Irie R, Noda K, Ito Y, Matsumoto N, Katsuki T. Tet. Lett., 1990; 31: 7345.

19. Lee NH, Lee CS, Jung DS. Tetrahedron Lett., 1998; 39: 1385.

20. Omura K, Uchida T, Irie R, Katsuki T. Chem. Commun., 2004; 2060.

21. Mc Gilvra JD, Rawal VH. Synlett., 2004; 2440.

22. Shin CK, Kim SJ, Kim GJ. Tetrahedron Lett., 2004; 45: 7429.

23. Maeda T, Takeuchi T, Furusho Y, Takata T. J. Polym. Sci., Part A: Polym.

Chem., 2004, 42: 4693.

24. Kim SS, Rajagopal G. Synthesis., 2003: 2461.

25. Patel KS, Rinehart KL, Bailar JC. Jr. Org. Mass Spectrom., 1970; 4: 441.

26. Lacey MJ, Macdonald CG, Shannon JS. Org. Mass Spectrom., 1978; 13: 188.

27. Rohly KE, Heffren JS, Douglas BE. Org. Mass Spectrom., 1984; 19: 398.

28. Huang SK, Rood MH, Zhao SH. J. Am. Soc. Mass spectrom., 1997; 8: 996.

29. Miller JM. Adv. Inorg. Chem. Radiochem., 1984; 28: 1.

30. Bruce MI, Liddell MJ. Appl. Organomet. Chem., 1987; 1: 191.

31. Dale MJ, Dyson PJ, Suman P, Zenobi R. Organometallics, 1997; 16: 197.

32. Dale MJ, Dyson PJ, Johnson BFG, Langridge-Smith PRR, Yates HT. J. Chem.

Soc. Dalton Trans., 1996; 771.

33. Dale MJ, Dyson PJ, Johnson BFG, Martin CM, Langridge-Smith PRR, Zenobi

R. J. Chem. Soc. Chem. Commun., 1995; 1689.

34. Huc V, Boussaguet P, Mazerolles P. J. Organomet. Chem., 1996; 521: 253.


33

35. Cole RB (ed.), "Electrospray Ionisation Mass Spectrometry, Fundamentals,

Instrumentation and Applications,” Wiley Interscience, New York, 1997.

36. Madhusudanan KP, Katti SB, Vijayalakshmi R, Nair BU. J. Mass spectrom.,

1999; 34: 880.

37. Lee SW, Chang S, Kossakovski D, Cox H, Beauchamp JL. J. Am. Chem. Soc.,

1999; 121: 10152.

38. Gaskell SJ. J. Mass Spectrom., 1997; 32: 677.

39. Siddall TL, Miyaura N, Huffman JC, Kochi JK. J. Chem. Soc., Chem. Commun.,

1983; 1185.

40. Samsel EG, Srinivasan K, Kochi JK. J. Am. Chem. Soc., 1985; 107: 7606.

41. Srinivasan K, Kochi JK. Inorg. Chem., 1985; 24: 4671.

42. Hu P, Loo JA. J. Am. Chem. Soc., 1995; 117: 11314.

43. Gatlin CL, Turecek F, Vaisar T. J. Mass Spectrom., 1995; 30: 1605.

44. Gatlin CL, Turecek F, Vaisar T. J. Mass Spectrom., 1995; 30: 1617.

45. Gatlin CL, Rao RD, Turecek F, Vaisar T. Anal. Chem., 1996; 68: 263.

46. Vaisar T, Gatlin CL, Turecek F. Int. J. Mass Spectrom. Ion Processes, 1997;

162: 77.

47. Waters T, O’Hair RAJ, Wedd AG. J. Am. Chem. Soc., 2003; 125: 3384.

48. Katta V, Choudhury SK, Chait BT. J. Am. Chem. Soc., 1990; 112: 5348.

49. Calligaris M, Randaccio L. in “Comprehensive Coordination Chemistry,”

Wilkinson G, Mc Cleverty JA. (Eds.), Vol. 2, Peramon, Oxford, 1987, chap.

20.1, pp. 715.

50. Cai Y, Cole RB. Anal. Chem., 2002; 74: 985.


34

51. Premsingh S, Venkatramanan NS, Rajagopal S, Mirza SP, Vairamani M,

Sambasivarao P, Valavan K. Inorg. Chem., 2004; 43: 5744.

(CH2)n

X

X

M+

X = NH2, OH; M+ = H+, Li+, Na+ and K+

PPRROOTTOONN AANNDD AALLKKAALLII MMEETTAALL IIOONN AAFFFFIINNIITTIIEESS OOFF BBIIDDEENNTTAATTEE BBAASSEESS::

SSPPAACCEERR CCHHAAIINN LLEENNGGTTHH EEFFFFEECCTTSS

Chapter 2.1 Proton and alkali metal ion interactions of Diamines..

35

CCCHHHAAAPPPTTTEEERRR 222 PPPAAARRRTTT---111

1. PROLOGUE

rotonated species are central to many chemical and biological processes,

such as acid-base phenomena, astrochemistry, radiation chemistry, mass

spectrometry, catalysis, surface chemistry, protein conformation, membrane

transport and enzyme catalysis.1 Since, the proton is having almost comparable

attributes with the alkali metal ions, the hydrogen has been placed in the first

group of the periodic table. Alkali metal ions are one of the most abundant ions

in biological systems, where they are involved in a variety of processes,

including osmotic balance, the stabilization of biomolecular conformations and

information transfer through ion pumps and ion channels.2-6 They interact with

poly functional molecules, like peptides and proteins to perform such

regulatory and structural functions.2,7 Thus, the knowledge of proton and alkali

metal ion binding interactions with polyfunctional biomolecules is an important

step in understanding the biochemical processes.8,9 Good correlations exist

between the metal ion and proton binding affinity to the bases, though the

proton affinities are much higher.10,11 Alkali metal ion binding interactions to

small model ligands bearing the heteroatom (oxygen or nitrogen) functional

group (binding sites) in the gas phase provides intrinsic information necessary

P

PPrroottoonn aanndd aallkkaallii mmeettaall iioonn aaffffiinniittiieess ooff αα,,ωω --DDiiaammiinneess::

SSppaacceerr cchhaaiinn lleennggtthh eeffffeeccttss


36

for better understanding of the interaction of the metal ions with biologically

active macromolecules.

The proton/alkali metal ion interactions decrease the space occupied in

three dimensional structures wherever possible, and they can opt to the

regulation of enzymatic activity, protein folding and functioning and stability of

biological systems.2,7 A common structural feature of the proton/alkali metal ion

bound complex is the presence of interactions between multiple functional

groups. For instance, in the protonated polyfunctional ions, protonated part of

the molecule may interact with an unprotonated group to form intramolecular

hydrogen bonds. Many organic reactions also proceed through protonated

intermediates or involve direct hydrogen bonding such as those involved in

protein or DNA complexes. Such hydrogen bondings greatly influence the

structure and the properties of organic compounds. In particular,

intramolecular hydrogen bonds are often responsible for determining the

predominant conformers in solution12 as well as in the gas phase.13-17

Intramolecular solvation of protonated functional groups influence the gas-

phase basicities of polyfunctional molecules. Occurrence of intramolecular

solvation in protonated species was characterized by several authors from a

series of di- and polyfunctional ions such as diols,17-21 diamines,10,13,14,16,20-28

diethers,15,16 diketones,15,16,29 amino acid derivatives,30 cyclic and acyclic

polyethers, and open chain and cyclic diols31 and amino alcohols using both

theoretical and experimental studies.14 The intramolecular hydrogen bond


37

stabilizes the ion by up to 20 kcal mol-1, and thereby increases the proton

affinity of several bi or polyfunctional compounds.20/21

In biological systems, especially in proteins, several basic motifs exist,

separated by varying chainlengths. Polyamines found to be present in the cells

of microorganisms and animal organisms, contribute to the stabilization of the

structure and activity of tRNA and DNA.32 It is well known that polyamines, such

as putrescine (1,4-diaminobutane), spermidine and spermine, are present in

millimolar concentrations in most tissues and microorganisms. Other

polyamine derivatives including cadaverine (1,5-diaminopentane) and 1,3-

diaminopropane are also found in some living cells. Although there were

several reports that describe the effects of the polyamines on the higher order

structure of DNA, the mechanism of the action of polyamines on DNA molecules

has not been clarified yet.33

α,ω-alkanediamines are compounds of interest in various domains of

organic and organometallic chemistry because these are bifunctional, can

cyclize after protonation (Scheme 1).13,22 These are also known as chelating

bidentate ligands in coordination chemistry, as reactants in industrial

polymerization processes, and as synthetic enzymes for complex formations

with target substrates through hydrogen bonding.34 Thermochemical

properties of α,ω-diamines have been studied by several

researchers.10,13,14,16,20-27


38

H2N (CH2)n NH2H+

NH2

(CH2)n

H2N

H+

Scheme 1

Estimation of thermo chemical properties to the monofunctional

molecules is straight forward, whereas to that of molecules with two or more

functional groups is intresting to investigate, because of possible internal

hydrogen bonding between the two like or unlike functional groups. Molecules

with two or more functional groups may have more proton affinity, greater than

that expected for either of the individual groups, due to the internal hydrogen

bond formation by favorable molecular geometry. Intramolecular hydrogen

bonding and the consequent chelating ring size were found to be the key

factors controlling the stability of the protonated complexes.10,13,14,16,20-28 The

first examples and interpretations of this phenomenon were explicated by Aue

et al.22 and followed by Yamdagni and Kebarle,13 who found that the proton

affinities of α,ω-diamines are significantly higher than those of monoamines

with the same alkyl chain length. The protonated diamines were proposed to

have cyclic structures, and the ring strain present in the structures was

evaluated with reference to the strain-free structures of proton bound dimers of

monoamines. Both groups of workers noted that the proton affinity of

H2NCH2CH2NH2 was substantially less than that of its higher diamino analogues,

which was attributed to the large strain energy expected for a five membered


39

ring (the assumption being that for maximum stability, the N--H+--N bond will

be linear as in proton-bound dimers of monoamines). Later, Bouchoux et al.

extensively studied protonation thermo chemistry of α,ω−diamines. Mass

spectrometric methods and computational techniques were extensively used

for the protonation studies on α,ω-diamines. 10,13,14,16,20-28

Though there were a numerous reports on the protonation

thermochemistry of α,ω-diamines, studies towards the alkali metal ion affinities

of the diamines are scarce. There was only one report on ab initio molecular

orbital (MO) calculations on the stabilities and binding energies of bidentate

ethylene diamine with alkali metal (Li+ and Na+) ions.10 The computed binding

energies of Li+ and Na+ ions with ethylene diamine are 66.3 and 42.3 kcal mol-1,

respectively. Thermochemical data obtained in the gas phase are of particular

value both for understanding the nature of metal ion-basic component

interactions in condensed phase and for explaining solvent phenomenon.35

The solvent-free environment of the mass spectrometer provides an

ideal medium for measuring the intrinsic properties such as proton/metal ion

affintity in the absence of interfering solvent effects. The kinetic method

developed by Cooks et al.36-39 has been used to estimate thermochemical data

for a wide range of organic and biological molecules for more than 25 years

and has often been reviewed.36


40

1.1. THE KINETIC METHOD

The best-known application of the kinetic method is for the

determination of proton affinities, gas phase acidities, metal and chloride ion

affinities, and electron affinities.20,21,36-48 This method37,40,41 is an effective

method for estimating the relative binding energies of two similar bases that

bind to a central ion, typically a proton/metal ion. Several series of bidentate

molecules, such as diols, diethers and diamines, have been studied by this

method for the determination of their proton affinities. The method starts with

the generation of proton/metal ion bound dimer between two bases and is

subjected to tandem mass spectrometric experiments to obtain the

corresponding proton/metal ion bound monomeric bases. The ratio of the

relative abundances associated with two competitive dissociation channels

(heterodimers) is then measured to estimate the relative binding energies. The

logarithmic value of the relative abundance is proportional to the logarithm of

the relative rate of dissociation of the two reaction channels.

For example, the dissociation of a proton/metal (M) bound heterodimer

of L1 and L2 leads to M+ bound monomers (equation 1 and 2)

[L1- - -M+- - -L2]

L1 + L2M+

L2 + L1M+ (2)

(1)k1

k2


41

Here k1 and k2 are the rate constants for the competitive dissociations of

the cluster ion to yield L1M+ and L2M+, respectively. Based on transition state

theory, 49 the natural logarithm of rate constant ratio is given by the equation 3.

ln(k1/k2) = ln(Q2*/Q1*) + [εo(1) - εo(2)]/RTeff (3)

In which Q1* and Q2* are the partition functions of the activated

complexes of reaction 1 and 2, respectively; εo(1) and εo(2) are the

corresponding activation energies; R is the gas constant and Teff is the effective

temperature, a parameter in temperature units that reflects the internal energy

of the dissociating heterodimer. Assuming that the abundances reflect rate

constants37, 40,41 and that no reverse activation barriers exist equation 3 tends

to,

ln([L2M+]/[L1M+]) = ln(Q2*/Q1*) - ∆HML1/RTeff + ∆HML2/RTeff (4)

Where, ∆H°M is the ∆H of the dissociation reaction LM+ L + M+ or the

metal ion affinity of L. If L1 and L2 are structurally similar, as expected with used

ligands, ∆(∆S M+) should be close to zero, i.e. Q2* ≈ Q1* and fragmentation of

L1M and L2M proceed by simple bond cleavages from the loosely bound

complex L1--M+--L2, the reverse activation energies for channels L1M and L2M

should be negligible. In such a case, the difference in proton/metal ion

affinities between the two amino acids of interest would be nearly equals to the

binding energy (∆E) of those amino acids: then the above equation is simplified

further to (equation 5).


42

ln([L2M+]/[L1M+]) (∆HML2 - ∆HML1)/RTeff ∆EM/RTeff (5)~~ ~~

∆EM is the binding energy of central proton/metal ion between the two

heterodimers, where L1M and L2M are the relative abundances of the two

reactions from the fragmentation of the M+ bound heteodimer, and Teff is the

effective temperature of the activated precursor cluster ion. Thus, Teff is a

measuring parameter for internal energy of dissociating cluster ion and

primarily depends on the structure and lifetime of the ion. Several

investigations have shown that different dimer ions (of chemically similar

molecules), generated under identical experimental conditions and found to

have the same lifetime, also have fairly similar Teff, independent of the central

ion holding them. Hence, Teff of [L1--M+--L2] can be approximated by the

effective temperature of the corresponding H+-bound heterodimers.

Using the above assumptions Cerda and Wesdemiotis39 semi-

quantitatively evaluated the relative Cu+ ion binding energies of α-amino acids.

In these experiments, Teff values for [AA1-Cu-AA2]+ was approximated to the Teff

for H+- bound hetero amino acids, coproduced in the same sample. Application

of the equation 5 yielded Teff value, and this value was further used to convert

ln(k2/k1) values in the estimation of Cu+ binding energies of all the 20 common

amino acids. In a similar way, Lee et al.41 also constructed relative Ag+ ion

binding energy ladder for essential α-amino acids using the kinetic method.

These binding energies were compared with their relative H+ and Cu+ ion

binding energies. However, there is no systematic study on the alkali metal


43

affnities of homologues series of α,ω-diamines, and the effect of spacer chain

length on their binding efficiency.


The literature reports clearly demonstrate the enhancement of the

proton affinities of α,ω-diamines with respect to the primary amines. This was

readily explained by the formation of a strong internal hydrogen bond in the

protonated form of the diamines. Among homologues series of α,ω-diamines,

1,4-butane diamine depicts highest proton affinity, owing to the seven

membered ring stabilized structure after protonation. However, there were no

systematic studies on the alkali metal ion affinities of α,ω-diamines. Hence, we

undertook a systematic experimental and computational study on the

measurement of relative gas phase affinity of alkali metal ions (Li+, Na+ and K+)

with a series of α,ω-diamines and compared them with the corresponding

proton affinities. In this part, the kinetic method and quantum chemical

calculations are employed to address the following points.

What are the variations in the relative binding affinities of proton and alkali

metal ions in the given series?

What is the nature of bridging interactions the alkali metal ion complexes

have?

What are the structural differences between the proton and alkali metal ion

complexes of diamines?


44


It is well known that electrospray ionization technique is the best method

to study the interactions between the metal ions and various systems. We have

used the kinetic method for evaluating the relative alkali metal ion [Li+, Na+ and

K+] affinities for a series of seven homologues α,ω-diamines, namely 1,2-

diaminoethane (1), 1,3-diaminoproane (2), 1,4-diaminobutane (3), 1,5-

diaminopentane (4), 1,6-diaminohexane (5), 1,7-diaminoheptane (6), and 1,8-

diaminooctane (7).

Figure 1: CID mass spectra of Li+ bound heterodimer of compounds 1 and 2.

The study was initiated with Li+ ion binding of 1-7. The typical ESI mass

spectrum recorded for a methanol/water solution containing two different

diamines from 1-7 (DA1 and DA2) and lithium chloride show the H+ and Li+

bound mono and dimeric cluster ions. The spectrum recorded for a mixture of

1 and 2 in the presence of Li+ is shown in Figure 1 as an example. The Li+

40 60 80 100 120 140 160 180 200m/z0

100

%

81

67

141

[Li-1]+

[Li-2]+

[1-Li-2]+

CID with Ar gas

[1-Li-2]+


45

bound heterodimeric ions, [DA1+Li+DA2]+ formed with various combinations of

diamines are mass selected by MS1, and dissociated in the collision cell under

similar experimental conditions. The heterodimeric ions dissociate by

competitive elimination of neutral diamines yielding two fragment ions

corresponding to [DA1+Li]+ and [DA2+Li]+ (equations 6 and 7). The relative

abundances of the resulted Li+ bound monomers, viz., I(Li+-DA1) and I(Li+-DA2)

vary and reflect the Li+ ion affinity of individual diamine. The diamine that has

more affinity results in higher abundance of its Li+ bound monomer than that of

with less affinity. For example, the CID spectrum of [1+Li++2] (Figure 1) shows

higher abundance of [1+Li+] than [2+Li+], which confirms higher Li+ ion affinity

of 1 when compared to that of 2.

[DA1--Li+--DA2]

DA2 + DA1Li+ (6)

DA2Li+ + DA1 (7)

3.1. Li+ ION AFFINITY LADDER CONSTRUCTION

The CID spectra were recorded for all possible lithiated heterodimers of

diamines (1-7). The spectra of fifteen out of twentyone heterodimers resulted in

both Li+ bound monomers, [Li-DA1]+ and [Li-DA2]+ with considerable

abundance. The other spectra are dominated with only one of the lithiated

monomer being the other monomer negligible due to large difference in their

Li+ ion affinities. Hence, we consider only those spectra, which resulted in both


46

fragment ions, for constructing the Li+ ion affinity order by the kinetic method.

The relative abundance ratio of two fragment ions, i.e., I(Li+-DA2)/I(Li+-DA1)

ratio values are calculated from the CID spectra of all possible heterodimers,

where the Li+ ion affinity of DA2 is higher than DA1. The natural logarithm of

I(Li+-DA2)/I(Li+-DA1) ratio values are used to construct relative Li+ ion affinity

ladder. A metal ion binding ladder can be constructed with ln[I(Li+-DA2)/I(Li+-

DA1)] values in which the ligand of lowest affinity is considered as reference.

The experimentally measured ln[I(Li+-DA2)/I(Li+-DA1)] values are summarized

in a relative Li+ affinity ladder shown in Figure 2. In this ladder construction,

most of the diamines are compared to at least three others.

Figure 2: Measured ln[I(Li+-DA2)/I(Li+-DA1)] values for Li+-bound heterodimers of

diamines (1–7). The data presented under the heading ln[I(Li+-DA2)/I(Li+-

1)] are average cumulative values expressed relative to ethylene diamine

(1). The numbers given in parentheses are estimated errors resulting from

the measurement of abundance ratios.


47

The ln[I(Li+-DA2)/I(Li+-DA1)] values calculated for the successful

combinations are found to be reproducible. The ln[I(Li+-DA2)/I(Li+-DA1)] values

are internally consistent for the Li+ bound heterodimers of diamines. For

example, the value for [4·Li·1]+ is 3.51, a very similar value is obtained by

adding the ln[(ILi+-DA2/I(Li+-DA1)] values of three intermediate steps, viz.

ln[I(Li+-3)/I(Li+-1)] + ln[I(Li+-2)/I(Li+-3)] + ln[I(Li+-4)/I(Li+-2)] = 1.91 + 0.08 +

1.58 = 3.57. The ln[I(Li+-DA2)/I(Li+-DA1)] values for other pairs are also

consistent internally with a difference not more than 0.2. Similarly results are

also obtained when the experiments were performed at different collision

energy values (2, 4, 6, and 8 eV). This accord confirms that entropic effects,

which tend to be non-additive, are indeed negligible with the diamines

studied.

From Li+ ion affinity ladder, the relative Li+ ion affinity order for α,ω-

diamines can be drawn as, 1Li+ < 3Li

+ ≤ 2Li

+ < 4Li

+ < 6Li

+ < 5Li

+ ≤ 7Li

+. In the

Li+ affinity order for α,ω-diamines, the deviation of compound 2 and 5 in the

order indirectly suggests that the structure of the lithiated diamine may be

playing a role. There are two possible structures for the resulted lithiated

species. One possibility is acyclic structure in which Li+ ion is bound with one

of the amine group. The other is a cyclic structure where both the amine groups

in diamine coordinate to the Li+ ion. If the lithiated diamines were acyclic, one

would expect gradually increase in Li+ ion affinity order as the chain length of

diamine increased due to increase of positive inductive effect with increase in


48

the number of methylene groups attached to the amine groups in diamine.

However, the observed Li+ ion affinity order shows the possibility of cyclic

structures. The formation of cyclic structures for protonated diamines was

studied in detail. As mentioned in the introduction, the high proton affinity for 3

among series of primary α,ω-diamines (1-7) is due to its stable cyclic structure

on protonation. Similarly, the higher Li+ ion affinity for compound 2 and 5 when

compared to their respective higher homologous diamine may also be due to

the stability of the resulted cyclic lithiated species. In fact, the formation of

bicoordinated lithium complexes is known in the literature.50,51

3.2. Na+ AND K+ ION AFFINITY LADDERS CONSTRUCTION

We have extended the experiments towards Na+ and K+ ion affinity order

determination for diamines (1-7) by performing similar experiments as we

applied to lithium, to study the effect of metal ion size in the stabilization of

metal bound diamines. For this purpose, we have generated all possible

heterodimers of Na+/K+ ion bound diamines, [DA1-M+-DA2], where M=Na or K.

The same fifteen pairs of diamines that are used for Li+ are also successful for

both the Na+ and K+ experiments. The CID spectra of these [DA1-M+-DA2] ions

are recorded, and the relative abundances of the Na+/K+ bound monomers

(i.e., M+-DA1 and M+-DA2 formed during the dissociation) correlated with the

relative Na+/K+ ion affinities of the two bases. The natural logarithm of

abundance ratio, ln[I(M+-DA2)/I(M+-DA1)] values are calculated from the CID

spectra of all possible heterodimers at similar experimental conditions, where


49

the M+ ion affinity of DA2 is higher than DA1, and are used to obtain the Na+ and

K+ ion affinity ladders. The relative Na+ and K+ affinity ladder is shown in

Figure 3 and Figure 4, respectively.

Figure 3: Measured ln[I(Na+-DA2)/I(Na+-DA1)] values for Li+-bound

heterodimers of diamines (1–7). The data presented under the

heading ln[I(Na+-DA2)/I(Na+-1)] are average cumulative values

expressed relative to ethylene diamine (1). The numbers given in

parentheses are estimated errors resulting from the measurement of

abundance ratios.

From these ladders, the relative Na+ ion affinity order can be given as

1Na+ < 2Na+ < 3Na+ < 4Na+ < 5Na+ < 6Na+ < 7Na+, and is not similar when

compared to that obtained for lithium. The sodium ion affinity towards diamines

increases as the number of methylene groups in diamine is increased. This


50

observation suggests that the resulted sodiated diamines either have linear

structure, or do not reflect the ring size effect on their stabilization if there are

cyclic. However, from the present data we are unable to propose the correct

structure for the sodiated diamines.

Figure 4: Measured ln[I(K+-DA2)/I(K+-DA1)] values for Li+-bound heterodimers

of diamines (1–7). The data presented under the heading ln[I(K+-

DA2)/I(K+-2)] are average cumulative values expressed relative to

propane diamine (2). The numbers given in parentheses are

estimated errors resulting from the measurement of abundance

ratios.

The relative K+ ion affinity orders of the diamines (1-7) were also

determined and can be given as 2K+ < 1K+ < 3K+ < 4K+ < 6K+ < 5K+ < 7K+. As in

the case of Li+ ion affinity ladder, the ln[I(M+-DA2)/I(M+-DA1)] values for Na+


51

and K+ ions calculated for successful combinations are found to be

reproducible and values are internally consistent for the Na+/K+ bound

heterodimers of diamines. The K+ ion affinity order is closely comparable to

that obtained for sodium ion, except 1 and 2 where the potassium ion affinity of

1 is higher than 2. The higher affinity of compound 1 compared to that of 2 may

be explained assuming cyclic structures for the potassiated diamines.

3.3. PROTON AFFINITY LADDER CONSTRUCTION

The present study on the relative affinity of a series of diamines towards

Li+, Na+ and K+ shows that the affinity order is affected by the size of metal atom

and diamines. The discrepancies in the Li+ ion affinity order of diamines may be

explained through cyclic structures and their stability. However, the sodium

and potassium ion affinity order of diamines cannot be explained in a similar

way. Though all the alkali metals used are known to be bi-dentate in binding

with ligands, the present experimental results does not give much information

on the structures of the ions.

With a view to understand the differences between the alkali metal ion

affinity order of the studied compounds and the proton affinity order, we have

also constructed proton affinity ladder. We applied similar method that was

followed for alkali metal ions, for construction of proton affinity ladder by

replacing alkali metal ion with proton. The obtained proton affinity ladder is

given in Figure 5. From the ladder, the relative proton affinity order can be

given as 1H+ < 2H+ < 7H+ < 6H+< 5H+ < 4H+ < 3H+. The proton affinity order


52

obtained in the present method is in good agreement with the literature

values.52

Figure 5: Measured ln[I(H+-DA2)/I(H+-DA1)] values for H+-bound heterodimers

of diamines (1–7). The data presented under the heading ln[I(H+-

DA2)/I(H+-1)] are average cumulative values expressed relative to

propane diamine (2). The numbers given in parentheses are

estimated errors resulting from the measurement of abundance

ratios.

3.4. RELATIVE ALKALINE METAL ION BINDING ENERGY CALCULATIONS

It is well known that, for chemically similar compounds, the natural

logarithm of abundance (I) ratio values are directly proportional to the alkali

metal ion binding energy difference (∆E) (equation 8) between the used

diamines, where the entropy term is close to zero.40,41,53,54


53

ln[I(M+- DA2) /I(M+- DA1)] ~ ∆E /RTeff (8)~

Attempts were made to convert relative alkali metal ion affinity orders into

relative binding energies by measuring the Teff of the dissociating cluster

ions.40,41,53,54 It was already shown in the literature that when experiments are

performed at identical conditions, Teff is fairly similar for dimeric ions of

chemically similar molecules, irrespective of the central ion holding the two

molecules. For the measurement of Teff, the dissociation of proton bound

heterodimers of diamines was studied at different collision energies (2, 4, 6, and

8 eV). For successful measurement of Teff value in this method, heterodimers of

each diamine with atleast three other diamines should be studied. The diamine 1

and 3 could not be used for this purpose because of their extreme low or high

proton affinity values when compared to the other diamines. The left out

diamines 2, 4, 5 and 6 also could not be used for this study because the

difference in the proton affinity values among the three diamines (4, 5 and 6) is

very less (± 0. 5 kcal mole-1). This restricts the number of good references

needed for the measurement of a reliable Teff value. Consequently, we could not

obtain reliable Teff values due to the non availability of enough number of

references among the studied diamines. Hence, the present study is limited to

the relative alkali metal ion affinity orders.

3.5. COMPARISON BETWEEN PROTON AND ALKALI METAL ION AFFINITY ORDERS

Inspection of the relative orders of proton affinities and alkali metal ion

affinities of primary α,ω-diamines (1-7) reveals that the proton affinity order is


54

substantially different from the alkali metal ion affinity order. In the case of

proton affinities of diamines, the diamine 3 has higher proton affinity due to its

stable seven membered cyclic structure for the protonated 3. The diamine 1

has the least proton affinity due to unstable five membered ring formation after

protonation, and the proton affinity of diamine 2 is a little higher than 1 with a

relatively stable six membered structure for the protonated 2. The proton

affinity of diamines 4-7 are in between 3 and 2, and gradually decrease from 4

to 7, which could be due to gradual increase in the ring strains. Whereas, the

relative alkali metal ion affinity is always high for diamine 7 for all the alkali

metal ions studied. Although there are few differences among the alkali metal

ion affinity orders i.e., between 2 and 3; 5 and 6 in Li+ order, and 1 and 2 in K+

order, overall the metal ion affinity order decreased from 7 to 1. It suggests that

the positive inductive effect is playing major role in stabilization of the

metallated diamine than those of ring strains. The minor differences among the

relative orders of alkali metal ions may be due to the size of alkali metal atom.

We seek to explain the observed contrasting order for H+ and Li+ ion

affinities of α,ω-diamines through quantum chemical calculations.

3.6. THEORETICAL STUDIES

The H+ and Li+ ion affinities are estimated using the equations 9 and 10,

respectively. B3LYP/6-31G* method is used for the geometry optimizations and

obtaining the thermochemical data. All the structures considered are

characterized as minima on the potential energy surface. This is followed by

single point calculations at MP2/6-311++G** level. Counterpoise method was


55

used to calculate the basis set super position error (BSSE). In our studies all the

calculations were done using the Gaussian 9855 suite of program.

Metal ion affinity (∆H298) = ∆Eele + ∆Ethermal + T∆S - BSSE (9)

Proton affinity (∆H298) = ∆Eele + ∆Ethermal + 5RT/2 (10)

The relative binding affinity orderings of the computed results are in

excellent agreement with the experimental observations for both proton and

Li+ ion affinities, except the change of proton affinity order between 4 and 5.

Theoretically obtained H+ and Li+ ion affinity orders can be given as 1H+ < 2H+

< 7H+ < 6H+ ≤ 4H+ < 5H+ < 3H+ and 1Li+ < 3Li+ ≤ 2Li+ < 4Li+ < 6Li+ < 5Li+ ≈ 7Li+

respectively. Figure 6 depicts the optimized geometries of the Li+ and

protonated complexes. All the Li+ complexes are virtually symmetrically

bridged, and as the length of the spacer chain increases Li+ is going into the

cavity of the molecule. In agreement with the previous studies,2 computations

reveal that the Li+ ion affinities are less than one third of the proton affinities to

the diamines. The non-linearity of the relative binding affinities of Li+ ions can

be clearly traced to the subtle and intricate conformational changes in the Li+

complexed cyclic structures. In addition, higher energy mono-dentate

minimum energy structures where the cation is bound to the acyclic isomers

are obtained. Systematic conformational analyses of neutral diamines reveal

that the open chain linear structures are global minima besides several other

local minima with warped on cyclic structures. The energy difference between

the acyclic and cyclic neutral conformation (∆E1), the conformation


56

reorganization energy upon complexing with the cation (∆E2), and the relative

energy differences for the mono- and bi-dentate complexation of the cations

(∆E3) are given (Table 1). The latter two quantities (∆E2 and ∆E3) aid in

dispersing exact differences in the ion bridging to the diamines, while ∆E1

granges the conformational flexibility of the diamines. One salient feature is

that while the complexation energy going from acyclic to cyclic structures

increases by about 5-11% for proton, it is more than 80% for Li+ ion. Therefore,

although metal ions have much smaller magnitudes of affinity to diamines, their

gain in going to bidentate ligation is substantial as reflected in the

corresponding higher ∆E3 for Li+. In contrast, ∆E2 is consistently higher for H+

shows that proton induces higher strain in the diamine skeleton upon

complexation. Thus, although Li+ ion gains substantially due to the coordination

of the second amine group, the corresponding Li+ ion cyclic complex has a less

strained diamine motif compared to that of proton complex. The larger size of

Li+ ion as well as its non-covalent nature of interaction is responsible for a

highly flexible complexation, as reflected in smaller ∆E2.

While the lithium bridging is virtually symmetrical in all cases, the

proton bridge is highly unsymmetrical. The contrasts in the trends of the

relative stability orderings are due to interplay of intricate conformational

energetics during the formation of metal ion chelate ring. Thus, while both

proton and metal ion α,ω-diamine complexes prefer cyclic conformations, the

nature of bridging and the energy differences between the mono and bidentate


57

complexes are quite different. Thus, the computational study on the binding

affinities of Li+ and H+ ions to α,ω-diamines highlights the disparities between

the complexation energetics and structures.

Structure 1 2 3 4 5 6 7

6-31G* 75.4 78.8 77.6 83.9 83.8 83.1 82.7

6-31++G** 65.7 69.3 67.6 71.6 74.6 74.4 B3LYP/

6-311++G**a 65.3 69 67.3 71.1 74.1 73.7 73.1

MP2/ 6-311++G**a

61.7 65.2 63.9 68.3 72.3 71.7 72

6-31G* 234 242.4 247.1 246.7 244.2 242.9 242

6-31++G** 229.1 237 241.3 238.1 237.2 236.8 235.3 B3LYP/

6-311++G**a 228.3 236 240.2 237.2 237.2 235

MP2/ 6-311++G**b 228.8 236.7 240.9 238.5 238.4 238.9 239.2

Table 1: Calculated metal ion (Li+) and proton affinity values:

a) Single point calculation on B3LYP/6-31G* optimized geometries and thermal

corrections were taken from B3LYP/6-31G* level.

b) Single point calculation on B3LYP/6-31++G** optimized geometries and

thermal corrections were taken from B3LYP/6-31G* level.

116.9

113.3

109.3

109.3

109.3

113.3

1.976

1.976

1.500

1.500

1.533

1.533

109.4

106.5

126.5

111.6

116.1

117.9

112.8

1.989

1.981

1.4991.530

1.542

1.542

1.503

115.8112.1

115.9113.3

110.5 110.4

149.7

109.7

2.006

2.011

1.5001.532

1.538

1.544

1.531

1.503

156.4

106.0

111.6

115.4

115.0

118.1116.2

112.6

116.4

106.62.002

1.997

1.4961.531

1.541

1.540

1.538

1.541 1.5361.502

121.4

112.6

115.3

116.4

112.7

115.9115.1

111.6109.1

134.8

115.32.021

2.031

1.4971.533

1.537

1.539

1.537

1.535

1.541 1.538

1.504

115.7

115.1

178.9

109.0

115.1 115.7111.9

111.9109.1

111.9

2.011

2.011

1.4981.532

1.540

1.538

1.540

1.532 1.498

101.3

110.3

101.3

92.6110.3

2.0021.495

1.527

1.4952.002

123.6

101.6107.7

106.186.1

1.877

1.0611.514

1.542

1.471

150.3

113.1

104.3

99.8110.41.651

1.097

1.520

1.531

1.537

1.494

109.4

165.6

107.8111.0

117.1

117.1112.3 106.4

1.545

1.133

1.5101.531

1.544

1.533

1.493

169.5

113.3112.1

114.5

116.8

116.9112.0

105.1

1.129

1.576

1.4931.533

1.541

1.552

1.5291.515

173.5

111.6

111.4

117.7

117.4

118.7

115.0

112.0

118.0

1.610

1.119

1.5131.528

1.541

1.542

1.544

1.5401.499

173.8112.3

109.9116.1

114.9

113.7

115.0 115.3112.4

121.9

114.2

1.674

1.108

1.5141.5291.541

1.545

1.549

1.545

1.537

1.532

1.502

170.5113.6

113.1115.2

114.2

116.1116.1

113.2123.9

1.108

1.664

1.503

1.538

1.541

1.535

1.540

1.541

1.528

1.517

114.3

Lithium Nitrogen Carbon Hydrogen

1Li+ 2Li+ 3Li+ 4Li+ 5Li+ 6Li+ 7Li+

1H+ 2H+ 3H+ 4H+ 5H+ 7H+6H+

Figure 6: B3LYP/6-311++G** optimised geometries of cyclic H+ and Li+ ion complexes of diamines. Bond lengths in Å and bond

angles in degrees.


59

4. CONCLUSION

Relative alkali metal ion (Li+, Na+ and K+) affinity ladders of α,ω-diamines

are successfully constructed by using the Kinetic method. From these ladders,

the obtained relative Li+, Na+ and K+ ion affinity orders are 1Li+ < 3Li+ ≤ 2Li+ <

4Li+ < 6Li+ < 5Li+ ≤ 7Li+; 1Na+ < 2Na+ < 3Na+ < 4Na+ < 5Na+ < 6Na+ < 7Na+ ; 2K+

< 1K+ < 3K+ < 4K+ < 6K+ < 5K+ < 7K+, respectively. We have also constructed

relative proton affinity ladders of the same diamines by the Kinetic method, and

the obtained relative proton affinity order is 1H+ < 2H+ < 7H+ < 6H+< 5H+ < 4H+

< 3H+. Contrasting orders were found for relative proton and alkali metal ion,

when compared them with their relative affinity orders. Proton, being smaller

than alkali metal ions, shows the contrasting order with 1-7. Whereas alkali

metal ions are having the almost similar trend, i.e. the affinity order increasing

with respect the chain length of the diamine, except for small discrepancies for

2 and 3 & 5 and 6 for Li+, 1 and 2 for K+. These small differences may be due to

the strain effects of metallated diamines. It is worth mentioning that the kinetic

method amplifies even small differences between similar compounds in the

measurement of important thermodynamic parameters. The observed

contrasting orders for H+ and Li+ are explained by quantum chemical

calculations. The contrasts in the trends of the relative stability orderings are

found to be due to interplay of intricate conformational energetics during the

formation of metal ion chelate ring. Though both proton and metal ion α,ω-

diamine complexes prefer cyclic conformations, the nature of bridging and the


60

energy differences between the mono and bidentate complexes are quite

different.

Chapter 2.2 Proton and alkali metal ion interactions of Diols..

61

CCCHHHAAAPPPTTTEEERRR 222 PPPAAARRRTTT 222

1. PROLOGUE

as-phase proton/metal ion energetics of bidentate bases is of

fundamental interest particularly because of their sensitivity to the

existence of intramolecular weak bonding interactions in the

protonated/metallated form.10,13,14,16-28,56 In the part one of this chapter, we

measured the relative alkali metal ion affinity orders of the homologues series

of α,ω-diamines by applying the Kinetic method. The alkali metal ion affinity

orders of α,ω-diamines were compared with their proton affinity order and

found that the affinity orders depend on the size of the central ion used as well

as the spacer chain length of α,ω-diamine. The differences in alkali metal ion

affinity orders when compared to the proton affinity orders were explained

with the support of theoretical calculations. It is always ideal to extend such

kind of gas phase ion studies to other bifunctional group molecules for better

understanding of their multiple interactions with proton/metal ions.

The homologues series of α,ω-diols are the best choice of bifunctional

molecules because they are also involved in various chemical and biochemical

processes.20,21 Looking at the series of diol structures, raises many interesting

questions: If we change the functional groups of bifunctional molecule, what

G

PPrroottoonn aanndd aallkkaallii mmeettaall iioonn aaffffiinniittiieess ooff αα,,ωω--ddiioollss::

SSppaacceerr cchhaaiinn lleennggtthh eeffffeeccttss


62

would be the alkali metal affinity orders and what kind of spacer chain length

effects can be seen in α,ω-diols? During the last few years, only the first

members of the series of α,ω-diols viz., 1,2-ethane diol, 1,3-propane diol, and

1,4-butane diol, have been experimentally and theoretically studied for their

proton energetics.17-21 In these experiments, the authors used different types of

techniques like high pressure mass spectrometry, ion-cyclotron resonance

mass spectrometry and ab intio calculations to measure accurate proton affinity

values. The authors proposed the formation of intramolecular hydrogen bond

to result in cyclic structures to explain the proton affinities of the studied α,ω-

diols, which was similar to that reported for α,ω-diamines. On the other hand,

Stone et al.57 studied the thermodynamics of the association of ammonium ion

with several α,ω-diols in high pressure mass spectrometers, where formation of

two hydrogen bonds between ammonium ion and α,ω-diols was reported to

lead a stable cyclic structure. However, there is no systematic study on the

metal ion energetics for α,ω-diols. Hence, we have undertaken the study of

proton and alkali metal ion affinity orders of α,ω-diols.


The Li+, Na+ and K+ ion affinity order of a series of α,ω-diols (HO-(CH2)n-

OH, n= 2-10, 8-16) can be measured by the Kinetic method. The proton

affinities of only a few α,ω-diols (8-10) are known in the literature.17,20,21 Here

we plan to measure the relative proton affinity order for the selected series of

α,ω-diols by applying the kinetic method, and these proton affinity orders will


63

then compared to their relative alkali metal ion affinity orders (Scheme 2).

Comparison of the proton affinity order of α,ω-diols with alkali metal ion affinity

order will reveal the effect of spacer chain length of α,ω-diols and the size of

central metal ion. The effect of functional groups (-OH Vs –NH2) can also be

divulged by comparison of affinity orders of α,ω-diols with those already

obtained for α,ω-diamines in part one of this chapter.

HO (CH2)n OHM+

OH

(CH2)n

HO

M+

Scheme 2

3. RESULTS AND DISCUSSIONS

The Kinetic method that we used for measuring proton/metal affinity

orders for α,ω-diamines is applied here also to obtain relative proton and alkali

metal ion affinity ladders for α,ω-diols.

3.1. PROTON AFFINITY LADDER CONSTRUCTION

To determine the proton affinity ladder, proton bound heterodimers

[Diol1-H+-Diol2], where Diol1 and Diol2 are two different diols of 8-16, viz., HO-

(CH2)n-OH, n= 2-10, are produced in the electrospray ionization source, and

their dissociation spectra are recorded. As expected, the spectra result in two

proton-bound monomers (H+-Diol1 and H+-Diol2). The CID spectrum recorded

for the proton bound heterodimer of 10 and 11 is given in Figure 7a as an


64

example. The relative abundances of the H+ bound monomers directly reflect

the relative proton affinities of the two diols in the heterodimer. The CID

spectra of 16 different heterodimers were evaluated to derive the relative

proton affinity order of 8-16.

Figure 7: Collision induced dissociation spectra of [10-M+-11] heterodimers,

where M+ is a) H+ b) Li+ c) Na+ d) K+ ions at 2 eV collision energy.

The natural logarithm of relative abundance ratio of formed proton

bound monomers, ln[I(H+-Diol2)/I(H+-Diol1)] values are calculated from the CID

spectra of all possible H+ bound heterodimers under similar experimental

conditions, and are used to obtain the proton affinity ladder. In the ladder

construction, most of the diols are compared to at least three other diols. The

ln[I(H+-Diol2)/I(H+-Diol1)] values calculated for successful combinations are


65

found to be reproducible. Based on this data, the H+ affinity ladder of 8–16 is

constructed (Figure 8). From the ladder it can be understood that the diols 8

and 9 have lower proton affinities when compared to other diols. We could not

measure ln[I(H+-Diol2)/I(H+-Diol1)] value from the CID spectrum of the

heterodimer formed from 8 and 9, because the spectrum contains only one

product ion, 9H+. Absence of the fragment ion due to 8H+ indicates a large

proton affinity difference between the diols 8 and 9. The reported proton

affinity values,52 of the diols of 8 and 9 differ by < 7.5 kcal mol-1, and the present

experimental results show that this difference is too large to be estimated by

the kinetic method. The ln[I(H+-Diol2)/I(H+-Diol1)] values are found to be

internally consistent for the proton bound heterodimers of diols. For example,

the value for [13-H+-10]+ is -1.63 (Figure 8), and a very similar value is

obtained by adding the ln[I(H+-Diol2)/I(H+-Diol1)] values of three intermediate

steps, viz. ln[I(H+-13/I(H+-12)] + ln[I(H+-12)/I(H+-11)] + ln[I(H+-11)/I(H+-10)] = -

0.48 - 0.98 - 0.35 = -1.81. The ln[I(H+-Diol2)/I(H+-Diol1)] values for other pairs

are also consistent internally with a difference not more than 0.2. Similar results

were also obtained when the experiments were performed at different collision

energy values (2, 4, 6, and 8 eV). This accord confirms that entropic effects,

which tend to be non-additive, are indeed negligible with the diols studied.

The cumulative natural logarithm ratios of all the diols are used as a measure of

relative affinity order. The relative proton affinity order can be given as 8H+<<

9H+<< 14H+≈ 13H+< 12H+< 11H+< 10H+< 15H+< 16H+, in which the order for 8-

11 is inline with the reported proton affinities.52


66

Figure 8: Measured ln[I(H+-Diol2)/I(H+-Diol1)] values for H+-bound

heterodimers of diols (8–16). The data presented under the heading

ln[I(H+-Diol2)/I(H+-14)] are average cumulative values expressed

relative to octane diol (14).

It is well known from the previous reports that 1,4-butane diol (10) was

having more proton affinity than other diols (8-11) due to attaining a stable

seven membered cyclic structure after protonation. The relative proton affinity

order of diols 8-14 is exactly correlating with the order that obtained for

corresponding diamines 1-7 (refer part 1). The studied series of diols includes

two higher homologues (15 and 16), which show higher proton affinity than the

diol 10.


67

3.2. Li+, Na+ AND K+ ION AFFINITY LADDERS

In a similar way we have constructed relative Li+, Na+ and K+ affinity

ladders by replacing proton with alkali metal ion. The constructed ladders of

Li+, Na+ and K+ are shown in Figure 9, 10, and 11, respectively. The natural

logarithm of abundance ratio value cannot be measured for lithium bound

dimer of 8 and 9, as found for their proton bound heterodimer, due to a large

difference in the Li+ ion affinity of 8 and 9. This problem is not encountered for

measuring the logarithm of abundance ratio values for the Na+ and K+

heterodimers of 8 and 9. The affinity order is identical for all alkali metal ions

used, and the obtained order can be given as 8M+<< 9M+< 10M+< 11M+<

12M+< 13M+< 14M+< 15M+< 16M+. From this metal ion affinity order of the

diols, it can be noticed that the relative metal ion affinity gradually increased

with the increase in the spacer chain length (number methylenes) of the diol.


68

Figure 9: Measured ln[I(Li+-Diol2)/I(Li+-Diol1)] values for Li+-bound


ln[I(Li+-Diol2)/I(Li+-9)] are average cumulative values expressed

relative to propane diol (9).


69

Figure 10: Measured ln[I(Na+-Diol2)/I(Na+-Diol1)] values for Na+-bound

heterodimers of diols (8–16). The data presented under the

heading ln[I(Na+-Diol2)/I(Na+-8)] are average cumulative values

expressed relative to ethane diol (8).

Since all the diols are having chemically similar structures, the natural

logarithm of abundance (I) ratio values are directly proportional to the binding

energy difference (∆M+) (equation 8) between the used diols with M+ ion,

where the entropy term is close to zero. Attempts were made to convert

ln[I(M+-Diol2)/I(M+-Diol1)] order into absolute proton/metal ion affinities by

measuring the Teff of the dissociating cluster ions. However, due to the non-


70

availability of enough number of references among the studied diols we could

not obtain reliable Teff values. Hence, the present study is limited to the relative

metal ion affinity order.

Figure 11: Measured ln[I(K+-Diol2)/I(K+-Diol1)] values for K+-bound


ln[I(K+-Diol2)/I(K+-8)] are average cumulative values expressed

relative to ethane diol (8).

From the relative proton and alkali metal ions affinity orders of 8-16, it

can be noticed that the relative proton affinity order of 8-16 is different when

compared to their alkali metal ion affinity orders. Like in the case of diamines,

the diols might have attained cyclic structures after proton (unsymmetrical

E


71

bridging)/metal ion (symmetrical bridging) addition. The observed contrasting

orders between proton and alkali metal affinity orders must be due to the size

of the used central ion.


72

4. CONCLUSIONS

In this study we have arrived at the relative proton and alkali metal ion

(Li+, Na+ and K+) affinity orders of nine α,ω-diols (HO-(CH2)2-10-OH, 8-16) by

applying the kinetic method under ESI conditions. The relative affinity order for

proton is 8H+<< 9H+<< 14H+ ≈ 13H+< 12H+< 11H+< 10H+< 15H+< 16H+; where

as for alkali metal ions the affinities are in the order of 8M+<< 9M+< 10M+<

11M+< 12M+< 13M+< 14M+< 15M+< 16M+, irrespective of alkali metal ion used.

From the relative affinity ladders of the proton and alkali metal ions of α,ω-

diols, we can conclude that the proton is tend to bind as cyclic structure and

hence show difference in the affinity order with respect to the spacer chain

length. Whereas, such differences are not observed in alkali metal ion affinity

orders.

The overall proton/alkali metal ion affinity orders of diols is almost similar to

that obtained for diamines, except some dissimilarities for the Li+ ion affinity

order of diamines. This suggests that the diols show similar structural effects to

that of diamines and there are no significant functional group effects on the

proton/metal ion affinity order among the homologues series. Furthermore, the

proton affinity of diols/diamines does not increase gradually with an increase

in the number of methylene groups, whereas their alkali metal ions affinities do

increase, which reflects less correlation between the proton and the alkali

metal ion affinity.

Chapter 2 Proton and alkali metal ion interactions of bidentate bases..

73

5. EXPERIMENTAL

All experiments were carried out using a triple quadrupole mass

spectrometer equipped with an external electrospray ionization (ESI) source

(micromass, Manchester, UK) and the data were acquired using Masslynx

software (version 3.2). Proton or alkali metal ion bound heterodimer complex

ions were formed from a methanol/water (50:50) solution of diamines (1-7)

diols (8-16) (10–4 mol L-1) in H2O–metal salts (10–4 mol L-1). This solution was

injected into the ESI source at a flow rate of 10µL min-1 using a syringe pump

(Harvard Apparatus, Kent, UK). The Electrospray ionization capillary voltage

was maintained between 3.5 to 4.0 kV, and the cone voltage was kept at 5–10 V.

Nitrogen was used as desolvation and nebulization gas. The source and

desolvation temperatures were kept at 100o C. Argon was used as the collision

gas for all CID experiments and the collision cell pressure was maintained at

1.5-3.0 X 10-4 mbar by using four different collision energies (2, 4, 6 and 8 eV).

These values were minimum energies that provide sufficiently abundant

product ions from the diamine and diol heterodimers to give accurate relative

abundance ratios. Using such conditions, consecutive fragmentations were

avoided and the influence of entropy effects was limited.


74

6. REFERENCES

1. Meot-Ner M. Int. J. Mass Spectrom. 2003; 227: 525.

2. Stryer L, Biochemistry, Freeman WH, New York, 1988.

3. Kaim W, Schwederski B. “Bioinorganic Chemistry: Inorganic Elements in the

Chemistry of Life,” John Wiley & Sons: Chichester, U.K., 1994.

4. Lippard SJ, Berg JM. “Principles of Bioinorganic Chemistry,” University

Science Books: Mill Valley, CA, 1994.

5. Aidley DJ, Stanfield PR, “Ion Channels: Molecules in Action,” Cambridge

University Press, Cambridge, 1996.

6. Cowan JA, “Inorganic Biochemistry: An Introduction,” 2nd ed., Wiley-VCH,

New York, 1997.

7. Hughes, M. N. “The Inorganic Chemistry of Biological Processes,” 2nd ed.;

John Wiley & Sons: Chichester, 1981.

8. Eller K, in Organometallic Ion Chemistry, Freiser BS, ed., Kluwer,

Amsterdam, 1996. pp 123.

9. Metal Ions in Biology and Medicine, Etienne JC, Khassanovc Z, Maymard I,

Collery P, Khassanova L, eds., Eurotext: Paris, 2002. Vol. 7.

10. Shigeru I. Chem. Phys., 1986; 108: 441.

11. Shoeib T, Siu KWM, Hopkinson AC. J. Phys. Chem. A, 2002; 106: 6121.


75

12. Crupi V, Jannelli MP, Magazu S, Maisano G, Majolino D, Migliardo P, Sirna

D. Mol. Phys., 1995; 84: 645.

13. Yamdagni R, Kebarle P. J. Am. Chem. Soc., 1973; 95: 3504.

14. Meot-Ner M, Hamlet P, Hunter EP, Field FH. J. Am. Chem. Soc., 1980; 102:

6393.

15. Meot-Ner M. J. Am. Chem. SOC., 1983; 105: 4906.

16. Yamabe S, Hirao K, Wasada H. J. Phys. Chem., 1992; 96: 10261.

17. Chen QF, Stone JA. J. Phys. Chem., 1995; 99: 1442.

18. Bouchoux G, Choret N, Flammang R. J. Phys. Chem. A, 1997; 101: 4271.

19. Bouchoux G, Jezequel S, Penaud-Berruyer F. Org. Mass Spectrom., 1993; 28:

421.

20. Bouchoux G, Djazi F, Gaillard F, Vierezet D. Int. J. Mass Spectrom., 2003;

227: 479.

21. Bouchoux G, Buisson DA, Bourcier S, Sablier M. Int. J. Mass Spectrom., 2003;

228: 1035.

22. Aue DH, Webb HM, Bowers MT. J. Am. Chem. Soc., 1973; 95: 2699.

23. Bowers MT. Gas Phase Ion Chemistry; Academic Press: New York, 1979; Vol.

2.

24. Shigeru I, Nomura O. J. Mol. Structure (Theochem), 1987; 152: 315.

25. Chang YP, Su TM, Li TW, Chao I. J. Phys. Chem. A, 1997; 101: 6107.


76

26. Wang Z, Chu IK, Rodriquez CF, Hopkinson AC, Siu KWM. J. Phys. Chem. A,

1999; 103: 8700.

27. Bouchoux G, Choret N, Berruyer-Penaud F. J. Phys. Chem. A, 2001; 105:

3989.

28. Poutsma JC, Andriole EJ, Sissung T, Morton TH. Chem. Commun., 2003;

2040.

29. Bouchoux G, Hoppilliard Y, Houriet R. New J. Chem., 1987; 11: 225.

30. Meot-Ner M. J. Am. Chem. Soc., 1984; 106: 278.

31. Sharma RB, Blades AT, Kebarle P. J. Am. Chem. SOC., 1984; 106: 510.

32. Wolfe SL. Biologia Molecolare e Cellulare; Edises: 1994.

33. Yoshikawa Y, Yoshikawa K. FEBS Let., 1995; 361: 277.

34. Lee SJ, Mhin BJ, Cho SJ, Lee JY, Kim KS. J. Phys. Chem., 1994; 98: 1129.

35. Uggerud E. Mass Spectrom. Rev. 1992; 11: 389.

36. Cooks RG, Kruger TL. J. Am. Chem. Soc., 1977; 99: 1279.

37. Cooks RG, Patric JS, Kotiaho T, McLuckey SA. Mass Spectrom. Rev., 1994;

13: 287.

38. Cooks RG, Wong PSH. Acc. Chem. Res., 1998; 31: 379.

39. Cooks RG, Koskinen JT, Thomas PD. J. Mass Spectrom., 1999; 34: 85.

40. Cerda BA, Wesdemiotis C. J. Am. Chem. Soc., 1995; 117: 9734.


77

41. Lee VWM, Li H, Lau TC, Guevremont R, Siu KWM. J. Am. Soc. Mass

Spectrom., 1998; 9: 760.

42. Shugang Ma, Feng W, Cooks RG. J. Mass Spectrom., 1998; 33: 943.

43. Feng WY, Gronert S, Lebrilla CB. J. Am. Chem. Soc., 1999; 121: 1365.

44. Tsang Y, Siu FM, Ma NL, Tsang CW. Rapid Commun. Mass Spectrom., 2002;

16: 229.

45. Kish MM, Ohanessianb G, Wesdemiotis C. Int. J. Mass Spec., 2003; 227: 509.

46. Mirza SP, Krishna P, Prabhakar S, Vairamani M, Giblin D, Gross ML. Int. J.

Mass Spec., 2003; 230: 175.

47. Tsang Y, Siu FM, Ho CS, Ma NL, Tsang CW. Rapid Commun. Mass Spectrom.,

2004; 18: 345.

48. Ng KM, Li WK, Wo SK, Tsang CW, Ma NL. Phys. Chem. Chem. Phys., 2004; 6:

144.

49. Robinson PJ, Holbrook KA. Unimolecular Reactions; Wiley Interscience:

London, 1972.

50. Cerda BA, Wesdemiotis C, J. Am. Chem. Soc., 1996; 118: 11884.

51. Marino T, Russo N, Toscano M. J. Phys. Chem. B, 2003; 107: 2588.

52. Hunter EPL, Lias SG. J. Phys. Chem. Ref. Data, 1998; 27: 413.

53. Chen LZ, Miller JM. Org. Mass Spectrom., 1992; 27: 883.

54. Chen LZ, Miller JM. Organomet. Chem., 1993; 448: 225.


78

55. Gaussian 98, Revision A. 11. 2, Frisch MJ, Trucks GW, Schlegel HB, Scuseria

GE, Robb MA, Cheeseman JR, Zakrzewski VG, Montgomery JA, Jr.,

Stratmann RE, Burant JC, Dapprich S, Millam JM, Daniels AD, Kudin KN,

Strain MC, Farkas O, Tomasi J, Barone V, Cossi M, Cammi R, Mennucci B,

Pomelli C, Adamo C, Clifford S, Ochterski J, Petersson GA, Ayala PY, Cui Q,

Morokuma K, Rega N, Salvador P, Dannenberg JJ, Malick DK, Rabuck AD,

Raghavachari K, Foresman JB, Cioslowski J, Ortiz JV, Baboul AG, Stefanov

BB, Liu G, Liashenko A, Piskorz P, Komaromi I, Gomperts R, Martin RL, Fox

DJ, Keith T, Al-Laham MA, Peng CY, Nanayakkara A, Challacombe M, Gill

PMW, Johnson BG, Chen W, Wong MW, Andres JL, Gonzalez C, Head-

Gordon M, Replogle ES and Pople JA, Gaussian, Inc. , Pittsburgh PA, 2001.

56. Cerda BA, Hoyau S, Ohanessian G, Wesdemiotis C. J. Am. Chem. Soc., 1998;

120: 2437.

57. Stone JA, Carter MD. Int. J. Mass Spectrom., 1998; 179/180: 1.

-

X

H

Ion-Molecule ?

Ortho, meta and paraX = -COO, -O and -SO3

XH

COOH

- ESI

Carbanion/oxide ion

SID reactions with CO2

GGEENNEERRAATTIIOONN OOFF RREEGGIIOOSSPPEECCIIFFIICC CCAARRBBAANNIIOONNSS UUNNDDEERR

EELLEECCTTRROOSSPPRRAAYY IIOONNIIZZAATTIIOONN CCOONNDDIITTIIOONNSS AANNDD

CCHHAARRAACCTTEERRIIZZAATTIIOONN BBYY IIOONN--MMOOLLEECCUULLEE RREEAACCTTIIOONNSS WWIITTHH

CCAARRBBOONN DDIIOOXXIIDDEE

Chapter 3.1 Regiospecific Carbanions…

79

CCChhhaaapppttteeerrr IIIIIIIII

PPPaaarrrttt 111

1. PROLOGUE

he study on generation and reactions of organic anions in the gas phase1-

12 is a subject of interest, since it provides information about the intrinsic

reactivity, and other fundamental properties of ions in the absence of solvent

effects. Among the anions, carbanions execute a broad and substantial role as

reactive intermediates in organic reaction chemistry. The formation of a

carbon-carbon bond by the addition of a carbanion to a carbonyl group is of

considerable magnitude as a synthetic method. The reaction mechanism of

such addition reactions viz., Claisen, Aldol, Dieckmann condensations,

Grignard reactions etc. have been extensively investigated.2,4,13 An extensive

and separate literature on the carbanionic species has been developed. The

survival of gas phase anions is well established and a good deal seems to be

known about their structure.7,8,10 While the manifold achievements in gas-phase

carbanion chemistry have been encouraging, but there is a substantial lack of

information on the property of probably the greatest interest, namely the

T

GGEENNEERRAATTIIOONN OOFF RREEGGIIOOSSPPEECCIIFFIICC CCAARRBBAANNIIOONNSS FFRROOMM AARROOMMAATTIICC

HHYYDDRROOXXYY AACCIIDDSS AANNDD DDIICCAARRBBOOXXYYLLIICC AACCIIDDSS


80

reactivity. Reactivity data have, for the most part, been obtained in a rather

indirect manner.

Studying physicochemical properties and reactions of carbanions in the

gas phase is of great significance for organic chemists to understand better the

intrinsic reaction mechanism in the absence of solvent effects.2,4,10 Qualitative

studies on the formation and reactivity of gas phase carbanions have provided

useful chemical strategies for elucidating carbanion structures, reaction

mechanisms and their stereochemistry.3,10 As Squires pointed out in a review,8

carbanions in solution are generally tightly ion-paired and often form

aggregates; therefore, the gas phase provides a unique opportunity to study

their intrinsic reactivity. Gas phase carbanions have been known since 1930

when they were first observed by electron ionization (EI) of methane and

larger hydrocarbons.1 A major challenge in this work has been the selective

formation of the carbanion and the verification of its structure. A range of gas

phase approaches have been developed to generate and characterize

carbanions.

1.1. THE GENERATION OF CARBANIONS IN THE GAS PHASE

To produce carbanions, one may have to localize the negative charge on

the carbon atom in a molecule. Various mass spectrometric techniques have

been used to generate the gas phase carbanions and most of these

experiments were carried out using flowing afterglow (FA) and Fourier

transform ion-cyclotron resonance (FTICR) spectrometers. Mostly, the gas


81

phase carbanions have been generated by three different pathways, they are

(i) gas-phase proton abstraction,3,10,14-24 (ii) fluoride-anion-induced

desilylation,10,25-40 (iii) collision-induced dissociation (CID) of carboxylic

acids.8,10,41-55

1.1.1. Proton abstraction Method

The first method consists of proton abstraction from R–H by use of a

strong base B (Scheme 1): 3,10,14-24

R-H +B- R- + B-H

(B- = OH-, NH2-, etc).

Scheme 1

FTICR has been shown to be the suitable instrument for studies on such

proton transfer reactions and H/D exchange processes in the gas phase.20,21,23

The deprotonation step towards formation of carbanions suffers from two

crucial limitations. First, the precursor must be sufficiently acidic so that it can

be deprotonated by an easily formed, gas phase base. The NH2 is one of the

strong and commonly used bases for this method; therefore, deprotonation is

limited to molecules with proton affinities (PA) less than 404 kcal/mol. This PA

value of base is sufficient to generate the carbanions stabilized by resonance or

electron-withdrawing groups, but not sufficient for localized hydrocarbons.

Second, problem comes when more than one type of acidic proton is present in

the precursor molecule, in such cases deprotonation is not selective and results

in a mixture of isomeric anions. Consequently, it is ideal to select the precursor


82

with the site of interest significantly more acidic than others; however, in

practice it is difficult to have such systems. To avoid these problems, two other

approaches have been widely used.

1.1.2. Fluorodesilylation Method

Taking advantage of analogies to condensed phase chemistry, DePuy

and co-workers developed fluorodesilylation reactions for the formation of

carbanions and hence it has become popular as the DePuy reaction.25-28 In this

method, the trimethylsilyl group of an appropriate trimethylsilyl derivative is

removed by a fluoride anion. This reaction proceeds via an addition-

elimination mechanism in which a penta-coordinate siliconate is produced

(Scheme 2).

RSi(CH3)3 + F - R- + (CH3)3SiFSi

CH3

FR

CH3H3C

Scheme 2

The displacement reaction of fluoride ion with trimethylsilyl derivatives

allows the specific production of a single isomeric species in the absence of

other anions or acidic neutrals; this method, therefore, will greatly facilitate the

detailed correlation of structure and reactivity of gas phase organic anions.

Further, this methodology is extremely powerful and has been

successfully applied in a number of instances. Although this desilylation

method is used for selective generation of carbanions, its routine use is


83

inadequate, because it does have several significant drawbacks: (i) molecular

fluorine (F2) must be used, which is highly toxic and corrosive, and can be

harsh on vacuum operating systems, (ii) the reaction leads to many side

products, in part, because the product ions are unstable with respect to F2, (iii)

F- is not reactive enough to generate basic anions (PA > ∼400 kcal/mol) at room

temperature and (iv) the requisite trimethylsilyl precursors must be

synthesized.

1.1.3. Collision induced Decarboxylation method

The elimination of small, stable, and neutral molecules is classically

observed in the gas-phase fragmentation of several cationic/anionic species.

The collision induced decarboxylation method employs decarboxylation of the

gas-phase carboxylate anions, which easily eliminate CO2 upon collisions with

neutral gas molecules yielding the corresponding carbanions (Scheme 3)::

R-COOH R¯+ CO2- ESI CID

R

O

O-

Scheme 3

This method has been used extensively, by Graul and Squires in a

flowing afterglow-triple quadrupole mass spectrometer.44,45,48,49 Like the

desilylation method, it enables selective generation of a carbanion with the

required structure. In some cases, however, the carboxylate anions abstract a

proton intramolecularly, e.g. from the α-carbon in aliphatic acids, inhibiting

decarboxylation.44 To overcome this problem for the generation of gas-phase


84

carboxylate anions, Graul and Squires used mainly fluorodesilylation of the

trimethylsilyl esters of appropriate acids44 (Scheme 4).

R-COO-SiMe3 + F - R-COO- + Me3SiF

Scheme 4

All of the above discussed carbanions were generated in the solvent free

environment, i.e. FTICR, FA etc. The carbanions are generally more basic in

nature; however, the problem comes when these carbanions are generated by

using the above methods in the solvent environment, for e.g. in the source of

electrospray ionization technique. More recently, Bienkowski and

Danikiewicz53 adopted the decarboxylation method using source induced

dissociation to produce phenide anions from the substituted benzoic acids

under the negative ESI conditions.

1.2. CHARACTERIZATION OF CARBANIONS

The carbanions are known to undergo nucleophilic addition reactions in

the solution phase (Scheme 5).

R- + O=C=O R-C-O-O

Scheme 5

The same reactivity of carbanions is applied to characterize the

carbanions in the gas-phase. The generated gas-phase carbanions in the mass

spectrometer are allowed to react with a suitable reagent molecule (ion-


85

molecule reactions). Since early work with mass spectrometers, it has been

recognized that reactions occur readily in these instruments. For example, the

development of chemical ionization methods centers on ion/molecule reactions

taking place in the source of a mass spectrometer. To study these reactions in

more detail under highly controlled conditions, a number of instruments and

experimental methods have been developed. These include instruments such

as Fourier transform mass spectrometers (FTMS) and quadrupole ion trap mass

spectrometers (QITMS), as well as triple quadrupole instruments.

Figure 1: Ion-molecule reactions of Phenide anions (Ph-) with carbondioxide in

the collision cell.

The product ions formed in the ion-molecule reactions between

carbanions and reagent neutral molecules are then mass analyzed to

characterize the carbanions. Almost all of the authors used ion-molecule

reactions for the characterization of carbanions. The determination of

-

CO2

m/z 77


86

carbanions by ion-molecule reactions relies on their reactivity towards the

used molecule in the experiment. Indeed, the selective reactivity of carbanions

with neutral triatomic neutral molecules, i.e. CO2, CS2, N2O, and COS, in the

gas-phase has been reported by several investigators.3,10,15,25,28,47,51,53,56-63

1.3. STABILITY STUDIES OF CARBANIONS

The characterization of gas-phase carbanions using mass spectrometry

solely depends on their stability as well as reactivity with reagent molecule in

ion-molecule reactions. Therefore, it is necessary to have an idea about the

structure and stability of the resulting carbanions. The effect of adjacent

substituents has been a major subject on the stability, structure, and reactivity

of carbanions.13,64,65

Stabilization of carbanions by α-substitutents has been a continual topic

in physical organic chemistry in the last decades.66 Nibbering et al.66 studied

the stabilization of α-substituted carbanions in the gas phase with substituents

having π-electrons (-Ph, -CN, -NO2, -COOMe). Due to the strong electron

withdrawing effect of these groups, the negative charge on carbon is

delocalized and the stability of these carbanions is enhanced. The stabilization

energy data have been analyzed in terms of polarizability, field/inductive, and

resonance effects. Further, it follows that inductive stabilization is more

effective in the substituted methyl, X-CH2¯ (X = Ph, CN, NO2, COOMe) than in

the substituted cycloalkyl (c), X-c-CnH2n-2¯ (n = 4, 5, 6 and 7) carbanions. This

result suggests that inductive stabilization is counteracted by the electron


87

releasing effect of alkyl groups. Resonance stabilization is significantly more

effective in the substituted cycloalkyl, X-c-CnH2n-2¯ (n = 4,5,6,7), than in the

substituted methyl, X-CH2¯ carbanions, which suggests that in contrast to

inductive stabilization, resonance stabilization is assisted by the electron

releasing effect of alkyl groups. In cycloalkyl carbanions, it appears that

substitutent stabilization in the geometrically restricted substituted cyclopropyl

carbanions, X-C3H4¯ is dramatically less effective than in the corresponding

geometrically unrestricted larger substituted cycloalkyl carbanions, X-c-C2H2n-

2¯ (n =4,5,6,7). The linear regression analyses of the substituted cycloalkyl

carbanions indicate that reduction of the stabilization energy is caused not

exclusively by geometrically hindered resonance stabilization, but also to a

smaller extent by a less efficient inductive stabilization in the substituted

cyclopropyl carbanions.

Very recently, Cooks et al. studied gas-phase reactions of three typical

carbanions -CH2NO2, -CH2CN, and -CH2S(O)CH3 with chloromethanes CH2Cl2,

CHCl3, and CCl4. These anions examined by tandem mass spectrometry,

showed a novel hydrogen/chlorine exchange reaction. Due to the strong

electron withdrawing effect of the nitro, cyano and sulfinyl groups (i.e.

delocalized carbanions) these carbanions are found to be stable enough to

react with chloromethanes (Scheme 6).

-CH2NO2 + CCl4 [CH2ClNO2.CCl3-] -CCl3 + CH2ClNO2~20%

~80%

-CHClNO2 + CHCl3

Scheme 6


88

All these studies reveal that carbanions get stabilized in the gas phase

due to substituents at α-positions having strong elelctron withdrawing effect,

i.e. inductive or mesomeric effect and destabilized by acidic groups, i.e. like

OH, NH2 by proton transfer. For the compounds with substituents containing

acidic hydrogen atoms, like OH and NH2 groups, Bienkowski and Danikiewicz33

have failed to generate the phenide anions due to either an intra- or an

intermolecular (mediated by the molecules of methanol or water) proton

transfer from the more acidic position to the benzene ring.

1.4. GENERATION AND CHARACTERIZATION OF SPECIFIC CARBANIONS

The ability to form a carbanion center in a molecule with position control

(regiospecificity) and spatial control (steriospecificity) is a major objective in

synthetic organic chemistry. In the gas phase analogous developments would

open up new avenues of exploration and would enable important condensed

phase problems to be addressed in a solvent free environment. Absolute

configuration assignment of isomers such as cis-/trans- pairs or ortho-/meta-

/para- sets based solely on mass spectrometric (MS) data is a very challenging,

most often unfeasible task. Ideally, for unequivocal MS structural elucidation,

the ionized molecule of each isomer should dissociate to form a unique,

structure diagnostic fragment ion. Otherwise, either the neutral or the ionized

molecule of each isomer should react to form a unique, structure diagnostic

product ion. These are, however, indirect structural assignments, since we

must know in advance the configuration of each positional isomer (as


89

determined by other techniques), and the whole set of isomers must be

analyzed and their spectra compared under the same MS conditions to ensure

the reliability and to establish a comparative MS method for the distinction of a

specific set of isomers.

Production of isomeric carbanions by proton abstraction method is often

ambiguous and unsatisfactory, because there may be several abstractable

protons in the target molecule, so that mixtures of anions may be formed. When

the relative acidity of the two types of allylic protons is similar, for example

from 2-pentene, the proton abstraction method form both the expected anions.

In some cases, where the proton acidities are expected to differ substantially,

for example in propyne, formation of both isomeric anions were found by

proton abstraction method in a flowing afterglow system, however,25 major ion

was the less stable anion as shown below (Scheme 7):

OH - + CH3C CD CH3C C - + HOD

OH - + CH3C CD -CH2C CD + H2O

<40%

>60%

Scheme 7

The more serious complication in proton abstraction method is the

probable base promoted rearrangement of certain molecules in the ion-

molecule complex.7 Double-bond migration and cis-trans isomerization may

occur during such rearrangements, hence it may lead to anions of unknown or

unexpected structure.


90

By the introduction of fluorodesilylation method, DePuy et al. reported

efficient generation of isomeric carbanions unambiguously by the reaction of

their trimethylsilyl derivatives with fluoride ion. For example, the two possible

C3H3- isomeric anions (for which proton abstraction method failed to generate

selectively) were separately generated by applying the fluorodesilyation

method by choosing the appropriate trimethylsilyl derivatives (Scheme 8):

F - + CH3C CSi(CH3)3 CH3C C- + FSi(CH3)3

F - + HC CCH2Si(CH3)3 HC CCH2- + FSi(CH3)3

Scheme 8

The two isomeric anionic products, viz., methyl acetylide and propargyl

(Scheme 8), show substantial difference in their reactivity, confirming their

isomeric structures. For example, the propargyl anion reacts with molecular

oxygen seven times more rapidly than does the methyl acetylide anion8 and

large differences in rates or products are observed with D2O and other

reagents. They extended this method to generate anions from the trimethylsilyl

derivatives of propylene, cis- and trans-2-butene and toluene.

Kass et al. exploited the fluorodesilylation method for the selective

formation of many carbanions in a flowing afterglow device.31,32,36,38,39,62 They

performed experiments at variable temperatures to study the stability of

carbanions. The results indicated that more basic and less accessible anions

could be prepared at elevated temperatures. A typical example was the

production of isomeric (E- and Z-) vinyl anions stereo selectively by treatment


91

of the appropriate trimethylsilyl precursor with fluoride in a flowing afterglow

tube at elevated temperatures, 200o C (Scheme 9).

C C

Si(CH3)3

HH3C

HF -

C C -

HH3C

H

+ (CH3)3SiF

C C

Si(CH3)3

H

H3C

HF -

C CH

H3C

H

+ (CH3)3SiF-

E- E-

Z- Z-

Scheme 9

They found that the generated isomeric 1-vinyl anions from propene give

different product distributions in their reactions with N2O. Each ion produced

four significant products with N2O and likely mechanisms for the formation of

the characteristic products are outlined in Scheme 10. The formation of this set

of diverse products gave a hint at the wide range of pathways that are available

with this simple reagent.

C CH-

H3C

HN2O

HO- + CH3C C- + -CHN2 + CH2=CHC-=N2

trans 17% 42% 9% 31% cis 21% 11% 5% 62%

Scheme 10: Vinyl anions in the presence of N2O gas.

In the cis- anion, the major product was a diazo anion, a type of product

that is commonly formed in the reactions of carbanions with N2O. Initial attack

of the carbanion on the terminal nitrogen of N2O gives an oxygen-centered


92

anion that can undergo an intramolecular proton transfer with the carbon that

had been the original carbanion. The resulting vinyl anion collapses with the

expulsion of hydroxide ion to give a diazo intermediate which can be

deprotonated by the hydroxide ion in the complex. Alternatively, the

hydroxide ion may escape the complex and be the observed ionic product.

In the trans- ion, the same pathway was found, but it is also possible for

the oxide anion of the initial addition product to attack the vinylic proton on

carbon-2. Once C-2 is deprotonated, the ion can breakdown with the expulsion

of N2 and HO-, leaving propyne behind. Deprotonation of propyne by HO¯

leads to the observed acetylide product. This pathway would be unlikely in the

cis- ion (the oxide anion would be trans- to the vinylic proton on carbon-2), and

therefore the acetylide product must be formed by another, less efficient route

for this stereoisomer.

In another study, the same group generated steriospecific cis- and trans-

β-formylcyclopropyl carbanions from cis- and trans-2-(trimethylsilyl)-

cyclopropanecarboxaldehyde, respectively, in a flowing-afterglow device by

using fluorodesilylation method (Scheme 11).

CHO + F -90%

10%

Trace

TMS CHO-

CHOTMS

CHOFMe2Si -

-

+ TMSF

+HF

CHO + F -90%

5%

5%

TMS

CHO-

CHO

TMS

CHO

FMe2Si -

-

+ TMSF

+HF

Scheme 11


93

They distinguished these isomeric anions by using the ion-molecule

reaction with a number of chemical reagents (N2O, CO2, COS, CS2and O2). The

products from the ion-molecule reaction of isomeric anions with CS2 are

summarized in Scheme 12. The authors concluded that the isomeric ions did

not interconvert or underwent ring-opening isomerization at 25 oC. Theoretical

calculations, with Hartree-Fock (HF) level using the 6-31+G(d), also suggested

that these isomeric anions do not undergo interconversion at room

temperature, hence the authors were able to differentiate these two species.

These results clearly demonstrated that stereochemistry can be successfully

dealt within the gas phase and that it will be possible to obtain mechanistic

information.

CHO- + CS2 HS-

HCS2-

CHO-S

CHO-S2C

m/z 33

m/z 77

m/z 101

m/z 145

CHO

-+ CS2 HS-

HCS2-

CHO

CHO-S2C

m/z 33

m/z 77

m/z 101

m/z 145

-S

42%

19%

37%

2%

8%

47%

44%

1%

Scheme 12

Squires et al.43 introduced the regiospecific generation of gas-phase

carbanions by means of collision-induced decarboxylation of organic

carboxylate anions in an FT-ICR. They had adapted this method to use in

flowing afterglow-triple quadrupole instruments and explored the scope of the

method with respect to the kinds of carbanions that can be produced. They


94

described an experimental evaluation of the absolute stabilities of alkyl

carbanions, as well as the observation of several new species synthesized by

collision induced dissociation reaction. The procedure involves the generation

of the carboxylate anion in the flow tube by means of fluorodesilylation of the

appropriate trimethylsilyl ester. The carboxylate ion is extracted from the flow

tube and mass-selected with the first quadrupole. Decarboxylation is affected

by means of collision-induced dissociation (CID) in the second quadrupole and

the resulting fragment ions are mass-analyzed with the third quadrupole. By

monitoring the intensity of carbanion (R-) as a function of the carboxylate ion

axial kinetic energy they obtained experimental threshold energy for the same

reaction.

Isomeric allyl (CH2CHCH2¯), 2-propenyl (CH3C¯=CH2), 1-propenyl

(CH3CH=CH¯), and cyclopropyl (CH2)2CH¯) anions, as well as the parent

unsubstituted vinyl anion (C2H5¯) were generated in the gas phase by CID of

the corresponding carboxylate anions by using a FTMS. Ion-molecule reactions

of each C3H5- ion with selected neutral reagents were described. The reactions

were distinct and indicated non-interconverting isomeric ion structures. Each

of the vinylic anions and the cyclopropyl anion abstracted a proton from

ammonia and D2O, while the more stable allylic isomer, CH2CHCH2¯, was

unreactive with the former and underwent four hydrogen-deuterium exchanges

with the latter. Sulfur abstraction from CS2 was observed with each of the vinyl

anions and the cyclopropyl anion, while the allylic ion reacted by addition

followed by loss of H2S. Four unique sets of products were produced by each of


95

the C3H5¯ isomers in the presence of N2O which were consistent with the

expected structures. Ab initio SCF-MO calculations for each C3H5¯ ion were

also described, which provided additional insight into their structures and

relative energies.

Bienkowski and Danikiewicz53 reported generation of carbanions in the

ESI source by source fragmentation of corresponding carboxylate anions. They

attempted to generate the regiospecific carbanions from isomeric

hydroxyl/amino benzoic acids in the ESI source. However, based on their

experimental results in ion-molecule reactions with CO2 they concluded that

the isomeric carbanions could not be generated from isomeric hydroxy/amino

benzoic acids by this method. The authors claimed that failure to generated the

isomeric carbanions was due to the presence of more acidic hydrogen like -OH

and -NH2 groups in the precursors, which induce intra- or intermolecular

(mediated by the molecules of methanol or water) proton transfer from the

more acidic position to the benzene ring.


As described in the introduction, the carbanions could be generated by

proton abstraction, fluorodesilylation, and collision induced decarboxylation

and the later two methods produced site selective carbanions. The generated

carbanions were unambiguously characterized by using selective ion-molecule

reactions with neutral reagent gases such as CO2, COS, CS2, N2O etc.

Bienkowski and Danikiewicz53 showed the successful generation of carbanions


96

in the ESI process by source-induced decarboxylation method and

characterized the carbanions by ion-molecule reactions with CO2. Though

fluorodesilylation and collision induced decarboxylation methods were shown

to be successful in the regiospecific generation of carbanions, there were only

limited number of reports covering the regiospecific formation of carbanions

and study of their structure and reactivity through ion-molecule reactions in the

collision cell. We therefore selected a series of geometrical and positional

isomers of dicarboxylic acids and hydroxy acids with a view to study the

generation of isomeric carbanions in the presence of additional acidic

functional groups and their stability and propensity for rearrangement. The

present study includes the generation of decarboxylated product ions from

regioisomeric carboxylic acids in the ESI process and to study of their ion-

molecule reactions with CO2 to confirm their structures. Quantum chemical

calculations are also performed to estimate the stabilities of the some of the

isomeric carbanions and the isomerized products.


The molecular structures of the isomeric substituted carboxylic acids (1-

17) under study are given in Scheme 13. These include one set of geometrical

isomers (1 and 2) and five sets of positional isomers (3-17), viz. two sets of

dicarboxylic acids (phthalic acid isomers, 3-5; and pyridinedicarboxylic acid

isomers, 6-9) and three sets of hydroxy acids (hydroxy benzoic acid isomers,

10-12; hydroxy nicotinic acid isomers, 13-14; and hydroxy phenylacetic acid

isomers, 15-17). The negative ion ESI mass spectra of all the isomeric


97

compounds (1-17) showed abundant [M-H]¯ ion at lower cone voltage (10 V). At

higher cone voltage (30V), all the compounds resulted in the decarboxylated

product ion, [(M-H)-CO2]¯ (1-17) as the major ion. In addition to [(M-H)-CO2]¯

ion, the dicarboxylic acids (3-9) showed [(M-H)-2CO2]¯ ions formed due to loss

of another CO2 molecule from [(M-H)-CO2]¯, but the compounds 1 and 2 failed

to result in [(M-H)-2CO2]¯ ions even at relatively higher cone voltages (>30 V).

However, the CID spectra of [M-H]¯ of all the dicarboxylic acids (1-9) show

both [(M-H)-CO2]¯ and [(M-H)-2CO2]¯ as the product ions. This clearly suggests

that the [(M-H)-2CO2]¯ ions from 1 and 2 are not stable enough in the source

environment. The CID spectra of [M-H]¯ from hydroxybenzoic acids (10-17), as

expected, exclusively resulted in the formation of [(M-H)-CO2]¯ ion due to the

presence of one carboxylic group in them.

The [(M-H)-CO2]¯ ions are formed easily by the decarboxylation of

carboxylate anions under ESI conditions, and we primarily focused the present

study on the structure of [(M-H)-CO2]¯ ions of all the studied compounds (1–17).

Since all these compounds possess a functional group at different positions,

which can transfer a proton to the carbanion, the initially generated carbanions

may isomerize to the corresponding aromatic carboxylate/oxide anions if such

proton transfer occurs. In the present investigation, we apply selective reaction

of CO2 with carbanions in ion-molecule reactions to address whether isomeric

carbanions are generated or not.


98

OH

O

O

OH

O

OH

HO

O

1 2

COOH

COOH

COOH COOH

COOH

COOH

NHOOC COOHN COOHN COOHN COOH

COOH

COOH

HOOC

NHOOC OHN OH

COOH

COOH

OH

COOH COOH

OH

OH

CH2-COOH

OH

CH2-COOH CH2-COOH

OHOH

3 4 5

6 7 8 9

13 1410 11 12

15 16 17

Geometrical isomers

Positional isomers

i) Dicarboxylic acids

ii) Hydroxy acids


99

Scheme 13

3.1. GEOMETRICAL ISOMERS (1-2)

The [(M-H)-CO2]¯ ions generated from 1 and 2 possess an acidic

hydrogen on the remaining carboxylic acid group that is prone to be

transferred easily. If the proton transfer occurs from -COOH to carbanion site

by an intramolecular reaction, the resulted [(M-H)-CO2]¯ ion would be a

carboxylate ion. However, The initial location of the anionic center in the [(M-

H)-CO2]¯ anions will be the same as the location of the carboxylate group in its

precursor.53 If this is the case, the decarboxylation products from 1 and 2 might

lead to isomeric structures differing in the orientation of the carbanion site, viz.

the structure of decarboxylation product ions from 1 and 2 may be assumed as

cis- (1C), and trans- (2C), (Scheme 14) respectively, with respect to the

remaining -COOH group.

OH

O

O

OH

O-

O

O

OH

O

O

HH-

O

O-

H H

O

OH

H-

O

OH

-O

O

O

OH

HO

O

1

2

m/z 115 1C, m/z 71

2C, m/z 71m/z 115

I, m/z 71

H

H


100

Scheme 14: Source fragmentation phenomenon for fumaric and maleic acids

(1 and 2) under Negative ESI conditions.

Vinyl anions are known to be highly reactive and they easily capture

proton from other sources such as solvent (methanol) or undergo

intramolecular proton transfer from –COOH to vinylic carbanion to produce

vinyl carboxylate anion in the present case, because the basicity of the vinyl

carbanion (proton affinity, PA ≈ 400 kcal/mol)67 is more than the vinyl

carboxylate anion (PA ≈ 346 kcal/mol) 67 and methoxide anion (PA ≈ 382 kcal/

mol).67 Hence, the survival of vinylic carbanion (1C and 2C) generated in

compounds 1 and 2 may be difficult and may undergo proton transfer from

either carboxylic acid or from solvent to the carbanion to form an acrylate

anion. However, it was shown by Bienkowski and Danikiewicz53 that the

carbanions formed upon decarboxylation of [M-H]¯ ions from carboxylic acids

under ESI conditions in the solvent free region retain their identity. To address

whether the intramolecular proton transfer from –COOH group affects the

survival of 1C and 2C, we performed ion molecule reactions on [(M-H)-CO2]¯

ions generated from 1 and 2. In the ion-molecule experiment, the [(M-H)-CO2]¯

ions are selected in MS1 and allowed to react with CO2 in the collision cell. The

experiments showed that the [(M-H)-CO2]¯ ions from 2 reacted with CO2 and

resulted in the corresponding carboxylate anions, and that from 1 did not react

(Figure 2).


101

Figure 2: The ion-molecule reaction mass spectra of [M-H-CO2]¯ ions from 1

and 2 (a and b) with CO2 in the collision cell.

This selective reaction reveals that the [(M-H)-CO2]¯ ion from 2 is

surviving as carbanion (2C), whereas the [(M-H)-CO2]¯ ion of 1 must be

undergoing rapid isomerization of the initially generated carbanion 1C to the

corresponding carboxylate anion (I) through intramolecular proton transfer. In

order to prove the selective reactivity of CO2, the acrylate anion (I) was

generated from acrylic acid by deprotonation and performed similar ion-

O

O

HH-

m/z 71

CO2

O

O

H-

m/z 71

CO2H


102

molecule experiments with CO2. The [M-H]¯ ion from acrylic acid failed to react

with CO2 confirming the selectivity of CO2 in reacting with carbanions. The

survival of carbanion 2C from 2 suggests that the proton transfer is kinetically

unfavorable in 2C by the fact that the carbanion and the -COOH group are in

trans- position. These results clearly reflect the geometrical isomeric

differences between the decarboxylated product ions of 1 and 2 and their

difference in their internal stability.

3.2. POSITIONAL ISOMERS

3.2.1. Aromatic dicarboxylic acids (3-9)

We also generated [(M-H)-CO2]¯ ions from all the aromatic dicarboxylic

acids (3-9). When ion-molecule reactions were performed on the [(M-H)-CO2]¯

ions of 3-9, the [(M-H)-CO2]¯ ions from meta- (4, 7 and 9) and para- isomers (5

and 8) reacted with CO2, whereas that of ortho- isomers (3 and 6) did not

(Figures 3 and 4). The expected structures of initially generated carbanions

from 3-9 can be given as 3C-9C (Scheme 15). The reactivity of [(M-H)-CO2]¯

ions from meta- and para- with CO2 substantiates the survival of initially

generated carbanions (4C-5C, 7C-9C). Whereas, the carbanions from ortho-

isomers (3C and 6C from 3 and 6, respectively) must be isomerizing rapidly

through a proton transfer, as found in the case of ion 1C from 1, to result

carboxylate anion (II and III, Scheme 15), which is unreactive with CO2. The

lack of reactivity of the aromatic carboxylate anion is further confirmed by

performing ion-molecule reactions of [M-H]¯ ion generated from benzoic acid


103

with CO2, which failed to react with CO2. The above results suggest that the

proton migration is more favorable when the carbanion is located at ortho-

position with respect to the carboxylic group but not at meta- or para- position.

M [M-H]- [M-H-CO2]--Ve ESI -CO2

Y

HOOC

XHX = COO, Y = CH, 5X = COO, Y = N, 8X = O, Y = CH, 12

5C/8C/12CY XH

-

XH

-OOC

Y

Y

HOOC

X-

II/III/IV

Y XHHOOC

X = COO, Y = N, 9X = O, Y = N, 14

9C/14CY XH-OOC Y XH

-

III/VY X-HOOC

COOH

XHYX = COO, Y = CH, 4X = COO, Y = N, 7X = O, Y = CH, 11

COOH

X-YII/III/IV

X = COO, Y = CH, 3X = COO, Y = N, 6X = O, Y = CH, 10X = O, Y = N, 13

Y

COOH

XH

Y

COO-

XH

3C/6C/10C/13C

YH

X

-

X = COO, Y = CH, IIX = COO, Y = N, IIIX = O, Y = CH, IVX = O, Y = N, V

Y

COOH

X-Y X-

COO-

XHY4C/7C/11C

YH

X

-

Scheme 15: Negative ESI source fragmentation phenomenon of compounds 3-

14.


104

The ion-molecule reactions of CO2 with the [(M-H)-CO2]¯ ions from 1-9

clearly illustrate that the carbanions generated after decarboxylation of [M-H]¯

ion of a cis- (1) or an ortho- isomer (3 and 6) immediately isomerize to the

corresponding carboxylate ion through a proton transfer from the carboxylic

acid group to the carbanion site.

FIGURE 3:. THE ION-MOLECULE REACTION MASS SPECTRA

OF [M-H-CO2]¯ IONS FROM PHTHALIC ACID

ISOMERS (3-5, A-C) WITH CO2 IN THE COLLISION

CELL.

C-O

OH

CO2

m/z 121

-

O

O

H CO2

m/z 121

OOH

-

CO2

m/z 121


105

FIGURE 4: THE ION-MOLECULE REACTION MASS SPECTRA

OF [M-H-CO2]¯ IONS FROM PYRIDINE

DICARBOXYLIC ACID ISOMERS (6-9, A-D) WITH CO2

IN THE COLLISION CELL.

With a view to understand the possibility of such proton transfer from

other functional group, like -OH, we selected isomeric hydroxy acids (10-17) to

generate [(M-H)-CO2]¯ ion and studied the isomeric structures by their ion-

molecule reactions with CO2.

N

C

C O 2

- N

-O

O

H O

O

H

m /z 122

N

CO 2

- N

-

O

O

H

O OH

m /z 122

N

CO 2

- N

-

O

O

HO

OH

m /z 122

N

CO 2

- N- O

O

H

O

OH

m /z 122


106

3.2.2. Aromatic hydroxy acids (10-17)

m/z 151

m/z 151

m/z 151

m/z 151

m/z 151

m/z 151

M [M-H] - [M-H-CO 2]--Ve ESI -CO2

15

CH2COO-

OH

O-

CH2COOH

16C, m/z 107

17C, m/z 107

CH2COOH

OH

16

17

CH2COOH

OH

OH

CH2COOH

CH2COOH

O-

CH3

O-

CH2-

OH

O-

CH3

15C, m/z 107

CH2COO-

OH

CH2-

OH

CH2COOH

O-

CH3

O-

OH

CH2COO-

OH

CH2-

VIa, m/z 107

VIb, m/z 107

VIc, m/z 107

Scheme 16

All the isomeric hydroxy acids (10-17) formed abundant [(M-H)-CO2]¯

ions under negative ion ESI conditions, and ion-molecule experiments were

performed with CO2 on them. The behavior of this isomeric set of [(M-H)-CO2]¯

ions with CO2 is similar to that observed in dicarboxylic acids. The [(M-H)-


107

CO2]¯ ions generated from ortho- isomers (10, 13 and 15) did not react with

CO2, but that from meta- (11, 14 and 16) and para- (12 and 17) isomers reacted

with CO2 and consequently formed the corresponding carboxylate anions

(Figures 5, 6 and 7). Since hydroxy group is also an acidic site in the hydroxy

benzoic acids, the initial deprotonation of hydroxy acids (10-17) results in two

isomeric [M-H]¯ ions in each case as shown in Schemes 15 and 16, wherein

deprotonation occurs at the -COOH group and the other at -OH group.

Consequently, the decarboxylation of these [M-H]¯ ions lead to two different

structures (Schemes 15 and 16), i.e., carbanions (10C-17C from 10-17) and

oxide anions (IV, V, VIa-c), respectively. The unreactivity of [(M-H)-CO2]¯ ion

from compound 10, 13 and 15 could be due to isomerization of their initially

generated carbanions to the corresponding oxide anion (IV, V and VI) through

a proton transfer as discussed earlier (Schemes 15 and 16). It is known that

aromatic oxide anions do not react with CO2.53 The [(M-H)-CO2]¯ ions of meta-

and para- isomers (11, 12, 14, 16 and 17) are surviving as carbanions (11C,

12C, 14C, 16C and 17C) and hence they reacted with CO2 during ion-molecule

reactions. It is important to note that Bienkowski and Danikiewicz53 generated

[(M-H)-CO2]¯ ions of hydroxy benzoic acids (10-12) under ESI conditions, and

studied their ion-molecule reactions with CO2. They could not see any peaks

due to addition of CO2 molecules, and based on these results they concluded

that the carbanions generated from these isomers are unstable under

experimental conditions they used. Hence, existence of [(M-H)-CO2]¯ ions of 11

and 12 as carbanions under present experimental conditions suggests that


108

experimental conditions under which the carbanions are generated are crucial

for studying the stability and reactivity of carbanions under ESI conditions. The

main dissimilarities of the present experimental parameters to that of

Bienkowski and Danikiewicz53 is use of high desolvation temperature and

different collision gas pressure/collision energies in the present study.

FIGURE 5: THE ION-MOLECULE REACTION MASS SPECTRA

OF [M-H-CO2]¯ IONS FROM HYDROXYBENZOIC

ACID ISOMERS (10-12, A-C) WITH CO2 IN THE

COLLISION CELL.

-

CO2

O

H

m/z 93

-

CO2

OH

m/z 93

-

CO2

O H

m/z 93


109

Figure 6: The ion-molecule reaction mass spectra of [M-H-CO2]¯ ions from

hydroxyl pyridine carboxylic acid isomers (13-14, a-b) with CO2

in the collision cell.

Figure 7: The ion-molecule reaction mass spectra of [M-H-CO2]¯ ions from

hydroxyl phenyl acetic acid isomers 15-17 (a-c) with CO2 in the

collision cell.

N

-CO2

OH

m/z 94

N-

CO2

O

H

m/z 94

CH2-

O

H

CO2

m/z 107

CH2-

OH

CO2

m/z 107

CH2-

OH

CO2

m/z 107


110

Among the studied meta- and para- isomers, it can be noted that the ion-

molecule reaction spectra obtained for [M-H-CO2]¯ of meta- and para- isomers

of dicarboxylic acid isomers (4C and 5C, 7C-9C) are similar, which show no

difference in their reactivity with CO2. Such is not the case with the hydroxy

carboxylic acid isomers. The relative abundance of CO2 adduct ion with [M-H-

CO2]¯ ions from meta- isomers of hydroxy acids (11C and 12C, 14C, 16C and

17C) is 10 times higher than that of corresponding para- isomer that could be

due to formation of predominant phenoxide anion over carboxylate anion

during initial deprotonation. In fact, it was reported in the case of para-

hydroxybenzoic acid that the deprotonation takes place in the phenolic group

rather than carboxylic group (Scheme 17).68,69 The same phenomenon explains

the difference in the reactivity of [(M-H)-CO2]¯ from 5 and 12 having –COOH

and –OH groups.

O-

HO O

O

HO O-

Scheme 17

3.2.3. EFFECT OF DESOLVATION TEMPERATURE

It is a known fact that an increase in the desolvation and source

temperatures suppress ion-molecule reactions of generated ions with solvent

molecules in the ESI source.70,71 Hence, we performed ion-molecule reactions


111

at increased desolvation/source temperatures. For this purpose, we generated

the [(M-H)-CO2]¯ ions from 1-17 operating the source and desolvation

temperature at 150o C and 300o C, respectively, instead of 100o C (see

experimental section); these ions were allowed to undergo ion-molecule

reactions with CO2 and the results are compared with those obtained at 100o C

of source/desolvation temperature. Under high temperatures, the ion currents

of [M-H]¯ and [(M-H)-CO2]¯ ions increased several orders of magnitude when

compared to the currents obtained at lower source/desolvation temperatures.

In addition, the relative abundance of CO2 adduct ions increased by 2-25% in

all cases (Table 1, Figure 8), when experiments are performed at increased

source/desolvation temperatures.

Compound % increase in yield

1 8.5

3 25.3

4 12.9

11 17.2

12 2

Table 1: The percentage enhancement of carbon dioxide adduct for

carbanions with increasing the source and desolvation temperature

from 100/100 to 150/3000 C.

This phenomenon clearly confirms increased ion formation in the solvent

free region at higher source/desolvation temperatures due to reduced proton


112

exchange reactions between the generated anions and solvent molecules. The

[(M-H)-CO2]¯ ions of cis-/ortho- isomers failed to react with CO2 even at this

conditions confirming that the carbanions in these isomers are very unstable

and rapidly isomerizing to the corresponding carboxylate/oxide anions by

intramolecular proton transfer.

Figure 8: CO2 ion-molecule reactions spectra of compound 4 at source and

desolvation temperatures of a. 100/1000C and b. 150/300oC.

3.3. THEORETICAL CALCULATIONS

Quantum chemical calculations were employed to corroborate the

observed experimental results. Calculations done on the model systems 1C

and 2C yielded virtually identical answers at the B3LYP, MP2 and CCSD(T)

levels with different basis sets (Table 2 and 3).


113

Method/basis set 1C 1C-TS PR-(1C-I) I-TS I

B3LYP/6-311+G** -266.64 -266.64 -266.70 -266.60 -266.63

B3LYP/cc-pVTZ -266.66 -266.65 -266.71 -266.61 -266.64

B3LYP/6-311++G** -266.64 -266.64 -266.70 -266.60 -266.63

MP2/6-31+G** -265.82 -265.82 -265.88 -265.77 -265.80

CCSD(T)/6-31+G** -265.87 -265.87 -265.93 -265.82 -265.85

CCSD(T)/6-311++G** -265.99 -265.98 -265.93 -265.93 -265.97

Table 2: The total energies of the systems (A, B) at various levels.

Structure 1C I

∆E# 0.8 18.1 6-311+G**

∆ER -33 -43

∆E# 0.7 19.4 B3LYP

cc-pVTZ ∆ER -31 -42

∆E# 0.12 20.4 MP2 6-31+G**

∆ER -36 -48

∆E# 1.2 22 6-31+G**

∆ER -35 -45

∆E# 1.3 21.4 CCSD(T)

6-311++G** ∆ER -31.2 -41

Table 3: The proton transfer energy barriers (∆E#) kcal/mole and reaction

energies (∆ER) kcal/mole of the given systems (A and B) at, B3LYP/6-311+G**,

B3LYP/cc-pVTZ, MP2/6-31+G**, CCSD(T)/6-31+G**, CCSD(T)/6-311++G**

levels.

For consistency, results obtained at the B3LYP/6-311++G** level are

given uniformly for the eight systems considered in Table 4 (Scheme 18). All


114

the calculations were done using the Gaussian 9872 program package.

Geometry optimizations and frequency calculations were performed on all the

structures considered at the B3LYP/6-311++G** level. For computations,

representative examples of dicarboxylic acids (3–5) and hydroxybenzoic acids

(10–12) were taken in addition to maleic (1) and fumaric (2) acids. Pyridine

analogues (6–9, 13 and 14) of aromatic hydroxy acids (15–17) have shown

virtually identical results as those of dicarboxylic acids and hydroxybenzoic

acids, and therefore essentially similar trends are expected for these molecules

in the computations. This prompted us to restrict our detailed calculations to

the eight isomeric carbanions (1C, 2C, 3C, 4C, 5C, 10C, 11C and 12C) and

their three corresponding carboxylate ions (PR-I, PR-II, PR-IV), which were

found to be at a minimum on the potential energy surface (Scheme 18).

Subsequently, the corresponding eight proton transfer transition states were

located and characterized through frequency calculations and examining the

normal mode direction of the imaginary frequency.

Structure 1C 2C 3C 4C 5C 10C 11C 12C

∆E# 0.8 18.1 0.01 37.5 50.1 18.3 58.3 105.4

∆ER -33.3 -42.9 -34.4 -53.4 -51.3 -40.0 -51.2 -53.8

Table 4: The proton transfer energy barriers (∆E#) and reaction energies (∆ER)

in kcal/mole of the given systems at B3LYP/6-311++G** level.


115

= O= = HC

1C-TS PR-I 2C-TS

1.353

1.224

1.0221.837 1.346

1.230

1.1421.546

1.362

1.221

0.9781.320

1.222

1.0151.334

1.5381.259

1.254

IC 2C

1.488 1.3481.506 1.322 1.508 1.346

2.130 2.334

1.480 1.372

1.412

1.408

1.400

1.394

1.398

1.400

1.507

1.330

1.067

1.407

1.4011.400

1.394

1.3971.509

1.2241.327

1.084

1.646

3C 3C-TS

1.223

1.411

PR -II

1.5541.399

1.395

1.397

1.396

1.399

1.396

1.253

1.435

1.380

1.4311.431

1.560

1.304

1.4671.435

1.410

1.393

1.4231.424

1.391

1.4091.456

1.223

1.383

0.967

5C-TS 5C−

3.726

1.217

1.4871.414

1.410

1.4161.404

1.390

1.403

1.212 1.2911.521

1.465

1.438

1.4101.398

1.414

1.515

2.400

4C-TS4C

1.372

1.369

1.401

1.404

1.4101.394

1.401

1.3961.399

1.385

1.4211.392

1.409

1.3981.342

1.282

1.479

10-C 10C-TS

0.972

1.417

PR -IV

1.2691.447

1.388

1.4041.404

1.388

1.447

−

1.392

1.408

1.4131.415

1.407

1.3901.403

1.462

1.381

1.4571.4571.381

1.462

1.3221.367

1.650

12C-TS 12C

0.963

1.4021.399

1.411

1.4201.399

1.400

1.391

1.401

2.714

1.439

1.416 1.408

1.405

1.3961.321

11C-TS11C

0.963

1.452

Scheme 18: Optimized geometries at B3LYP/6-311++G** level

When comparing the anions generated from the maleic and fumaric

acids, 1C and 2C, the former, with a very small barrier, undergoes

spontaneous proton transfer resulting in the generation of the corresponding

carboxylate ion. 2C, despite having a higher exothermicity for the proton


116

transfer reaction, has a substantial barrier of about 18 kcal/mol and thus the

collapse to the carboxylate ion is not spontaneous. Similarly, in the anions

generated from phthalic/hydroxy acid isomers, the ortho- isomer undergoes

the proton transfer in a virtually barrier less fashion. In the anions of the

phthalic acid isomers, the barrier for the proton transfer is substantial for the

meta- isomer 4C, and this increases further for the para- isomer 5C.

Interestingly, the meta- and para- isomers are less stable than the ortho- isomer

and also have higher exothermicity for the proton transfer. Thus, the stability of

the carbanion is exclusively controlled by the proton transfer energy barriers,

which in turn depends on the disposition of the migrating proton with respect to

the carbanion orientation. A similar situation has been observed for the

carbanions generated from the hydroxy acids 10C, 11C, and 12C. Even here

the proton transfer barrier for the ortho- isomer is much less than those of the

meta- and para- isomers, indicating a facile proton transfer only in the case of

10C. Where proton migration is difficult, the compounds have tight transition

states, 2C-TS, 4C-TS, 5C-TS, 11C-TS and 12C-TS. In contrast, the transition

structures with low barriers, 1C-TS, 3C-TS, and 10C-TS, have reactant-like and

loose transition states (Table 4). The above results unambiguously

demonstrate that facile proton transfer is responsible for the non observation of

the carbanions in compounds.

Interestingly, proton transfer propensity is exclusively controlled by the

structural disposition of the migrating proton and is independent of the

exothermicity of the reaction. In fact, in all three cases, the more exothermic


117

proton transfer pathways have higher barriers! In the anions of hydroxybenzoic

acids the preference for CO2 addition to the meta- isomer is significant

compared with that to the para isomer (CO2 addition to 11C over 12C). In

contrast, in the anions of phthalic acids, the situation is reversed, albeit in small

magnitude. In this case, CO2 has a slightly higher preference to add to 5C than

to 4C. Thus the quantum chemical calculations have excellently supported the

experimental results.


118

4. CONCLUSION

The intricacy in the formation of carbanion in the presence of internal

acidic groups has been studied under ESI mass spectrometry. For this purpose,

the [M-H-CO2]¯ ions from different isomeric carboxylic acids (1-17) were

generated in the ESI process, and performed ion-molecule reactions with CO2

in the collision cell of a triple quadrupole mass spectrometer. The ion-molecule

reaction with CO2 reveals the difference in the internal proton transfer in

geometrical (cis-/trans-) and positional (ortho-/meta-/para-) isomeric

compounds and also explains the survival of carbanion in some isomers (trans-

/meta-/para-). This data also provides the information that the immediate

location of the anionic centre upon decarboxylation of [M-H]¯ ions from

aromatic carboxylic acids is at the position of the leaving CO2 group. Clear

support comes from the computational study, which shows that the

intramolecular proton transfer is responsible for precluding the observation of

the one of the twin isomeric carbanions.


119

CCCHHHAAAPPPTTTEEERRR IIIIIIIII PPPAAARRRTTT 222

1. PROLOGUE

n the part 1 of this chapter, we have successfully generated regiospecific

carbanions from compounds containing acidic substituents by the source

induced decarboxylation of the corresponding carboxylic acids. Further,

these carbanions were characterized by ion-molecule reactions with CO2 in

the collision cell. With a view to explore the study of other compounds having

acidic substituents after decarboxylation, we have selected isomeric

sulfobenzoic acids and disulfonic acids. Aromatic sulfonic acids are well-

known surfactants used as solubilizers for dyes and emulsifiers for liquid

paraffins and fuels.73 Some of the aromatic sulfonic acids are also used as key

intermediates in the syntheses of antiviral drugs.74 The sulfonic acid

derivatives of chalcones are used as fungicides.75 Aromatic sulfonates are also

used as intermediates in the production of ion-exchange resins, pesticides

and as wetting agents.76 Detection of aromatic sulfonic acid compounds by

using mass spectrometry is also well known.77-79 Analysis of sulfonic acids

using traditional mass spectral techniques such as EI and CI is difficult due to

their involatile nature, however, they have been analyzed by FAB77,79 and FD78

techniques. Recently, the ion-pair chromatography–ESI mass spectrometry

approach has been used to analyze isomeric aromatic sulfocarboxylates that

I

GGeenneerraattiioonn ooff rreeggiioossppeecciiffiicc ccaarrbbaanniioonnss ffrroomm SSuullffoobbeennzzooiicc aacciiddss


120

originate from the degradation of sulfophthalocyanine dyes by white-rot

fungi.80


The objective of the present investigation is to study the substituent

effect on the stability, structure, and reactivity of carbanions generated in

aromatic sulfonic acids. Recently, Weisz et al.54 studied isomeric 2-, 3- and 4-

sulfobenzoic acids from an ion trap mass spectrometer under negative ion

electrospray ionization (ESI) conditions. In their study, the [M-H]¯ ions

generated by ESI process were subjected to further experiments in the ion

trap. The [(M-H)-CO2]¯ ion formed by the decomposition of [M-H]¯ ion in MS2

was subjected to decompose in MS3 experiment. Based on the MS3 spectra of

[M-H-CO2]¯ ions from isomeric 2-, 3- and 4-sulfobenzoic acids, they showed

that the [M-H-CO2]¯ ions from 2-sulfobenzoic acid existed as benzenesulfonate

anion due to acidic proton transfer to the ortho- position, whereas the [M-H-

CO2]¯ ions from 3- and 4-sulfobenzoic acids survived as carbanions due to

forbidden proton transfer in these isomeric carbanions. But, they have not

applied ion-molecule reactions to characterize the carbanion products. By

using the source induced decarboxylation method which was applied

successfully for aromatic carboxylic acids as described in the part 1 of this

chapter, we took the isomeric 2-, 3- and 4- sulfobenzoic acids, to generate the

carbanion and followed by their characterization using ion-molecule reactions

with CO2.


121

3. RESULTS AND DISCUSSIONS

The molecular structures of the studied isomeric sulfobenzoic acids are

given in Scheme 19, which include isomers of sulfobenzoic acid (18-20) and

disulfobenzoic acids (21-22). The negative ion ESI mass spectra of all the

isomeric sulfobenzoic acids (18-22) showed abundant [M-H]¯ ions at lower

cone voltage (10 eV). While increasing the cone voltage (>30 eV), the

compounds showed common fragment ions at m/z 157 corresponding to [(M-

H)-CO2]¯ (18-20) or [(M-H)-SO3]¯ (21 and 22), m/z 93 (C6H5O-) and 80 (SO3¯·),

with varied relative abundances (Figure 9). In addition to these, the spectra of

19 and 20 include specific fragment ions i. e., the ions at m/z 109 and 81 in 19,

and the ion at m/z 121 [(M-H)-SO3]¯ and 137 [(M-H)-SO2]¯ in 20. Similar

fragment ions are observed in the CID spectra of [M-H]¯ ions from 18-22

(Figure 10). The observed fragment ions in 18-22 are similar to the reported

MSn (n = 2 and 3) experiments on [M-H]¯ ions and m/z 157 from 18-22 by

Weisz et al.54 The fragment ions obtained from the CID spectra of [(M-H)-

CO2]¯ ions from 18-20 are similar to the reported MS3 experiments in ion trap,

except the absence of the ion at m/z 109 in the case of 19, however,

differences are observed in the relative abundances of the fragment ions.

These differences can be attributed to the data generated from two types of

instruments i.e., an ion trap and a triple quadrupole mass spectrometer.81


122

COOH

SO3H

COOH COOH

SO3H

SO3H18 19 20

SO3H

SO3H

SO3H

SO3H

21 22

Scheme 19

2. ISOMERIC SULFOBENZOIC ACIDS (18-20)

In order to confirm whether the [(M-H)-CO2]¯ ions from 18-20 generated

in the ESI process survive as carbanions, we have performed ion-molecule

reactions of these ions with CO2. In the ion-molecule experiment, the [(M-H)-

CO2]¯ ions are selected in MS1 and allowed to react with CO2 in the collision

cell. The experiments showed that the [(M-H)-CO2]¯ ions from 19 and 20

reacted with CO2 and resulted in the corresponding carboxylate anions, and

that from 18 did not react (Scheme 20, Figure 11).


123

-Ve ESI

- CO2

18

19 20

SO3-

COOH

SO3H

COO-

SO3HCOOH

SO3-

COOH

SO3-

SO3H

COOH

SO3-

COOH SO3H

COOHSO3H

COO-

SO3-

COOH

SO3H

COO-

SO3H

COO-

Ion-Molecule Reactions with CO2

SO3H

-

SO3H

-

SO3HCOO-

-Ve ESI

-Ve ESI

-Ve ESI-Ve ESI

-Ve ESI - CO2

- CO2 - CO2

- CO2- CO2

m/z 201

m/z 201

m/z 201

m/z 201

m/z 201

m/z 201

m/z 201 m/z 201

I, m/z 157

A, m/z 157 B, m/z 157

Scheme 20: The mechanism for the decarboxylation of compounds 18-20 and

the ion-molecule reactions of generated carbanions with CO2.

This selective reaction reveals that the [(M-H)-CO2]¯ ions from 19 and 20

are surviving as carbanions, whereas that of 18 is undergoing isomerization

through exchange of acidic proton from COOH to carbanion. In order to

confirm the selective reactivity of CO2, we used p-toluenesulfonic acid and

generated the p-toluenesulfonate ion by deprotonation and performed similar

ion-molecule experiments with CO2. The [M-H]¯ ion from p-toluenesulfonic acid

failed to react with CO2 confirming selectivity of CO2 in reacting only with

carbanions. The survival of [(M-H)-CO2]¯ ions from 19 and 20 as carbanion

suggests that the proton exchange is kinetically unfavorable and this is in good

agreement with the Weisz et al.54 observations. From Figure 11, It can be noted


124

that the abundance of the CO2 adduct ion of [(M-H)-CO2]¯ ion is relatively

higher in 19 than in 20, which suggests relatively higher stability of the

carbanion from 19.


125

Figure 9: The ESI mass spectra of compounds 18-22 (a-e) at the cone

voltage of 50 V. (* specific fragment ions)


126

Figure 10: The CID mass spectra of [M-H]¯ ion for compounds 18-22 (a-e).

(*specific fragment ions)


127

Figure 11: The ion-molecule reaction mass spectra of [M-H-CO2]¯ ions from 18-

20 (a-c) with CO2 in the collision cell.

3. ISOMERIC BENZENEDISULFONIC ACIDS (21 AND 22)

We extended the ion-molecule reactions of CO2 on [(M-H-SO3]¯ ions from

ortho- and meta- benezenedisulfonic acids (21 and 22). The fragment ion at m/z

157 corresponding to [(M-H)-SO3]¯ ions from 21 and 22 were generated in the

source (Scheme 21) and further performed ion-molecule reactions with CO2.

SO O

OH

-CO2

m/z 157

SO O

OH

-

CO2

m/z 157

SO O

OH

-

CO2

m/z 157


128

m/z 237m/z 237 I, m/z 15721 22

-Ve ESI -Ve ESI

SO3HSO3H

SO3-

SO3HSO3

- SO3-

SO3H

SO3H

SO3H- SO3 - SO3

Scheme 21: The mechanism for the desulfonation of compound 21 and 22.

Interestingly, the [(M-H)-SO3]¯ ions from both 21 and 22 did not react

with CO2, which confirms that the generated [(M-H)-SO3]¯ ions from 21 and 22

are benzenesulfonate anions, but not carbanions. These experimental results

are similar to the results of Weisz et al.54, wherein the MS/MS spectra of the ion

m/z 157 due to expulsion of SO3 from [M-H]¯ from 21 and 22 were identical and

similar to that of m/z 157 from benzenesulfonate anion. Formation of

benzenesulfonate anion in the case of 22 presumably due to the elimination of

SO3 from the [M-H]¯ ion involves the SO3H group with a 1,3-proton transfer from

oxygen to carbon atom rather than the SO3¯ moiety. This conclusion is drawn

based on the fact that the SO3 expulsion is absent from the ion of m/z 157, and

the ion of m/z 121 from 20 by the loss of SO3 failed to react with CO2 in ion-

molecule experiment, which can only be explained by the expulsion of SO3

involving SO3H as explained in Scheme 22.


129

3

SO3H

COOHSO3H

COO-

SO3-

COOH

-Ve ESI

-Ve ESI

COO-

-SO3

-SO3m/z 121

m/z 201

m/z 201

Scheme 22: The mechanism for the desulfonation of compounds 20.


130

4. CONCLUSION

In summary, the selective ion-molecule reactions of CO2 with carbanions are

successfully used as a probe to understand formation of isomeric carbanions in the case

of 2-, 3- and 4-sulfobenezoic acids upon decarboxylation in the negative ESI process.

Based on the ion-molecule reactions of CO2, it was concluded that the [M-H-CO2]¯ ion

from 2-sulfobenzoic acid is isomerizing to a benzenesulfonate anion, whereas that from

3- and 4-sulfobenzoic acids do survive as carbanions. However, the present study does

not rule out the presence of benezenesulfonate anions, which do not react with CO2 in the

ion-molecule reaction experiments. The process of desulfonation from the [M-H]¯ ion of

1,2- and 1,3-benzenedisulfonic acids involves the SO3H group with a 1,3-proton transfer

from oxygen to carbon atom rather than the SO3¯ moiety; hence, corresponding

carbanions could not be formed on decarboxylation process.

Chapter 3 Regiospecific Carbanions…

131

5. EXPERIMENTAL

a. MATERIALS

The compounds 1-22 were purchased from Sigma-Aldrich (St. Louis,

USA) and were used without further purification. HPLC grade solvents (water

and methanol) were purchased from Merck (Mumbai, India).

b. MASS SPECTRAL ANALYSIS

All the experiments described in this paper were carried out on Quattro

LC, a triple quadrupole mass spectrometer equipped with an ESI source

(Micromass, Manchester, UK) operated in negative ion mode. Spectra were

acquired using Masslynx software (version 3.2). The typical operating

conditions were: capillary voltage, 3.5 kV; cone voltage, 10-30 V; the source

housing and desolvation temperatures, 100 oC, unless otherwise stated.

Nitrogen was used as nebulization and desolvation gas. Stock solutions (1 mM)

of all the compounds were made in 50:50 (vol/vol) water:methanol and working

standards were prepared by diluting the stock solutions with appropriate

volumes of methanol to achieve a final concentration of 100 µM. The sample

solutions were introduced using an infusion pump (Harward apparatus, Kent,

UK) at a flow rate of 10 µL/min. We have performed the ion-molecule reactions

by introducing the CO2 gas into the collision cell, and the pressure inside the

collision cell was maintained at 9 x 10-4 mbar at the collision energy of zero eV.

For collision-induced dissociation (CID) experiments argon was used as


132

collision gas, and the pressure was maintained at 3 x 10-4 mbar. All the

experimental data reported here are an average of 50 spectra.


133

5. REFERENCE

1. Smith LG. Phy. Rev. 1937; 51: 263.

2. Buncel E, Durst T. “Comprehensive Carbanion Chemistry,” Part A: Structure

and Reactivity, Elsevier: Amsterdam, 1980.

3. DePuy CH, Bierbaum VM. Acc. Chem. Res. 1981; 14: 146.

4. Bates RB, Ogle CA. “Carbanion Chemistry,” Springer-Verlag: Berlin, 1983.

5. Nibbering NMM. Adv. phys. Org. Chem. 1988; 24: 1.

6. Graul ST, Squires RR. Mass Spectrom. Rev., 1988; 7: 263.

7. Bowie JH. Mass Spectrom. Rev. 1990; 9: 349.

8. Squires RR. Acc. Chem. Res. 1992; 25: 461.

9. DePuy CH. Int. J. Mass Spectrom., 2000; 200: 79.

10. Gronert S. Chem. Rev., 2001; 101: 329.

11. DePuy CH. J. Org. Chem,. 2002; 67: 2393.

12. Waters T, O’Hair RAJ. Annu. Rep. Prog. Chem., Sect. B, 2002; 98: 433.

13. D. J. Cram. “Fundamentals of Carbanion Chemistry”, Academic Press,

New York, N.Y., 1965.

14. Sullivan SA, Beauchamp JL. J. Am. Chem. SOC. 1977; 99: 5017.

15. Bierbaum VM, DePuy CH, Shapiro RH. J. Am. Chem. Soc., 1977; 99: 5800.

16. Bartmess JE. J. Am. Chem. SOC., 1980; 102: 2483.


134

17. Wight CA, Beauchamp JL. J. Am. Chem. SOC., 1981; 103: 6499.

18. Squires RR, DePuy CH, Bierbaum VM. J. Am. Chem. Soc., 1981; 103: 4256.

19. Grabowski JJ, DePuy CH, Bierbaum VM. J. Am. Chem. SOC., 1983; 105:

2565.

20. Ingemann S, Nibbering NMM. J. Org. Chem., 1983; 218: 183.

21. Kleingeld JC, Nibbering NMM. Tetrahedron, 1983; 209: 4193.

22. Lee HS, DePuy CH, Bierbaum VM. J. Am. Chem. Soc. 1996; 118: 5068.

23. Buker HH, Nibbering NMM, Espinosa D, Mongin F, Schlosser M.

Tetrahedron Lett., 1997; 208: 8519.

24. Chen H, Cooks RG, Meurer EC, Eberlin MN. J. Am. Soc. Mass. Spectrom.,

2005; 186: 2045.

25. DePuy CH, Bierbaum VM, Flippin LA, Grabowski JJ, King GK, Schmitt RJ. J.

Am. Chem. Soc. 1979; 101: 6443.

26. DePuy CH, Bierbaum VM, Flippin LA, Grabowski JJ, King GK, Schmitt RJ,

Sullivan SA. J. Am. Chem. SOC., 1980; 102: 5012.

27. Squires RR, DePuy CH. Org. Mass Spectrom. 1982; 17: 187.

28. DePuy CH, Bierbaum VM, Damrauer R, Soderquist JA. J. Am. Chem. Soc.,

1985; 107: 3385.

29. O’Hair RA. J, Gronert S, DePuy CH, Bowie JH. J. Am. Chem. Soc., 1989; 18:

3105.


135

30. Kass SR, Guo H, Dahlke GD. J. Am. Soc. Mass Spectrom. 1990; 18: 366.

31. Rabasco JJ, Kass SR. J. Am. Soc. Mass Spectrom., 1992; 20: 91.

32. Chou PK, Kass SR. J. Am. Chem. Soc., 1991; 113: 4357.

33. Chou PK, Dahlke GD, Kass SR. J. Am. Chem. Soc., 1993; 115: 315.

34. Sachs RK, Kass SR. J. Am. Chem. Soc. 1994; 116: 783.

35. Wenthold PG, Hu J, Squires RR. J. Am. Chem. Soc., 1994; 116: 6961.

36. Baschky MC, Peterson KC, Kass SR. J. Am. Chem. Soc., 1994; 116: 7218.

37. Hu J, Squires RR. J. Am. Chem. Soc., 1996; 118: 5816.

38. Merrill GN, Dahlke GD, Kass SR. J. Am. Chem. Soc., 1996; 118: 4462.

39. Hare M, Emrick T, Eaton PE, Kass, SR. J. Am. Chem. Soc. 1997; 119: 237.

40. Wenthold PG, Hu J, Hill BT, Squires RR. Int. J. Mass Spectrom., 1998; 180:

173-183.

41. Kondrat RW, McClusky GA, Cooks RG. Anal. Chem. 1978; 50: 1222.

42. McClusky GA, Kondrat RW, Cooks RG. J. Am. Chem. Soc. 1978; 100: 6045.

43. Froelicher SW, Freiser BS, Squires RR. J. Am. Chem. SOC., 1986; 108:

2853.

44. Graul ST, Squires RR. J. Am. Chem. Soc., 1988; 110: 607.


46. Siu KMW, Gardner GJ, Berman SS, Org. Mass. Spectrom. 1989; 1921: 931.


136

47. Maas WPM, van Veelen PA, Nibbering NMM. Org. Mass Spectrom. 1989;

1921: 546.



50. Bachrach SM, Hare M, Kass SR, J. Am. Chem. Soc. 1998; 120: 12646.

51. Visser R, Maas WPM, Nibbering NMM. Rec. Trav. Chim. Pays-Bas 1990;

1809: 248.

52. Gronert S, Feng WY, Chew F, Wu W. Int. J. Mass Spectrom., 2000, 195/196,

251.

53. Bienkowski T, Danikiewicz W. Rapid Commun. Mass Spectrom. 2003; 187:

697.

54. Ben-Ari J, Etinger A, Weisz A, Mandelbaum A. J. Mass. Spectrom. 2005;

210: 1064.

55. Teng-Yi H, Emory JF, O’Hair RAJ, McLuckey SA. Anal. Chem. 2006; 78:

7387.

56. Stewart JH, Shapiro RH, DePuy CH, Bierbaum VM. J. Am. Chem. Soc., 1977;

99: 7650.

57. Bierbaum VM, Grabowski JJ, DePuy CH. J. Chem. Phys., 1984; 88: 1389.

58. Liang JY, Lipscomb WN. J. Am. Chem. SOC., 1986; 108: 5051.

59. Hunt DF, Sethi SK. J. Am. Chem. Soc., 1980; 102: 6953.


137

60. Robinson MS, Breitbeil FW. Int. J. Mass. Spectrom., 1992; 117: 647.

61. Nibbering NMM. Recl. Trav. Chim. Pays-Bas, 1981; 100: 297.

62. Kass SR, Filley J, Doren JMV, DePuy CH. J. Am. Chem. Soc., 1986; 108:

2849.

63. O’Hair RAJ, Gronert S, DePuy CH. Eur. Mass Spectrom., 1995; 18: 429.

64. Bartmess JE, Caldwell G, Rozeboom MD, J. Am. Chem. SOC., 1983; 105:

340.

65. Bell RP. 'The Proton in Chemistry", 2nd ed.; Cornell University Press:

Ithaca, New York, 1973.

66. Peerboom RAL, de Koning LJ, Nibbering NMM. J. Am. Soc. Mass Spectrom.,

1994; 5: 159.

67. NIST Chemistry Web Book. Available: http://webbook. nist. -

gov/chemistry/.

68. McMahon TB, Kebarle P. J. Am. Chem. Soc., 1977; 99: 2222.

69. Notario R. J. Mol. Str., 2000; 556: 245.

70. Asbury GR, Hill HH. Int. Soc. Ion Mob. Spec., 1999; 19: 1, 1-8, p. 2

71. Chen H, Talaty NN, Takats Z, Cooks RG. Anal. Chem., 2005; 77: 6915.

72. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman

JR, Zakrzewski VG, Montgomery JA Jr, Stratmann RE, Burant JC, Dapprich

S, Millam JM, Daniels AD, Kudin KN, Strain MC, Farkas O, Tomasi J, Barone


138

V, Cossi M, Cammi R, Mennucci B, Pomelli C, Adamo C, Clifford S,

Ochterski J, Petersson GA, Ayala PY, Cui Q, Morokuma K, Rega N,

Salvador P, Dannenberg JJ, Malick DK, Rabuck AD, Raghavachari K,

Foresman JB, Cioslowski J, Ortiz JV, Baboul AG, Stefanov BB, Liu G,

Liashenko A, Piskorz P, Komaromi I,Gomperts R, Martin RL, Fox DJ, Keith

T, Al-Laham MA, Peng Y, Nanayakkara A, Challacombe M, Gill PMW,

Johnson B, Chen W, Wong MW, Andres JL, Gonzalez C, Head-Gordon M,

Replogle ES, Pople JA. Gaussian 98, Revision A. 11. 3. Gaussian:

Pittsburgh, PA, 2002.

73. Asakura H, Muramoto Y, Negoro K. Yakagaku, 1978; 28: 294.

74. Bauer DJ. Ann. NY Acad. Sci. 1965; 130: 110.

75. Wurm G. Arch. Pharm. 1975; 308: 142.

76. Hashm MA, Kulandai J, Hassan RS. J. Chem. Technol. Biotechnol. 1992; 54:

207.

77. Mathias A, Williams AE, Games DE, Jackson AH. Org. Mass Spectrom.,

1976; 11: 266.

78. Monaghan JJ, Barber M, Bordoli RS, Sedgwick RD, Tyler AN. Org. Mass

Spectrom., 1982; 187: 529.

79. Lin HY, Gonyea G, Killen S, Chowdhury SK. Rapid Commun. Mass

Spectrom., 2000; 14: 520.

80. Reemtsma T. Journal of Chromatography A, 2001; 919: 289.


139

81. Bristow AWT, Webb KS, Lubben AT, Halket J. Rapid Commun. Mass

Spectrom., 2004; 18: 1447.

O-

.

O-

R - ve

ESI/APCI

m/z 92ortho-, meta- and para-Distonic dehydrophenoxide

radical anion

phenols/nitrobenzenesSubstituted SID/CID

GGEENNEERRAATTIIOONN OOFF DDIISSTTOONNIICC DDEEHHYYDDRROOPPHHEENNOOXXIIDDEE RRAADDIICCAALL

AANNIIOONNSS UUNNDDEERR EELLEECCTTRROOSSPPRRAAYY AANNDD AATTMMOOSSPPHHEERRIICC PPRREESSSSUURREE

CCHHEEMMIICCAALL IIOONNIIZZAATTIIOONN CCOONNDDIITTIIOONNSS

Chapter 4.1 Phenoxide radical anions under ESI…

140


PPPAAARRRTTT 111

1. PROLOGUE

he chemistry of radical anions is an important and well established field

of research as indicated by the numerous studies of reactions involving

these species in the condensed phase and in the gas phase.1,2 Most of the

studies on condensed phase reactions were concerned with the involvement of

radical anions either as primary reactants or as transient intermediates.3-7 One

of the first radical anions formed in the condensed phase was undoubtedly the

benzil semidione radical anion (Scheme 1), which was recognized by the

appearance of deep blue color upon the addition of potassium hydroxide to a

solution of benzil.8 The radical anions generated in the condensed phase have

been frequently studied with the use of Electron Spin Resonance (ESR)

spectroscopy, electrochemical methods etc.5,6

O

O e-

O-

O.

K+

OH-

Scheme 1

T

GGeenneerraattiioonn ooff DDiissttoonniicc ddeehhyyddrroopphheennooxxiiddee rraaddiiccaall aanniioonnss

ffrroomm ssuubbssttiittuutteedd pphheennoollss uunnddeerr EElleeccttrroosspprraayy iioonniizzaattiioonn ccoonnddiittiioonnss


141

Though the gas-phase studies are well established, reports on the gas

phase radical anions are not many. Mass spectrometry has become a

remarkably versatile tool for studying molecular structures, reactivities and

thermochemistry of various radical anions. In general, these ions are energetic

and reactive species, and react with stable neutrals or atomic radicals.

Studying physico chemical properties and reactions of the radical anions in the

gas phase is of great importance for organic chemists, because most of the

synthetically useful condensed phase reactions involve these species as

intermediates.3-7 The most important physicochemical characteristic of radical

anions are electron affinity and the gas phase acidity of conjugate acid.

Comparison of gas-phase studies with the condensed-phase results helps to

understand better intrinsic reaction characteristics and solvent effects. The

studies on the gas phase radical anions are quite impressive; some recent

comprehensive reports were available.9-12

1.1. FORMATION OF RADICAL ANIONS IN THE GAS PHASE

With the aid of mass spectrometry and photoelectron spectroscopy, it

has become apparent that radical anions may be formed as stable species in

the gas phase. When the radical anions are formed under mass spectral

conditions, in principle the formation of radical anions may involve a number of

processes. The most important processes are: (i) electron attachment

(dissociative), (ii) electron transfer, and (iii) ion/molecule reactions.

1.1.1. Electron attachment (dissociative)


142

The first method is the simplest one to generate the radical anion, i.e.,

the attachment of a free low-energy electron to the molecule. Initially, the

attachment of an electron to a molecule (A) will result in an excited species (A-.)

which may spontaneously detach an electron. The observation of a stable

radical anion requires, therefore, that the lifetime of the excited species (A-•)*

is sufficiently long for stabilization to occur by collisions with a bath gas (M)

that is present at a relatively high pressure13 (Scheme 2) and/or by the

emission of photons.14

A (A.-)* A.-e M-M*

Scheme 2

Undoubtedly one of the most-studied examples is the formation of the

radical anion of the oxygen atom by dissociative electron attachment to N2O

(Scheme 3). This process has proven to be extremely useful in mass

spectrometry, as indicated by the large number of studies concerned with the

reactions of the O-• radical anion.9,15

N2O N2 + O-.e

Scheme 3

Though this method is simple to generate radical anions, it has got one

drawback, i.e. electron attachment may involve dissociation of the transient

radical anion (A˙¯)*. For a large number of species, the process involves the

loss of a radical with the formation of an anion.

1.1.2. Electron transfer


143

Electron transfer from one ionic species to a neutral molecule is another

convenient method for the formation of radical anions and has also been

studied extensively.16 In most instances, these studies have involved an

electron transfer between aromatic species as exemplified in Scheme 4, where

formation the cinnoline radical anion is shown by electron transfer from the

hexafluorobenzene radical anion (C6F6-.).

C6F6-. +

NN

C6F6 +

NN

-.

Cinnoline

Scheme 4

This process may not include necessarily a radical anion as the donor,

because the generation of radical anions has also been noticed by a facile

electron transfer process in the reactions of some anions with neutral molecules

in the gas phase. As an example, the molecular oxygen radical anions (O2-.

ions) are formed in high yields in the reaction of the allyl anion with O2

(Scheme 5).16

CH2=CH-CH2- + O2 CH2=CH-CH2

. + O2¯.

allyl anion

Scheme 5

In fact, the O2-. radical anion can be used to generate radical anions from

appropriate neutral molecules. For example, most of the chlorofluoromethane

radical anions were generated by using this method (Scheme 6).


144

O2-. + CFCl3

CFCl3-. + O2

.CFCl2 + Cl- + O2

Scheme 6

1.1.3. Ion-molecule reactions

Another promising method to generated radical anions is by ion-

molecule reactions. A common ion-molecule reaction method is subjecting the

analyte for chemical ionization with N2O as the reagent gas. Under N2O/CI, the

dissociative electron attachment to N2O molecule yields atomic oxygen radical

anion, O¯˙ (Scheme 3). This ion (O-.) reacts with an organic compound (MH2) by

the following ways: (i) Proton abstraction, (ii) hydrogen atom abstraction, and

(iii) formal H2+. abstraction (Scheme 7). The reaction proceeded by H2

+. –

abstraction is of particular interest, because this process leads to the formation

of radical anions.

O-. + MH2

HO. + MH-

HO- + MH.

H2O + M-.

H+ abstraction

H abstraction

H2+. abstraction

Scheme 7

The reactions of the O-. ion with organic molecule has also proven to be a

valuable method for the formation of radical anion, for example the formation of

alkenes or cycloalkene radical anions (Scheme 8). Normally, radical anions of


145

alkenes or cyclo-alkenes are not formed in the gas phase because of the

negative electron affinity of these species.17

+ O-.

-.

+ H2O

Scheme 8

1.2. DISTONIC RADICAL ANIONS

The term “distonic ion” first coined by Yates and co-workers,18,19 refers

to an ion with formally separated charge and radical sites.20-24 The structure,

stability and unimolecular dissociations of radical ions have been studied

extensively,20-24 whereas insight into the gas-phase chemistry of distonic ions is

somewhat more limited. Distonic ions, which possess distinct, spacially

separated charged and radical sites have the potential to undergo both ionic

and radical reactions. These species are of fundamental interest for several

reasons: for use in mass spectrometric studies of free radical chemistry, for use

in investigating the acid-base properties of radicals, and, most significantly, for

use in negative ion photoelectron spectroscopic measurements of the electron

affinities and singlet-triplet energy gaps in the corresponding neutral

biradicals. Less work has been reported on the reactions of distonic radical

anions, but some fascinating species have been identified. The distonic radical

anions have been generated for chemical and physical studies in the gas phase

by CID experiments.25


146

OH

R

O-

R

O-

.OH -

-R.

R = H/Me/Et/i-Pr m/z 92Distonic dehdyro phenoxide

radical anions

Scheme 9

Bowie et al.25 produced isomeric distonic dehydrophenoxide radical

anions (m/z 92) in the CID experiments on one or two alkyl (R = methyl, ethyl

etc.) substituted phenoxide anions. In these experiments, the major

fragmentation process involves the loss of radical (R.) (Scheme 9).

Other than the general CID methods, there have been relatively new

methods used to generate distonic radical anions. One of them is the reaction of

a trimethylsilyl (TMS)- substituted carbanion (generated from corresponding

TMS derivative in the presence of strong a base) with fluoride anion under

flowing afterglow conditions. For example, .CH2COO- radical anion has been

generated from H3C-C(O)OSi(CH3)3 (Scheme 10).

H3C

O

OSi(CH3)3

OH -

H2C

O

OSi(CH3)3-

.CH2COO-F -

Scheme 10

Squires and co-workers presented several applications of the above

method to generate isomeric distonic radical anions.26-28 Most importantly, they

reported a remarkable synthetic pathway for forming gas phase distonic


147

radical anions from bis-trimethylsilyl compounds. Treatment of bis-

trimethylsilyl compounds with a combination of fluoride and molecular fluorine

leads to the sequential loss of the trimethylsilyl groups as (CH3)3SiF and yield

regioselective distonic radical anions (Scheme 11). In this mechanism, the

process is initiated by fluoride-induced desilylation to give a carbanion.

Reaction of this carbanion with molecular fluorine leads to formation of a

carbon-centered radical by dissociative electron transfer. A second fluoride

induced desilylation gives the distonic radical anion. This method is general

and a wide range of distonic ions were reported by using this method.

Si(CH3)3

-(CH3)3SiFSi(CH3)3

Si(CH3)3

- -

.

F -

- (CH3)3SiF

-F -, -F .

F2

m/z 76o-, m- and p-

Scheme 11

Kass et al. recently reported another new method for the generation of

distonic radical anions from aromatic mono and dicarboxylic acids.29,30 In the

case of mono carboxylic acids, the initially generated singly charged

carboxylate anions are subjected for decarboxylation by SORI (sustained off-

resonance irradiation). Further SORI decomposition of the decarboxylated

product produces distonic radical anion upon the loss of NO· radical. Formation

of isomeric distonic radical anions (m/z 92) by the SORI-CID of ortho-, meta-,

and para-nitrobenzoate involving the sequential loss of carbon dioxide and


148

nitric oxide is shown in Scheme 12.31 This sequence violates the “even-

electron rule” and affords radical anions at m/z 92.32

O--CO2

NO2

CO2-

-NO.

SORI-CID SORI-CIDNO2

- .

EI/ESINO2

CO2H

m/z 92 o-, m- and p-Nitrobenzoic Acid

Scheme 12

In the case of dicarboxylic acids, doubly charged carboxylate anions are

initially generated under ESI. Upon the SORI-CID, the resulted dicarboxylated

anions undergo the sequential loss of two carbon dioxides to produce distonic

isomeric benzyne radical anions (Scheme 13).

COOH

COOH

COO-

COO- COO-

.

-

.

F -

HF

SORI

CO2

SORI

o-, m- and p-benzenedicarboxylic acid

-CO2

Scheme 13

1.3. CHARACTERIZATION OF RADICAL ANIONS

In order to characterize the structures of the radical anions formed by

various routes, several common mass spectrometric methods were used, which

include isotopic labeling, specific ion-molecule reactions, CID, and collision

induced charge reversal processes.1,33 Among them, characterization by the

ion-molecule reactions is important because these are specific and the results


149

reflect deep insights into the fundamental physical and chemical behavior of

gas phase radical anions. The structures of the ortho-, meta-, and para-benzyne

distonic radical anions were confirmed by converting them to corresponding

nitrobenzoates in a two-step process. Initially addition of CO2 produces a

distonic benzoate species (i.e., reaction at the anionic site) and followed by

addition of NO2 (to the radical site) yields nitrobenzoate (Scheme 14).

Interestingly, the reagents cannot be added in the reverse order because the

anionic site undergoes a charge transfer reaction with NO2 faster (produce

NO2-) than the radical site can add NO2. Conversion of the phenyl anion to the

carboxylate eliminates the charge transfer pathway because the carboxylate

has a higher electron binding energy. This behavior again points to the dual

reactivity of distonic species. Comparison of the gas phase basicities of the

nitrobenzoates formed in the ion-molecule reactions with those of authentic

samples confirmed the structural assignments of the ortho-, meta-, and para-

benzyne radical anions.34,35

.

CO2

.

CO2-

NO2

CO2-

NO2

-

Scheme 14

In a recent communication, Kass et al. successfully differentiated distonic

isomeric dehydrophenoxide radical anions by ion-molecule reactions with

several reagent gasses, such as CCl4, MeSSMe, SO2, NO2, COS, etc. All the


150

three isomers of distonic dehydrophenoxide radical anions were differentiated

on the basis their selective reactivity with the used regent gasses.


From the introduction, it is clear that most of the experiments regarding

the generation of radical anions were done in the flowing-afterglow or FT-ICR

instruments and a very few experiments were known for the formation of

radical anions in the collision cell. However, to the best of our knowledge the

electrospray ionization technique has not been applied for the generation of

distonic radical anions. This brings up a question whether or not isomeric

distonic radical anions can exist and their identity can be detected, if they are

generated under ESI conditions. In Chapter 3, we have shown that isomeric

carbanions do survive in the ESI process and selectively react with CO2 when

ion-molecule reactions are performed on these carbanions in the collision cell.

This encouraged us to extend the same method to study generation of isomeric

dehydrophenoxide radical anions from suitably substituted nitrobenzoic acids

and phenols, and studying their ion-molecule reactions with CO2 in the

collision cell.


3.1. NITROBENZOIC ACIDS (1-3)

For initial experiments, we have undertaken isomeric

nitrobenzoic acids (1-3) to study their source fragmentation under ESI


151

conditions. The molecular structures of the compounds 1-3 are shown in

Scheme 15.

COOH

NO2

COOH

NO2

COOH

NO2

1 2 3

Scheme 15

The negative ion ESI mass spectra of all the isomeric compounds (1-3)

show abundant [M-H]- (m/z 166) ions at lower cone voltage (10 V). The typical

spectrum of 3 is shown in Figure 1a. While increasing the cone voltage from 20

to 50 eV, the compounds 1-3 show fragment ions at m/z 122, 92 and 46 (Figure

1b-d) with varied relative abundances. The primary fragment ion is the ion at

m/z 122 corresponding to loss of CO2 from [M-H]- ion, and the loss of CO2 is a

common fragmentation process observed for aromatic carboxylic acids in the

negative ions mode. The other major fragment ion observed is the ion at m/z 46

corresponding to NO2-, which is characteristic for nitroaromatic compounds.

The fragment ion of interest is found at m/z 92 corresponds to loss of 74 Da from

[M-H]- ion.

In order to confirm the formation of m/z 92 ion from [M-H]- ion, CID

experiments are performed on the ions m/z 166 and 122 from 1-3. Typical CID

spectra from 3 are shown in Figure 2.


152

Figure 1: Negative ion electrospray ionization spectra of 3 at different cone

voltage values: (a) 10, (b) 20, (c) 30 and (d) 50 V.


153

Figure 2: CID mass spectra of (a) [3-H]- (m/z 166) at 20 eV collision energy, (b)

[3-H-CO2]- (m/z 122) at 20 eV collision energy.

The CID spectrum of [M-H]- ion shows the loss of CO2 as the major

fragmentation pathway to give the ion at m/z 122. The CID spectrum of the [(M-

H)-CO2]- ion (m/z 122) show the fragment ion at m/z 46 as the major fragment in

addition to the ion at m/z 92. The CID experiments clearly confirm that the ion

at m/z 92 is resulted from [(M-H)-CO2]- ion by the loss of NO.. During this

process, the NO must leave as a radical and hence the ion m/z 92 could be a

radical anion. Proposed mechanism for the formation of radical anions from 1-3

is shown in Scheme 16. Kass et al. proposed similar mechanism for the

formation of the ion m/z 92 from nitrobenzoic acids by SORI-CID experiments in

FT-ICR cell. Existence of the m/z 92 in the ESI spectra of 1-3 reveals that the

radical anions can also be generated in the ESI process.

*

*


154

COOH

NO2

COOH

NO2

COOH

NO2

COO-

NO2

COO-

NO2

COO-

NO2

NO2

NO2

NO2

-

-

-

O-

O-

O-

.

.

.

1

2

3

I, m/z 92

II, m/z 92

III, m/z 92m/z 166 m/z 122

m/z 122

m/z 122

m/z 166

m/z 166

-ve ESI

-ve ESI

-ve ESI

-NO.

-NO.-CO2

-NO.-CO2

-CO2

Scheme 16

3.2. Isomeric hydroxytoluenes (3-6)

Further, we extended the ESI experiments to generate radical anions

from substituted phenols. Bowie et al. reported generation of radical anions

from substituted phenols in the CID experiments on the corresponding [M-H]-

ions generated under chemical ionization conditions. Hence, we have selected

the same set of isomeric cresols (Scheme 17), viz., ortho-, meta-, para-


155

hydroxyl toluenes (4-6) to check whether the radical anions can be generated

under ESI conditions. The negative ion ESI spectra of 4-6 recorded at higher

cone voltage (>50 eV) show the expected ion at m/z 92 in addition to other

fragment ions at m/z 93 and 77 (Figure 3, Scheme 18). It can be observed

formation of the fragment ion at m/z 92 is less in 4-6, when compared to 1-3.

The same is reflected in the CID spectra of [M-H]- ion from 4-6. HRMS data

confirmed the elemental composition of the ion at m/z 92 from 4-6 also, and

formation of this ion from 4-6 involves direct loss of methyl radical from [M-H]-

ion. In fact, Bowie et al. studied a series of alkyl phenols by using CID

experiments, wherein it was shown that under CID, the [M-H]- ions of alkyl

phenols readily lose an alkyl radical.

OH

CH3

OH

CH3

OH

CH34 5 6

Scheme 17

[M-H]- m/z 92 Other ions Compound

Cone CID Cone (50 eV) CID (20 eV) Cone (50 eV) CID (20 eV)

4 138 (76)

138 (35)

[M-H-NO2.]-.

(6) [M-H-NO2

.]-.

(22) 108, [M-H-NO]- (100);

46, NO2-. (89)

108, [M-H-NO]- (53); 46, NO2

-. (100)

5 138 (74)

138 (41)

[M-H-NO2.]-.

(4) [M-H-NO2

.]- .

(12) 108, [M-H-NO]- (100);

46, NO2-. (15)

108, [M-H-NO]- (100); 46, NO2

-. (22)

6 138 (59)

138 (23)

[M-H-NO2.]-.

(5) [M-H-NO2

.]-. (5) 108, [M-H-NO]- (56);

46, NO2-. (100)

108, [M-H-NO]- (30); 46, NO2

-. (100)


156

Table 1: Ions detected in the cone and CID fragmentation under

negative ESI conditions for compounds 4-6. The numbers which shown in

parenthesis ( ) are relative abundances.

Figure 3: a) source fragmentation of [6-H+]- at 50 eV cone voltage, b) CID

spectra of [6-H+]- at 20 eV collision energy of compound 6.

*

*


157

OH

R

OH

OH

R

R

-Ve ESI

-Ve ESI

-Ve ESI

O-

R

O-

O-

R

R

-R.

-R.

-R.

O-

.

O-

O-

.

R = -CH3 (4) ; -NO2(7)

R = -CH3 (5) ; -NO2 (8); -CHO(11)

R = -CH3 (6) ; -NO2 (9); -CHO (12)

.

Scheme 18

3.3. Isomeric nitrophenols (7-9) and hydroxy benzaldehydes (10-12)

With these encouraging results, we selected other derivatives of

phenols (Scheme 19), viz isomeric nitrophenols (7-9) and isomeric hydroxy

benzaldehydes (10-12).


158

OH

CHO

OH

CHO

OH

CHO

OH

NO2

OH

NO2

OH

NO27 8 9

10 11 12

Scheme 19

All these isomeric compounds also show abundant [M-H]- ions in the ESI

spectra recorded at lower cone voltages. At higher cone voltages (>30 V), all

the compounds yield a common ion at m/z 92, C6H6O-. (except in 10) (Table 2,

Figure 4a). The ion at m/z 92 is formed by the direct loss of .NO2 radical from

[M-H]- ion in the case of 7-9. In addition to this ion, the spectra of 7-9 include

the fragment ions at m/z 108 and 46, which correspond to [(M-H)-NO.]- and

NO2-, respectively. Similar fragment ions are observed in the CID spectra of

[M-H]- ions from compounds 7-9 (Table 2). and typical mass specta shown for

[M-H]- ion of compound 12 (Figure 4b).


159

Figure 4: a) source fragmentation of [9-H]- at 50 eV cone voltage, b) CID

spectra of [9-H]- at 20 eV collision energy of compound 9.

[M-H]- m/z 92 Other ions Compound

Cone CID Cone (50 eV) CID (20 eV) Cone (50 eV) CID (20 eV)

7 121 (44)

121 (15)

- - 93, [M-H-CO]- (100) 93, [M-H-CO]- (100)

8 121 (100)

121 (100)

[M-H-CHO.]-.

(15) [M-H-CHO.]-.

(30) 93, [M-H-CO]- (31) 93, [M-H-CO]- (33)

9 121 (100)

121 (100)

[M-H-CHO.]-.

(24) [M-H-CHO.]-.

(37) 93, [M-H-CO]- (12) 93, [M-H-CO]- (30)

10 107 (100)

107 (100)

[M-H-CH3.]-.

(2) [M-H-CH3

.]-.

(15) 93, [M-H-CH2]- (8) -

11 107 (100)

107 (100)

[M-H-CH3.]-.

(44) [M-H-CH3

.]-.

(53) 93, [M-H-CH2]- (12) -

12 107 (100)

107 (100)

[M-H-CH3.]-.

(59) [M-H-CH3

.]-.

(10) 93, [M-H-CH2]- (3) -

Table 2: Ions detected in the cone and CID fragmentation under negative ESI

conditions for compounds 7-12. The numbers which shown in

parenthesis ( ) are relative abundances.

In the case of hydroxyl benzaldehydes (11-12), the ion at m/z 92 is

formed by the direct loss of .CHO radical from the [M-H]- ion. The compound 10

*

*


160

does not yield the expected ion at m/z 92, instead it shows the ion at m/z 93

corresponding to the loss of CO from [M-H]- ion due to ortho-effect ((Table 2,

Figure 5b).36 Similar behaviour is observed in the CID spectra of [M-H]- ion

from compounds 10-12 (Table 2) and typical mass specta shown for [M-H]- ion

of compound 12 (Figure 5b).

Figure 5: a) source fragmentation of [12-H+]- at 50 eV cone voltage,

b) CID spectra of [12-H]- at 20 eV collision energy of compound 12.

The high resolution (10000 FWHM) spectrum recorded for 10-12

confirmed that the ion at m/z 92 correspond to the elemental composition of

C6H4O (Figure 6). The high resolution data suggests that the ion at m/z 92

might be formed by the loss of (.CHO) from [M-H]- ion of compound 12 (Figure

6).

*

*


161

Figure 6. ESI-high resolution mass spectrum of compound 12.

3.1. ION MOLECULE REACTIONS IN THE COLLISION CELL WITH CO2

In chapter 3, we have demonstrated that the carbanions can be

characterized by their selective ion-molecule reactions with CO2 in the

collision cell of a triple quadrupole mass spectrometer. In the ion-molecule

experiments the ions of interest are selected using MS1 and allowed to undergo

ion-molecule reactions in the collision cell with the ligand of interest

introduced into the collision cell. The resultant product ions are then analyzed

by scanning MS2. Hence, we decided to perform the ion-molecule reactions of

radical anions with CO2 to check whether the radical anions selectively react

with CO2 as found with the carbanions.

*


162

Figure 7: The Ion-molecule reactions mass spectra of m/z 92, [(12-CHO)-NO]-.

with CO2 in the collision cell.

Wenthold et al.26 used ion-molecule reactions with CO2 as a part to

characterize ortho-, meta-, and para- benzyne radical anions, in which the

carbanion part of radical anion reacting with CO2 to produce corresponding

distonic benzoate species keeping the radical site intact. However, in the

present systems, distonic dehydrophenoxide radical anions (m/z 92) are

generated wherein the negative charge is on oxygen and the radical site is on

benzene ring. If this is the structure, when ion-molecule reactions are

performed on these radical anions (m/z 92), it should not react with CO2 by the

known fact that only the carbanions react with CO2 but not phenoxide anions.

To check this proposal we have performed ion-molecule reactions of the ion at

m/z 92 generated from all the sources (1-9 and 11-12) using CO2 as the

collision gas. Surprisingly, all the ions at m/z 92 react with CO2 with varied

yields of the products corresponding to the addition of CO2. Typical ion-


163

molceule reaction mass spectrum of [12-H-CHO.]-., m/z 92, is shown in Figure

7. Addition of CO2 to the studied dehydrophenoxide radical anions suggests

that to some extent these ions exist as phenide anion (negative charge on

benzene ring) keeping the radical site on oxygen (Scheme 20).

-ve ESI

HO

OH

HO

O-O-

.

O.

-

12 m/z 121

m/z 92 m/z 92

Scheme 20

However, it is not possible to prove the existence of dehydrophenoxide

radical anion with the available experimental data. This is because of there are

no reagents that can react selectively with radical site of dehydrophenoxide

radical anions. The conversion of dehydrophenoxide radical anion to phenide

type structure may be possible during the ESI process, however, further

experiments need to be done to confirm this process.


164

4. CONCLUSIONS

The intricacy in the formation of distonic dehydrophenoxide radical

anions in the presence of solvent molecules has been studied by electrospray

ionization mass spectrometry. We could generate the distonic

dehydrophenoxide radical anions by the ESI process from suitably substituted

aromatic compounds at higher excitation energies (source induced

dissociation). This method is an alternative method to synthesize distonic

dehydrophenoxide radical anions from isomeric nitrobenzoic acids as well as

from substituted phenols. Formation of these distonic radical anions are also

observed in the collision induced dissociation of the [M-H]- ions. Ion-molecule

reaction experiments with CO2 are also attempted to check the reactivity of

radical anions with CO2. The results show that the radical anions do react with

CO2 to some extent, and that can only be explained by the structure where the

negative charge is on the benzene ring and radical site on phenol oxygen.

Chapter 4.2 Radical anions under ESI and APCI…

165


PPPAAARRRTTT 222

1. Prologue

he expulsion of NO, CHO, CH3 and NO2 as radicals led to a variety of

distonic radical anions in the electrospray ionization experiments from

deprotonated anion, [M-H]- ion of benzoic acids and substituted phenols.

Interestingly, nitrophenols yield distonic dehydrophenoxide radical anion by

the loss of NO2. from [M-H]- ion under ESI conditions, whereas the same

dehydrophenoxide radical anions can also be generated from nitrobenzoic

acids by the sequential loss of CO2 and NO. radical from [M-H]- ion. However,

there are no reports on the generation of dehydrophenoxide radical anions

from other substituted nitrobenzenes. Hence, we moved to perform ESI

experiments on isomeric nitrobenzaldehydes and nitroacetophenones towards

generating dehydrophenoxide radical anions. But when ESI experiments are

performed on these compounds, they did not yield any characteristic ion. It

suggests that ESI not amenable for these compounds; that could be due lack of

acidic proton in the selected analytes. There have been reports in the literature

covering the successful analysis of such nitroaromatic compounds under APCI

T

GGeenneerraattiioonn ooff DDiissttoonniicc ddeehhyyddrroopphheennooxxiiddee rraaddiiccaall aanniioonnss

ffrroomm ssuubbssttiittuutteedd pphheennoollss uunnddeerr aattmmoosspphheerriicc pprreessssuurree cchheemmiiccaall

iioonniizzaattiioonn ccoonnddiittiioonnss


166

conditions, where ESI is not amenable. Hence we moved to study the

substituted nitrobenzoic acids under APCI conditions.

1.1. ATMOSPHERIC PRESSURE CHEMICAL IONIZATION (APCI)

Atmospheric pressure chemical ionization (APCI) is one of atmospheric

pressure ionization techniques used in mass spectrometry (Figure 8). In this

ionization the analyte solution is passed through a heated capillary surrounded

with nitrogen flow used for nebulization, and here the analyte and solvent

molecules are vaporized. The mixture of vaporized solvent and analyte flows

towards the ion formation region. In this region, there is a metallic needle

called ‘corona discharge needle’, which is at potential of a few kilovolts. This

needle causes partial discharge around a conductor and leads to ionization and

electrical breakdown of the atmosphere immediately to the surrounding

conductor.

Figure 8

This effect is known as ‘corona discharge’. Since the atmosphere

surrounding the corona electrode consists mainly in the vapour generated from


167

the solvent, nitrogen gas, and the analyte molecules, the corona discharge

initiates chemical ionization at atmospheric pressure using the vaporized

solvent as the reagent gas. Both positive and negative coronas rely on electron

avalanches.

The APCI technique can be applied to volatile and thermally labile

compounds. Non-polar compound can also be analyzed by this technique.

APCI is a soft ionization technique, but unlike ESI, APCI usually does produce

some degree of fragmentation that is useful for structural characterization.

Although use of APCI is not yet as widespread as ESI, the number of reported

applications of APCI-MS is burgeoning. The number of recent applications

reported in the literature warrants an updated chronicle of APCI-MS for

analysis of lipids, where ESI-MS is not amenable. In the absence of acidic

hydrogens, the analytes might undergo dissociate electron capture reaction

under negative APCI conditions and result in a fragment ions formed by the

loss of a radical.37 Here, the APCI interface served as a source of low-energy

thermal electrons which are generated when electrons from the corona

discharge needle interact with the nebulizer gas. However, the negative ion

APCI analysis of nitroaromatic compounds result molecular radical anion,

which can only be explained on the basis of a non-dissociative electron capture

process. The electron capture occurs owing to the strongly electronegative

nitro group of the analyte.



168

Though ESI technique is not amenable to study the isomeric

nitrobenzaldehydes and nitroacetophenones, they can be analyzed under

negative ion APCI conditions. Loss of NO˙ from the molecular ions of

nitroaromatic compounds generated under EI conditions was reported using a

tandem sector mass spectrometer.38,39 Hence, the isomeric nitrobenzaldehydes

and nitroacetophenones could give molecular radical anions under APCI

conditions and upon further loss of NO radical they may form the carbonyl

substituted phenoxide anions, the precursor for dehydrophenoxide radical

anion. Hence, in this study we have attempted to ionize these isomeric

compounds to study whether they can produce dehydrophenoxide radical

anions. It is also decided to extend APCI experiments on the compounds that

are successful under ESI conditions to check whether the ions of interest

formed under ESI process can also be formed under APCI conditions.


In this part, two groups of isomeric substituted nitrobenzenes (13-18), i.e.

ortho-, meta- and para- nitrobenzaldehydes (13-15) and ortho-, meta- and para-

nitroacetophenones (16-18) were selected to study their source fragmentation

under APCI conditions. The molecular structures of the compounds 13-18 are

shown in Scheme 21.


169

NO2

CHO

NO2

CHO

NO2

CHO

NO2

COCH3

NO2

COCH3

NO2

COCH3

13 14 15

16 17 18

Scheme 21

3.1. ISOMERIC NITROBENZALDEHYDES (13-15)

The negative ion APCI mass spectra of 13-15 show abundant molecular

anions M¯˙ at low cone voltage (10 eV). While increasing the cone voltage the

mass spectra of all the compounds include characteristic fragment ions (Figure

9). Formation of the NO2- ion (m/z 46) is found to be dominant in all the

compounds. In addition the compounds 13-15 show the fragment ion at m/z 121

corresponding to the loss of NO. radical from molecular anion. Such loss of NO.

radical is reported earlier in the collision-induced dissociation studies on

nitroaromatic compounds. The compound 14 and 15 show specific ion at m/z 92

corresponding to the loss of CHO. radical from the ion at m/z 121. Whereas, the

compound 13 shows a specific fragment ion at 93, which is formed by the loss

of CO from the ion at m/z 121 due to ortho- effect as observed in the case of

ortho-hydroxyl benzaldehyde (Chapter 4, part 1). To support the formation of


170

the above characteristic fragment ions from molecular anions, we have

performed CID experiments on M-. ions of 13-15. The CID spectra are shown in

Figure 10. The results obtained in CID experiments are similar to those

obtained in the source fragmentation. Appearance of the specific fragment ion

at m/z 92 from 14 and 15 confirms the formation of dehydrophenoxide radical

anion (Scheme 22). When ion-molecule reactions with the carbondioxide are

performed in the collision cell for the ion at m/z 92, it resulted in product ion

corresponding to addition of CO2. Hence, it can be confirmed that the ion at

m/z 92 generated from 14-15 under APCI conditions is similar to that observed

from substituted phenols and nitrobenzoic acids under ESI conditions.


171

Figure 9: Negative ion APCI spectra of para- Nitrobenzaldehyde (15) at

different cone voltage values: (a) 10, (b) 20, (c) 30 and (d) 50 eV.

*

*


172

Figure 10: CID mass spectra of [M]-. of a) 13 b) 14 and c) 15 at 20 eV collision

energy.

*

*


173

NO2

NO2

CHO

CHO

-Ve APCI

-Ve APCI

O-

O-

CHO

CHO

-CHO.

O-

.

O-

.

NO2

NO2

CHO

CHO

-.

-.

-NO.

-NO.

14

15

m/z 92

m/z 92

-CHO.

Scheme 22

3.2. ISOMERIC NITROACETOPHENONES (16-18)

Like nitrobenzaldehydes, the negative ion APCI mass spectra of

nitroacetophenones (16-18) show abundant molecular anions M¯˙ at low cone

voltage (10 eV), and include fragment ions in the spectra recorded at higher

cone voltage (Figure 11). All the spectra show the ion at m/z 46 corresponding

to NO2- as the major fragment ion. All the compounds yield [M-NO]- ion at m/z

135. The compound 17 and 18 show specific ion at m/z 92 corresponding to the

loss of .COCH3 radical from the [M-NO]- ion. As expected, the ortho- compound

16 show a specific fragment ion at 93, which is formed by the loss of COCH2

from the [M-NO]- ion due to ortho-effect.36 Formation of the above characteristic

fragment ions from molecular anions is further supported by the CID spectra of

M-. ions of 16-18 (Figure 12). The CID spectra resulted in similar fragment ions


174

as observed in their source fragmentation. Formation of dehydrophenoxide

radical anion 17 and 18 could be confirmed by the appearance of the specific

fragment ion at m/z 92 and their ion-molecule reactions with the carbondioxide

in the collision cell.

With these encouraging results from nitrobenzaldehydes and

nitroacetophonones towards successful generation of distonic

dehydrophenoxide radical anions under APCI conditions, we moved to extend

the APCI experiments to test whether such radical anions can be generated

from the nitrobenzoic acids (1-3) and substituted phenols (4-12) (Chapter 4,

part 1) that produced the dehydrophenoxide radical anions under ESI

conditions. Unlike nitrobenzalidehydes and nitroacetophenones, these

compounds exclusively produce [M-H]- ions at lower cone voltage. As

expected, the source spectra recorded at higher cone voltage and the CID

spectra of [M-H]- ion show similar set of fragment ions as that found under ESI

conditions. It suggests that the behavior of these compounds under APCI

conditions is similar to that observed under ESI conditions.


175

Figure 11: Negative ion APCI spectra of para- Nitroacetophenone (18) at

different cone voltage values: (a) 10, (b) 20, (c) 30 and (d) 50 eV.

*

*


176

Figure 10: CID mass spectra of [M]-. of a) 16 b) 17 and c) 18 at 20 eV collision

energy.


177

4. CONCLUSION

The APCI mass spectrometry has been successfully used to study

nitrobenzaldehydes and nitroacetophenones, which are not amenable under

ESI conditions. Under APCI conditions the studied compounds form M-. ion, and

upon source fragmentation/CID they result in [M-NO]- ion. Further

fragmentation of the [M-NO]- of ortho-isomers specifically show loss of a neutral

(CO or COCH2) to yield the fragment ion at m/z 93. The [M-NO]- of meta- and

para- isomers further show a radical loss (.CHO or .COCH3) to generate

dehydrophenoxide radical anion (m/z 92). The gas-phase ion-molecule

reactions of these distonic radical anions with CO2 in the collision cell show

their reactivity to yield the expected product ions. The dehydrophenoxide

radial anions from nitrobenzoic acids and substituted phenols can also be

generated under APCI conditions from their [M-H]- ions as obtained in the ESI

experiments.

Chapter 4 Radical anions under ESI and APCI…

178

5. EXPERMENTAL

All the studied compounds (1-18) are commercially available (Sigma-

Aldrich, St. Louis, USA) and were used without further purification. HPLC grade

solvents (water and methanol) were purchased from Merck (Mumbai, India).

All the experiments were carried out on a Quattro LC, triple quadrupole mass

spectrometer (Micromass, Manchester, UK) equipped with an ESI and APCI

source. Spectra were acquired using Masslynx software (version 3.2). The

typical operating conditions were: capillary voltage, -3.5 kV; cone voltage, 10-

50 V. Based on our previous experience in increasing the survival of the anions

from solvent environment in ESI process, we set higher source and dissolvation

temperature (1500 and 3000 C, respectively). Nitrogen was used as the

nebulization and desolvation gas. Stock solutions (1mM) of all the compounds

were made in 50:50 (vol/vol) water:methanol and working standards were

prepared by diluting the stock solutions with appropriate volumes of 50:50

(vol/vol) water:methanol to achieve a final concentration of 100µM. The sample

solutions were introduced using an infusion pump (Harward apparatus, Kent,

UK) at a flow rate of 10µL/min. For ion-molecule reaction experiments, CO2 gas

was used as the collision gad and introduced into the collision cell so that the

pressure inside the collision cell was 9 x 10-4 mbar; the collision energy set at

zero eV. For collision-induced dissociation (CID) experiments argon was used

as collision gas, and the pressure was maintained at 3 x 10-4 mbar. All the

experimental data reported here are an average of 20 scans. High resolution

mass spectra were recorded on a QSTAR XL mass spectrometer (Applied


179

Biosystems/MDS Sciex,CA, USA) operating capillary voltage 4-5 kV;

declustering potential, 50 V. Nitrogen was used as the curtain and collision gas

for CID experiments.


180

6. REFERENCE

1. Born M, Ingemann S, Nibbering NMM, Mass Spectrometry Rev., 1997; 16:

181.

2. Gronert S . Chem. Rev., 2001; 101: 329.

3. Ashby EC. Acc. Chem. Res., 1988; 21: 414.

4. Bunnett JF. Acc. Chem. Res., 1978; 11: 413.

5. Eberson L. “Electron Transfer Reactions in Organic Chemistry,” Springer:

Berlin, 1987.

6. Rossi RA, Rossi RH. ‘‘Aromatic substitution by the SRN1 mechanism,’’ ACS

Monograph 178; American Chemical Society: Washington, D.C., 1983.

7. Saveant, JM. Acc. Chem. Res., 1993; 26: 455.

8. Laurent A. Ann. Chem. 1836; 17: 88.

9. Lee J, Grabowski J. J. Chem. Rev., 1992; 92: 1611.

10. Bowie JH, Williams BD. ‘‘Negative ion mass spectrometry of organic,

organometallic and coordination compounds,’’ Phys. Chem. Ser. Two (Mass

Spectrometry), Maccoll, A., Ed.; Butterworths: London, 1975, Vol. 5, pp. 89.

11. Bowie JH. Mass. Spectrom. Rev. 1984; 3: 161.

12. Echinger PCH, Dua S, Bowie JH. 1997; 11: 1996.

13. Kebarle P, Chowdhury S. Chem. Rev. 1987; 87: 513.

14. Christophorou LG, Siomos K. ”Electron-Molecule Interactions and Their

Applications” Ed. Christophorou LG., Ed.; Academic Press: New York, 1984,

Vol. 2, Chapt. 4, pp. 222.

15. Nibbering NMM. Recl. Trav. Chim. Pays-Bas, 1981; 100: 297.


181

16. McDonald RN, Chowdhury AK. J. Am. Chem. Soc. 1985; 107: 4123.

17. Jordan KD, Burrow PD. Chem. Rev. 1987; 87: 557.

18. Yates BF, Bouma WJ, Radom L. J. Am. Chem. Soc., 1984; 106: 5805.

19. Yates BF, Bouma WJ, Radom L. Tetrahedron, 1986; 42: 6225.

20. Hammerum S. Mass Spectrom. Rev., 1988; 7: 123.

21. Bouchoux G. Mass Spectrom. Rev. 1988; 7: 1.

22. Stirk KM, Kiminkinen LKM, Kenttamaa HI. Chem. Rev., 1992; 92: 1649.

23. Kenttamaa HI. Org. Mass Spectrom., 1994; 1: 1.

24. Smith RL, Chou P, Kenttamaa HI. “In The Structure, Energetics and Dynamics

of Organic Ions,” Baer T, Ng CY, Powis I. Eds.; Wiley: New York, 1996.

25. Reeks LB, Eichinger PCH, Bowie JH. Rapid Commun. Mass Spectrom.1993; 7:

286.

26. Wenthold PG, Squires RR. J. Am. Chem. Soc., 1994; 116: 6401.


28. Wenthold PG, Hu J, Hill BT, Squires RR. Int. J. Mass Spectrom., 1998; 180:

173.

29. Reed DR, Hare MC, Fattahi A, Chung G, Gordon MS, Kass SR. J. Am. Chem.

Soc. 2003; 125: 4643.

30. Reed DR, Hare MC, Kass SR. J. Am. Chem. Soc. 2000; 122: 10689.

31. Hunt DF, Shabanowitz J, Giordani AB. Anal. Chem., 1980; 52: 386.

32. Crews P, Rodriguez J, Jaspars M. “Organic Structure Analysis” Oxford

University Press: Oxford, 1998, pp. 246.

33. Wenthold PG, Hu J, Squires RR. J. Mass Spectrom. 1998; 3: 796.


182


35. Wenthold PG, Squires RR. Int. J. Mass Spectrom. Ion Proc., 1998; 175: 215.

36. Attyagalle A, Ruzicka J, Varughese D, Sayed J. Tet. Lett, 2006; 47: 4601.

37. Song Y, Cooks RG. Rapid Commun. Mass Spectrom. 2006; 20: 3130.

38. Bowie JH, Blumenthal T , Walsh I. Org. Mass Spectrom., 1971; 5: 777.

39. McLuckey SA, Glish GL, Org. Mass Spectrom., 1987; 22: 224.

183

ABSTRACT

i

The thesis entitled “Gas phase studies of metal complexes, isomeric

carbanions and distonic radical anions under soft ionization mass spectral

conditions” is organized into four chapters.

CHAPTER 1: Gas phase ion Chemistry of CrIII(Salen) complex under

electrospray ionization conditions.

CHAPTER 2: Proton and alkali metal ion affinities of bidentate bases:

spacer chain length effects.

Part 1: Proton and alkali metal ion affinities of α,ω-diamines: spacer chain

length effects.

Part 2: Proton and alkali metal ion affinities of α,ω-diols: spacer chain length

effects.

CHAPTER 3: Generation of regiospecific carbanions under electrospray

ionization conditions and characterization by ion-molecule reactions with

carbon dioxide.

Part 1: Generation of regiospecific carbanions from aromatic hydroxy acids

and dicarboxylic acids.

Part 2: Generation of regiospecific carbanions from sulfobenzoic acids.

CHAPTER 4: Generation of distonic dehydrophenoxide radical anions

under electrospray and atmospheric pressure chemical ionization

conditions.

Part 1: Generation of distonic dehydrophenoxide radical anions from

substituted phenols under electrospray ionization conditions.

Part 2: Generation of distonic dehydrophenoxide radical anions from

substituted nitrobenzenes under atmospheric pressure chemical ionization

conditions.

ii

CHAPTER 1: Gas phase ion Chemistry of Chromium-Salen complex under

electrospray ionization conditions.

The gas-phase coordination behavior of the [CrIII(Salen)]PF6 complex at

the free axial positions has been studied in the presence of amines

anddiamines under ESI conditions. The positive ion ESI mass spectrum of

[CrIII(Salen)]+ complex in acetonitrile (ACN) shows stable five- and six-

coordinated complex ions, [CrIII(Salen)(ACN)]+ and [CrIII(Salen)(ACN)2]+ ions,

at m/z 359 and 400 respectively. This result prompted us to study the

coordination chemistry of the [CrIII(Salen)]+ complex with mono- and bi-dentate

ligands (amines and diamines) in detail. The ESI mass spectrum of the

[CrIII(Salen)]+ complex in the presence of propylamine (PA) clearly

demonstrate the displacement of solvent molecules present in the axial

positions by the stronger ligand. Due to cone fragmentation of the

[CrIII(Salen)(PA)2]+ to [CrIII(Salen)(PA)]+ and [CrIII(Salen)]+ that immediately

picks up one molecule of acetonitrile, the surrounding solvent molecule, to

result in a stable six-coordinated complex ions.

In the ESI mass spectra of [CrIII(Salen)]+ and diamines [H2N-(CH2)n-NH2, n

= 2-8 (1-7)] the relative abundances of ligated complex ions, fragment ions,

and solvent adducts of fragment ions in the ESI mass spectra, were found to

depend on the ligand size used for recording the spectrum at 30 V cone

voltage. With a view to study the stability of diamine complexes with

[CrIII(Salen)]+, we also performed ligand-pickup experiments with acetonitrile

in the collision cell and collision induced dissociation (CID) experiments using

iii

Pc/Pd ratios on [CrIII(Salen)(DA)]+ ions. In all these experiments the obtained

order of stabilities of [CrIII(Salen)(DA)]+ complexes for diamines 1-7 can be

given as 2>1>3>4,7>6>5.

CHAPTER II: Proton and alkali metal ion affinities of bidentate

bases: spacer chain length effects.

Part I: Proton and alkali metal ion affinities of α,ω-diamines: spacer chain

length effects.

To evaluate the relative alkali metal ion [Li+, Na+ and K+] affinities for a

series of seven homologues α,ω-diamines, H2N-(CH2)n-NH2 (n = 2-8, 1-7) we

have used the kinetic method. The heterodimeric ions, [DA1+Li+DA2]+,

dissociate by competitive elimination of neutral diamines yielding two

fragment ions corresponding to [DA1+Li]+ and [DA2+Li]+. The relative

abundances of the resulted Li+ associated monomers, viz., I(Li+-DA1) and I(Li+-

DA2) vary and reflect the Li+ ion affinity of individual diamine.

[DA1--Li+--DA2]

DA2 + DA1Li+

DA2Li+ + DA1

Alkali metal ions and proton affinity ladder constructions

The ln[I(Li+-DA2)/I(Li+-DA1)] ratio values are used to construct relative

Li+ ion affinity ladder. In this ladder construction, most of the diamines are

compared to at least three others. From Li+ ion affinity ladder, the relative Li+

ion affinity order for α,ω-diamines can be given as, 1Li+ < 3Li

+ ≤ 2Li

+ < 4Li

+ <

iv

6Li+ < 5Li

+ ≤ 7Li

+. In the lithium affinity order for α,ω-diamines, the deviation of

compound 2 and 5 in the order indirectly suggesting that structure of the

lithiated diamine may be playing a role.

We have extended similar experiments to construct Na+, K+ and H+ ion

affinity ladders of diamines by performing similar experiments. From these

ladders, the relative Na+ ion affinity order can be given as1Na+ < 2Na+ < 3Na+ <

4Na+ < 5Na+ < 6Na+ < 7Na+, for K+ ion 2K+ < 1K+ < 3K+ < 4K+ < 6K+ < 5K+ < 7K+

and for proton 1H+ < 2H+ < 7H+ < 6H+< 5H+ < 4H+ < 3H+. Contrasting orders

were found for relative proton and alkali metal ion, when compared them with

their relative affinity orders. The observed contrasting orders for H+ and Li+ are

explained by quantum chemical calculations.

Part II: Proton and alkali metal ion affinities of α,ω-diols: spacer chain length effects.

However, we have also applied the kinetic method to measure the

relative proton and alkali metal ion affinity orders for the selected series of α,ω-

diols (HO-(CH2)n-OH, n = 2-10, 8-16). The obtained relative proton affinity

order can be given as 8H+<< 9H+<< 14H+≈ 13H+< 12H+< 11H+< 10H+< 15H+<

16H+, in which the order for 8-12 is inline with the reported proton affinities.

The relative proton affinity order of diols 8-16 is exactly correlating with the

order that obtained for corresponding diamines 8-16.

In a similar way we have constructed relative Li+, Na+ and K+ (M) affinity

ladders by replacing proton with alkali metal ion. The obtained orders can be

given as 8M+<< 9M+< 10M+< 11M+< 12M+< 13M+< 14M+< 15M+< 16M+. From

v

this metal ion affinity order of the diols, it can be noticed that the relative metal

ion affinity gradually increased with the increase in the spacer chain length

(number methylenes) of the diol. The overall proton/alkali metal ion affinity

orders of diols is almost similar to that obtained for diamines, except some

dissimilarities for the Li+ ion affinity order of diamines.

Chapter III: Generation of regiospecific carbanions under electrospray

ionization conditions and their characterization by ion-molecule reactions

with carbon dioxide.

Part 1: Generation of regiospecific carbanions from aromatic hydroxy acids and

dicarboxylic acids.

We have generated [(M-H)-CO2]- anions under high cone voltage

negative electrospray ionization mass spectral conditions, which are highly

unstable as carbanions, from isomeric carboxylic acids studied (1-17) i.e., one

set of geometrical isomers [maleic acid (1), fumaric acid (2)] and five sets of

positional isomers i) aromatic dicarboxylic acids [phthalic acid isomers, 3-5;

and pyridinedicarboxylic acid isomers, 6-9], ii) aromatic hydroxy acids,

[hydroxy benzoic acid isomers, 10-12; hydroxy nicotinic acid isomers, 13-14;

and hydroxy phenylacetic acid isomers, 15-17], [M-H]- ion by the loss of CO2

under electrospray ionization conditions. These decarboxylation products are

unstable due to the internal proton exchange from acid/hydroxy group or

external with solvent is expected. These compounds possess carbanions at

different positions (cis/trans or o/m/p) and hence, we have performed ion-

molecule reactions with CO2 in the collision cell to confirm the survival of

carbanion.

vi

The ion-molecule experiments showed that the [(M-H)-CO2]− from 2, 4, 5,

7-9, 11, 12, 14, 16 and 17 reacted with CO2 and resulted in the corresponding

carboxylate anion, and that from 1, 3, 6, 10, 13, 15 did not react. The survival of

carbanion suggests that the proton exchange is kinetically unfavorable by the

fact that the carbanion and the -COOH group are in trans/meta/para position.

These results clearly reflect the geometrical isomeric differences between the

decarboxylation product ions of 1 to 17 and their difference in internal stability.

The relative abundance of CO2 adduct ions increased by 2-25 %, when

experiments are performed at increased source/desolvation temperatures.

This phenomenon clearly confirms increased ion formation in the solvent free

region at higher source and desolvation temperatures due to reduced proton

exchange reactions between the generated carbanion and solvent molecules.

Clear support comes from the computational studies, which show that

intramolecular proton transfer is responsible for precluding the observation of

the one of the twin isomeric carbanions.

Part 2: Generation of regiospecific carbanions from sulfobenzoic acids.

We have extended our study with isomeric sulfobenzoic acids (18-20)

and disulfonic acids (21-22) to explore the stability studies of carbanions in the

presence of internal acidic protons after decarboxylation. The negative ion ESI

mass spectra of all these isomeric compounds showed common fragment ions

at m/z 157 corresponding to [(M-H)-CO2]¯ (18-20) or [(M-H)-SO3]¯ ions (21 and

22), at 30 eV cone voltage.

vii

The ion-molecule experiments showed that the [(M-H)-CO2]¯ ions from

19 and 20 reacted with CO2 and resulted in the corresponding carboxylate

anions, and that from 18 did not react. This selective reaction reveals that the

[(M-H)-CO2]¯ ions from 19 and 20 are surviving as carbanions, whereas that of

18 is undergoing isomerization through exchange of acidic proton from

COOH to carbanion. Interestingly, the [(M-H)-SO3]¯ ions from both 21 and 22

did not react with CO2, which confirms that the generated [(M-H)-SO3]¯ ions

from 21 and 22 are benzenesulfonate anions, but not carbanions.

Chapter IV: Generation of distonic dehydrophenoxide radical anions

under electrospray and atmospheric pressure chemical ionization

conditions.

Part I: Generation of distonic dehydrophenoxide radical anions from substituted

phenols under electrospray ionization conditions.

The distonic ions, which contain the charge and radical sites at different

positions, have been used in mass spectrometric studies of free radical

chemistry. In this work, we have used cone fragmentation phenomena towards

the generation of distonic isomeric dehydrophenoxide radical anions using the

cone fragmentation method.

The negative ion ESI spectra of nitro benzoic acids (1-3) showed abundant

specific fragment ions at m/z 92. With the aid of CID experiments, the fragment

ion at m/z 92 confirmed as the [(M-H)-CO2-NO]-.. Kass et al observed similar

fragment ions in the two stage SORI CID experiments on these [M-H]- ions.

viii

Appearance of the ion at m/z 92 in the source spectra of 1-3 clearly suggests

that radical anions did form by the ESI source fragmentation also.

We have also generated m/z 92 ion from substituted isomeric phenols (4-

12), i.e. o-, m- and p- isomers of nitro phenols (4-6), hydroxy benzaldehydes (7-

9) and cresols (10-12). All these isomeric compounds at high cone voltages

generating ion at m/z 92, C6H6O-. except in the case of 7, by loosing R which are

corresponds to [(M-H)-R.]-, R= -CH3 (4-6), -NO2 (8-9), -CHO (10-12),

respectively. We have also confirmed the formation of ion at m/z 92 in the CID

spectra of [M-H]- ion from the compounds 4-6, 8-12. The composition of the m/z

92 dehydrophenoxide radical anion is also supported by the high-resolution

measurements. The C6H4O-. radical anions from 2-6, 8-12 reacted with CO2 in

the ion-molecule experiments and resulted in the corresponding radical anions.

Part II: Generation of distonic dehydrophenoxide radical anions from substituted

nitrobenzenes under atmospheric pressure chemical ionization conditions.

In this part, we have studied o-, m- and p- substituted (R) nitro benzenes [R =

-CHO (13-15) and -COCH3 (16-18)]. Since nitro aromatic compounds are not

amenable in negative electrospray conditions we have gone for negative

atmospheric chemical ionization source conditions. In all cases, APCI mass

spectra, M-. molecular anions are appeared at low cone voltage (10 eV). While

increasing the cone voltage the mass spectra of all the compounds yeild

characteristic fragment ions. The compound 14 and 15; 17 and 18 show specific

ion at m/z 92 corresponding to the loss of CHO. and COCH3. radical from the ion

at m/z 121 and 135, respectively. The results obtained in CID experiments are

ix

similar to those obtained in the source fragmentation. To test these produced

dehydrophenoxide radical anions, we have gone for ion-molecule reactions

with the corbondioxide in the collision cell. As expected, similar results

obtained as in the ESI experiments.

x

List of Publications:

This thesis is based on the following papers, which are refered to in the text by

their roman numbers:

1. Co-ordination chemistry of chromium-salen complexes studied by

electrospray ionization mass spectrometry.

M. Kiran Kumar, S. Prabhakar, M. Ravi Kumar, T. Jagadeshwar Reddy, S.

Premsingh, S. Rajagopal and M. Vairamani.

Rapid Commun. mass spectrom. 2004; 10: 1103.

2. The effect of spacer chain length on ion binding to bidentate α,ϖ-diamines:

Contrasting ordering for H+ and Li+ ion affinities

M. Kiran Kumar, J. Srinivasa Rao, S. Prabhakar, M. Vairamani and G.

Narahari Sastry.

Chem. Commun., 2005: 1420.

3. Generation of regiospecific carbanions under electrospray ionisation

conditions and their selectivity in ion-molecule reactions with CO2

M. Kiran Kumar, B. Sateesh, S. Prabhakar, G. Narahari Sastry and M.

Vairamani

Rapid Commun. Mass Spectrom. . 2006; 20: 987.

4. Auxiliary approach to evaluate the isomeric decarboxylated anions from 2-,

3- and 4-sulfobenzoates in the gas phase by using ion-molecule reactions

with carbon dioxide in the collision cell.

M. Kiran Kumar, S. Prabhakar, M. Ravi Kumar and M. Vairamani

Rapid Commun. Mass Spectrom. 2006; 20: 1045.

5. Generation of dehydrophenoxide radical anions under ESI and APCI mass

spectral conditions.

M. Kiran Kumar, S. Prabhakar and M. Vairamani, Manuscript under

preparation.

xi

Papers not included in the theses:

6. Negative ion electrospray ionization mass spectral study of dicarboxylic

acids in the presence of halide ions.

M. Ravi Kumar, S. Prabhakar, M. Kiran Kumar, T. Jagadeshwar Reddy, M.

Vairamani.

Rapid Commun. Mass spectrom. 2004; 10: 1109.

7. Dissociation of gas-phase dimeric complexes of lactic acid and transition-

metal ions formed under electrospray ionization conditions; the role of

reduction of the metal ion.

M. Ravi Kumar, S. Prabhakar, M. Kiran Kumar, T. Jagadeshwar Reddy and

M. Vairamani


8. Electrospray ionization studies of transition-metal complexes of 2-

acetylbenzimidazolethiosemicarbazone using collision-induced dissociation

and ion-molecule reactions

G. Bhaskar, M. Adharvana Chary, M. Kiran Kumar, K. Syamasundar, M.

Vairamani and S. Prabhakar


xii

Poster Presentations:

“Generation and Reaction of Isomeric Carbanions under Electrospray

Ionization Mass Spectrometry” presented at CRSI, Kolkatta, February-

2005.

“The Effect on H+, Li+, Na+, K+ ions affinity orders with α,ϖ-Diols by

using ESI mass spectrometry” presented at ISMAS (Indian Society for

Mass Spectrometry), Munnar, Kerala, January-2006.

Documents

Thesis Title: Gas phase studies of metal complexes, isomeric carbanions and distonic radical anions under soft ionization mass spectral conditions