MASS SPECTROMETRIC STUDIES ON PETROLEUM ASPHALTENES …

Purdue UniversityPurdue e-Pubs

Open Access Dissertations Theses and Dissertations

January 2015

MASS SPECTROMETRIC STUDIES ONPETROLEUM ASPHALTENES ANDORGANOSULFUR COMPOUNDS, ONFUNCTIONAL-GROUP SELECTIVE ION-MOLECULE REACTIONS AND ON GAS-PHASE REACTIVITY OF META-BENZYNESTOWARD AMINO ACIDSWeijuan TangPurdue University

Follow this and additional works at: https://docs.lib.purdue.edu/open_access_dissertations

This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] foradditional information.

Recommended CitationTang, Weijuan, "MASS SPECTROMETRIC STUDIES ON PETROLEUM ASPHALTENES AND ORGANOSULFURCOMPOUNDS, ON FUNCTIONAL-GROUP SELECTIVE ION-MOLECULE REACTIONS AND ON GAS-PHASEREACTIVITY OF META-BENZYNES TOWARD AMINO ACIDS" (2015). Open Access Dissertations. 1199.https://docs.lib.purdue.edu/open_access_dissertations/1199

https://docs.lib.purdue.edu?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F1199&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F1199&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/etd?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F1199&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F1199&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations/1199?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F1199&utm_medium=PDF&utm_campaign=PDFCoverPages

PURDUE UNIVERSITY GRADUATE SCHOOL

Thesis/Dissertation Acceptance

To the best of my knowledge and as understood by the student in the Thesis/Dissertation Agreement, Publication Delay, and Certification/Disclaimer (Graduate School Form 32), this thesis/dissertation adheres to the provisions of Purdue University’s “Policy on Integrity in Research” and the use of copyrighted material.

Weijuan Tang

MASS SPECTROMETRIC STUDIES ON PETROLEUM ASPHALTENES ANDORGANOSULFUR COMPOUNDS, ON FUNCTIONAL-GROUP SELECTIVE ION-MOLECULEREACTIONS AND ON GAS-PHASE REACTIVITY OF META-BENZYNES TOWARD AMINO ACIDS

Doctor of Philosophy

Hilkka I. Kenttämaa

Scott A. McLuckey

Mahdi Abu-Omar

Hilkka I. Kenttämaa

Chengde Mao

R. E. Wild 04/30/2015

MASS SPECTROMETRIC STUDIES ON PETROLEUM ASPHALTENES AND

ORGANOSULFUR COMPOUNDS, ON FUNCTIONAL-GROUP SELECTIVE ION-

MOLECULE REACTIONS AND ON GAS-PHASE REACTIVITY OF META-

BENZYNES TOWARD AMINO ACIDS

A Dissertation

Submitted to the Faculty

of

Purdue University

by

Weijuan Tang

In Partial Fulfillment of the

Requirements for the Degree

of

Doctor of Philosophy

May 2015

Purdue University

West Lafayette, Indiana

ii

To my husband, Huaming Sheng

To my parents, Yi Tang and Yali Gan

To my parents in law, Tiansheng Sheng and Tanni Wang

iii

ACKNOWLEDGEMENTS

First and foremost, I would like to express my deepest gratitude to my advisor,

professor Hilkka I. Kenttämaa. She not only provided me a wonderful opportunity to join

her research group, but also gave me guidance, inspiration, encouragement and support

throughout my PhD study. I was able to follow my passion and explore a wide variety of

exciting research projects in mass spectrometry. More importantly, I learned how to think,

how to solve problems, and how to better communicate and collaborate with different

people. She helped me grow into an independent scientist who is ready to tackle

challenges ahead. I am forever grateful to be her student, and I truly appreciate all her

help along the way. Without her guidance and support, I would have never become who I

am today.

Special thanks go to Dr. John Nash for his help in quantum chemical calculations

on the mono- and biradicals studied in my thesis and all the great discussions. I would

also like to thank Mark Carlsen, Dr. Hartmut Hedderich, and members of the Jonathan

Amy Facility, for helping me repair an FT-ICR instrument (NEL) that was down for two

years. Many thanks go to my thesis committee members, Dr. Scott A. McLuckey, Dr.

Mahdi Abu-Omar and Dr. Chengde Mao, for valuable discussions. I also want to thank

Dr. McLuckey for writing me recommendations for a PhD internship.

iv

I would like to express my gratitude to the past and present group members in Dr.

Kenttämaa’s group. Special thanks go to Dr. Jinshan Gao and Dr. Benjamin Owen who

trained me on ICR and LQIT instruments when I first joined the group; to Dr. Matthew

Hurt and Dr. David Borton who mentored me on petroleum projects; to Dr. Huaming

Sheng and Dr. James Riedeman who provided lots of great suggestions to my research

projects and helped with mechanistic studies. I also want to thank Dr. Ashley Wittrig, Dr.

Tiffany Jarrell, Dr. Peggy Williams, Dr. Fanny Widjaja, Dr. Enada Archibold, Dr. Linan

Yang, Dr. Nelson Vinueza, Dr. Vanessa Gallardo, Alex Dow, Chris Marcum, Guannan Li,

Priya Murria, John Degenstein, Chunfen Jin, Hanyu Zhu, Joann Max, John Kong,

Xueming Dong, Xin Ma, Mingzhe Li, Ravikiran Yerabolu, Mark Romanczyk, Laurance

Cain, Raghu Kotha, Babu Mistry, Rashmi Kumar, and Yuyang Zhang. Without their

feedback, friendship and generous help, my graduate study would not have been so

enjoyable.

I would like to thank Dr. Daniel Raftery for introducing me to the interesting area

of metabolomics. I am grateful for all the mentorship and help that I received from him

and the group members. I also want to extend my thanks to the people whom I have

worked with during my internship, including Dr. Vincent Asiago, Dr. Jan Hazebroek, Dr.

Cathy Zhong, Dr. Bruce Orman, Chris Vlahakis and Teresa Harp.

Lastly, but most importantly, I would like to thank my loving family for always

being there to support me. My wonderful husband, Huaming Sheng, has given me endless

support, understanding and encouragement. I am very grateful to have met him in China

because of our common goal to pursue a graduate degree at Purdue. It has been enjoyable

and fruitful five years in graduate school. The most sincere thanks go to my dear parents,

v

Yi Tang and Yali Gan, for all the years of unconditional love, caring, guidance and

encouragement. I also wish to thank my parents-in-law, Tiansheng Sheng and Tanni

Wang, who constantly encourage me and treat me as their own daughter.

vi

TABLE OF CONTENTS

Page

LIST OF TABLES ............................................................................................................. ix

LIST OF FIGURES ........................................................................................................... xi

LIST OF SCHEMES........................................................................................................ xiv

ABSTRACT ................................................................................................................... xvi

LIST OF PUBLICATIONS ............................................................................................. xix

CHAPTER 1. INTRODUCTION AND OVERVIEW ....................................................... 1

1.1 Introduction .............................................................................................................. 1 1.2 Overview .................................................................................................................. 3 1.3 References ................................................................................................................ 4

CHAPTER 2. THEORY, INSTRUMENTATION AND EXPERIMENTAL ASPECTS OF LINEAR QUADRUPOLE ION TRAP (LQIT) AND FOURIER TRANSFORM ION CYCLOTRON RESONANCE (FT-ICR) MASS SPECTROMETRY ................................................................. 6

2.1 Introduction .............................................................................................................. 6 2.2 Ionization Methods .................................................................................................. 7

2.2.1 Electron Ionization (EI) .................................................................................. 8 2.2.2 Chemical Ionization (CI) ................................................................................ 8 2.2.3 Electrospray Ionization (ESI) ......................................................................... 9 2.2.4 Atmospheric Pressure Chemical Ionization(APCI) ...................................... 11

2.3 Linear Quadrupole Ion Trap (LQIT) Mass Spectrometry ...................................... 14 2.3.1 Introduction ................................................................................................... 14 2.3.2 Instrument Overview .................................................................................... 14 2.3.3 Ion Motions in LQIT ..................................................................................... 18

2.3.3.1 Radial Motion ..................................................................................... 19 2.3.3.2 Axial Motion ....................................................................................... 23

2.3.4 Ion Excitation and Detection ........................................................................ 24

vii

Page

2.3.5 Tandem Mass Spectrometry ......................................................................... 27 2.3.5.1 Ion Isolation ........................................................................................ 27 2.3.5.2 Collision-activated Dissociation (CAD) ............................................. 29

2.3.6 Ion-molecule reactions in LQIT .................................................................... 30 2.4 Fourier-Transform Ion Cyclotron Resonance (FT-ICR) Mass Spectrometry ........ 32

2.4.1 Introduction ................................................................................................... 32 2.4.2 Instrument Overview .................................................................................... 32 2.4.3 Ion Motions in FT-ICR ................................................................................. 36

2.4.3.1 Cyclotron Motion ................................................................................ 36 2.4.3.2 Trapping Motion ................................................................................. 38 2.4.3.3 Magnetron Motion .............................................................................. 40

2.4.4 Ion Manipulations in FT-ICR ....................................................................... 43 2.4.4.1 Ion Transfer ......................................................................................... 45 2.4.4.2 QuadrupolarAxialization (QA) ........................................................... 46 2.4.4.3 Ion Excitation and Detection .............................................................. 47 2.4.4.4 Ion Isolation ........................................................................................ 50 2.4.4.5 Collision-activated dissociation (CAD) in FT-ICR ............................ 52

2.5 Fundamental Aspects of Gas-phase Ion-Molecule Reactions ............................... 53 2.5.1 Brauman’s Double-Well Potential Energy Surface ...................................... 53 2.5.2 Kinetics of Ion-Molecule Reactions ............................................................. 56

2.6 References .............................................................................................................. 60

CHAPTER 3. STRUCTURAL COMPARISON OF ASPHALTENES OF DIFFERENT ORIGINS BY USING MULTIPLE-STAGE TANDEM MASS SPECTROMETRY ...................................................... 65

3.1 Introduction ............................................................................................................ 65 3.2 Experimental Section ............................................................................................. 67 3.3 Results and Discussion .......................................................................................... 68 3.4 Conclusions ............................................................................................................ 77 3.5 References .............................................................................................................. 78

CHAPTER 4. CHARACTERIZATION OF ORGANOSULFUR MODEL COMPOUNDS RELEVANT TO FOSSIL FUELS BY USING HIGH-RESOLUTION TANDEM MASS SPECTROMETRY ................. 82

4.1 Introduction ............................................................................................................ 82 4.2 Experimental Section ............................................................................................. 85 4.3 Results and Discussions ......................................................................................... 86 4.4 Conclusions .......................................................................................................... 106 4.5 References ............................................................................................................ 107

viii

Page

CHAPTER 5. GAS-PHASE ION/MOLECULE REACTIONS FOR THE IDENTIFICATION OF SULFONE FUNCTIONALITIES IN PROTONATED ANALYTES IN A LINEAR QUADRUPOLE ION TRAP MASS SPECTROMETER ........................................................... 110

5.1 Introduction .......................................................................................................... 110 5.2 Experimental Section ........................................................................................... 112 5.3 Results and Discussions ....................................................................................... 114 5.4 Conclusion ........................................................................................................... 123 5.5 Reference ............................................................................................................. 124

CHAPTER 6. GAS-PHASE REACTIVITY OF META-BENZYNES TOWARD AMINO ACIDS ....................................................................................... 126

6.1 Introduction .......................................................................................................... 126 6.2 Experimental Section ........................................................................................... 130 6.3 Results and Discussion ........................................................................................ 134 6.4 Conclusions .......................................................................................................... 155 6.5 References ............................................................................................................ 157

VITA ............................................................................................................................... 161

ix

LIST OF TABLES

Table .............................................................................................................................. Page

Table 3.1 MWD and AVG MW of Molecules in the Six Asphaltenes Samples, and Structural Information for the Eight Selected Ions .......................................... 75

Table 4.1 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Thiophenes ....................................................................................................... 98

Table 4.2 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Thiols ............................................................................................................... 99

Table 4.3 MS2 and MS3 CAD Product Ions (with Relative Abundances) for ................ 100

Table 4.4 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Sulfides ........................................................................................................... 101

Table 4.5 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Disulfides ....................................................................................................... 102

Table 4.6 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 CAD Product Ions for Ionized Thiols ........................................... 103 Table 4.7 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm)

of MS2 and MS3 CAD Product Ions for Ionized Polyaromatic Sulfur Compounds .. 103 Table 4.8 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 and MS3 CAD Product Ions for Ionized Sulfides ................................................ 104 Table 4.9 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 and MS3 CAD Product Ions for Ionized Disulfides ............................................ 105

Table 5.1 Reaction products (m/z values and branching ratios) and efficiencies for reactions of protonated sulfones and sulfoxides with TMP (PA = 222.2 kcal/mola). ...................................................................................................... 118

x

Table Page

Table 5.2 Reaction products (m/z values and branching ratios) and efficiencies for reactions between protonated N-oxides, ketones, hydroxylamines, carboxylic acids, aliphatic and aromatic amines with TMP (PA = 222.2 kcal/mola). ...................................................................................................... 120

Table 5.3 Reaction products (m/z values and branching ratios) and efficiencies for reactions of protonated sulindac and sulindac sulfone with TMP (PA = 222.2 kcal/mola). ............................................................................................ 122

Table 6.1 Reaction efficiencies (Eff.) and product branching ratios for biradiucals a – d upon reaction with tetrahydrofuran, allyl iodide, dimethyl disulfide, tert-butyl isocyanide, and cyclohexane; secondary products are noted as (2o) and are listed after the primary products that produce them ................... 148

Table 6.2 Reaction efficiencies (Eff.) and product branching ratios for monoradicals upon reaction with L-glycine, L-leucine, L-proline, L-lysine, DL-lysine-ε-15N. ................................................................................. 150

Table 6.3 Reaction efficiencies (Eff.) and product branching ratios for biradiucals a – d upon reaction with glycine, L-leucine, L-lysine, and DL-lysine-ε-15N. .. 151

Table 6.4 Reaction efficiencies (Eff.) and product branching ratios for biradiucals a – d upon reaction with L-methionine and L-cysteine. ................................... 153

Table 6.5 Reaction efficiencies (Eff.) and product branching ratios for biradiucals a – d upon reaction with L-proline and L-phenylalanine. ................................ 154

xi

LIST OF FIGURES

Figure ............................................................................................................................. Page

Figure 2.1 Examples of two chemical reactions used to ionize a radical precursor (R). ..................................................................................................................... 9

Figure 2.2 Depiction of the ESI process in positive ion mode. ........................................ 11

Figure 2.3 Depiction of the APCI process in positive ion mode. ..................................... 12

Figure 2.4 Schematic of the Thermo Scientific LQIT mass spectrometer with operating pressure indicated for each region of the instrument. ...................... 15

Figure 2.5 Components of the API stack and ion guides to which a downhill ................. 17

Figure 2.6 Schematic of the Thermo Scientific LQIT mass analyzer. .............................. 18

Figure 2.7 The oscillating RF potentials applied to the x- and y-axis pairs of rods ......... 19

Figure 2.8 The most well-defined stability region of Mathieu stability diagram for LQIT. The circles of different sizes represent ions of different masses. Ions that fall outside of this region have unstable motions in x and/or y directions as indicated. ..................................................................................... 22

Figure 2.9 Diagram of DC potential well for trapping ions in the center along the z-axis. .................................................................................................................. 23

Figure 2.10 Illustration of ion ejection from LQIT at (a) = 0.908 during mass selective instability scan, and (b) = 0.880 during resonance ejection. ........ 25

Figure 2.11 Depiction of the ion detection system in Thermo Scientific LQIT. .............. 26

Figure 2.12 The sequence of events for ion isolation and fragmentation by collision-activated dissociation (CAD). The ion of interest is represented by the green circle. ........................................................................................... 28

Figure 2.13 Tailored RF waveform for ion isolation. ....................................................... 29

xii

Figure ............................................................................................................................. Page

Figure 2.14 Schematic of an LQIT with a manifold set-up to introduce neutral reagent for ion-molecule reactions. .................................................................. 31

Figure 2.15 Schematic of the 3-Tesla dual-cell FT-ICR mass spectrometer utilized in this research. ................................................................................................ 33

Figure 2.16 Details of the dual-cell in FT-ICR. Reproduced with permission from Nicolet FT-MS 2000 instruction manual. Copyright 1985 Thermo Fisher Scientific Inc. ................................................................................................... 35

Figure 2.17 Cyclotron motion of a positively charged ion in the FT-ICR. ...................... 38

Figure 2.18 Depiction of the cyclotron motion and trapping motion of a positively charged ion in an ICR cell ................................................................................ 40

Figure 2.19 Depiction of an ion's cyclotron motion (little circles) and magnetron motion (large circles). It should be noted that the magnetron radius depicted here is exaggerated. ........................................................................... 42

Figure 2.20 The sequence of events for ionization of a radical precursor, generation of radical sites and ion-molecule reaction in a dual-cell FT-ICR. ................... 44

Figure 2.21 Illustration of (a) trapping a positively charged ion; (b) transfer of the ion into the other side of a dual-cell FT-ICR ................................................... 46

Figure 2.22 Depiction of (a) quadrupolar excitation for quadrupolar axialization (QA) and (b) dipolar excitation for ion detection. ........................................ 47

Figure 2.23 Illustration of ion excitation and detection in an FT-ICR cell. (a) Ions are kinetically excited to move coherently as ion packets. (b) An ion packet passes by the detection plates and induces an image current, which is converted to frequency domain spectrum and finally mass spectrum. .......................................................................................................... 49

Figure 2.24 Comparison of chirp excitation (left) and SWIFT excitations (right). .......... 51

Figure 2.25 Comparison of potential energy surfaces for ion-molecule reactions in the gas phase (top) and in solution (bottom). ................................................... 55

Figure 2.26 Brauman double-well potential energy surface illustrating the entropy constraints for a gas-phase ion-molecule reaction. .......................................... 56

Figure 2.27 A semi-logarithmic plot of the relative abundances of a reactant ion and its products versus time ........................................................................................ 59

xiii

Figure ............................................................................................................................. Page

Figure 3.1 APCI mass spectrum showing the MWD and AVG MW of the Maya asphaltene sample. ........................................................................................... 69

Figure 3.2 Fragmentation pattern of molecular ions of m/z 634 ± 1 of Surmont asphaltene sample. ........................................................................................... 70

Figure 3.3 MS2 CAD mass spectrum of ions of m/z 500 ± 1derived from the Maya asphaltene sample, with the maximum total number of carbons in alkyl chains and the estimated aromatic core size indicated. .................................... 71

Figure 3.4 MS3 CAD mass spectrum of [M-CH3]＋ fragment ions of m/z 485 ± 1

that were formed from molecular ions of m/z 500 ± 1derived from the same Maya asphaltenes sample (Figure 3.3).................................................... 72

Figure 3.5 General trend for the approximate maximum total number of carbons in alkyl chains as a function of MW of the molecules derived from the six asphaltene samples. .......................................................................................... 76

Figure 3.6 General trend for the approximate aromatic core size as a function of MW of the molecules derived from the six asphaltene samples. ..................... 76

Figure 4.1 Two forms of the molecular ion of benzenethiol ........................................... 89

Figure 4.2 MS2 spectrum measured for the molecular ion of 2,2’-bithiophene, and MS3 spectra measured for its three fragment ions ................................................... 90

Figure 5.1 A mass spectrum measured after 100 ms reaction of protonated dibenzothiophene sulfone with TMP in LQIT (*secondary products of protonated TMP). ........................................................................................... 116

Figure 5.2 Mass spectra measured after 300 ms reaction of protonated sulindac (top) and sulindac sulfone (bottom) with TMP in LQIT (*secondary products of TMP adducts; ** secondary products of protonated TMP) ......... 117

Figure 6.1 Structures of ortho- (1), meta- (2), and para-benzyne (3) ............................ 128

Figure 6.2 Structures of the meta-benzyne analogues (a-d) and related monoradicals (e-g) studied ........................................................................... 135 Figure 6.3 Relative energy versus dehydrocarbon atom separation for

meta-benzyne analogues a-d .......................................................................... 136

xiv

LIST OF SCHEMES

Scheme Page

Scheme 4.1 Fragmentation pathways for the molecular ions of (a) benzenethiol, and (b) benzyl mercaptan upon multiple-stage CAD. ................................... 88 Scheme 4.2 Fragmentation pathways for the molecular ion of diphenyl sulfide upon multiple-stage CAD. ............................................................................. 92 Scheme 4.3 Fragmentation pathways for the molecular ion of benzyl sulfide upon multiple-stage CAD. ...................................................................................... 93 Scheme 4.4 Fragmentation pathways for the molecular ion of phenyl disulfide upon multiple-stage CAD. ............................................................................. 95 Scheme 4.5 Fragmentation pathways for the molecular ion of dicyclohexyl disulfide upon multiple-stage CAD.. ............................................................................ 96 Scheme 4.6 Fragmentation pathways for the molecular ion of dibenzyl disulfide upon multiple-stage CAD.. ............................................................................ 97 Scheme 4.7 Fragmentation pathways for the molecular ion of butyl disulfide upon multiple-stage CAD.. ............................................................................ 98 Scheme 5.1 The proposed mechanism for the formation of a stable [TMP adduct- MeOH] product ion when a protonated sulfone reacts with TMP. ...............117 Scheme 6.1 Proposed mechanism for the formation of adduct – COOH for biradical d upon reaction with lysine .......................................................................... 143 Scheme 6.2 Proposed mechanism for HSCH3 abstraction from methionine by biradical c.. ....................................................................................................144

Scheme 6.3 Proposed radical mechanism for H2O abstraction from proline by biradical c. .....................................................................................................145

xv

Scheme Page

Scheme 6.4 Proposed nonradical mechanism for H2O abstraction from proline by biradical c.. .................................................................................................. 146

Scheme 6.5 Proposed mechanism for H2O abstraction from proline by biradical d. ..... 147

Scheme 6.6 Proposed mechanism for formation of adduct and adduct-CO2 for biradical d upon reaction of with proline. .....................................................147

xvi

ABSTRACT

Weijuan Tang. Ph.D., Purdue University, May 2015. Mass Spectrometric Studies on Petroleum Asphaltenes and Organosulfur Compounds, on Functional-Group Selective Ion-Molecule Reactions and on Gas-phase Reactivity of meta-Benzynes toward Amino Acids. Major Professor: Hilkka I.Kenttämaa.

Mass spectrometry has found a wide variety of applications in many fields of

study, such as fundamental chemistry, biological science, food and fuels, advanced

materials, etc. Due to its high sensitivity, selectivity and speed, mass spectrometry

provides an invaluable tool for direct mixture analysis. When coupled with separation

methods, such as gas chromatography or high performance liquid chromatography,

analysis of minor components in complex mixtures is possible. In addition to the

molecular weight information, mass spectrometers can provide structural information for

the ionized analyte molecules. However, mass spectrometric analysis of complex

mixtures is not without challenges, such as suitable evaporation/ionization methods are

not readily available for different types of samples. For example, because of such

limitations, little is known about the molecular weight or structural information of

asphaltenes, which are the heaviest components of crude oil and one of the most complex

mixtures in nature. Characterization of asphaltenes at the molecular level can alleviate

some of the problems they cause to petroleum industry and facilitate the discovery of

beneficial uses for asphaltenes.

xvii

Multiple-stage tandem mass spectrometry (MSn) based on collision-activated

dissociation (CAD) is usually a method of choice for structural elucidation of unknown

compounds. However, this method alone does not always unambiguously identify the

functional groups in an unknown analyte. Therefore, tandem mass spectrometry (MS/MS)

based on ion-molecule reactions was developed and implemented in a linear quadrupole

ion trap (LQIT) mass spectrometer for functional group identification. This method has

great potential for rapid identification of unknown drug metabolites in the pharmaceutical

industry.

Gas-phase ion-molecule reactions are also very useful in study of reaction kinetics

and mechanisms. The intrinsic chemical properties of such highly reactive molecules as

radicals can be studied in the gas phase, which are otherwise difficult to access by other

experimental approaches. Knowledge on the reactivity of aromatic carbon centered σ,σ-

type biradical intermediates is desirable as they are associated with the biological activity

of a naturally occurring enediyne antitumor agents. Of particular interest is the reactivity

of 1,3-biradical species (meta-benzynes) because of its therapeutic importance. In this

thesis, the reactivity of four meta-benzyne analogues towards eight amino acids was

examined by using “distonic ion approach” in a Fourier-transform ion cyclotron

resonance (FT-ICR) mass spectrometer.

The experiments described in this thesis were aimed to provide more detailed

structural information of mixture components by using different mass spectrometry based

methods. Chapter 2 briefly describes the theory, instrumentation, and experimental

aspects of the two instruments used for these studies. Chapter 3 focuses on structural

comparisons of asphaltenes of different origins by using multiple-stage tandem mass

xviii

spectrometry. Chapter 4 describes structural characterization of organosulfur model

compounds related to fossil fuels by using high-resolution tandem mass spectrometry.

Chapter 5 focuses on development of gas-phase ion-molecule reactions for the

identification of the sulfone functionality in drug metabolites. Chapter 6 is devoted to the

study of gas-phase reactivity of pyridine, quinoline, and isoquinoline based meta-

benzynes towards various amino acids.

PUBLICATIONS

xix

xx

xxi

xxii

xxiii

xxiv

xxv

xxvi

xxvii

xxxviii

xxix

xxx

xxxi

xxxii

xxxxiii

xxxxiv

xxxv

xxxxvi

xxxxvii

xxxxviii

xxxxix

xl

xli

xlii

xliii

1

CHAPTER 1. INTRODUCTION AND OVERVIEW

1.1 Introduction

With more than a century of development, mass spectrometry (MS) has

undergone tremendous technological improvements. It has become one of the most

powerful, sensitive, selective, and versatile analytical techniques.1,2 Mass spectrometric

analysis includes three key steps: sample evaporation and ionization, separation of ions

by their mass-to-charge (m/z) ratios, and detection of the ions.3After sample evaporation,

the neutral analyte molecules are converted to ions. Numerous efforts have been

dedicated to the development of methods that allow ionization of different analytes

effectively, from small organic molecules to large biomolecules.3 The resulting ions are

then separated based on their m/z ratios, which can be done using various mass analyzers,

including those that utilize a combination of electric and magnetic fields under vacuum

conditions. After the ions are detected, mass spectra are generated, which show a plot of

the relative abundances of the ions as a function of their m/z ratios. If a molecular ion or

pseudo-molecular ion is formed for each molecule, the molecular weight information of

the analyte can be obtained. Different isotopes of a given element can also be easily

distinguished.4 It should be noted that high-resolution mass spectrometers can measure

accurate masses of the ions, which provides elemental composition for an unknown ion.

2

In addition to molecular weight information, mass spectrometers are capable of

providing structural information for the ionized analyte molecules. Multiple-stage tandem

mass spectrometry (MSn) is an important approach to achieve this.5 By isolating and

subjecting the ion of interest to collision-activated dissociation (CAD), it often generates

characteristic fragmentation products that provide structural information for the analyte.6

As another alternative to probe the structure of the ionic analytes, ion-molecule reactions

have been explored extensively.7-11 The ion of interest can be allowed to react with

selected neutral reagents, producing diagnostic products that facilitate identification of

different functional groups in the analyte molecules. This is particularly useful when

CAD alone does not provide enough structural information for the analytes. Isomer

differentiation also often gets easier when using ion-molecule reactions in mass

spectrometers.12,13

Because of the uniquely valuable information MS can provide, it has found a wide

range of applications in qualitative and quantitative analysis of both small molecules and

big polymers.14 MS has become an indispensable analytical tool in such areas as drug

discovery, clinical analysis, biological science, environmental chemistry, geological

study, and many others.15 MS has been successfully coupled with chromatographic

separations for the analysis of minor components in complex mixtures, which are

difficult to analyze by other techniques, such as nuclear magnetic resonance.16-18

3

1.2 Overview

This dissertation focuses on the structural characterization of petroleum

asphaltenes and organosulfur compounds in them, development of methods for drug

metabolite identification based on functional group selective ion-molecule reactions, and

exploring gas-phase reactivity of meta-benzynes towards amino acids. Chapter 2 briefly

describes the theory, instrumentation, and experimental aspects of the two instruments

used for these studies. They are linear quadrupole ion trap (LQIT) and Fourier transform

ion cyclotron resonance (FT-ICR) mass spectrometers. Chapter 3 focuses on structural

comparisons of asphaltenes of different origins by using multiple-stage tandem mass

spectrometry. Chapter 4 describes structural characterization of organosulfur model

compounds related to fossil fuels by using high-resolution tandem mass spectrometry.

Chapter 5 focuses on the development of gas-phase ion-molecule reactions for the

identification of the sulfone functionality in drug metabolites. Chapter 6 is devoted to the

study of gas-phase reactivity of pyridine, quinoline, and isoquinoline based meta-

benzynes towards various amino acids.

4

1.3 References

1. McLuckey, S.; Wells, J. Chem. Rev.2001,101, 571.

2. Perry, R. H.; Cooks, R. G.; Knoll, R. J. Mass Spectrom. Rev. 2008, 27, 661.

3. de Hoffman, E.; Stroobant, V. Mass Spectrometry: Principles and Applications, 2nd ed.; John Wiley and Sons, Ltd.: New York, 2002.

4. McLafferty, F. W. Interpretation of Mass Spectra, 4th ed.; University Science Books: Sausalito, CA, 1993.

5. Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry/Mass Spectrometry:

Techniques and Applications of Tandem Mass Spectrometry; VCH Publishers: New York, 1988.

6. McLuckey, S.J. Am. Soc. Mass Spectrom. 1992, 3, 599.

7. Brodbelt, J. S. Mass Spectrom. Rev. 1997, 16, 91.

8. Eberlin, M. N. J. Mass Spectrom. 2006, 41, 141.

9. Watkins, M. A.; Price, J. M.; Winger, B. E.; Kenttämaa, H. I. Anal. Chem.2004, 76, 964.

10. Habicht, S.; Vinueza, N.; Archibold, E.; Duan, P.; Kenttämaa, H. I. Anal. Chem. 2008, 80, 3416.

11. Sheng, H.; Williams, P. E.; Tang, W.; Riedeman, J. S.; Zhang, M.; Kenttämaa, H. I. J. Org. Chem. 2014, 79, 2883.

12. Schwartz, J.; Wade, A.; Enke, C.; Cooks, R., Anal. Chem.1990,62, 1809.

13. Fu, M.; Duan,P.; Li,S.; Habicht, S. C.; Pinkston, D. S.; Vinueza, N. R.; Kenttämaa, H. I. Analyst, 2008,133, 452.

14. Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse,C. M. Science1989, 246, 64.

15. Gross, J. H., Mass Spectrometry, a textbook. Springer: New York, 2004.

5

16. McCormack, A. L.; Schieltz, D. M.; Goode, B.; Yang, S.; Barnes, G.; Drubin, D.; Yates, J. R. 3rd. Anal Chem. 1997, 69, 767.

17. Ermer, J.; Vogel, M. Biomed. Chromatogr.2000,14, 373.

18. Loegel, T. N.; Danielson, N. D.;Borton, D. J.; Hurt, M. R.;Kenttämaa, H. I. Energy Fuels, 2012, 26, 2850.

6

CHAPTER 2. THEORY, INSTRUMENTATION AND EXPERIMENTAL ASPECTS OF LINEAR QUADRUPOLE ION TRAP (LQIT) AND FOURIER TRANSFORM

ION CYCLOTRON RESONANCE (FT-ICR) MASS SPECTROMETRY

2.1 Introduction

With its unique attributes in sensitivity, selectivity and versatility, mass

spectrometry (MS) has become an invaluable analytical tool across a broad range of

applications.1-4 Mass spectrometric analysis involves three key steps: sample evaporation

and ionization, separation of ions by their mass-to-charge (m/z) ratios, and detection of

the ions. In the ion source, the analyte molecules are evaporated and ionized by addition

or loss of proton(s), cation(s), anion(s) or electron(s).1 Mass analyzers, which separate the

ions based on their m/z ratios, are the core components of mass spectrometers. Generally,

they can be categorized into two types, trapping and scanning.2 In trapping mass

analyzers, the ions are manipulated at different points of time while confined in the same

space. In scanning mass analyzers, the selected ions fly to different regions of the

instrument for mass spectrometric manipulations. Finally, the separated ions reach the

detector where they are detected.

In this thesis, two mass spectrometers with trapping mass analyzers were

employed: linear quadrupole ion trap (LQIT) mass spectrometer and Fourier-transform

ion cyclotron resonance (FT-ICR) mass spectrometer. The theory, instrumentation, and

7

experimental aspects of these two mass spectrometers are discussed in the following

sections.

2.2 Ionization Methods

There are different ways to ionize the analyte molecules in mass spectrometry,

such as electron ionization (EI),5 chemical ionization (CI),6 electrospray ionization

(ESI),7,8 atmospheric pressure chemical ionization (APCI),9,10 atmospheric pressure

photoionization (APPI),11 desorption electrospray ionization (DESI),12 direct analysis in

real time (DART),13 matrix-assisted laser desorption ionization (MALDI),14 inductively

coupled plasma (ICP)ionization,15 fast atom bombardment (FAB),16 field desorption /

field ionization (FD/FI),17 and laser-induced acoustic desorption (LIAD) / ionization,18,19

among others. Each ionization method has its own advantages and disadvantages, and a

choice is often made depending on the nature of the molecules to be analyzed. Four

ionization methods, EI, CI, ESI, and APCI, were employed for the work in this

dissertation and will be discussed in detail in the following sections.

8

2.2.1 Electron Ionization (EI)

Electron ionization (EI), introduced by Dempster in 1918, is the oldest ionization

method in mass spectrometry.5 Ionization is achieved by bombarding the analyte

molecule with a beam of energetic electrons (typically ~70 eV) in the gas phase.20

Molecular ion of the analyte molecule can be generated if the kinetic energy of the

electrons is greater than the ionization energy of the analyte. However, EI can deposit

more energy than needed for ionizing the molecule, thus fragmentation may occur.

Although molecular weight information is not retained in this case, it can be useful for

structural elucidation of the analyte based on reproducible fragmentation patterns

produced by EI.

2.2.2 Chemical Ionization (CI)

Chemical ionization (CI) is a soft ionization method, which can generate pseudo-

molecular ion of the analyte molecule with minimum fragmentation.6 Therefore,

molecular weight information can be obtained. The analyte of interest is ionized through

chemical reactions with reagent ions, such as electron and/or group transfer.21,22 Figure

2.1 gives two examples of chemical reactions to ionize a radical precursor (R). Figure 2.1

(a) shows how a protonated radical precursor [R+H]+ is generated via self-chemical

ionization (self-CI) and CI. First, an acetone molecule undergoes self-CI with an acylium

ion that was generated by EI of another acetone molecule, forming protonated acetone

ion as the final CI reagent ion. Then the neutral radical precursor is allowed to react with

the CI reagent ion to generate protonated radical precursor. This proton transfer reaction

has to be exothermic to occur. Typically, a CI reagent with PA slightly lower than that of

th

sh

un

ca

an

an

ca

co

ra

d

ot

m

he analyte m

hows how

ndergoes EI

ation to the n

Figure 2.1

Electr

nalysis.7,8 It

nalytes, such

an form mu

orresponding

anges. Proto

eprotonation

ther ionizati

most basic or

molecule is c

a methylate

I and self-CI

neutral radic

Examples o

rospray ioniz

has been su

h as proteins

ultiply charg

g analytes, w

onation typic

n generally

ion methods

r acidic analy

chosen, such

ed radical

I to form dim

cal precursor

f two chemi

2.2.3 Elec

zation (ESI)

uccessfully a

s, oligonucle

ged gas-pha

which enabl

cally occurs

occurs for m

, ESI has lim

ytes in a com

h as acetone

precursor [

methyl iodid

r, subsequen

cal reactions

ctrospray Ion

is a soft ion

applied to ge

eotides, lipid

ase ions, th

les their det

s for molecu

molecules in

mitation in te

mplex mixtur

e or methan

[R+CH3]+ is

de cation, w

ntly generatin

s used to ion

nization (ES

nization tech

enerate gase

ds, etc.23,24 U

herefore redu

tection in in

ules in posi

n negative i

erms of ioni

re are ionize

nol. Similarly

s generated

which then tr

ng a methyla

nize a radica

I)

hnique for m

eous ions for

Upon ESI, la

ucing the m

nstruments w

itive ionizat

onization m

ization bias,

ed.

y, Figure 2.

d. Methyl io

ransfers a m

ated molecul

l precursor (

ass spectrom

r thermally l

arge biomole

m/z ratios o

with limited

tion mode, w

mode. As wit

because onl

9

.1 (b)

odide

methyl

le.

(R).

metric

labile

ecules

of the

mass

while

th all

ly the

10

In electrospray ionization, ions are preformed in solution, and they are transferred

from solution to gas phase through a stepwise process.26 As shown in Figure 2.2, the

solution containing the analyte of interest first passes through a high voltage needle (±3-5

kV), from which a fine mist of charged droplets is ejected.24 With the assistance of

nebulizing gas (typically nitrogen), the solvent molecules keep evaporating and the

droplets shrink in size, which causes the charges to move towards each other. Once the

charges are concentrated to a critical point, known as the Rayleigh stability limit, the

droplet explodes to produce smaller droplets. This process occurs as the Coulombic

repulsion overcomes the surface tension of the droplet, and it repeats until singly or

multiply charged gas-phase ions are released and reach the mass spectrometer inlet.26

io

a

g

ca

re

so

to

co

Atmo

onization tec

method of

enerate ionic

an be obtain

elies on a ser

olution conta

o several hu

ommonly us

Figure 2.2

2.2.4 At

spheric pres

chnique.9,10 A

choice for

c molecules

ned, althoug

ries of reacti

aining the an

undred degr

sed as sheat

Depiction of

mospheric P

ssure chemi

APCI is often

analytes wi

with no or l

gh this is not

ions to ioniz

nalyte of int

rees of Cels

th and auxili

f the ESI pro

Pressure Che

cal ionizatio

n used to ion

ith low to m

little fragmen

t always the

ze the analyt

terest flows i

sius (300 °C

iary gas, als

ocess in posi

emical Ioniza

on (APCI)

nize thermal

medium pola

ntation, thus

e case.27 AP

te of interest

into the ion

C - 500 °C

so assists wi

itive ion mo

ation(APCI)

is also cons

lly stable mo

arity. APCI

s molecular m

PCI is simila

t. As shown

source, whe

C) and vapo

ith the solut

de.

)

sidered as a

olecules. It is

can be tun

mass inform

ar to CI in t

in Figure 2.3

ere it is heate

orized. Nitro

tion nebuliza

11

a soft

s also

ned to

mation

that it

3, the

ed up

ogen,

ation,

cr

h

b

so

m

to

re

pr

reating a fin

igh-voltage

elow) occur

A typ

olvent.28 Firs

molecules) to

o form reage

eagent ions,

rotonated m

ne mist of dr

corona disch

in the zoom

Figure 2.3 D

pical APCI r

st, the coron

o produce the

ent ions H2O

H3O+. The

molecules in p

oplets. Whe

harge needle

med plasma r

Depiction of

reaction seri

na discharge

e primary io

O+, which th

final reagen

positive ion

en the mist le

e, where a se

egion.

the APCI pr

ies is illustr

needle ioniz

ons. These pr

hen undergo

nt ions then

mode. It sh

eaves the va

equence of c

rocess in pos

rated below,

zes the carri

rimary ions

ion-molecul

ionize the a

hould be note

aporizer heat

chemical rea

sitive ion mo

where wat

er gas N2 (th

react with s

le reactions

analyte mole

ed that the s

ter, it passes

actions (discu

ode.

er is used a

he most abun

solvent mole

to form the

ecules, gener

specific choi

12

s by a

ussed

as the

ndant

ecules

final

rating

ice of

so

th

olvent, such

he types of io

as carbon d

ons formed i

disulfide, tolu

in APCI.29,30

uene, or chlo

0

oroform, as well as sheaath gas can a

13

affect

14

2.3 Linear Quadrupole Ion Trap (LQIT) Mass Spectrometry

2.3.1 Introduction

Linear quadrupole ion trap (LQIT) mass spectrometers have found a wide range

of applications since they were first introduced in 2002.31,32 These 2-dimensional (2-D)

ion traps operate by trapping the ions in a confined space while performing ion

manipulations based on time. The underlying principles of 2-D LQIT are similar to those

of 3-D quadrupole ion traps (QIT).33 However, LQITs have improved ion trapping

capacity and efficiency, which results in higher sensitivity.31 Compared to the traditional

QIT, LQIT has five-fold lower detection limit.31 In order to combine tandem mass

spectrometry capabilities of LQIT with high-resolution measurements, several hybrid

instruments emerged not long after the stand-alone LQIT was introduced, such as LQIT-

TOF,34 LQIT-Orbitrap,35 and LQIT-FT-ICR.36 The latter two hybrid instruments were

used to perform accurate mass measurements for part of the work discussed here.

2.3.2 Instrument Overview

LQIT experiments were performed using a Thermo Scientific LTQ linear

quadrupole ion trap mass spectrometer.31,37 A general schematic is shown in Figure 2.4.

The LQIT is equipped with an atmospheric pressure ionization source, such as ESI or

APCI. In these instruments, the ionized molecules are drawn into the API stack region,

whose pressure is maintained at approximately 1 Torr by two Edwards E2M30 rotary-

vane mechanical pumps (10.8 L/s). The ions then travel through a series of ion optics and

reach the mass analyzer. The pressure of the ion trap is maintained at approximately 0.5-

1

tu

pu

ef

m

st

g

(7

.0 x 10-5 T

urbomolecul

umps (10.8

fficiencies a

multipole Q0

Figure 2.4 S

Comp

tack compri

enerated, a n

760 Torr, 3-

Torr by on

lar pump, wh

L/s). Anot

are 25 L/s an

region (1 m

Schematic ofpressu

ponents of th

ises a trans

negative pre

5 kV) to the

ne inlet (40

hich is back

ther two inl

nd 300 L/s, p

mTorr) respec

f the Thermoure indicated

he API stack

sfer capillar

essure gradie

e heated tran

00 L/s) of

ked by two E

lets of the

pump down t

ctively.

o Scientific Lfor each reg

k and ion gu

ry, tube len

ent and large

nsfer capillar

a Leybold

Edwards E2M

turbomolecu

the multiple

LQIT mass sgion of the in

uides are sh

ns, and skim

e potential d

ry (1 Torr, ±

d TW220/15

M30 rotary-

ular pump,

Q00 region

spectrometernstrument.

hown in Figu

mmer cone.

decrease fro

±20 V) draw

50/15 triple

-vane mecha

whose pum

(500 mTorr

r with opera

ure 2.5. The

. After ion

m the ion so

w the ions int

15

e-inlet

anical

mping

r) and

ating

e API

s are

ource

to the

16

API stack. A potential of 0 to ±10 V (positive for positive ions and negative for negative

ions) is applied to the transfer capillary to aid in focusing ions into a concentrated beam.

Additionally, the transfer capillary is normally heated to 250-300 °C to aid in ion

desolvation. A mass-dependent potential is applied to the tube lens to focus the ions into

a tighter packet for transmission. It should be noted that the voltage on tube lens can

increase the kinetic energy of the ions, which may induce fragmentation. The slightly off-

axis skimmer is a grounded lens (0 V) used to remove neutral molecules and act as a

vacuum baffle to the lower pressure ion guide region.

After exiting the API stack, ions are guided through a series of lenses and

multipoles into the mass analyzer. Q00 and Q0 are quadrupole assemblies with square

rods; multipole Q1 is an octupole assembly with round rods. RF potentials are applied to

the rods of these multipoles so that the ions are confined in the x-y plane. A downhill DC

potential gradient is applied to the lenses and multipoles so that the ions gain kinetic

energy that facilitates their transmission further into the ion trap.

ro

m

sl

an

FigureDC p

The m

ods. Each of

mm) sections

lits (0.25 mm

nd an electro

e 2.5 Compopotential grad

mass analyze

f the four ro

s, as shown i

m × 30 mm

on multiplier

nents of the dient was ap

er of the LQ

ds is divided

in Figure 2.6

dimension)

r is on each

API stack applied to aid

QIT mass sp

d into front

6. At the x-a

which allow

side of the tr

and ion guidein ion transf

pectrometer

(12 mm), ce

axis of the c

w for ion eje

rap for ion d

es to which afer in the z-d

r consists of

enter (37 mm

enter two ro

ection. A co

detection.

a downhill direction.

f four hyper

m), and bac

ods, there are

onversion dy

17

rbolic

k (12

e two

ynode

cr

p

ro

to

tr

Fig

Ions a

reates an o

otential (±5

ods so that th

o all three se

rapped axiall

gure 2.6 Sch

are trapped

scillatory io

kV, 1.2 MH

he ions are t

ections of th

ly (z-directio

hematic of th

2.3.3

in LQIT b

on motion.33

Hz) with opp

trapped radia

he mass ana

on).

he Thermo S

Ion Motions

y a combin

3,38-41 As sh

posite polarit

ally (x-y pla

alyzer to cre

Scientific LQ

s in LQIT

nation of RF

hown in Fig

ty is applied

ane). A statio

ate a potent

QIT mass ana

F and DC p

gure 2.7, an

d to the x- an

onary DC po

tial well, so

alyzer.

potentials, w

n oscillating

nd y-axis pa

otential is ap

that the ion

18

which

g RF

airs of

pplied

ns are

2

p

ca

w

th

el

ar

th

Figure 2.

.3.3.1 Radia

In ord

otentials is a

an be describ

wherein is

he angular fr

lectric field

re subject to

hat trap the i

.7 The oscill

al Motion

der to trap

applied to th

bed as show

the DC pot

requency of

between the

o a potential

ons in the x

lating RF pofor trapping

the ions ra

he four hyper

wn in Equatio

Ф

ential, is t

the RF field

e four rods. W

in the x-y p

and y direct

tentials applg ions radial

adially (x-y

rbolic rods o

on 2.1.

the zero-to-p

d, and is tim

When the io

lane, which

tions are sho

lied to the x-lly (x-y plan

plane), a c

of the LQIT

cos Ω

peak amplitu

me. These po

ons are trans

is described

own in Equat

- and y-axis ne).

combination

mass analyz

ude of the R

otentials crea

sferred into t

d by Equatio

tions 2.3 and

pairs of rods

n of RF and

zer. The pote

Equatio

RF potential

ate a quadru

the ion trap,

on 2.2. The f

d 2.4.

19

s

d DC

ential

on 2.1

l,Ω is

upolar

, they

forces

20

Ф , Ф Equation 2.2

Ф Equation 2.3

Ф Equation 2.4

wherein is the mass of the ion, is the elementary charge, is the number of charges

of the ion, and is the radius of the circle inscribed within the ion trap. Rearrangements

of the above equations result in Equations 2.5 and 2.6 respectively.

cosΩ 0Equation 2.5

cosΩ 0Equation 2.6

These two equations are similar to the general form of Mathieu equation (Equation 2.7).

2 cos 2ξ 0Equation 2.7

The Mathieu equation can be rearranged into Equation 2.8 by defining as ,

cos Ω 0 Equation 2.8

21

Therefore, the radial motion of an ion can be described as shown in Equations 2.9 and

2.10.

Equation 2.9

Equation 2.10

wherein and are Mathieu stability parameters. An ion will have a stable trajectory

in the LQIT only when its and values fall within the stability region. The most

well-defined stability region is shown as the overlap area in Figure 2.8. Ions that fall

outside of this region have unstable motions in x and/or y directions as indicated.

Generally, the LQIT operates with 0 so that a broad range of ions can be trapped

simultaneously. Since value is inversely proportional to the ion's m/z ratio, larger ions

have lower values while smaller ions have higher values. This is illustrated by the

circles of different sizes in Figure 2.8, which represent ions of different masses.

F

se

w

fr

Figure 2.8 ThThe circles o

this

At a g

ecular (reso

wherein

requency

he most wellof different s

s region have

given RF pot

nant) freque

of an ion is

l-defined stasizes represee unstable m

tential, each

ency, which

. Since th

s one half of

ability regionent ions of d

motions in x a

h ion with a d

is described

e maximum

f the RF frequ

n of Mathieudifferent masand/or y dire

different m/z

d in Equation

m value of

uency.38

u stability disses. Ions thaections as ind

z ratio oscill

n 2.11,

is 1, the m

agram for Lat fall outsiddicated.

lates at a spe

Equation

maximum se

22

QIT. de of

ecific

n 2.11

ecular

2

p

H

th

T

in

es

d

en

F

.3.3.2 Axial

In ord

otentials are

Higher poten

hat a potenti

Trapping ions

n the x-rods

Addit

ssential for i

irection and

nergy.

Figure 2.9 D

Motion

der to trap

e applied to t

ntials are app

al well is cre

s in the cent

, but also re

ionally, the

ion trapping

d collisions w

Diagram of D

ions axially

the front, ce

plied to the

eated, which

ter not only

educes fring

helium buf

g.31 Ions gain

with the buff

DC potential

y (z-directio

enter and bac

front and ba

h traps the io

facilitates ef

e field effec

ffer gas, wh

n kinetic ene

fer gas help

well for trap

on) in the m

ck sections o

ack sections

ons in the ce

fficient ion e

cts created b

hich is prese

ergy while b

to cool the

pping ions in

mass analyz

of the four h

s than the m

enter, as show

ejection thro

by the front

ent in LQIT

being transfe

ions and red

n the center a

zer, differing

hyperbolic ro

middle sectio

wn in Figure

ough the two

and back le

at ~3 mTo

erred along t

duce their ki

along the z-a

23

g DC

ods.31

on, so

e 2.9.

o slits

enses.

orr, is

the z-

inetic

axis.

24

2.3.4 Ion Excitation and Detection

An ion's radial motion in quadrupole fields can be expressed mathematically in

terms of the Mathieu equations. As shown in equations 2.9 and 2.10, is proportional to

the DC voltage while is proportional to the amplitude of the applied RF voltage. The

LQIT normally operates at = 0. By scanning the amplitude of the RF voltage linearly,

ions of increasing m/z ratios will be pushed to the instability boundary at = 0.908. At

this point, ions’ motion becomes unstable and the ions are ejected through the slits in the

x-rod for detection. This process is known as the “mass selective instability scan”.42 This

technique suffers from low mass spectral resolution because not all ions of a specific m/z

ratio are ejected from the trap simultaneously.

In order to increase mass resolution, another technique known as "resonance

ejection" is often employed.43 Using this technique, ions are scanned out at = 0.880, as

shown in Figure 2.10. This is achieved by applying a small supplemental RF voltage of

fixed frequency during the ramp of the main RF voltage. When an ion is about to be

ejected from the trap by the main RF voltage, it is brought to resonance with the

supplemental RF voltage. This facilitates ion ejection from the trap and improves mass

resolution and sensitivity.

F

d

si

co

ap

g

el

T

Figure 2.10 I

After

etection syst

ide of the i

onversion dy

pplied to it.

enerated. Po

lectrons, and

These second

Illustration oinstability s

ions are ejec

tem, which

on trap.31 A

ynode with

When an ion

ositively cha

d negatively

dary particle

of ion ejectioscan, and (b)

cted through

consists of a

As shown in

either -15 k

n hits the su

arged ions g

y charged io

es are focus

on from LQI) = 0.880

h the slits in

a conversion

n Figure 2.1

kV (for posi

urface of the

generate one

ons generate

sed by the c

IT at (a) =during reson

the x-rods,

n dynode and

11, the eject

itive ions) o

dynode, mu

e or more n

e one or m

concave surf

= 0.908 durinnance ejectio

they are dire

d electron m

ted ions are

or +15 kV (f

ultiple secon

negatively c

more positive

face of the

ng mass seleon.

ected toward

multiplier on

e attracted t

for positive

dary particle

charged ions

ely charged

dynode, an

25

ctive

ds the

n each

o the

ions)

es are

s and

ions.

d are

ac

th

m

th

th

fr

ccelerated to

he dynode a

multiplier, wh

he electron m

he number o

rom the trap.

Figure

oward an ele

and the elec

hich generat

multiplier. F

of secondary

.

2.11 Depicti

ectron multip

tron multipl

tes a cascad

Finally, a me

y particles, a

ion of the ion

plier becaus

lier.31 Each

de of electro

easurable cur

and thus pro

n detection s

e of a large

secondary p

ons that cont

rrent is crea

oportional to

system in Th

potential di

particle will

tinue to strik

ated, which i

o the numbe

hermo Scien

fference bet

l hit the ele

ke the surfa

is proportion

er of ions ej

ntific LQIT.

26

tween

ectron

ace of

nal to

ected

27

2.3.5 Tandem Mass Spectrometry

Multiple-stage tandem mass spectrometry (MSn) is a powerful technique to probe

the structure of an ionized unknown compound.44 After the ion of interest is isolated, it

can be subjected to collision-activated dissociation (CAD) to generate characteristic

fragment ions; or it can be allowed to undergo ion-molecule reactions to form diagnostic

product ions, both of which shed light on the structure of the ionic parent molecule. Since

LQIT is a trapping instrument, tandem mass spectrometry events occur in the same space

but sequentially in time.

2.3.5.1 Ion Isolation

As shown in Figure 2.12 I-IV, a series of events occur before an ion of interest

(represented by the green circle) is isolated and subjected to CAD. The ion of interest is

first placed at = 0.880 by ramping the RF voltage, during which process ions of lower

masses are ejected from the trap (I-II). Next, a tailored broadband RF waveform is

employed to eject all remaining unwanted ions, except the ion of interest (III). This

isolation waveform is similar to what is employed for ion excitation and detection as

described above. It consists of a 5–500 kHz multi-frequency waveform with sine

components spaced every 0.5 kHz, with a notch at q = 0.83 so that the ion of interest

remains in the trap (Figure 2.13). Finally, the RF voltage is decreased to move the ion of

interest to q = 0.25 (IV). Upon CAD, the parent ion dissociates into fragment ions (IV),

which can be efficiently trapped for detection or further dissociation reactions.

Figure 2.12activated d

2 The sequedissociation

nce of event(CAD). The

ts for ion isoe ion of inter

olation and frrest is repres

fragmentationsented by the

n by collisioe green circl

28

on-le.

2

w

bu

in

k

su

ap

m

en

R

am

"n

.3.5.2 Collis

In the

wherein ions

uffer gas ov

nterest is iso

inetically e

upplemental

pproximately

multiple colli

nergy and ev

RF voltage ap

In The

mount of k

normalized

Figure

sion-activate

e LQIT, col

undergo fr

ver a relative

olated, the q

excited by

l RF voltage

y 10-100 m

isions with t

ventually cau

pplied to the

ermo Scient

kinetic energ

collision en

e 2.13 Tailor

ed Dissociati

lision-activa

agmentation

ely long activ

value of 0.2

resonance

e (~ 1 V) ma

ms. During

the buffer g

uses the ion

e x-rods is re

tific LQIT, t

gy an excit

ergy" on a

red RF wave

ion (CAD)

ated dissocia

n after many

vation time.

25 is genera

excitation.

atching the s

this process

as, which co

s’ dissociati

ferred as "tic

wo paramet

ted ion gain

scale from

eform for ion

ation (CAD)

y low-energy

45,46 As disc

ally chosen.

This is ac

secular frequ

s, the kinet

onverts their

ion into frag

ckle voltage

ters are usua

ns during C

0% to 100%

n isolation.

) is a slow

y collisions

cussed above

The isolated

ccomplished

quency of the

tically excit

r kinetic ene

gment ions. T

e".

ally adjusted

CAD. One p

%. The high

heating me

with the he

e, once the i

d ion can the

by applyi

e isolated io

ed ions und

ergy into int

The supplem

d to determin

parameter i

her the value

29

ethod,

elium

on of

en be

ing a

on for

dergo

ternal

mental

ne the

s the

e, the

30

more energy is deposited into the parent ion. The exact collision energies are unknown.

Another factor is the q value. At higher q values, ions oscillate at a higher frequency and

have higher kinetic energies, which may result in more fragmentation upon collisions

with helium. Typically, a q value of 0.25 is chosen to activate the ion of interest. This

value allows most of the fragment ions to be efficiently trapped during CAD.31 For ions

that require higher energy to dissociate, a q value that is greater than 0.25 can be used.

However, this limits the mass range of fragment ions that can be trapped.

2.3.6 Ion-molecule reactions in LQIT

In addition to performing CAD experiments on the ion of interest, tandem mass

spectrometry experiments based on ion-molecule reactions are also of great use for

structural investigations.47 This is particularly true for isomer differentiation because

CAD may yield the same fragmentation patterns while ion-molecule reactions often show

distinct products for isomeric ions.48 Figure 2.14 shows a schematic of an LQIT with a

manifold set-up that can be used to introduce neutral reagents for ion-molecule reactions.

In this example, a neutral reagent, trimethyl phosphite, is introduced into the ion trap. The

neutral reagent is then allowed to react with the analyte ion of interest for various periods

of time, ranging from 30 ms to 10 s.

FFigure 2.14 S

Schematic off an LQIT wion-

with a manifo-molecule re

old set-up to actions.

introduce neeutral reagen

31

nt for

32

2.4 Fourier-Transform Ion Cyclotron Resonance (FT-ICR) Mass Spectrometry

2.4.1 Introduction

In 1932, Ernest Lawrence and Stanley M. Livingston first discovered the principle

of ion cyclotron resonance (ICR).49,50 It was later introduced to mass spectrometry,

followed by the commercialization of the first ion cyclotron resonance (ICR) mass

spectrometer.51,52 Afterwards, the successful combination of Fourier transformation with

nuclear magnetic resonance (NMR) spectroscopy inspired Comisarow and Marshall to

couple Fourier transformation (FT) with ion cyclotron resonance (ICR). 53,54 In 1974, the

FT-ICR mass spectrometry was created. FT-ICR is a very powerful mass spectrometric

technique, offering ultra high resolution and high mass accuracy.55,56 Its many

applications include complex mixture analysis and accurate mass measurement, as

demonstrated in Chapters 3 and 4. Moreover, the high vacuum condition of FT-ICR

instrument allows the ions to be trapped for relatively long periods of time. Therefore,

gas-phase ion-molecule reactions and multiple stage MS/MS experiments can be

performed, as demonstrated in Chapter 6.

2.4.2 Instrument Overview

The gas-phase reactivity study of radicals described in this dissertation was

carried out in a ~3 Tesla dual-cell Extrel model FTMS 2001 mass spectrometer (Figure

2.15). The source side and analyzer side of the instrument are two differentially pumped

regions, the purpose of which is to maintain different pressures in each cell. The

instrument is equipped with an IonSpec data station running on IonSpec Omega 8

s

b

m

i

a

c

n

t

l

software. A

by two Pfeif

mechanical p

is measured

above the ce

convert the

neutral reage

Figure 2.1

Differ

to liquids an

liquids, such

high vacuum

ffer HiPace7

pump serves

by a Bayar

ell. Accordin

ion gauge re

ent concentr

15 Schemati

rent methods

nd gases. G

h as methyl

m with a nom

700 turbo pu

s as a back p

rd-Alpert ion

ngly, ion gau

eadings to r

ation in the c

ic of the 3-T

s can be use

aseous samp

iodide, can

minal baselin

umps, one on

pump for eac

nization gau

uge correcti

real pressure

cell (describ

esla dual-celin this resea

ed to introdu

ples, such a

be introduc

ne pressure o

n each side o

ch turbo pum

uge, which s

ion factors n

e, which allo

bed in section

ll FT-ICR march.

uce various

as helium an

ced via pulse

of ~1x10-9 T

of the instrum

mp. The pres

sits approxim

need to be m

ows for dete

n 2.5.2).

mass spectrom

samples ran

nd argon, as

ed valves. T

Torr is maint

ment. An A

ssure of each

mately one m

measured so

ermination o

meter utilize

nging from s

s well as vo

This instrume

33

ained

lcatel

h cell

meter

as to

of the

ed

solids

olatile

ent is

34

equipped with two sets of pulsed valves on the source side and one set on the analyzer

side. If a constant pressure rather than a pulse of the reagent is desired, batch inlets can

be employed for introduction of volatile liquids. For volatile solids and highly volatile

liquids that do not require heating, they can be introduced into the instrument by using

Varian leak valves. For nonvolatile and thermally stable solids, they are introduced into

the instrument by using solids probes, which can be heated up to ~250˚C to facilitate

sample desorption. Laser-induced acoustic desorption (LIAD) is an alternative method

of introducing nonvolatile and thermally labile samples.57 On either side of the

instrument, it is equipped with a heated solids probe, batch inlet and Varian leak valve.

Electron ionization filament, which emits electrons to induce ionization of the analytes,

is located in the analyzer side of the instrument.

The dual-cell consists of two cubic cells that are aligned collinearly within the

homogenous magnetic field. The two cells consist of a total of eleven stainless steel

plates, as shown in Figure 2.16. Three of them are perpendicular to the magnetic field.

They are trapping plates, which are used for trapping ions along the z-axis (discussed in

section 2.4.3.2). The trapping plate in the center, called the conductance limit, is shared

between the two cells.58 The three plates all have 2 mm diameter holes in the center,

which allow ions to be transferred from one cell into another, as well as allow the

electron beam to travel from the analyzer side into the source side for sample ionization.

The remaining eight stainless steel plates are parallel to the magnetic field. For each cell,

one pair of plates is used for ion excitation and another pair of plates for ion detection.

F

Figure 2.16 DFT-MS

Details of the2000 instruc

e dual-cell inction manual

n FT-ICR. Rl. Copyright

Reproduced wt 1985 Therm

with permissmo Fisher Sc

sion from Nicientific Inc.

35

icolet

36

2.4.3 Ion Motions in FT-ICR

In FT-ICR mass spectrometers, ions undergo three types of motion, cyclotron,

trapping, and magnetron motions.52,53 The electric fields applied to the trapping plates

confine the ions along the z-direction, whereas the homogeneous magnetic field confines

the ions in the x-y plane. The sum of the forces (F) an ion experience arises from the

electric and magnetic fields, and it can be described by Equation 2.12.53

Equation 2.12

wherein q is the ion's charge, E is the electric field, B is the strength of the magnetic field,

and is the velocity of the ion.

2.4.3.1 Cyclotron Motion

The most important ion motion, cyclotron motion, arises from the interaction of a

moving ion with the homogeneous magnetic field. The magnetic field exerts an inward-

directed Lorentz force, FLorentz, on the ion, as shown in Equation 2.13

FLorentz qvxyB Equation 2.13

wherein q is the ion's charge, vxy is the velocity of the ion in the x-y plane, and B is the

strength of the magnetic field. Furthermore, the ion also experiences an outward-directed

centrifugal force, FCentrifugal, as shown in Equation 2.14,

Equation 2.14

37

wherein m is the mass of the ion, vxy is the velocity of the ion in the x-y plane, and r is the

radius of its cyclotron motion.

As an ion moves circularly in a unidirectional magnetic field, the Lorentz force

and centrifugal force counter balance each other (Figure 2.17). Therefore Equations 2.13

and 2.14 are equal, resulting in Equation 2.15, which can also be rearranged to Equation

2.16. This equation shows that an ion's cyclotron radius ) is proportional to the mass-

to-charge ratio of the ion and its velocity in the xy-plane while inversely proportional to

the strength of the magnetic field.

Equation 2.15

Equation 2.16

Since frequency ѡ can be expressed as shown in Equation 2.17, the ion's

cyclotron frequency (ѡ ) in FT-ICR can be described as shown in Equation2.18,

ѡ Equation2.17

ѡ Equation2.18

An ion's cyclotron frequency is independent of its velocity or kinetic energy, and

proportional to the strength of the magnetic field and inversely proportional to the ion’s

mass to charge ratio. FT-ICR detects ions based on their cyclotron frequencies, which are

unique to the m/z ratios of the ions. Therefore, FT-ICR has higher mass accuracy and

re

sp

2

o

p

tr

is

esolution co

pectrometer

Figu

.4.3.2 Trapp

The m

f the cell if

otential (+2

rapping plate

s shown in F

ompared to

which meas

ure 2.17 Cycl

ping Motion

magnetic fiel

f no constra

V for posi

es to constra

Figure 2.18. T

o other m

sure ions' ma

lotron motio

d constrains

aint is impos

itive ions an

ain the ions

The trapping

, ,

ass spectro

asses based o

on of a positi

s ions in the

sed in the z

nd -2 V for

along the z-

g potential is

ometers, su

on their velo

ively charge

xy-plane, ye

z-direction. T

r negative i

-axis. Trappi

s shown in E

2

uch as tim

ocities.53

d ion in the

et the ions ca

Therefore, a

ions) is app

ing motion i

Equation 2.19

e-of-flight

FT-ICR.

an freely dri

a small repu

plied to the

in the z-dire

9.

Equation

38

mass

ft out

ulsive

three

ection

n 2.19

39

wherein is the voltage applied to the trapping plates, and and are constants related

to the geometry of the cell. For a cubic cell, = 0.3333 and = 2.77373. For a trapping

voltage of 2 V, the trapping plates have the maximum value of 2 V, whereas the center of

the cell has a minimum value of 0.67 V ( ). The frequency at which the ions oscillate

harmonically between the trapping plates (along z-direction) can be shown in Equation

2.20.

Equation 2.20

wherein is the charge of the ion, is the voltage applied to the trapping plates, is a

constant related to the cell's geometry, is the mass of the ion, and is the distance

between the trapping plates.

2

tr

m

th

un

in

ce

Fig

.4.3.3 Magn

Ions a

rapping moti

magnetron m

he electric fi

niform elect

n equation 2

enter of the c

gure 2.18 Dep

netron Motio

are constrain

ion, respecti

motion, canno

eld along the

tric field gen

.21. It oppos

cell, causing

piction of tha positively

on

ned in the x

ively. Howe

ot be neglect

e z-axis is no

nerates an ou

ses the Lore

g them to orb

F

he cyclotron charged ion

xy-plane an

ever, the pre

ted (Figure 2

ot uniform d

utward-direc

enz force and

bit at larger r

F r

motion and n in an ICR c

nd z-axis by

esence of an

2.19). This m

due to the fin

cted radial fo

d pushes the

radii in x-y p

trapping mocell.

y their cyclo

n undesirable

motion is du

nite size of th

orce (F , wh

e ions radiall

plane.

otion of

otron motion

e ion motion

ue to the fac

he cell. The

hich is illust

ly away from

Equation

40

n and

n, the

ct that

e non-

trated

m the

n 2.21

41

wherein is a constant related to the cell's geometry ( = 1.39 cm for a cubic cell), is

the charge of the ion, is the voltage applied to the trapping plates, is the distance

between the trapping plates, and is the radius of the magnetron motion.

The frequency of the magnetron motion ) is defined by Equation 2.22,

wherein is the cell geometry factor, is the voltage applied to the trapping plates, is

the distance between the trapping plates, and is the strength of the magnetic field. It

should be noted that the magnetron frequency is independent of an ion's m/z ratio, hence

all ions will have the same magnetron frequency. Generally, the magnetron frequencies

are not detected, because they are much smaller than the cyclotron frequencies.53 It

should be noted that magnetron motion can adversely affect ion transfer, sensitivity,

resolution, and mass accuracy.58 Quadrupolar axialization (described in section 2.4.4.2)

can be used to counter these effects.

Equation 2.22

F(

Figure 2.19 D(large circles

Depiction ofs). It should b

f an ion's cycbe noted tha

clotron motiat the magnet

ion (little cirtron radius d

rcles) and madepicted her

agnetron moe is exagger

42

otion ated.

43

2.4.4 Ion Manipulations in FT-ICR

Similar to the LQIT mass spectrometer, FT-ICR is a trapping instrument. A series

of events occurs in the same space but at different times. The high vacuum of FT-ICR

enables the ions to be trapped for a relatively long period of time. Therefore, FT-ICR

provides an ideal environment for studying ion-molecule reactions in the gas phase. The

event sequence used for studying radical reactions in a dual-cell FT-ICR is shown in

Figure 2.20. It includes desorption and ionization of the radical precursor, transfer of the

ions into a clean cell, generation of radical ions by sustained off-resonance irradiation

collision-activated dissociation (SORI-CAD), isolation of the desired radical ions, and

allowing them to react with a neutral reagent, followed by excitation and detection of the

reaction products.

Figure 2.200 The sequenradical site

nce of eventss and ion-mo

s for ionizatiolecule react

ion of a radiction in a dua

cal precursoal-cell FT-IC

r, generationCR.

44

n of

45

2.4.4.1 Ion Transfer

In the dual-cell FT-ICR, ion transfer is an essential step needed to perform ion-

molecule reactions in a clean environment. Protonated or methylated radical precursors

are generated in the source cell and transferred into the clean analyzer cell for reactions.

Ion transfer is achieved by grounding the conductance limit plate for a certain period of

time, as shown in Figure 2.21. The optimal transfer time is dependent on the m/z ratio of

the ion, and can be calculated using Equation 2.23.

⁄ 10μ Equation 2.23

In preparation for ion transfer, the voltage for the remote trapping plate is changed

from +2.0 V to -3.5 V for 15 ms, so as to eject all ions from the analyzer cell. This step

ensures that the charged radical precursors will enter a clean analyzer cell for reactions.

During transfer, ions gain kinetic energy. Thus, they need to be cooled before performing

further experiments.59,60 This is accomplished by allowing the ions to undergo energy

dissipation through collisions with neutral molecules or atoms present in the analyzer cell

for 1 – 5 s. After transfer, the voltage on the conductance limit plate is brought back to +2

V, so that the transferred ions remain trapped in the analyzer cell. In the work described

in this dissertation, argon was pulsed in as cooling gas. Collisions with argon gas convert

the ion's kinetic energy to internal energy which can be released by IR emission.

F

2

th

p

T

em

m

an

co

m

F

tr

p

igure 2.21 Il

.4.4.2 Quadr

Due t

he radius of

late, the tran

This results in

mployed to o

Quadr

motion. It wa

nd detection

ollisions wit

motion increa

inally, the i

ransfer.

Quadr

otential mat

llustration of

rupolarAxial

o the presen

the magnetr

nsferring ion

n loss of ion

overcome th

rupolar excit

as achieved b

n plates of th

th a collision

ases slowly,

ions are kin

rupolar axia

tched the cy

f (a) trappingthe other s

lization (QA

nce of magn

ron motion i

ns will collid

n signal. Qua

his disadvant

tation works

by applying

e source cel

n gas, usuall

, while the

etically coo

alization is

yclotron freq

g a positivelide of a dual

A)

netron motio

s larger than

de with the co

adrupolar axi

tage.61-63

s by convert

an RF poten

l, which gen

ly helium (~

radius of th

led and axia

mass select

quency of t

ly charged iol-cell FT-IC

n, ion transf

n the 2-mm h

onductance

ialization (Q

ting ions' ma

ntial of oppo

nerated a qua

~10-5 Torr), t

heir cyclotro

alized into t

tive. The fr

the ions of

on; (b) transfCR.

fer is not alw

hole in the c

limit and be

QA) is a tech

agnetron mo

osite phases

adrupolar ele

the radius of

on motion d

the center o

requency of

interest. On

fer of the ion

ways efficie

conductance

ecome neutra

hnique that c

otion to cycl

s to the excit

ectric field. U

f ions' magn

decreases rap

of the cell b

f the applied

nly the ions

46

n into

ent. If

limit

alized.

an be

lotron

tation

Upon

netron

pidly.

before

d RF

with

cy

ax

ex

se

2

in

Io

m

yclotron freq

xialized into

xcitation and

ection 2.4.4.

Figure 2.22

.4.4.3 Ion Ex

Ion ex

n Section 2.4

on detection

move on orbi

quency that i

o the center

d dipolar ex

3) respective

2 Depiction oan

xcitation and

xcitation is a

4.4.4) and co

is discussed

its of small r

is in resonan

r of the cell

xcitation for

ely.

of (a) quadrund (b) dipola

d Detection

an essential

ollision-activ

d here. The i

radii in the c

nce with the

l. Figure 2.2

the purpose

upolar excitaar excitation

step for ion

vated dissoc

ions' kinetic

cell. In order

applied qua

22 shows a

es of QA and

ation for quafor ion dete

detection as

ciation (discu

energies are

r for them to

adrupolar RF

a comparison

d ion detect

adrupolar axection.

s well as iso

ussed in Sec

e typically ve

o be detected

F potential w

n of quadru

tion (discuss

ialization (Q

olation (discu

ction 2.4.4.5

ery low, and

d, they need

47

will be

upolar

sed in

QA)

ussed

).53,64

d they

to be

48

kinetically excited to larger radii so that they move closer to the detection plates.

Moreover, ions move in random phases in FT-ICR. They need to be excited to move

coherently as ion packets to yield detectable signal. Therefore, a dipolar RF voltage is

applied to the excitation plates to kinetically excite the ions. When the cyclotron

frequency of an ion is in resonance with the RF frequency, it will absorb energy from the

field, accelerate, and move to a larger orbit. An ion's radius after excitation is given by

Equation 2.24,

Equation 2.24

wherein is the peak-to-peak RF voltage, is the excitation time, is the distance

between the two excitation plates, and is the strength of the magnetic field. It should be

noted that this radius is independent of an ion's m/z ratio, thus all ions will be excited to

the same radius with the same RF amplitude. Typically, a frequency sweep, known as

"chirp", is employed to excite ions of different m/z ratios to larger orbits. In the

experiments discussed here, a chirp excitation with a bandwidth of 2.7 MHz and sweep

rate of 3200 Hz/μs was used to excite ions for detection.

When the excited ion packet passes by the detection plates, a small current called

image current is induced.64-66 The image current contains frequency and amplitude

information, which correspond to ion's cyclotron frequency and relative abundance,

respectively. The image current is converted to voltage, digitalized, and recorded as a

function of time (transient). Fourier transformation generates a frequency domain

spectrum, which can be subsequently converted to a mass spectrum by using Equation

2.18. The process of ion excitation and detection is described in Figure 2.23. Image

cu

in

a

re

m

urrent detect

n the FT-ICR

mass spect

esolution, o

molecules pre

Figure 2.23kinetically

detection pl

tion is a non

R cell after d

trum with h

therwise ion

esent in the c

3 Illustrationexcited to mates and ind

n-destructive

detection. A

igher resolu

n signal wi

cell.

n of ion excitmove coherenduces an imag

spectrum a

e technique,

transient obt

ution. High

ill decay du

tation and dently as ion page current, wand finally m

which mean

tained for lo

vacuum is

ue to collis

etection in aackets. (b) A

which is conmass spectrum

ns that ions c

onger period

also critical

sions with n

an FT-ICR ceAn ion packenverted to frem.

can still be s

of time resu

l to achieve

neutral atom

ell. (a) Ions aet passes by equency dom

49

stored

ults in

high

ms or

are the

main

50

2.4.4.4 Ion Isolation

In the process of generating desired radical ions, unwanted ions are always also

generated. As mentioned in the previous section, ion's kinetic excitation can also be used

for ion isolation. Different from ion detection, where all ions are kinetically excited to

radii that are slightly smaller than the dimensions of the cell, ion isolation is achieved by

exciting unwanted ions to radii that are larger than the cell dimensions. The unwanted

ions can be ejected when colliding with the cell plates.

The broadband frequency sweep, known as chirp excitation, can be employed to

excite and eject ions with a broad mass range.53,67 However, this method suffers from two

major limitation. First, the RF amplitude varies throughout the frequency domain,

resulting in non-uniform ion excitation. Therefore, not all unwanted ions will be ejected

from the cell. Second, tailing in both ends of the RF chirp may result in ejection of

desired ions and hence poor ion signal. These drawbacks can be avoided by the use of

stored-waveform inverse Fourier transform (SWIFT) excitation. 67

SWIFT is a tailored excitation method.68-70 A frequency-domain waveform is pre-

defined. The frequencies are determined by the mass-to-charge ratios of the selected ions

that are to be ejected. An inverse Fourier-transformation is applied to this waveform to

generate a time-domain excitation waveform, which is then used to excite all the selected

ions simultaneously. The processes of chirp and SWIFT excitation are shown in Figure

2.24. In the work described in this dissertation, chirp was used for ion excitation and

detection, while SWIFT was used for isolation of charged radicals.

Figure 2.24 Compariison of chirpp excitation ((left) and SWWIFT excitations (right)

51

.

52

2.4.4.5 Collision-activated dissociation (CAD) in FT-ICR

As discussed above, kinetic excitation is not only necessary for ion detection and

isolation, but is also used for ion dissociation. In FT-ICR, a single-frequency RF voltage

corresponding to the cyclotron frequency of the ion of interest can be applied to the

excitation plates to kinetically excite the desired ion. In the presence of an inert gas, the

kinetic energy of the excited ion is converted into internal energy through multiple

collisions. When sufficient amount of internal energy is accumulated by the ion,

fragmentation occurs. This process is called on-resonance collision-activated dissociation

(CAD).67 The characteristic fragment ions formed through CAD can provide structural

information for the parent ion, which is particularly useful in structural elucidation of an

unknown compound.

In the case of generating radical ions, off-resonance irradiation collision-activated

dissociation (SORI-CAD) is often used.71,72 As its name suggests, an off-resonance RF

voltage is applied to the excitation plates to excite the ion of interest. Typically, an RF

frequency 1000 Hz higher or lower than the ion's cyclotron frequency is used, which

causes the ion's cyclotron radius to increase and decrease repeatedly. In the course of

SORI-CAD, the ion of interest gains less kinetic energy as compared to on-resonance

CAD. Through collisions with inert gas, the ion's internal energy increases slowly,

resulting in its dissociation via the lowest energy pathway.

In this dissertation, SORI-CAD was employed to generate radical ions for gas-

phase reactivity studies. Argon (typically at the pressure of ~10-5 Torr) was used as the

inert gas for collisions. The radical precursors selected have weakly bound nitro- or iodo-

substituents, which can be cleaved off via homolytic cleavages to generate radical sites

53

upon off-resonance CAD. The number of SORI-CAD events needed is dependent on the

number of radical sites desired, being one for monoradicals and one or two for biradicals.

After the radical ions were successfully generated, they were cooled through collisions

with neutral molecules or atoms before their ion-molecule reactions were examined.

2.5 Fundamental Aspects of Gas-phase Ion-Molecule Reactions

Reactions between ions and neutral molecules, which are most often studied in

mass spectrometers, have received considerable interest for a long time.70,73-76 They can

be used as an alternative method to probe the structure of an ion of interest, because

sometimes dissociation reactions do not give much useful structural information.

Moreover, ion-molecule reactions are proven to be a powerful tool for studying

thermodynamics and kinetics of reactions. Compared to solution, gas phase provides an

ideal environment to investigate the intrinsic chemical properties of highly reactive

intermediates without interference of solvent. For the work discussed in this dissertation,

gas-phase ion-molecule reactions were used to study the chemical properties of carbon-

centered ơ-type mono- and biradicals that cannot be easily generated in solution. This

section introduces the fundamental and experimental aspects of gas-phase ion-molecule

reactions in FT-ICR and LQIT.

2.5.1 Brauman’s Double-Well Potential Energy Surface

The potential energy surface for gas-phase ion-molecule reactions is different

from that in solution (Figure 2.25). In solution, both endothermic and exothermic

54

reactions can occur. In the case of endothermic reactions, the reactants acquire energy

from outside of the system to overcome the reaction barrier.77 In gas phase in high

vacuum, however, energy is conserved, and the overall reaction must be exothermic to

occur. In order to explain the rates of gas-phase ion-molecule reactions, Brauman

proposed a double-well potential energy surface model.78-80 Based on this model,

reactions in the gas phase proceed through formation of a reactant complex and product

complex. The reactant complex is formed between the ion and a neutral reagent molecule

due to long-range ion-dipole and/or ion-induced dipole forces.81 These attractive forces

lower the potential energy of the reactant complex, which is known as solvation energy.

The magnitude of this energy is dependent on the dipole moment and polarizability of the

neutral molecule. The solvation energy is available for the reactant complex to overcome

energy barriers along the reaction pathways. Ultimately, whether the reactant complex

can proceed to form products or dissociate back to separated reactants is determined by

the height of the reaction barrier. If this barrier is greater than the total energy of the

reactants, the reaction cannot occur. The rate at which the reaction occurs is largely

controlled by the energy difference (∆E) between the transition state and the separated

reactants.

Gas-phase ion-molecule reactions do not always occur, even when the net

reaction is exothermic and the system has enough energy to overcome the reaction barrier.

This is due to entropy constraints.79 As shown in Figure 2.26, the transition state leading

to product formation can be "tight" (low entropy) as the orientation of the complex is

specific, which has few rotational modes to it. However, the transition state leading to

separated reactants is "loose" (high entropy) as there are many rotational modes available.

T

b

F

Therefore, so

ack to separ

igure 2.25 C

ometimes it

ated reactan

Comparison o

is entropica

ts than to pr

of potential phase (top)

ally favored

oceed to form

energy surfa) and in solu

d for the rea

rm products.

aces for ion-ution (bottom

actant comp

-molecule rem).

plex to disso

actions in th

55

ociate

he gas

w

o

re

co

on

ra

Figure 2.2

In the

were investig

f interest. Th

eactions. It

onstant, and

n the param

ate constant

26 Braumanconst

2

e work desc

gated, becaus

he reaction

can be calcu

d is the

meterized traj

measuremen

n double-weltraints for a g

2.5.2 Kinetic

ribed in this

se they shed

efficiency is

ulated by E

theoretical c

ajectory theo

nts is estima

ll potential egas-phase io

cs of Ion-Mo

s thesis, rea

d light on the

s defined as

quation 2.25

collision rate

ory proposed

ated to be 5

energy surfacon-molecule

olecule Reac

action efficie

e intrinsic ch

the percent

5, wherein

e constant w

d by Su.82 T

50%, and the

ce illustratinreaction.

tions

encies and r

hemical prop

tage of collis

is the e

which can be

The accuracy

e precision i

ng the entrop

reaction pro

perties of th

sions that le

experimenta

calculated b

y of the rea

s less than

56

py

oducts

he ion

ead to

l rate

based

action

10%.

57

100Equation 2.25

The gas-phase ion molecule reactions studied here follow second-order kinetics as

shown in Equation 2.26, where indicates the reaction rate, is the second-order

experimental reaction rate constant, and are the concentrations of the neutral

reagent and the ion, respectively.

Equation 2.26

Under the experimental conditions used here, the neutral reagent is in great excess

as compared to the radical ions. Hence, the concentration of the neutral reagent is

presumed to be constant throughout the reaction. Therefore, the reactions can be assumed

to follow pseudo-first order kinetics as shown in Equation 2.27, where equals

. can be derived by using Equation 2.28.

Equation 2.27

Equation 2.28

In order to calculate , and need to be determined. As shown in

Figure 2.27, plotting ln versus time generates a line with a slope equal to - ,

wherein and are the relative abundances of the reactant ion at time t and time

zero. The concentration of the neutral reagent [N] is derived from its nominal pressure (P)

read by an ion gauge. A conversion factor (1 Torr = 3.239 × 1016 molecules/cm3) is used

58

to convert the pressure of the neutral reagent into its concentration. An ion gauge

correction factor (IGCF) is also necessary to correct for the location of the ion gauges,

because they are located about one meter above the cell where the reactions take place.

The ion gauge correction factor is obtained by measuring the reaction rate of an

exothermic reaction that is assumed to occur at collision rate, such as electron transfer or

proton transfer reaction. For instance, carbon disulfide radical cation is used to abstract

an electron from the neutral reagents, or protonated acetone or methanol is used to

transfer a proton to the neutral molecules. Additionally, the ion gauge's sensitivity

towards different neutral reagents (B/I) is corrected.83 In summary, the experimental

reaction rate constant can be experimentally determined by using Equation 2.29.

. /

Equation 2.29

m

ab

re

fr

th

F

pr

Figure 2.27

Besid

measured. Th

bundance by

elative abun

rom seconda

he primary p

igure 2.27,

roduct of pro

A semi-loga

es reaction

he branching

y the sum o

ndances with

ary products

products and

products 1

oduct 1.

arithmic plotpro

rates, the re

g ratio for a p

f all primary

hin short rea

. Secondary

d further mo

and 2 are

t of the relatoducts versu

eaction prod

primary prod

y products' a

action times,

products ca

onitoring the

primary pro

tive abundanus time.

ducts and th

duct is determ

abundances.

, primary pr

an also be co

eir reactions

oducts, whil

nces of a reac

eir branchin

mined by div

. By measur

roducts can

onfirmed by

s. For the ex

le product 3

ctant ion and

ng ratios are

viding its rel

ring the prod

be distingu

isolating ea

xample show

3 is a secon

59

d its

e also

lative

ducts'

uished

ach of

wn in

ndary

60

2.6 References

1. Gross, J. H. Mass Spectrometry, Springer: New York, 2004.

2. McLuckey, S. A.; Wells, J. M. Chem. Rev. 2001,101, 571.

3. Stephenson, J. L.; McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1998, 9, 585.

4. Cooks, R. G.; Wong, P. S. H. Acc. Chem. Res. 1998, 31, 379.

5. Dempster, A. Phys. Rev. 1918, 11, 316.

6. Munson, M.; Field, F. J. Am. Chem. Soc. 1966, 88, 2621.

7. Whitehouse, C.; Dreyer, R.; Yamashita, M.; Fenn, J. Anal. Chem. 1985, 57, 675.

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

9. Carroll, D.; Dzidic, I.; Stillwell, R.; Haegele, K.; Horning, E. Anal. Chem.1975, 47, 2369.

10. Carroll, D.; Dzidic, I.; Stillwell, R.; Horning, M.; Horning, E. Anal. Chem. 1974, 46, 706.

11. Raffaelli, A.; Saba, A., Mass Spectrom. Rev. 2003, 22, 318.

12. Takáts, Z.; Wiseman, J.; Gologan, B.; Cooks, R. Science 2004, 306, 471.

13. Cody, R.; Laramée, J.; Durst, H., Anal. Chem. 2005, 77, 2297.

14. Karas, M.; Bahr, U.; Gießmann, U. Mass Spectrom. Rev. 1991, 10, 335.

15. Houk, R.; Thompson, J. Mass Spectrom. Rev. 1988, 7, 425.

16. Barber, M.; Bordoli, R.; Sedgwick, R.; Tyler, A., J. Chem. Soc. Chem. Commun. 1981, 7, 325.

17. Beckey, H. Int. J. Mass Spectrom. Ion Phys. 1969, 2, 500.

18. Lindner, B.; Seydel, U. Anal. Chem. 1985, 57, 895.

19. Perez, J.; Ram rez-Arizmendi, L.; Petzold, C.; Guler, L.; Nelson, E.; Kenttämaa, H. I. Int. J. Mass Spectrom. 2000, 198, 173.

61

20. Mark, T. D. Electron Impact Ionization in Gaseous Ion Chemistry and Mass Spectrometry; John Wiley and Sons: New York, 1986.

21. Vestal, M. L. Chem. Rev. 2001, 101, 361.

22. Munson, M. S. B. Int. J. Mass Spectrom. 2000, 200,243.

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

24. Cech, N.; Enke, C. Mass Spectrom. Rev. 2001, 20, 362.

25. Nayak, R.; Knapp, D. Anal. Chem. 2010, 82, 7772.

26. de Hoffmann, E.; Stroobant, V. Mass Spectrometry: Principles and Applications. 3rd

Ed.; John Wiley & Sons, Ltd: West Sussex, England, 2007.

27. Thomson, B. A. J. Am. Soc. Mass Spectrom. 1998, 9, 187. 20.

28. Sunner, J.; Nicol, G.; Kebarle, P. Anal. Chem. 1988, 60, 1300.

29. Kostiainen, R.; Kauppila, T. J. J. Chromatogr. A 2009, 1216, 685.

30. Owen, B. C.; Gao, J.; Borton II, D. J.; Amundson, L. M.; Archibold, E. F.; Tan, X.; Azyat, K.; Tykwinski, R.; Gray, M.; Kenttämaa, H. I. Rapid Commun. Mass Spectrom. 2011, 5, 1924.

31. Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. J. Am. Soc. Mass Spectrom. 2002, 13, 659.

32. Hager, J. W. Rapid Commun. Mass Spectrom. 2002, 16, 512.

33. March, R. E. J. Mass Spectrom. 1997, 32, 351.

34. Collings, B.; Campbell, J.; Mao, D.; Douglas, D. Rapid Commun. Mass Spectrom. 2001, 15, 1777.

35. Makarov, A.; Denisov, E.; Kholomeev, A.; Balschun, W.; Lange, O.; Strupat, K.; Horning, S. Anal. Chem. 2006, 78, 2113.

36. Syka, J.; Marto, J.; Bai, D.; Horning, S.; Senko, M.; Schwartz, J.; Ueberheide, B.; Garcia, B.; Busby, S.; Muratore, T.; Shabanowitz, J.; Hunt, D. J. Proteome Res. 2004, 3, 621.

37. Bier, M. E.; Syka, J. E. P. US Patent No 5,420,425, 1995.

62

38. March, R. E. Quadrupole ion trap mass spectrometry. Wiley: Hoboken, NJ, 2005.

39. March, R. E., Int. J. Mass Spectrom. Ion Processes 1992, 118/119, 71.

40. March, R. E., Mass Spectrom. Rev. 2009, 28, 961.

41. Todd, J. Mass Spectrom. Rev. 1991, 10, 3.

42. Stafford, G. C.; Kelley, P. E.; Syka, J. E. P.; Reynolds, W. E.; Todd, J. F. J. Int. J.

Mass Spectrom. Ion Processes 1984, 60, 85.

43. Kaiser, R. E.; Cooks, R. G.; Stafford, G. C.; Syka, J. E. P.; Hemberger, P. H. Int. J. Mass Spectrom. Ion Processes 1991, 106, 79.

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

45. McLuckey, S. A.; Goeringer, D. E. J. Mass Spectrom. 1997, 32, 461.

46. Louris, J. N.; Cooks, R. G.; Syka, J. E. P.; Kelley, P. E.; Stafford, G. C.; Todd, J. F. J. Anal. Chem. 1987, 59, 1677.

47. S. Osburn, V. Ryzhov, Anal. Chem. 2013, 85, 769.

48. Schwartz, J.; Wade, A.; Enke, C.; Cooks, R. Anal. Chem. 1990, 62, 1809.

49. Lawrence, E. O.; Livingston, M. S. Phys. Rev. 1932, 40, 19.

50. Comisarow, M. B.; Marshall, A. G. J. Mass Spectrom. 1996, 31, 581.

51. Baldeschwieler, J. D. Science 1968, 159, 263.

52. Anders, L. R.; Beauchamp, J. L.; Dunbar, R. C.; Baldeschwieler, J. D. J. Chem. Phys. 1966, 45, 1062.

53. Comisarow, M. B.; Marshall, A. G. Chem. Phys. Lett. 1974, 25, 282.

54. Marshall, A. G.; Comisarow, M. B.; Parisod, G. J. Chem. Phys. 1979, 71, 4434.

55. He, F.; Hendrickson, C. L.; Marshall, A. G. Anal. Chem. 2001, 73, 647.

56. Marshall, A. G.; Hendrickson, C. L.; Stone, D. H. S. Anal. Chem. 2002, 74, 252A.

63

57. Borton, D. J.; Amundson, L. M.; Hurt, M. R.; Dow, A.; Madden, J. T.; Simpson, G. J. Kenttämaa, H. I. Anal. Chem. 2013, 85, 5720.

58. Amster, I. J. J. Mass Spectrom. 1996, 31, 1325.

59. Dunbar, R. C. J. Chem. Phys. 1989, 90, 7369.

60. Huang, F. S.; Dunbar, R. C. J. Am. Chem. Soc. 1989, 111, 6497.

61. Guan, S. H.; Kim, H. S.; Marshall, A. G.; Wahl, M. C.; Wood, T. D.; Xiang, X. Z. Chem. Rev. 1994, 94, 2161.

62. Guan, S. H.; Marshall, A. G. J. Chem. Phys. 1993, 98, 4486.

63. Savard, G.; Becker, S.; Bollen, G.; Kluge, H. J.; Moore, R. B.; Otto, T.; Schweikhard, L.; Stolzenberg, H.; Weiss, U. Phys. Lett. A 1991, 158, 247

64. Comisaro, M. B.; Marshall, A. G. Chem. Phys. Lett. 1974, 26, 489.

65. Marshall, A. G.; Roe, D. C. J. Chem. Phys. 1980, 73, 1581.

66. Wang, M.; Marshall, A. G. Int. J. Mass Spectrom. Ion Processes 1990, 100, 323.

67. Cody, R. B.; Burnier, R. C.; Freiser, B. S. Anal. Chem. 1982, 54, 96.

68. Marshall, A. G.; Wang, T. C. L.; Ricca, T. L. J. Am. Chem. Soc. 1985, 107, 7893.

69. Chen, L.; Wang, T. C. L.; Ricca, T. L.; Marshall, A. G. Anal. Chem. 1987, 59, 449.

70. Gronert, S. Chem. Rev. 2001, 101, 329.

71. Gauthier, J. W.; Trautman, T. R.; Jacobson, D. B. Anal. Chim. Acta 1991, 246, 211

72. Mirgorodskaya,E.; O'Connor, P. B.; Costello, C. E. J. Am. Soc. Mass Spectrom. 2002, 13, 318.


74. Eberlin, M. J. Mass Spectrom. 2006, 41, 141.

75. Heaton, A. L.; Armentrout, P. B. J. Am. Chem. Soc. 2008, 130, 10227.

76. Eismin,R. J.; Fu, M.; Yem, S.; Widjaja, F.; Kenttämaa, H. I. J. Am. Soc. Mass Spectrom. 2012, 23, 12.

64

77. Smith, M. B.; March, J. March's Advanced Organic Chemistry: Reactions, Mechanisms, and Structure. Wiley-Interscience: New York, 2001.

78. Farneth, W. E.; Brauman, J. I. J. Am. Chem. Soc. 1976, 98, 7891-7898.

79. Olmstead, W. N.; Brauman, J. I. J. Am. Chem. Soc. 1977, 99, 4219.

80. Pellerite, M. J.; Brauman, J. I. J. Am. Chem. Soc. 1980, 102, 5993.

81. Su, T.; Bowers, M. T. Gas Phase Ion Chemistry; Bowers, M. T., Ed.; Academic Press: New York, 1979.

82. Su, T.; Chesnavich, W. J. J. Chem. Phys.1982, 76, 5183.

83. Bartmess, J. E.; Georgiadis, R. M. Vacuum 1983, 33, 149.

65

CHAPTER 3. STRUCTURAL COMPARISON OF ASPHALTENES OF DIFFERENT

ORIGINS BY USING MULTIPLE-STAGE TANDEM MASS SPECTROMETRY

3.1 Introduction

Asphaltenes, the heaviest components in crude oil, are generally defined by its

solubility regime: insoluble in n-alkanes and soluble in aromatic solvents such as toluene,

benzene, or pyridine.1,2 They are extremely complex mixtures containing molecules with

multiple fused aromatic rings, alkyl chains, heteroatoms and metals.1Asphaltenes are

problematic to the petroleum industry since they result in reduced oil recovery,

precipitation in transport pipelines, adsorption on refinery equipments, fouling of

catalysts used in crude oil conversion2-4 Moreover, the depletion of conventional lighter

crude oils necessitates exploration of heavy crudes that have a high concentration of

asphaltenes.5 In order to address these emerging problems and potentially convert

asphaltenes to useful chemicals, an in-depth understanding of the molecules that

comprise asphaltenes is necessary.6

A wide range of analytical methods have been used to interrogate the bulk

properties of asphaltenes. Nuclear magnetic resonance (NMR) spectroscopy sheds light

on the molecular parameters such as carbon aromaticity and average size of asphaltenes;7

X-ray absorption near edge structure (XANES) measures the sulfur and nitrogen

functionalities.8,9

66

The molecular weight distribution of asphaltene remains controversial.10-12 Vapor

pressure osmometry, size-exclusion chromatography yielded molecular weight of several

thousand amu or even larger values.13-15 However, fluorescence depolarization indicated

an average molecular weight (MW) of 450-850 Da.16-20 Other literatures using different

techniques such as NMR and mass spectrometry summarized a range of 500-2000 Da.1,19-

27 The disparity in measurements was suggested as a result of asphaltene aggregation.27,28

In the past decades, two structural models are debated about the asphaltene

molecules: the island model and the archipelago model.16,19-22,29The island model

(sometimes referred as continental model30) has only one aromatic core with peripheral

alkyl chains, whereas the archipelago model has multiple aromatic cores that are bridged

by alkyl chains and may also containperipheral alkyl chains.20 Multiple experimental

methods, such as time-resolved fluorescence depolarization, Taylor diffusion, and NMR

spectroscopy provide strong support to the island model.16,18-23,31,32 However, the

presence of archipelago structures has also been demonstrated by NMR spectroscopy and

average structural parameter calculations,33 mass spectrometry,34 and thermal cracking of

asphaltenes.35,36 Which of these two motifs predominates asphaltene structure has not

reached a general consensus yet.

Mass spectrometry is an important analytical tool to characterize asphaltenes at

the molecular level, yet it faces many challenges. Asphaltenes have a tendency to degrade

and aggregate upon introduction into the gas phase.28,37-39 Compatible solvent system

needs to be carefully chosen for asphaltene desolvation.28,40 Ionization bias is another

concern because of the highly complex composition of asphaltene mixture. 41

Electrospray ionization (ESI) is especially suitable for ionizing polar constituents with

67

high heteroatom content; atmospheric pressure chemical ionization (APCI) and

atmospheric pressure photo ionization (APPI) are more suitable for ionizing nonpolar

hydrocarbons.37,42-43 In addition, a variety of desorption/ionization methods have been

used for asphaltene analysis, such as matrix-assisted laser desorption/ionization,14 field

desorption/field ionization,44,45 and laser-induced acoustic desorption/electron

ionization.39

In this work, positive ion mode APCI doped with carbon disulfide (CS2) was used

to study six petroleum asphaltenes samples of different geographical origins in a linear

quadrupole ion trap (LQIT) mass spectrometer. CS2 reagent has been demonstrated to

generate stable molecular ions for the asphaltenes based on previous experiments.25,46-47

Furthermore, multi-stage tandem mass spectrometry was employed to examine the

structures of asphaltene molecules by subjecting the selected molecular ions to

collisionally activated dissociation (CAD). In addition to the molecular weight

distribution (MWD) and average MW, structural information, including maximum

number of carbons in alkyl chains and minimal sizes of the aromatic cores, was obtained.

3.2 Experimental Section

Chemicals. The petroleum asphaltene samples from Bohai (China), Maya

(Mexico), Claire (UK), Surmont (Canada), Montana (US), and McKittrick (US) were

provided by ConocoPhillips. Carbon disulfide (>99.9 %) and heptane (>99.9 %) were

purchased from Sigma-Aldrich (St. Louis, MO) and used without further purification.

The samples were dissolved in heptane and sonicated for 1 hour, followed by filtration

and drying under inert gas flow to remove the maltene content.

68

Instrumentation. A Thermo Scientific linear quadrupole ion trap (LQIT) was

used for mass spectrometric analysis. The asphaltenes were dissolved in CS2 at a

concentration of 0.5 mg/mL. The sample solutions were introduced into the APCI source

via direct infusion from a Hamilton 500 μL syringe through the instrument’s syringe

pump at a flow rate of 20 µL/min and ionized via positive ion mode APCI (at 300°C) by

using CS2 as a dopant so that only molecular ions were generated. Molecular ions with

eight randomly selected mass-to-charge (m/z) ratios ranging from m/z 500 up to m/z

808were isolated using an isolation window of 2 Da (±1 Da), and subjected to CAD at an

energy of 35 arbitrary units. The use of an isolation window of 2 Da (±1 Da) results in

isolated ion populations that may contain isomeric and isobaric ions. This large window

was used due to the relatively low ion signals measured for these complex mixtures. This

is justified as it was previously demonstrated that the fragmentation patterns and main

fragment ions of ionized asphaltenes are independent of the size of the isolation window

as long as it is equal or less than 2 Da.25 The data were processed by using Thermo

Xcalibur software. All measurements were repeated four times. The averaged results or

ranges are reported in Table 3.1.

3.3 Results and Discussion

In this study, APCI doped with CS2 was used to ionize asphaltenes in the positive

ion mode so that only stable molecular ions were generated. 25,46-47 Based on the measured

mass spectra, the molecular weight distributions (MWD) were determined for six

petroleum asphaltene samples originating from Bohai, Maya, Claire, Surmont, Montana

and McKittrick. All of the studied asphaltenes from the American continent, Europe and

C

w

sa

w

an

(r

8

n

d

sh

China have s

weight (AVG

amples with

with an AVG

Fi

Struct

nd MS3 exp

randomly se

08) were iso

ominal colli

erived from

how a simila

similar MW

G MW), wh

h different g

G MW of ~61

igure 3.1 AP

tural features

periments of

elected) mas

olated and s

ision energy

the differen

ar decay patt

WD, ranging

hich was cal

geographical

15 Da is show

PCI mass speof the M

s of the asph

f the isolated

s-to-charge

subjected to

y. The fragm

nt asphaltene

tern (for exa

from 200 u

lculated usin

l origins (Ta

wn Figure 3

ectrum showMaya asphalt

haltenemolec

d ions. For

ratios (m/z

collision-ac

mentation pat

e samples w

ample, see Fi

up to 1400

ng Equation

able 3.1). A

.1 for Maya

wing the MWtene sample.

cules were e

each sampl

500, 515, 6

ctivated diss

tterns of mo

were compar

igure 3.2 for

Da. The av

n 1,25 varies

A typical mo

a asphaltenes

WD and AVG

examined by

e, molecular

606, 626, 63

sociation (C

olecular ions

red. All the

r the fragmen

verage mole

s slightly am

onomodal M

s.

G MW

y performing

r ions with

34, 704, 736

AD) at the

of the same

ions consist

ntation patte

69

ecular

mong

MWD

g MS2

eight

6, and

same

e m/z

tently

ern of

th

ra

on

T

F

±

[M

ar

fr

m

fr

du

he molecula

adical loss, l

n, with the

The reason fo

Figure 3.2 Fr

For co

± 1 derived f

M-CH3]＋ fra

re presented

ragment ion

molecular ion

ragment ion

ue to the [M

ar ions of m

less favored

larger alkyl

or such aspha

ragmentation

omparison p

from the Ma

agment ions

d below. Int

ns derived f

ns themselv

s have great

M-CH3]＋ ion

m/z 634 + 1

ethyl radica

radical loss

altene behav

n pattern of m

purpose, MS

aya asphalten

of m/z 485

terestingly, M

from the m

es (Figure 3

ter abundanc

n population

of Surmon

al loss, even

ses always b

viors is still u

molecular iosample.

2 CAD mass

nes sample,

± 1 that wer

MS3 mass s

molecular ion

3.3) although

ces than slig

n being a som

nt asphaltene

n less favore

being less fa

under invest

ons of m/z 63

s spectrum o

as well as M

re formed fr

spectrum (F

ns show sim

h not quite

ghtly larger

mewhat sim

es), with a d

ed propyl rad

avored than t

tigation.

34 ± 1 of Sur

of molecular

MS3 CAD m

rom the abov

Figure 3.4) o

milar decay

as smooth,

fragment io

mpler mixtur

dominant m

dical loss, an

the smaller

rmont aspha

r ions of m/z

mass spectru

ve molecular

of the [M-C

y patterns a

as some sm

ons. This ma

e of isobaric

70

methyl

nd so

ones.

altene

z 500

um of

r ions

CH3]＋

as the

maller

ay be

c and

is

io

le

co

io

su

someric ions

ons formed

engths. Hen

ommonly ob

ons fragmen

urprising as

Figure 3.3asphaltene

s than the m

from the [M

nce, these fr

beyed by fra

nt to yield e

aromatic ion

3 MS2 CAD sample, with

molecular ion

M-CH3]＋ions

ragmentation

agmentations

even-electro

ns are the mo

mass spectrh the maximestimated ar

ns. Similar to

s correspond

ns do not o

s observed i

on and not o

ost common

rum of ions omum total num

romatic core

o molecular

d to losses o

obey the Ev

in mass spec

odd-electron

n species sho

of m/z 500 ±mber of carbe size indicat

ions, all the

of alkyl rad

ven Electro

ctrometry (i.

n ions). Thi

owing this ty

± 1derived frbons in alkylted.

e major frag

dicals of diff

on Rule48 th

.e., even-ele

is is not en

ype of behavi

rom the Mayl chains and

71

gment

ferent

hat is

ectron

ntirely

ior.48

ya the

F

m

ex

io

th

4

ob

m

io

fr

d

m

igure 3.4 Mformed fro

Comp

molecular io

xperiment) r

ons of m/z 5

he isolated io

71 and m/z

btained for

molecular ion

ons of m/z

ragment ion

emonstrated

molecules.24

S3 CAD masom molecula

parison of th

ns (MS2 ex

reveals that

00 yield maj

ons of m/z 4

457 (Figur

the molecu

ns and not v

457 are for

ns of m/z 4

d that this is

ss spectrum ar ions of m/z

sam

he m/z valu

xperiment) a

they differ b

jor fragment

485 forms fr

re 3.4). Thes

ular ions in

via further f

rmed directl

485 or m/z

also true for

of [M-CH3]z 500 ± 1dermple (Figure

es of the fr

and those o

by one mass

t ions of m/z

ragment ions

se observati

n MS2 expe

fragmentatio

ly from mo

471, as sh

r losses of a

＋ fragment irived from the 3.3).

ragment ions

obtained fro

s unit. For e

z 485, m/z 4

s of m/z 470

ions demons

eriments are

on of larger

lecular ions

hown in Fig

alkyl radicals

ions of m/z 4he same May

s obtained d

om the [M-

example, wh

471 and m/z

0 and m/z 45

strate that th

e formed d

fragment io

s of m/z 50

gure 3.3. A

s from proto

485 ± 1 that ya asphalten

directly from

CH3]＋ions

hile the mole

457 (Figure

56 instead o

he fragment

directly from

ns. For exam

00 instead o

A previous

onated aspha

72

were nes

m the

(MS3

ecular

e 3.3),

f m/z

t ions

m the

mple,

of the

study

altene

73

The MS2 and MS3 results discussed above indicate that the fragmenting ion

populations in all cases contain many isomeric and possibly isobaric ions with alkyl

chains of differing lengths since no single structure can undergo losses of so many

different alkyl radicals. Further, the results support the island model more than the

archipelago structural model due to the absence of facile losses of large aromatic moieties

that would be expected for archipelago structures. These findings are in agreement with

earlier results obtained by tandem mass spectrometry for both protonated and molecular

ions of asphaltenes.24,25

Based on MS2 CAD mass spectrum of the selected asphaltene molecular ions, the

maximum total number of carbons in alkyl chains was determined by counting the

number of carbons in the largest eliminated alkyl radical that formed the smallest

detectable fragment ion (considered to be the ion with approximately 1% relative

abundance from the most abundant ion in the mass spectrum). The aromatic core size was

estimated by counting the maximum number of fused aromatic rings possible for an ion

to reach the m/z value of this fragment ion (methylene functional groups possibly left on

the aromatic core after loss of the alkyl chains). For molecular ions of m/z 500 ±1 derived

from the Maya asphaltene sample as shown in Figure 3.3, the smallest detectable

fragment ion has an m/z value of 233, therefore it is determined that there are 4 fused

aromatic rings after cleavage of alkyl chains with at least 19 carbons. It should be noted

that the above determination is based on the assumption that all non-aliphatic carbons are

aromatic carbons, and all aromatic rings are fused. The accuracy of this determination

method is under investigation as the heteroatoms and metals are not taken into account,

yet the

74

above values would not be significantly changed. Overall, it provides insight into the

qualitative structural comparisons of asphaltene samples of different geographic origins.

The molecular weight and structural information for the eight selected molecular

ions derived from the six asphaltenes samples were summarized in Table 1. Since the

experiments were repeated for four times, a slightly different ion population may be

isolated and fragmented each time. Therefore, the maximum total number of carbons in

alkyl chains and the estimated aromatic core size may differ, which are represented by

ranges respectively. For visualization of the general trends, the number of aliphatic

carbons, as well as estimated aromatic core size, was plotted against a series of MWs

studied for the six asphaltene samples respectively, as shown in Figure 3.5 and Figure

3.6. The results that fall within a range are represented by the average value and standard

error.

For asphaltene molecules with MWs ranging from 500 to 808 Da, the maximum

total number of carbons in the alkyl chains ranges from 17 to 41, while the approximate

number of aromatic rings ranges from 3 to 7. These results are consistent with other

reports suggesting that asphaltene aromatic moieties contain 4-10 fused rings, and the

length of aliphatic chains covers a wide range up to 30-40 carbon atoms.49-50 Generally,

molecules of greater MWs were found to have more carbons in alkyl chains, yet they do

not always have more aromatic rings in the cores. For instance, molecules of smaller

MWs (from 500 to 626 Da) derived from Montana, Maya and Surmont asphaltenes have

fewer carbons in alkyl chains and larger aromatic cores compared to those molecules of

75

larger MWs. Additionally, molecules of MWs ranging from 634 to 808 Da have about the

same number of carbons in alkyl chains and aromatic cores regardless of their geographic

origins.

Table 3.1 MWD and AVG MW of Molecules in the Six Asphaltenes Samples, and Structural Information for the Eight Selected Ions

Bohai Maya Claire Surmont Montana McKittrick

MWD 250-1450 200-1400 300-1500 200-1350 300-1400 200-1360 AVG MW 702 615 681 575 664 571

Ion of m/z 500 Maximum number of

Carbons in chains 22 17-21 21-22 17-21 17 21-22

Estimated Core Size 3 4 3-4 4-5 5 3-4 Ion of m/z 515

Maximum number of Carbons in chains

23-24 18-19 23-24 16-18 13-14 22-23

Estimated Core Size 3 5 3 5 7 3-4 Ion of m/z 606

Maximum number of Carbons in chains

29-30 28-30 28-30 27-28 25-28 28

Estimated Core Size 3 3-4 3-4 4 4-5 4 Ion of m/z 626

Maximum number of Carbons inchains

30-31 31 31 30 28-31 30-31

Estimated Core Size 3-4 3 3 4 3-5 3-4 Ion of m/z 634


30-32 30-32 31-32 31-32 30 30

Estimated Core Size 3-4 3-4 3-4 3-4 4 4 Ion of m/z 704


35 34-35 35 34-35 35 34

Estimated Core Size 4 4 4 4 4 4 Ion of m/z 736


36-37 35-36 37 36 36 36-37

Estimated Core Size 4 4-5 4 4 4 4 Ion of m/z 808


39-40 40 40 40-41 40 39-41

Estimated Core Size 5 5 5 5 5 5

76

Figure 3.5 General trend for the approximate maximum total number of carbons in alkyl chains as a function of MW of the molecules derived from the six asphaltene samples.

Figure 3.6 General trend for the approximate aromatic core size as a function of MW of the molecules derived from the six asphaltene samples.

0

5

10

15

20

25

30

35

40

45

500 515 606 626 634 704 736 808

MW

Num

ber

of A

liph

atic

Car

bons

Bohai

Montana

Maya

Claire

Surmont

McKittrick

0

1

2

3

4

5

6

7

8

500 515 606 626 634 704 736 808

MW

Aro

mat

ic C

ore

Siz

e Bohai

Montana

Maya

Claire

Surmont

McKittrick

77

3.4 Conclusions

Examination of six petroleum asphaltene samples of different geographical

origins by using positive ion mode APCI doped with CS2 and multi-stage tandem mass

spectrometry allowed the determination of the molecular weight distribution (MWD),

average molecular weight (MW) as well as structural information for these samples. The

asphaltene samples studied have similar MWDs, ranging from 200up to 1450 Da. The

average MWs range from 570 up to 700 Da, and are dependent on the origin of the

samples. Eight randomly selected molecular ions with m/z values ranging from m/z 500

up to m/z 808, derived from the different asphaltenes, all show a similar fragmentation

pattern, providing support to the island structural model of asphaltenes. The fragmenting

ion populations in all cases contain many isomeric and possibly isobaric ions with alkyl

chains of differing lengths since no single structure can undergo losses of so many

different alkyl radicals. The maximum total number of carbons (ranging from 17 to 41) in

all alkyl chains generally increase with the increase of MWs of the asphaltene molecules

ranging from 500 to 808 Da. However, the number of aromatic rings (ranging from 3 to

7) in the cores does not reveal an obvious correlation with MWs of the asphaltene

molecules.

78

3.5 References

1. Mullins, O. C.; Sheu, E. Y.; Hammami, A.; Marshall, A. G. Asphaltenes, Heavy Oils, and Petroleomics; Springer: New York, 2007.

2. Adams, J. J. Energy Fuels 2014, 28, 2831.

3. Ancheyta, J.; Betancourt, G.; Centeno, G.; Marroquin, G.; Alonso, F.; Garciafigueroa, E. Energy Fuels 2002, 16, 1438.

4. Trejo, F.; Centeno, G.; Ancheyta, J. Fuel 2004, 83, 2169.

5. Rana, M. S.; Samano, V.; Ancheyta, J.; Diaz, J. A. I. Fuel 2007, 86,1216.

6. Akbarzadeh, K.; Hamami, A.; Kharrat A.; Zhang, D.; Stephan Allenson; Creek, J.; Kabir, S.; Jamaluddin, A.; Marshall, A. G.; Rodgers, R. P.; Mullins, O. C.; Solbakken, T. Oilfield Rev. 2007, 19, 22.

7. Ostlund, J. A.; Wattana, P.; Nydén, M.; Fogler, H. S. J. Colloid Interface Sci. 2004, 271, 372.

8. Zhang, L.; Wang, C.; Zhao, Y.; Yang, G.; Su, M.; Yang, C. J. Fuel Chem.Tech. 41 (11), 2013, 1328–1335.

9. Pomerantz, A. E.; Seifert, D. J.; Bake, K. D.; Craddock, P. R.; Mullins, O. C.; Kodalen, B. G.; Mitra-Kirtley, S.; Bolin, T. B. Energy Fuels 2013, 27, 4604.

10. Mullins, O. C.; Martinez-Haya, B.; Marshall, A. G. Energy Fuels 2008, 22, 1765.

11. Herod, A. A.; Bartle, K. D.; Kandiyoti, R. Energy Fuels 2008, 22, 4312.

12. Strausz, O. P.; Safarik, I.; Lown, E. M.; Morales-Izquierdo, A. Energy Fuels 2008, 22, 1156.

13. Harvey W. Yarranton, H. A., Jakher, R. Ind. Eng. Chem. Res. 2000, 39, 2916.

14. Trejo, F.; Ancheyta, J.;Morgan, T. J.; Herod, A. A.; Kandiyoti, R. Energy Fuels 2007, 21, 2121.

15. Anderson, S. I.; Speight, J. G. Fuel 1993, 72, 1343.

16. Groenzin, H.; Mullins, O. C. J. Phys. Chem. A 1999, 103, 11237.

17. Buch, L.; Groenzin, H.; Buenrostro-Gonzalez, E.; Anderson, S. I.; Lira-Galeana, C.; Mullins, O. C. Fuel 2003, 82, 1075.

79

18. Badre, S.; Goncalves, C. C.; Norinaga, K.; Gustavson, G.; Mullins, O. C. Fuel 2006, 85, 1.

19. Mullins, O. C. SPE J. 2008, 13, 48.

20. Mullins, O. C. Annu. Rev. Anal. Chem. 2011, 4, 393.

21. Mullins, O. C. Energy Fuels 2010, 24, 2179.

22. Mullins, O. C.; Sabbah, H.; Eyssautier, J.; Pomerantz, A. E.; Barre, L.; Andrews, A. B.; Ruiz-Morales, Y.; Mostowfi, F.; McFarlane, R.; Goual, L.; Lepkowicz, R.; Cooper, T.; Orbulescu, J.; Leblanc, R. M.; Edwards, J.; Zare, R. N. Energy Fuels 2012, 26, 3986.

23. Sabbah, H.; Morrow, A. L.; Pomerantz, A. E.; Zare, R. N. Energy Fuels 2011, 25, 1597.

24. Borton, D.; Pinkston, D. S.; Hurt, M. R.; Tan, X. L.; Azyat, K.; Scherer, A.; Tykwinski, R.; Gray, M.; Qian, K. N.; Kenttämaa, H. I., Energy Fuels 2010, 24, 5548.

25. Hurt, M. R.; Borton, D. J.; Choi, H. J.; Kenttämaa, H. I. Energy Fuels 2013, 27, 3653.

26. Loegel, T. N.; Danielson, N. D.; Borton, D. J.; Hurt, M. R.; Kenttämaa, H. I. Energy Fuels 2012, 26, 2850.

27. Tanaka, R.; Hunt, J. E.; Winans, R. E.; Thiyagarajan, P.; Sato, S.;Takanohashi, T. Energy Fuels 2003, 17, 127.

28. McKenna, A. M.; Donald, L. J.; Fitzsimmons, J. E.; Juyal, P.; Spicer, V.; Standing, K. G.; Marshall,A. G.; Rodgers, R. P. Energy Fuels 2013, 27, 1246.

29. Gray, M. R.; Tykwinski, R. R.; Stryker, J. M.; Tan, X. Energy Fuels 2011, 25, 3125.

30. Tukhvatullina, A. Z.; Barskaya, E. E.; Kouryakov, V. N.; Ganeeva, Y. M.; Yusupova, T. N.; Romanov, G. V. J Pet Environ Biotechnol 2013, 4, 152.

31. Andrews, A. B.; Edwards, J. C.; Pomerantz, A. E.; Mullins, O. C.; Nordlund, D.; Norinaga, K. Energy Fuels 2011, 25, 3068.

80

32. Majumdar, R. D.; Gerken, M.; Mikula, R.; Hazendonk, P. Energy Fuels 2013, 27, 6528.

33. Morgan, T.; Alvarez-Rodriguez, P.; George, A.; Herod, A.; Kandiyoti, R. Energy Fuels 2010, 24, 3977.

34. Podgorski, D. C.; Corilo, Y. E.; Nyadong, L.; Lobodin V. V.; Bythell B. J.; Robbins

W. K.; McKenna, A. M.; Marshall, A. G.; Rodgers, R. P. Energy Fuels 2013, 27, 1268.

35. Rueda-Velasquez, R. I.; Freund, H.; Qian, K.; Olmstead, W. N.; Gray, M. R. Energy Fuels 2013, 27, 1817.

36. Karimi, A.; Qian, K.; Olmstead, W. N.; Freund, H.; Yung, C.; Gray, M. R. Energy Fuels 2011, 25, 3581.

37. Gaspar, A.; Zellermann, E.; Lababidi, S.; Reece, J.; Schrader, W. Anal Chem. 2012, 84, 5257.

38. Wu, Q.; Pomerantz, A. E.; Mullins, O. C.; Zare, R. N. J Am Soc Mass Spectrom. 2013, 24, 1116.

39. Pinkston, D. S.; Duan, P.; Gallardo, V. A.; Habicht, S. C.; Tan,X.; Qian, K.; Gray, M.; Müllen, K.; Kenttämaa, H. I. Energy Fuels 2009, 23, 5564.

40. Kim, Y.; Kim, S. J Am Soc Mass Spectrom. 2010, 21, 386.

41. Rodgers, R. P.; McKenna, A. M. Anal. Chem. 2011, 83, 4665.

42. Hughey, C. A.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2002, 74, 4145.

43. Cunico, R. L.; Sheu, E. Y.; Mullins, O. C. Pet. Sci. Technol. 2004, 22, 787.

44. Qian, K. N.; Edwards, K. E.; Siskin, M.; Olmstead, W. N.;Mennito, A. S.; Dechert, G. J.; Hoosain, N. E. Energy Fuels 2007, 21, 1042.

45. Nyadong, L.; McKenna, A. M.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G. Anal. Chem. 2011, 83, 1616.

46. Owen, B. C.; Gao, J.; Amundson, L.; Archibold, E.; Tan, X.; Azyat, K.; Tykwinski, R. R.; Gray, M. R.; Kenttämaa, H. I. Rapid Commun. Mass Spectrom. 2011, 25, 1924.

47. Jarrell, T. M.; Jin, C.; Riedeman, J. S.; Owen, B. C.; Tan, X.; Scherer, A.; Tykwinski, R. R.; Gray, M. R.; Slater, P.; Kenttämaa, H. I. Fuel 2014, 133, 106.

81

48. Karni, M.; Mandelbaum, A. Org. Mass Spectrom. 1980, 15, 53.

49. Mullins, O. C., Sheu, E. Y. Structures and Dynamics of Asphaltenes; Plenum Press: New York, 1998.

50. Calemma, V.; Rausa, R.; D’Antona, P.; Montanari, L. Energy Fuels 1998, 12, 422.

82

CHAPTER 4. CHARACTERIZATION OF ORGANOSULFUR MODEL COMPOUNDS RELEVANT TO FOSSIL FUELS BY USING HIGH-RESOLUTION

TANDEM MASS SPECTROMETRY

4.1 Introduction

Sulfur-containing compounds in fossil fuels are of big concern to environment, as

they cause pollution by releasing toxic gases, such as H2S and SO2, upon combustion.1-3

Hydrodesulfurization is usually required to lower the sulfur content of petrochemical

products and vacuum residues below the legal limits in many countries.4,5 The chemistry

of desulfurization involved in processing crude oil is greatly dependent on the forms of

sulfur in the oil.6,7 Since sulfur exists in different chemical bonding environments in

crude oil, such as in thiophenes, thiols, sulfides, disulfides, and polyaromatic sulfur

compounds, they require different conditions to be removed from hydrocarbons.6-8

Therefore, better understanding of the molecular structures of organosulfur compounds in

crude oil is highly desirable to aid in rational improvement of the desulfurization

processes.

Many attempts have been made to determine the chemical bonding environments

of sulfur in complex geological samples. X-ray absorption near-edge structure

spectroscopy (XANES) and X-ray photoelectron spectroscopy (XPS) have been used for

detection and quantitation of sulfur compounds in petroleum and coal samples.9-11

Nuclear magnetic resonance (NMR) coupled with sulfur functionality derivatization has

83

been employed to identify nonvolatile sulfur compounds in crude oil.12 Mass

spectrometry (MS), coupled with separation methods, such as gas chromatography (GC)

and liquid chromatography (LC), has become an increasingly important analytical tool to

identify organosulfur components in complex hydrocarbon mixtures.13-16

Different ionization methods have been employed to ionize organosulfur

compounds in mass spectrometry, such as electron ionization (EI), chemical ionization

(CI), electrospray ionization (ESI), and atmospheric pressure chemical ionization

(APCI).17 Several studies focused on EI mass spectra of organosulfur compounds.18-20

However, EI is not an ideal ionization method for complex mixture analysis because it

tends to induce fragmentation.21 For CI experiments of sulfur-containing heterocycles,

more than one ion type is generated for each analyte, such as protonated molecule

[M+H]+ and nitric oxide complex [M+NO]+.22

Traditional ESI is not optimal for nonpolar compounds’ analysis because it is

biased towards polar compounds.23 Therefore, the influence of various dopants on ESI of

sulfur-containing heterocycles has been studied.24-26 For instance, addition of Pd2+ ions

has been used to enhance molecular ion formation for selected sulfur-containing

heterocycles in ESI.27,28 However, this method may have problems with samples of

unknown sulfur content since the concentration ratio of Pd2+ and organosulfur

compounds is crucial for successful molecular ion generation. Another approach used to

ionize sulfur-containing heterocycles in ESI is to convert them to S-methyl sulfonium or

S-phenyl sulfonium salts in solution.29 This method, however, leaves open the question as

to whether all forms of organosulfur compounds are derivatized equally efficiently.

84

Compared to the ionization methods mentioned above, APCI is a more generally

applicable approach. The chemistry of APCI ionization can be tuned to achieve

molecular ion generation by choosing a proper reagent with desired thermochemical

properties.30,31 Recent research has demonstrated the production of stable molecular ions

for non-polar and polar aromatic compounds by using APCI with carbon disulfide (CS2)

as the solvent.31 In this work, the APCI/CS2 method was explored for the ionization of

various organosulfur compounds. Stable molecular ions were generated for all the

compounds studied except for a few sulfides and disulfides that showed minor

fragmentation upon ionization.

In order to probe the structures of the ionized organosulfur compounds, multiple-

stage tandem mass spectrometry (MSn) based on collision-activated dissociation (CAD)

can be used.32-34 In this work, the fragmentation patterns of the molecular ions of

organosulfur compounds with various sulphur-containing functionalities, including

thiophenes, thiols, polyaromatic sulfur compounds, sulfides and disulfides, were

systematically studied, whereas many other studies have mainly focused on molecular

ions of protonated ions of polyaromatic sulfur compounds.35-40 The characteristic

fragment ions generated in MS2 and MS3 experiments provide clues for the chemical

bonding environment of sulfur atoms in the examined compounds.

85


All organosulfur compounds and carbon disulfide (>99.9 %) were purchased from

Sigma-Aldrich and used without further purification. The analytes were dissolved in CS2

at a concentration of approximately 1 mg/mL and sonicated for 20 min if dissolution did

not occur instantaneously. These sample solutions were directly infused into the ion

source at a flow rate of 20 μL/min by using a syringe pump.

Multiple-stage tandem mass spectrometry experiments were carried out using a

Thermo-Scientific linear quadrupole ion trap (LQIT) equipped with an atmospheric

pressure chemical ionization (APCI) source. The instrument parameters were as follows:

ion mode, positive; vaporizer temperature, 300 °C; source voltage and current, 4 kV and

4 μA, respectively; capillary temperature, 275 °C; nitrogen sheath gas and auxiliary gas

flow rates, 40 μL/min and 10 μL/min, respectively; tube lens voltage, 20-65 V. The

voltages for the ion optics were optimized for each individual analyte by using the tune

feature of the LTQ Tune Plus interface.

In MS2 experiments, molecular ions of the organosulfur compounds were isolated

using an isolation window of 1 Da (±0.5 Da), and subjected to collision-activated

dissociation (CAD) at a collision energy of 20-35 arbitrary units. In MS3 experiments,

fragment ions generated in MS2 experiments were isolated and subjected to CAD. A q

value of 0.25 was generally used for CAD, as small fragment ions can be efficiently

trapped when using this value. For those molecular ions requiring higher collision energy

to fragment, a higher q value of 0.4 was used so that more structural information on the

ions could be obtained. The relative abundances provided for the product ions in Tables

1-5 are the average values measured over 50 scans. A threshold of 1% was used, i.e., the

86

MSn products with a relative abundance lower than 1% relative to the base peak were

ignored. The data were processed by using Thermo Xcalibur software.

High-resolution experiments were performed using a Thermo-Scientific LQIT

coupled with a Fourier transform ion cyclotron resonance (FT-ICR; 7-T magnet) mass

spectrometer. The accurate masses of some ionic fragment ions were measured, based on

which molecular formulae can be assigned. This information helps to confirm the

identities of the neutral molecules lost during CAD process. The measured accurate

masses were mostly within 10 ppm from the expected values, as shown in Figure 4.6-4.9.

4.3 Results and Discussions

APCI(+)/CS2 generated stable molecular ions for all 19 organosulfur compounds

studied, with no or little fragmentation. Tables 4.1 – 4.5 summarize the CAD products of

ionized organosulfur compounds obtained in MS2 experiments and further CAD products

of those fragment ions obtained in MS3 experiments. Relative abundances of the ions are

listed. The sulfur containing neutral molecules that were lost during fragmentation are

highlighted in red color, including S (32 Da), HS● (33 Da), H2S (34 Da), CS (44 Da),

●CHS (45 Da), etc. Generally, losses of HS● and H2S were found to be associated with

aliphatic sulfur moieties, while losses of S, CS and ●CHS were more common for

polyaromatic sulfur compounds. Accurate mass measurements were performed for some

of the MS2 and MS3 CAD product ions, with their elemental compositions and mass

accuracy (ppm) shown in Tables 4.6 – 4.9. The proposed fragmentation pathways for

molecular ions of the selected organosulfur compounds are shown in Schemes 4.1 – 4.7,

with the major fragment ions highlighted by boxes.

87

Thiophenes. Three thiophene compounds with or without alkyl chains were

studied (Table 4.1). They form abundant molecular ions upon ionization via APCI(+).

The molecular ion of thiophene readily loses H and C2H2 upon CAD. No fragmentation

products were observed in MS3 experiments. For 2-methylthiophene molecular ion, H

loss is the most facile fragmentation, followed by C2H4 loss. Isolation of the [M-H]+

fragment ion followed by CAD results in C2H4 loss as well. The molecular ion of 2-

ethylthiophene fragments by alpha-cleavage, producing [M-CH3]+ ion that further

fragments by the loss of C2H4. None of these CAD products of the ionized thiophenes

provide much structural information of sulfur.

Thiols. Two thiol compounds were studied (Table 4.2). They form abundant

molecular ions upon ionization via APCI(+), except that benzyl mercaptan has minor [M-

HS]+ fragments. Their fragmentation pathways are proposed in Scheme 4.1. It was

noticed that benzenethiol molecular ion requires higher energy collisions for

fragmentation, thus a q value of 0.4 was used. It fragments by loss of H, C2H2 and CS.

Since isolation and further CAD of the [M-C2H2]+. fragment ion (m/z 84) results in

another acetylene loss, which is also observed for thiophene molecular ion, m/z 84 is

proposed to be thiophene radical cation. Elimination of CS possibly forms a five-

membered carbocyclic radical cation (m/z 66). This mass spectrometric fragmentation

product was also reported by Earnshaw et al. by using isotopic labeling and energetics

considerations.41 The molecular ion of benzenethiol was suggested to consist two forms,

as shown in Figure 4.1, and m/z 66 must only arise from a seven-membered ring

88

(expanded ring) structure.41 When m/z 66 is isolated and subject to further CAD, it loses

hydrogen atom and acetylene.

As for benzyl mercaptan molecular ion, the major fragmentation pathway is the

elimination of HS, producing a stable tropylium cation (m/z 91). Upon further isolation

and CAD, tropylium ion loses acetylene. In summary, losses of sulfur containing neutral

molecules, including CS and HS, were observed for molecular ions of thiols. This could

provide information about the presence of sulfur in the ionized molecules. Specifically,

when the molecular ions are subject to CAD, CS elimination appears to be associated

with aromatic compounds, while HS● loss is associated with aliphatic sulfur moieties.

Scheme 4.1 Fragmentation pathways for the molecular ions of (a) benzenethiol, and (b) benzyl mercaptan upon multiple-stage CAD.

89

Figure 4.1 Two forms of the molecular ion of benzenethiol

Polyaromatic Sulfur Compounds. Five polyaromatic sulfur compounds were

studied (Table 4.3). Abundant molecular ions are generated upon ionization via (+)APCI.

The molecular ions of benzothiophene and dibenzothiophene fragment by losses of S, CS

and CHS, apart from acetylene loss. In MS3 experiments, isolation and further CAD of

the above fragment ions readily loses H and C2H2 (Table 4.3). As for the molecular ion of

4,6-dimethyldibenzothiophene, in addition to H and CH3 losses upon CAD, HS loss was

observed as well (Table 4.3). The [M-HS]+ fragment ion (m/z 179) can be explained as a

facile hydrogen atom loss followed by sulfur atom elimination.39

For molecular ion of thianthrene, the major fragmentation pathway is the

elimination of S, while CHS and CS2 losses ware also observed (Table 4.3). Upon further

isolation and CAD of the predominant fragment [M-S] + ion (m/z 184), it readily loses

another S and CHS, along with H and C2H2 losses. This MS3 fragmentation behavior is

very similar to that of benzothiophene molecular ion, which indicates that the [M-S]+ ion

(m/z 184) generated during MS2 of ionized thianthrene has the same structure as that of

benzothiophene.

For another two sulfur atoms containing analyte, 2,2'-bithiophene molecular ion

fragments predominantly by losses of S, CS, CHS, and 2S (Table 4.3). Upon further

CAD of the fragment [M-S]+ ion (m/z 134), additional S, CS and CHS were lost besides

ac

fu

fo

fr

ob

w

b

b

cetylene los

urther CAD.

or its three fr

In sum

rom the m

bservations

which carbon

ecause carb

enzene, and

Figure 4.2 M

s. Moreover

The MS2 sp

ragment ions

mmary, sulfu

olecular ion

are consiste

n or hydrog

bon and hyd

polyaromati

MS2 spectrumspe

r, the [M-CH

pectra measu

s were show

fur containin

ns of poly

ent with sev

gen atoms a

drogen atom

ic sulfur com

m measured ectra measur

HS]+ fragme

ured for the m

wn in Figure 4

ng molecules

aromatic su

veral literatu

are eliminat

m scrambling

mpounds. 42,4

for the molered for its th

ent ion (m/z

molecular io

4.2 as an exa

s, including

ulfur compo

ure reports.27

ted during f

g readily oc

43

ecular ion ofhree fragmen

z 121) readil

on and MS3

ample.

S, CS and C

ounds durin

7,28,39 It shou

fragmentatio

ccurs for io

f 2,2’-bithiopnt ions.

ly loses CS

spectra mea

CHS, can be

ng CAD. T

uld be noted

on is not kn

onized thiop

phene, and M

90

upon

sured

e lost

These

d that

nown

hene,

MS3

91

Sulfides. Six sulfide compounds were studied (Table 4.4). Abundant molecular

ions were generated for most of the sulfide compounds via APCI. Upon ionization,

methyl phenyl sulfide is accompanied by minimal fragmentation while benzyl methyl

sulfide and benzyl sulfide form [M-H]+ fragment ions and/or other fragment ions.

The molecular ions of all sulfide compounds studied lose HS or H2S or both upon

CAD, with the exception of ionic benzyl phenyl sulfide. For example, methyl phenyl

sulfide molecular ion predominantly loses HS to form tropylium cation (m/z 91). CH3

and CH2S losses were also observed upon CAD. Ethyl phenyl sulfide molecular ion

fragments by cleavage of the alkyl chain, producing benzenethiol radical cation (m/z 110)

and methyl phenyl sulfide cation (m/z 123). At the meantime, minor [M-HS]+ products

(m/z 105) were observed. Further isolation and CAD of [M-C2H4]+. fragment ion (m/z

110) results in similar fragmentation behavior as that observed in benzenethiol molecular

ion. Further isolation and CAD of [M-CH3]+ fragment ion (m/z 123) results in loss of CS.

For molecular ion of benzyl methyl sulfide, the primary fragmentation pathway is -

cleavage, producing [M-CH3S]+ ion. Upon CAD, CH3, H2S and CH2S losses also

occurred. In MS3 experiments, only a hydrogen atom is lost upon CAD of [M-CH3]+

fragment ion (m/z 123).

For the molecular ion of diphenyl sulfide, two hydrogen atoms loss is the

dominant fragmentation pathway, as it drives the formation of conjugated aromatic rings,

producing m/z 184 (Scheme 4.2). Further isolation and CAD of this [M-2H]+. product ion

generates similar fragmentation behavior as that of dibenzothiophene molecular ion,

which supports the structure proposed for the fragment ion of m/z 184. Additionally,

diphenyl sulfide molecular ion fragments by the loss of H, CH3, HS and H2S. Further

92

isolation and CAD of [M-CH3]+ ion (m/z 171) results in loss of CS, CHS, H and C2H2,

suggesting that sulfur is present in the ring system for the fragment ion of m/z 171.

Further isolation and CAD of [M-H2S]+ ion (m/z 152) results in consecutive losses of

hydrogen atoms and acetylene, which is indicative of the aromatic ring structure

proposed for the fragment ion of m/z 152.

Scheme 4.2 Fragmentation pathways for the molecular ion of diphenyl sulfide upon multiple-stage CAD.

For the molecular ion benzyl sulfide, the primary fragmentation pathway is shown

in Scheme 4.3. H2S loss is observed upon CAD of the molecular ion, which can result in

the formation of a conjugated ring system for the fragment ion of m/z 180. Cleavage of

C-S bond yields two fragment ions, m/z 92 and m/z 91. Benzylic cleavage yields two

fragment ions, m/z 123 and m/z 122. Isolation and further CAD of [M-C7H8]+ ion (m/z

122) results in losses of H, CS and CHS, which suggests that sulfur is incorporated in the

93

ring system for m/z 122. Additionally, a benzene loss is observed upon CAD of ionic

benzyl sulfide parent ion (m/z 214). Isolation and further CAD of [M-C6H6] + ion (m/z

136) results in losses of H, 2xH, CS and CHS, suggesting that m/z 136 is likely to be a

sulfur-containing heterocyclic fragment ion.

Scheme 4.3 Fragmentation pathways for the molecular ion of benzyl sulfide upon multiple-stage CAD.

94

Based on the above observations, HS and H2S appear to be the characteristic

neutral molecules lost upon CAD of the sulfide molecular ions. However, benzyl phenyl

sulfide is an exception. Its molecular ion fragments predominantly by -cleavage and

produces tropylium cation (m/z 91).

Disulfides. Four disulfide compounds were studied (Table 4.5). Upon CAD,

phenyl disulfide molecular ion not only undergoes -cleavage, but also fragments by

losses of HS, H2S and 2 S atoms (Scheme 4.4). MS3 CAD experiments were conducted

for the two most abundant fragment ions generated in MS2 CAD. Isolation and further

CAD of [M-HS]+ fragment ion (m/z 185) results in consecutive hydrogen losses, as well

as HS, H2S and CH2S losses. It is reasonable that the fragmentation behavior of

protonated dibenzothiophene (m/z 185) is different than that of dibenzothiophene

molecular ion (m/z 184), which shows S, CS, and CHS losses upon MS2 CAD (Table

4.3). Isolation and further CAD of the [M-2S]+ ion (m/z 154) results in consecutive

hydrogen atom losses and acetylene loss.

95

Scheme 4.4 Fragmentation pathways for the molecular ion of phenyl disulfide upon multiple-stage CAD.

For dicyclohexyl disulfide molecular ion, however, no HS or H2S losses were

observed. The primary fragmentation pathway is α-cleavage, which generates two

product ions, m/z 148 and m/z 83 (Scheme 4.5). Isolation and MS3 CAD of [M-C6H10] +

fragment ion (m/z 148) results in losses of S2H, the product ion (m/z 83) of which has the

same structure as that of MS2 CAD product (m/z 83). This is because that isolation and

further CAD of both m/z 83 fragment ions results in the same fragmentation pattern

(acetylene loss). The difference in MS2 CAD behaviors between molecular ions of phenyl

disulfide and dicyclohexyl disulfide is due to the fact that ionized phenyl disulfide can

form a stabilized 3-membered aromatic ring upon HS and H2S loss (Scheme 4.4), while

ionized dicyclohexyl disulfide cannot.

96

Scheme 4.5 Fragmentation pathways for the molecular ion of dicyclohexyl disulfide upon multiple-stage CAD.

For dibenzyl disulfide molecular ion, the primary fragmentation pathway is the

loss of S2H besides -cleavages (Scheme 4.6). Further isolation and CAD of the [M-

S2H]+ ion (m/z 181) results in losses of methyl radical, acetylene and two hydrogen

atoms. Isolation and MS4 CAD of [M-S2H-CH3]+ ion (m/z 166) results in consecutive

hydrogen atom losses at higher collision energy (data not shown), which is typical

fragmentation behavior observed for conjugated aromatic rings. Such information

supports the resonance stabilized structures proposed for the fragmentation products of

m/z 166.

97

Scheme 4.6 Fragmentation pathways for the molecular ion of dibenzyl disulfide upon multiple-stage CAD.

Butyl disulfide molecular ion fragments by -cleavage of a C-S bond, generating

product ions of [M-C4H8]+ (m/z 122) and [M-C4H9S2]

+ (m/z 57). Meanwhile, S-S bond

cleavage products, ions of m/z 89 and m/z 88, were also observed, as shown in Scheme

4.7. Isolation and further CAD of the [M-C4H8]+ (m/z 122) fragment ion results in losses

of HS (m/z 89) and S2H (m/z 57). Further CAD of the fragment ions of m/z 89 results in

loss of H2S (m/z 55). This is consistent with what was observed for sulfide molecular

ions in previous section of this work. Upon CAD, HS and/or H2S losses are characteristic

for ionized organosulfur compounds with sulfur in an alkyl chain.

98

Scheme 4.7 Fragmentation pathways for the molecular ion of butyl disulfide upon multiple-stage CAD.

Table 4.1 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Thiophenes

Analyte (MW) MS (m/z)

MS2 CAD product ions (m/z) and their relative

abundance


abundance

Thiophene (84)

M+● (84) 84 – H (83) 2% 84 – C2H2 (58) 100%

No further fragmentation

2-Methylthiophene

(98)

M+● (98) 98 – H (97) 100% 98 – H – C2H4 (69) 11%

97– C2H4 (69) 100%

2-Ethylthiophene

(112)

M+● (112) 112 – CH3 (97) 100% 97 – C2H4 (69) 100% 97 – C2H2 (71) 2%

99

Table 4.2 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Thiols

Analyte (MW) MS (m/z) MS2 CAD product ions (m/z) and their relative

abundance


abundance

Benzenethiol

(110)

M+● (110) 110 – H (109) 10% 110 – C2H2 (84) 62% 110 – CS (66) 100%

84 – C2H2 (58) 100% 66 – H (65) 100% 66 – H –C2H2 (39) 67%

SH

Benzyl mercaptan

(124)

M+● (124) 100% [M-HS]+ (91) 11%

124 – HS (91) 100% 91 – C2H2 (65) 100%

Note: The sulfur containing neutral molecules that were lost during CAD are highlighted in red color.

100

Table 4.3 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Polyaromatic Sulfur Compounds


abundance


abundance

Benzothiophene

(134)

M+● (134)

134 – C2H2 (108) 100% 134 – S (102) 43% 134 – CS (90) 75% 134 – CHS (89) 60%

108 – C2H2 (82) 100% 102 – C2H2 (76) 100% 90 – H (89) 100% 90 – H – C2H2 (63) 27%

Dibenzothiophene

(184)

M+● (184)

184 – H (183) 3% 184 – C2H2 (158) 3% 184 – S (152) 100% 184 – CS (140) 6% 184 – CHS (139) 30%

152 – H (151) 30% 152 – 2 H (150) 100% 152 – C2H2 (126) 4% 139 – 2 H (137) 43% 139 – C2H2 (113) 100%

4,6-

Dimethyldibenzo thiophene (212)

M+● (212)

212 – H (211) 2% 212 – CH3 (197) 100% 212 – HS (179)a 14%

197 – 2H (195) 2% 197 – C2H2 (171) 5% 197 – S (165) 100% 197– C3H6 (155) 3% 197 – CS (153) 39% 197 – CHS (152) 11% 179 – H (178) 100% 179 – 2 H (177) 3% 179 – 3 H (176) 14% 179 – C2H2 (153) 2% 179 – H – C2H2 (152) 25%

Thianthrene (216)

M+● (216) 216 – S (184) 100% 216 – CHS (171) 5% 216 – CS2 (140) 3%

184 – H (183) 2% 184 – C2H2 (158) 3% 184 – S (152) 100% 184 – HS (151) 2% 184 – CHS (139) 33%

2,2’-Bithiophene

(166)

M+● (166)

166 – S (134) 96% 166 – CS (122) 16% 166 – CHS (121) 100% 166 – 2 S (102) 7%

134 – S (102) 40% 134 – CS (90) 61% 134 – CHS (89) 55% 134 – C2H2 (108) 100% 121 – CS (77) 100%

a Refer to explanation on the formation of this m/z 179 fragment ion (212 – HS) in the text.


101

Table 4.4 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Sulfides


abundance


abundance

Methyl phenyl sulfide (124)

M+● (124) 100% [M-HS]+ (91) 3% [M-CH2S]+ (78) 3%

124 – CH3 (109) 1% 124 – HS (91) 100% 124 – CH2S (78) 44%

91 – C2H2 (65) 100% 78 – H (77) 5% 78 – C2H2 (52) 100% 78 – H – C2H2 (51) 93% 78 – C2H4 (50) 15%

Ethyl phenyl sulfide (138)

M+● (138) 138 – CH3 (123) 50% 138 – C2H4 (110) 100% 138 – HS (105) 6%

123 – CS (79) 100% 110 – H (109) 8% 110– C2H2 (84) 63% 110 – CS (66) 100%

Benzyl methyl sulfide (138)

M+● (138) 100% [M-H]+ (137) 60% [M-CH3S]+ (91) 45%

138 – CH3 (123) 24% 138 – H2S (104) 9% 138 – CH2S (92) 12% 138 – CH3S (91) 100%

123 – H (122) 100% 104 – H (103) 26% 104 – C2H2 (78) 100% 104– H – C2H2 (77) 35% 91 – C2H2 (65) 100%

Diphenyl sulfide

(186)

M+● (186)

186 – H (185) 17% 186 – 2 H (184) 100% 186 – CH3 (171) 5% 186 – HS (153) 8% 186 – H2S (152) 11%

184 – C2H2 (158) 5% 184 – S (152) 100% 184 – CHS (139) 43% 171 – 2 H (169) 18% 171 –C2H2 (145) 8% 171 – CS (127) 100% 171 – CHS (126) 43% 152 – H (151) 90% 152 – 2 H (150) 100% 152 – 3 H (149) 80% 152 –C2H2 (126) 30%

Benzyl sulfide

(214)

M+● (214) 100% [M-H]+ (213) 40%

214 – H2S (180) 2% 214 – C6H6 (136) 15% 214 – C7H7 (123) 44% 214 – C7H8 (122) 100% 214 – C7H6S (92) 5% 214 – C7H7S (91) 4%

136 – H (135) 40% 136 – 2 H (134) 8% 136 – CS (92) 100% 136 – CHS (91) 62% 122 – H (121) 100% 122 – CS (78) 4% 122 – CHS (77) 3%

Benzyl phenyl sulfide (200)

M+● (200) 200 – C6H5S (91) 100% 91– C2H2 (65) 100%


102

Table 4.5 MS2 and MS3 CAD Product Ions (with Relative Abundances) for Ionized Disulfides


abundance


abundance

Phenyl disulfide

(218)

M+● (218) 100% [M-C6H6]

+ (140) 10%

218 – HS (185) 87% 218 – H2S (184) 4% 218 – 2 S (154) 100% 218 - C6H6 (140) 2% 218 - C6H5S (109) 1%

185 – H (184) 100% 185 – 2 H (183) 30% 185 – HS (152) 91% 185 – H2S (151) 2% 185 – CH2S (139) 3% 154 – H (153) 35% 154 – 2 H (152) 100% 154 – 3 H (151) 12% 154 –C2H2 (128) 3%

Dicyclohexyl

disulfide (230)

M+● (230) 230 – C6H10 (148) 100% 230 – C6H11S2 (83) 5%

148 – S2H (83) 100% 83 – C2H4 (55) 100%

Dibenzyl disulfide

(246)

M+● (246) 246 – S2H (181) 100% 246 – C7H7S2 (91) 28%

181 – 2 H (179) 7% 181 – CH3 (166) 100% 181 – C2H4 (153) 14% 91 – C2H2 (65) 100%

Butyl disulfide (178)

M+● (178) 100% [M-HS]+ (145) 4%

178 – HS (145) 1% 178 – C4H8 (122) 100% 178 – C4H9S (89) 4% 178 – C4H10S (88) 8% 178 – C4H9S2 (57) 3%

122 – HS (89) 14% 122 – S2H (57) 100% 89 – H2S (55) 100% 88 – C2H4 (60) 100% 88 – H2S (54) 40%


103

Table 4.6 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 CAD Product Ions for Ionized Thiols

Analyte (MW) Measured Mass Elemental Composition Mass Accuracy (ppm)

Benzenethiol

(110)

MS2: 110.01863 66.04693

MS2: C6H6S C5H6

MS2: 0.158 0.528

Benzyl

mercaptan (124)

MS2: 124.02839 91.05461

MS2: C7H8S C7H7

MS2: -9.042 0.383

Table 4.7 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 and MS3 CAD Product Ions for Ionized Polyaromatic Sulfur Compounds


Benzothiophene

(134)

MS2: 134.01792 108.00254 102.04611 90.04605 89.03808

MS2: C8H6S C6H4S

C8H6 C7H6 C7H5

MS2: -4.122 -2.615 -2.859 -3.907 -5.579

Dibenzothiophene

(184)

MS2: 184.03346 152.06154 140.06176 139.05408

MS2: C12H8S C12H8 C11H8 C11H7

MS2: -3.600 -3.366 -2.084 -1.056

4,6-

Dimethyldibenzo thiophene (212)

MS2: 212.06462 211.05697 197.04145 179.08488 MS3: 165.06920 153.06927 152.06144

MS2: C14H12S C14H11S C13H9S C14H11 MS3: C13H9 C12H9 C12H8

MS2: -3.785 -2.974 -2.525 -3.612 MS3: -4.101 -3.965 -4.024

Thianthrene (216)

MS2: 216.00548 184.03353 171.02595 140.06171 MS3: 152.06139 151.05383 139.05392

MS2: C12H8S2 C12H8S C11H7S

C11H8 MS3: C12H8 C12H7 C11H7

MS2: -3.302 -3.220 -2.032 -2.441 MS3: -4.352 -2.627 -2.206

104

Table 4.7, continued


2,2’-Bithiophene

(166)

MS2: 165.98980 134.01809 122.01803 121.01020 102.04607 MS3: 108.00240 102.04599 90.04595 89.03814 77.03804

MS2: C8H6S2 C8H6S C7H6S C7H5S

C8H6 MS3: C6H4S

C8H6 C7H6 C7H5 C6H5

MS2: -4.477 -2.854 -3.626 -3.698 -3.251 MS3: -3.911 -4.035 -5.017 -4.905 -6.967

Table 4.8 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 and MS3 CAD Product Ions for Ionized Sulfides


Methyl phenyl sulfide (124)

MS2: 124.03352 91.05379 78.04590

MS2: C7H8S C7H7 C6H6

MS2: -4.858 -4.797 -6.429

Ethyl phenyl sulfide (138)

MS2: 138.04920 123.02589 110.01810 105.06958 MS3: 79.05368

MS2: C8H10S C7H7S C6H6S

C8H9 MS3: C6H7

MS2: -4.148 -3.312 -3.386 -2.825 MS3: -6.917

Benzyl methyl sulfide (138)

MS2: 138.04476 123.02579 104.05714 92.05713 91.05379

MS2: C8H10S C7H7S

C8H8 C7H8 C7H7

MS2: -11.893 -4.125

-47.203 -53.465

-4.797

Diphenyl sulfide

(186)

MS2: 186.04843 171.02593 153.06502 152.06165 MS3: 152.06137 139.05390 127.05363 126.04580

MS2: C12H10S C11H7S

C12H9 C12H8

MS3: C12H8 C11H7 C10H7 C10H6

MS2: -7.216 -2.149

-31.731 -2.643 MS3: -4.484 -2.350 -4.697 -4.774

Benzyl sulfide

(214)

MS2: 214.07693 180.08818 92.05710 91.05377

MS2: C14H14S C14H12

C7H8 C7H7

MS2: -19.352 -28.719 -53.791

-5.017

Benzyl phenyl sulfide (200)

MS2: 200.06039 91.05378

MS2: C13H12S C7H7

MS2: -5.033 -0.447

105

Table 4.9 Measured Accurate Masses, Elemental Compositions, and Mass Accuracy (ppm) of MS2 and MS3 CAD Product Ions for Ionized Disulfides


Phenyl disulfide

(218)

MS2: 218.02111 185.04145 184.03368 154.07718 109.01034 MS3: 152.06151 151.05373 139.05379

MS2: C12H10S2 C12H9S C12H8S C12H10 C6H5S

MS3: C12H8 C12H7 C11H7

MS2: -3.363 -2.689 -2.405 -3.387 -2.820 MS3: -3.563

-3.0289 -3.141

Dicyclohexyl

disulfide (230)

MS2: 230.11458 148.03681 83.08500 MS3: 83.08494

MS2: C12H22S2 C6H12S2

C6H11 MS3: C6H11

MS2: -3.884 -4.616 -6.342 MS3: -7.064

Dibenzyl disulfide

(246)

MS2: 246.05226 181.10050

91.05381

MS2: C14H14S2 C14H13

C7H7

MS2: -3.590 -3.738 -4.578

Butyl disulfide (178)

MS2: 178.08367 145.10394 122.02118 89.04151 88.03369 57.06937 MS3: 89.04143 57.06931

MS2: C8H18S2 C8H17S

C4H10S2 C4H9S C4H8S

C4H9 MS3: C4H9S

C4H9

MS2: -4.344 -4.189 -5.436 -4.914 -4.913 -8.882 MS3: -5.813 -9.933

106

4.4 Conclusions

APCI(+) with CS2 as reagent can be used to generate stable molecular ions for all

the organosulfur compounds studied, with the exception of minor fragmentation for some

ionized sulfides and disulfides. Upon CAD, characteristic product ions were observed in

MS2 and MS3 experiments, which correspond to losses of S (32 Da), HS● (33 Da), H2S

(34 Da), CS (44 Da) and ●CHS (45 Da). These fragmentations indicate the presence of

sulfur in the ionized molecules. Losses of HS● and H2S were found to be associated with

aliphatic sulfur moieties, while losses of S, CS and ●CHS were more common for

heteroaromatic compounds. However, the reverse of the above statement is not true, i.e.,

not all molecular ions of organosulfur compounds show such characteristic losses. For

instance, molecular ions of thiophenes do not lose sulfur-containing molecules upon

CAD. Molecular ions of benzyl phenyl sulfide and some disulfides lose a sulfur atom

along with other parts of the molecules in different ways upon CAD.

In summary, knowledge of the chemical bonding environments of sulfur is

beneficial to improving the desulfurization process of fossil fuels in the petroleum

industry. High-resolution tandem mass spectrometry holds promise to determining the

presence and types of organosulfur compounds (aliphatic vs. aromatic sulfur) in complex

mixture. However, caution should be taken when making generalized conclusions about

the components in a mixture because not all molecular ions of organosulfur compounds

show characteristic sulfur losses upon CAD.

107

4.5 References

1. Lewan. M. D. Nature 1998, 391, 164. 2. Monticello. D. J. Chemtech. 1998, 28, 38. 3. Choudhary, T. V.; Parrott, S.; Johnson. B. Environ. Sci. Technol. 2008, 42, 1944. 4. Mochida, I.; Choi. K. J. Jpn. Petrol. Inst. 2004, 47, 145. 5. Muller, H.; Andersson, J. T. Anal. Chem. 2005, 77, 2536. 6. Orr, W. L.; Sinninghe Damsté, J. S. ACS Symposium Series 1990, 429, 2. 7. White, C. M.; Douglas, L. J.; Perry, M. B.; Schmidt, C. E. Energy Fuels 1987, 1, 222. 8. Nishioka. M. Energy Fuels 1988, 2, 214. 9. Pickering, I. J.; Prince, R. C.; Divers, T.; George, G. N. FEBS Letters 1998, 441, 11. 10. Pomerantz, A. E.; Seifert, D. J.; Bake, K. D.; Craddock, P. R.; Mullins, O. C.; Kodalen, B. G.; Kirtley, S. M.; Bolin, T. B. Energy Fuels 2013, 27, 4604. 11. Manceau, A.; Nagy, K. L. Geochimica et Cosmochimica Acta 2012, 99, 206. 12. Green, T. K.; Whitley, P.; Wu, K.; Lloyd, W. G.; Gan, L. Z. Energy Fuels 1994, 8, 244. 13. Van Stee, L. L. P.; Beens, J.; Vreuls, R. J. J.; Brinkman. U. A. T. J. Chromatogr. A 2003, 1019, 89.

14. F. Wang, C. Y.; Qian, K.; Green. L. A. Anal. Chem. 2005, 77, 2777.

15. Qian, K.; Hsu, C. S. Anal. Chem. 1992, 64, 2327.

16. Thomas, D.; Crain, S. M.; Sim, P. G.; Benoit, F. M. J. Mass Spectrom. 1995, 30, 1034.

17. Rodgers, R. P.; McKenna, A. M. Anal. Chem. 2011, 83, 4665.

18. Porter, Q. N. Aust. J. Chem. 1967, 20, 103.

19. Meyerson, S.; Fields, E. K. J. Org. Chem. 1968, 33, 847.

20. Boberg, F.; Bruns, W.; Mußhoff, D. Erdöl und Kohle, Erdgas. Petrochemie 1994, 47, 56.

108

21. Märk, T. D. Int. J. Mass Spectrom. Ion Phys. 1982, 45, 125.

22. Hunt, D. F. J. Shabanowitz. Anal. Chem. 1982, 54, 574.

23. Rosenberg, E. J. Chromatogr. A 2003, 1000, 841.

24. Bayer, E. G.; Gfrorer, P.; Rental C, C. Angew. Chem., Int. Ed. 1999, 38, 992.

25. Van Berkel, G. J.; Asano, K. G. Anal. Chem. 1994, 66, 2096.

26 .Roussis, S. G.; Prouix, R. Anal. Chem. 2002, 74, 1408.

27. Rudzinski, W. E.; Zhang, Y.; Luo, X. J. Mass Spectrom. 2003, 38, 167.

28 .Rudzinski, W. E.; Zhou, K.;X. Luo. Energy Fuels 2004, 18, 16.

29. Panda, S. K.; Andersson, J. T.; Schrader, W. Angew. Chem. Int. Ed. 2009, 48, 1788.

30. Kostiainen, R.; Kauppila, T. J. J. Chromatogr. A. 2009, 1216, 685.

31. Owen, B. C.; Gao, J.; Borton II, D. J.; Amundson, L. M.; Archibold, E. F.; Tan, X.; Azyat, K.; R. Tykwinski, M. Gray, H. I. Kenttämaa. Rapid Commun. Mass Spectrom. 2011, 5, 1924.


33. Sheng, H.; Williams, P. E.; Tang, W.; Riedeman, J. S.; Zhang, M.; Kenttämaa, H. I. J. Org. Chem. 2014, 79, 2883.

34. Sheng, H.; Williams, P. E.; Tang, W.; Zhang, M.; Kenttämaa, H. I. Analyst, 2014, 139, 4296.

35. Airiau, C. Y.; Brereton, R. G.; Crosby, J. Rapid Commun. Mass Spectrom. 2001, 15, 135.

36. Rudzinski, W. E.; Rai, V. Energy Fuels, 2005, 19, 1611.

37. Frache, G.; Krier, G.; Vernex-Loset, L.; Muller, J. F.; Manuelli, P. Rapid Commun. Mass Spectrom. 2007, 21, 2601.

38. Panda, S. K.; Schrader, W.; al-Haiji, A.; Andersson, J. T. Energy Fuels 2007, 21, 1071.

39. Herrera, L. C.; Ramaley, L.; Grossert, J. S. Rapid Commun. Mass Spectrom. 2009, 23, 571.

40. Hourani, N.; Andersson, J. T.; Möller, I.; Amad, M.; Witt, M.; Sarathy, S. M. Rapid Commun. Mass Spectrom. 2013, 27, 2432.

109

41. Earnshaw, D. G.; Cook , G. L.; Dinneen, G. U. J. Phys. Chem. 1964, 68, 296. 42. Cooks, R. G.; Howe, I.; Tam, S. W.; Williams, D. H. J. Amer. Chem. Soc. 1968, 90, 4064. 43. Cooks, R. G.; Bernasek, S. L. J. Am. Chem. Soc. 1970, 92, 2129.

110

CHAPTER 5. GAS-PHASE ION/MOLECULE REACTIONS FOR THE IDENTIFICATION OF SULFONE FUNCTIONALITIES IN PROTONATED

ANALYTES IN A LINEAR QUADRUPOLE ION TRAP MASS SPECTROMETER

5.1 Introduction

Rapid identification of drug metabolites, degradation products, and impurities is

crucial in the drug discovery and development process since some of them are toxic.1-3

Analytical techniques, such as NMR, FT-IR, and X-ray crystallography can be utilized to

obtain information on functional groups and elemental connectivity in the analyte.4-6

However, compounds that are present in only small quantities in complex mixtures are

difficult to identify using the above methods.

Tandem mass spectrometry (MSn) has evolved to be a powerful technique for

mixture analysis due to its high sensitivity, selectivity and speed.7,8 Single-stage mass

spectrometry (MS) can provide molecular mass and elemental composition of the ionized

analytes, while MSn utilizing collision-activated dissociation (CAD) can provide

structural information for the isolated, ionized unknown analytes.9 When the

functionalities of ionized analytes cannot be definitively determined by CAD alone, an

alternative MSn technique based on ion-molecule reactions can be utilized to obtain

structural information.10 Our group has successfully developed methods based on ion-

molecule reactions to identify different functional groups in ionized analytes.11-21 Several

neutral reagents have been investigated extensively, including boron- and carbon-

111

centered reagents with good leaving groups.11-21 However, phosphorous-centered neutral

reagents have been rarely studied. The only report appeared thus far focused on the

identification of the amino functionality as well as differentiation of primary, secondary

and tertiary protonated amino functionalities by using diethyl methylphosphonate and

hexamethylphosphoramide.22 In this case, proton transfer reaction and adduct formation

are the two major reaction pathways.22

Oxidized sulfur functionalities, such as sulfone and sulfoxide, are common in

drug metabolites.23 Only a few tandem mass spectrometry (MS/MS) experiments based

on CAD of ionized sulfones and sulfoxides have been published, yet none of them

showed sulfone or sulfoxide specific fragmentation patterns.24-27 In an effort to enable

identification of sulfur-containing functionalities in drug metabolites, we recently

reported a boron-centered reagent (trimethyl borate, TMB) that allows the identification

of protonated sulfone analytes19 and another carbon-centered reagent (2-methoxypropene,

MOP) for the identification of protonated sulfoxide analytes.20 TMB was found to yield a

diagnostic product ion, adduct-Me2O, upon reaction with protonated sulfone analytes.

MOP was found to yield the same MOP adducts for protonated sulfoxides and N-oxides,

yet the distinction of sulfoxides was based on their high reaction efficiencies. In order to

search for another reagent that would differentiate protonated sulfoxides from other

functionalities based on different reaction products, a phosphorous-containing reagent,

trimethyl phosphite (TMP), was examined. However, to our surprise, TMP was found to

allow for the differentiation of protonated sulfone functionality from many other

functional groups, including sulfoxide, hydroxylamino, N-oxide, aniline, amino, keto and

carboxylic acid functionalities. The reaction specificity was further demonstrated by

112

using a sulfoxide-containing anti-inflammatory drug, sulindac, as well as its metabolite

sulindac sulfone.


Chemicals. All chemicals were purchased from Sigma-Aldrich. Their purities were ≥

98%. All chemicals were used without further purification.

Instrumentation. All mass spectrometry experiments were performed using a Thermo

Scientific LTQ linear quadrupole ion trap (LQIT) equipped with an APCI source. The

analytes were dissolved in methanol with a final concentration of 0.01-1 mg/mL. The

sample solutions were introduced into the mass spectrometer through direct infusion at a

flow rate of 20 μL/min by using a syringe drive. The APCI source was operated in

positive ion mode. The temperatures for the vaporizer and transfer capillary were set at

300 °C and 275 °C, respectively. Nitrogen was used as the sheath gas and auxiliary gas,

with the flow rate maintained at 30 and 10 arbitrary units respectively. The voltages for

the ion optics were optimized for each individual analyte by using the tune feature of the

LTQ Tune Plus interface. The normal mass range (m/z 50-500) was used for all the

experiments, while the low mass range (m/z 20-200) was used for examination of the

exothermic proton-transfer reaction between protonated methanol and the reagent (TMP).

The type of the manifold that was used to introduce the reagent was first described by

Gronert.28,29 A diagram of the exact manifold used in this research was published by

Habicht et al.[13] TMP was introduced into the manifold via a syringe pump at the rate of

10 μL/h. A known flow of helium gas (0.8 L/h) was used to carry TMP into the mass

113

spectrometer. The syringe port and surrounding area were heated to ~70 °C to ensure

evaporation of TMP. A Granville-Phillips leak valve was used to control the amount of

the reagent introduced into the instrument, while another leak valve controlled the

amount of helium diverted to waste.30 A typical nominal pressure of TMP in the ion trap

during the experiments was 0.6×10-5 Torr.

Kinetics. After the analytes were ionized by protonation in the APCI source, the

protonated analytes were isolated by using an isolation window of 2 Da. The isolated ions

were allowed to react with the reagent TMP for variable periods of time. In the course of

ion/molecule reactions, the concentration of the neutral reagent is in great excess of that

of the ion of interest. Therefore, the pressure of TMP can be considered as a constant, and

the reactions follow pseudo-first-order kinetics. The reaction efficiency corresponds to

the fraction of ion/molecule collisions that lead to the formation of products. The reaction

efficiency, kreaction/kcollision, was calculated by measuring the rate of each ion/molecule

reaction (IM) and the rate of the highly exothermic proton-transfer reaction (PT) between

protonated methanol and the reagent (TMP) under identical conditions. The above rates

were measured by determining the relative abundances of the reactant ion and product

ions as a function of reaction time. In a semilogarithmic plot of the ion abundances as a

function of time, the decay slope of the reactant ion corresponds to the rate constant k

multiplied by the neutral reagent's concentration. Assuming that the exothermic proton-

transfer reaction (PT) between protonated methanol and TMP proceeds at collision rate

(kcollision; this can be calculated by using a parameterized trajectory theory31), the

efficiencies of the ion/molecule reactions can be obtained by using equation 1. The

114

reaction efficiency is based on the ratio of the slopes of the two reactions studied, i.e.,

kreaction[TMP] = slope (IM), kcollision[TMP] = slope (PT), wherein [TMP] = TMP

concentration. It is obvious that there is no need to measure the concentration of TMP as

it cancels out in the calculation. Additionally, the reaction efficiency is dependent on the

masses of the ion (Mi), neutral reagent (Mn), and methanol (M(PT)), as well as the pressure

read by an ion gauge for the reagent during the ion/molecule reaction (Pn(IM)) and the

proton-transfer reaction (Pn(PT)).

5.3 Results and Discussions

In an effort to search for a reagent that would allow the mass spectrometric

distinction of protonated sulfoxides from protonated sulfones, trimethyl phosphite (TMP)

was examined because its proton affinity (PA) (~220 kcal/mol32) is close to that of

sulfoxides (~220 kcal/mol19), and higher than that of sulfones (~205 kcal/mol19).

According to the reactivity that has been reported for a similar boron-centered reagents

TMB,11-13 TMP was expected to react with a protonated sulfoxide via proton abstraction

followed by replacement of a methanol molecule in the adduct ion of protonated TMP

and the neutral analyte molecule. Meanwhile, TMP was expected to react with protonated

sulfones mostly via proton abstraction.

115

TMP was allowed to react with protonated analytes containing sulfone or

sulfoxide functionality. As shown in Table 5.1, proton transfer was the major reaction for

both protonated sulfones and sulfoxides. However, contrary to the expectations, only

protonated sulfones showed the [TMP adduct-MeOH] product ion, not protonated

sulfoxides. A mass spectrum measured after 100 ms reaction of protonated

dibenzothiophene sulfone with TMP is shown in Figure 5.1 as an example. The most

abundant product ion (m/z 125) corresponds to proton transfer. The other product ion

(m/z 309) corresponds to [TMP adduct-MeOH], which was only observed for protonated

sulfones.

In order to probe the selectivity of the above reaction for protonated sulfones,

TMP was allowed to react with other protonated analytes containing various functional

groups, such as N-oxide, hydroxylamino, keto, carboxylic acid, and aliphatic and

aromatic amino. The reaction products and efficiencies are summarized in Table 5.2. The

main reactions were proton transfer, and minor addition reactions were observed for

some protonated analytes, yet no [TMP adduct-MeOH] product ion was observed. When

protonated sulindac and its metabolite sulindac sulfone were allowed to react with TMP,

only sulindac sulfone showed [TMP adduct-MeOH] product ion (Figure 5.2). This result

demonstrates the utility of this method for the identification of the sulfone functionality

in drug metabolites.

A mechanism is proposed for the formation of [TMP adduct-MeOH] ion for

protonated sulfones as shown in Scheme 5.1. The unique selectivity of TMP toward

sulfones can be rationalized by the six-membered transition state of the ion-neutral

complex of protonated sulfone and TMP (Scheme 5.1). Through this transition state, the

pr

at

O

pr

fu

d

roton could

ttack of the s

On the other

roton would

unctional gro

Figurdibenzothiop

be transferr

sulfone to th

hand, for oth

d most likely

oups could n

re 5.1 A masphene sulfon

ed to the les

he phosphoro

her protonat

y be transferr

not form the

ss spectrum mne with TMP

ss basic meth

ous center to

ted functiona

red to the m

six-member

measured aft in LQIT (*s

hoxy group,

o form [TMP

alities such a

more basic ph

red transition

fter 100 ms rsecondary pr

, which initi

P adduct-Me

as sulfoxide

hosphorous c

n state as sul

reaction of prroducts of p

ated nucleop

eOH] produc

and N-oxid

center since

lfone did.

rotonated rotonated TM

116

philic

ct ion.

e, the

these

MP).

S

Fs

Scheme 5.1 T

Figure 5.2 Msulindac sulf

The proposeproduct i

Mass spectra fone (bottom

s

d mechanismion when a p

measured afm) with TMPsecondary pr

m for the forprotonated su

fter 300 ms rP in LQIT (*roducts of pr

rmation of a ulfone react

reaction of p*secondary protonated TM

stable [TMPs with TMP

protonated suproducts of TMP).

P adduct-Me.

ulindac (top)TMP adducts

117

eOH]

) and s; **

118

Table 5.1 Reaction products (m/z values and branching ratios) and efficiencies for reactions of protonated sulfones and sulfoxides with TMP (PA = 222.2 kcal/mol a).

Reagent

(m/z of [M+H]+) PA

(kcal/mol)Product ions (m/z) and

branching ratiosReaction efficiency

(119)

206.3b Proton Transfer (125) 72%

Adduct – MeOH(309) 28%

90%

(157)


Adduct – MeOH (249) 26%

76%

(179)

203.7b Proton Transfer (125) 98% Adduct –MeOH (271) 2%

94%

(95)


Adduct – MeOH (187) 25%

98%

(121)

198.3b

Proton Transfer (125) 71% Adduct – MeOH (213) 29%

90%

(217)

205.0b

Proton Transfer (125) 94% Adduct – MeOH (309) 6%

59%

(197)

200.9c

Proton Transfer (125) 29% Adduct – MeOH (289) 7% Adduct (321) 64%

58%

(158)

202.9c


89%

S

O

O

S

O

O

SO O

119


Reagent (m/z of [M+H]+)

PA (kcal/mol)

Product ions (m/z) andbranching ratios

Reaction efficiency

(163)

220.1b

Proton Transfer (125) 96% Adduct (287) 4%

87%

(203)


Adduct (327) 5%

73%

(79)

211.3a


101%

(166)

227.9c Proton Transfer (125) 1%

Adduct (290) 99%

3%

aReference 32. b Reference 19. c Calculated at the B3LYP/6-31G++(d,p) level of theory.

H3CS

OH

OO

NH2

120

Table 5.2 Reaction products (m/z values and branching ratios) and efficiencies for reactions between protonated N-oxides, ketones, hydroxylamines, carboxylic acids, aliphatic and aromatic amines with TMP (PA = 222.2 kcal/mol a).


PA (kcal/mol)

Product ions (m/z) and branching ratios

Reaction efficiency

(96)


Adduct (220) 2%

94%

(146)

225.5b Proton Transfer (125) 66% Adduct (270) 34%

10%

(99)

201a Proton Transfer (125) 100%

110%

(183)

210.8b Proton Transfer (125) 98% Adduct (307) 2%

91%

(90)

218.6b Proton Transfer(125) 97%

Adduct (214) 3% 87%

(116)

215.9b


94%

(138)

206.7a Proton Transfer (125) 100% 71%

N

O

N

O

O

121



PA (kcal/mol)

Product ions (m/z) andbranching ratios

Reaction efficiency

(139)

200.7c Proton Transfer (125) 100% 78%

(180)

207.5c Proton Transfer (125) 98%

Adduct (304) 2% 95%

(74) 220.2b


30%

(100)

223.3a

Proton Transfer (125) 94%

Adduct (224) 6% 86%

(94)

210.9b Proton Transfer (125) 100% 53%

(170)

214.4c

Proton Transfer (125)

97%

Adduct (294) 3% 46%

aReference 32. bReference 20. cCalculated at the B3LYP/6-31G++(d,p) level of theory.

HO

HO

O

HN

122

Table 5.3 Reaction products (m/z values and branching ratios) and efficiencies for reactions of protonated sulindac and sulindac sulfone with TMP (PA = 222.2 kcal/mol a).


PAb (kcal/mol)

Product ions (m/z) and branching ratios

Reaction efficiency

Sulindac (373)

224 Proton Transfer (125) 89%

Adduct (481) 11%

53%

Sulindac sulfone (373)

203


78%

aReference 32. bReference 33.

HOOC

F

SO

HOOC

F

S

O

O

123

5.4 Conclusion

A method based on a functional group-selective ion/molecule reaction in a linear

quadrupole ion trap mass spectrometer has been demonstrated for the identification of the

sulfone functionality in protonated analytes. A phosphorous-centered neutral reagent,

trimethyl phosphite (TMP), can form characteristic [TMP adduct- MeOH] product ions

only when allowed to react with protonated sulfone analytes. All other protonated

compounds investigated in this study, with functionalities such as sulfoxide, N-oxide,

hydroxylamino, keto, carboxylic acid, aliphatic and aromatic amino, react with TMP via

proton transfer and/or addition. The selectivity of TMP toward sulfones can be

rationalized by the six-membered transition state of the ion-neutral complex of

protonated sulfone and TMP, while other functionalities cannot form such transition state.

The results obtained for sulindac and sulindac sulfone suggest that this method allows the

identification of sulfone functional group in drug metabolites even in the presence of

other functionalities.

124

5.5 Reference

1. F. I. Zuniga, D. Loi, K. H. Ling, D. D. Tang-Liu, Expert Opin. Drug Metab. Toxicol. 2012, 8, 467-485.

2. X. Chen, S. Hussain, S. Parveen, S. Zhang, Y. Yang, C. Zhu, Curr. Med. Chem. 2012, 19, 3578-3604.

3. M. D. Coleman, Human Drug Metabolism, An Introduction, 2nd Ed., John Wiley & Sons. 2010.

4. J. Caslavska, W. Thormann, J. Chromatogr. A. 2011, 1218, 588-601.

5. D. P. Bhave, Muse, Carroll, Infect. Disord. Drug Targets 2007, 7, 140-158.

6. D. A. Dibbern, A. Montanaro, Ann. Allergy Asthma Immunol. 2008, 100, 91-100.

7. S. A. McLuckey, J. M. Wells. Chem. Rev. 2001, 101, 571-606.

8. R. H. Perry, R. G. Cooks, R. J. Knoll. Mass Spectrom. Rev. 2008, 27, 661-699.

9. R. G. Cooks, K. L. Busch, G. L. Glish, Science, 1983, 222, 273-291.

10. S. Osburn, V. Ryzhov, Anal. Chem. 2013, 85, 769-778.

11. M. A. Watkins, B. E. Winger, R. C. Shea, H. I. Kenttamaa. Anal. Chem. 2005, 77, 1385-1392.

12. K. M. Campbell, M. A. Watkins, S. Li, M. N. Fiddler, B. Winger, H. I. Kenttämaa, J. Org. Chem. 2007, 72, 3159-3165.

13. S. C. Habicht, N. R. Vinueza, E. F. Archibold, P. Duan, H. I. Kenttämaa, Anal. Chem. 2008, 80, 3416-3421.

14. M. Fu, P. Duan, S. Li, S. C. Habicht, D. S. Pinkston, N. R. Vinueza, H. I. Kenttämaa, Analyst 2008, 133, 452-454.

15. P. Duan, T. A. Gillespie, B. E. Winger, H. I. Kenttamaa. J. Org. Chem. 2008, 73, 4888-4894.

16. P. Duan, M. Fu, T. A. Gillespie, B. E. Winger, H. I. Kenttämaa, J. Org. Chem. 2009, 74, 1114-1123.

17. J. Somuramasami, P. Duan, L. Amundson, E. Archibold, B. Winger, H. I. Kenttämaa, J. Am. Soc. Mass Spectrom. 2011, 22, 1040-1051.

125

18. R. J. Eismin, M. Fu, S. Yem, F. Widjaja, H. I. Kenttämaa, J. Am. Soc. Mass Spectrom. 2012, 23, 12-22.

19. H. Sheng, P. E. Williams, W. Tang, M. Zhang, H. I. Kenttämaa, J. Org. Chem. 2014, 79, 2883-2889.

20. H. Sheng, P. E. Williams, W. Tang, M. Zhang, H. I. Kenttämaa, Analyst 2014, 139, 4296-4302.

21. H. Sheng, W. Tang, R. Yerabolu, J. Y. Kong, P. E. Williams, M. Zhang, H. I. Kenttämaa. Rapid. Comm. Mass Spetrom. 2015, 29, 730-734.

22. M. Fu, R. J. Eismin, P. Duan, S. Li, H. I. Kenttämaa, Int. J. Mass Spectrom. 2009, 282, 77-84.

23. H. L. Holland, Nat. Prod. Rep. 2001, 18, 171-181.

24. J. A. Hibbs, F. B. Jariwala, C. S. Weisbecker, A. B. Attygalle. J. Am. Soc. Mass Spectrom. 2013, 24, 1280-1287.

25. W. Lam, R. Ramanathan, J. Am. Soc. Mass Spectrom. 2002, 13, 345-353.

26. X. Jiang, J. B. Smith, E. C. Abraham. J. Mass Spectrom. 1996, 31, 1309-1310.

27. J. M. Froelich, G. E. Reid. J. Am. Soc. Mass Spectrom. 2007, 18, 1690-1705.

28. S. Gronert, J. Am. Soc. Mass Spectrom. 1998, 9, 845-848.

29. S. Gronert, Mass Spectrom. Rev. 2005, 24, 100-120.

30. J. C. Schwartz, M. W. Senko, J. E. Syka, J. Am. Soc. Mass Spectrom. 2002, 13, 659-669.

31. T. Su and W. J. Chesnavich, J. Chem. Phys. 1982, 76, 5183–5185.

32. http://webbook.nist.gov/

33. H. Sheng, W. Tang, J. S. Riedema, R. Yerabolu, J. Max, R. R. Kotha, M. Zhang, H. I. Kenttämaa. In preparation.

126

CHAPTER 6. GAS-PHASE REACTIVITY OF META-BENZYNES TOWARDS

AMINO ACIDS

6.1 Introduction

Radicals are known to attack proteins, resulting in their denaturation, oxidation,

fragmentation and degradation, which is associated with aging and many human

diseases.1-4 Therefore, it is important to understand the processes involved. Numerous

solution studies have been carried out on reactions of oxygen-containing radicals with

proteins, peptides, and amino acids,5-8 whereas carbon-centered radicals are rarely

investigated. However, carbon-centered biradicals are of great interest as they are the

biologically active intermediates of the naturally occurring anticancer antibiotics called

enediynes, which are known to attack DNA.9-11 Additionally, proteins can also be

damaged by such carbon-centered radicals.12-14 The one report that has appeared thus far

focused on reactions of phenyl radicals toward glycine in solution.15 Specifically, a para-

benzoic acid radical (4-dehydrobenzoic acid) was found to abstract a deuterium atom

from the α-position of α,α-dideuterioglycine in solution.15 The study of carbon-centered

radicals is challenging, which is largely due to the difficulty of generating pure radicals in

solution. Alternatively, gas-phase experiments can be performed to generate the radicals

relatively easily, which also allow their intrinsic chemical properties to be examined in a

solvent-free environment.

127

In order to investigate the gas-phase reactivity of carbon-centered radicals toward

various substrates, our laboratory has advanced the “distonic ion approach” by using

mass spectrometry.16-20 A distonic radical ion can be formed by attaching a chemically

inert charged moiety to the radical of interest, which allows for its mass spectrometric

manipulation and detection. This approach has been applied to gas-phase reactions of

phenyl radicals with amino acids and peptides, the results of which reveal similar

reactivity to that of neutral phenyl radicals in solution.21-24 For instance, both a gas-phase

positively charged phenyl radical (e.g., N-(3-dehydrophenyl)pyridinium) and a neutral

phenyl radical in solution (e.g., 4-dehydrobenzoic acid) abstract a deuterium atom from

the α-position of α,α-dideuterioglycine.15,21 In addition to the hydrogen atom abstraction

(radical mechanism), other reaction pathways were observed for more electrophilic

phenyl radicals (e.g., N-phenyl-3-dehydropyridinium) in the gas phase.21,22 These

reactions include NH2 abstraction, SH and SCH3 abstractions from cystein and

methionine, addition to the aromatic ring and side-chain abstraction from aromatic amino

acids, which were assumed to occur via addition-elimination mechanisms due to the

greater electrophilicity of these radicals.21,22 The electrophilicity of a radical can be

quantified by the calculated vertical electron affinity (EA) of its radical site, which is

defined as the energy released upon addition of an electron to the radical site with no

geometry change.25 The greater the EA, the more polar the transition state, thus the faster

are both radical and nonradical reactions. This can be rationalized by the ionic avoided

curve crossing model proposed by Anderson et al.26 EA has been demonstrated to be an

important reactivity controlling factor not only for carbon-centered σ-monoradicals but

also for related σ,σ-biradicals.25-27

128

Among carbon-centered biradicals, the three isomeric didehydrobenzenes, ortho-,

meta-, and para-benzynes (Figure 6.1), have attracted great attention in the scientific

community.28-41 They are important reactive intermediates and fundamentally intriguing

molecules. The ortho-benzyne 1 has been thoroughly studied both experimentally and

computationally.28-30 It has been generated from various commercially available

precursors in solution, and characterized by IR, UV-Vis, and NMR spectroscopy.31-33

Interest in para-benzynes (for example, 3) has grown rapidly since the discovery of para-

benzyne intermediates that are responsible for the antitumor activity of enediynes.9-14

This is due to the fact that para-benzyne analogues can abstract a hydrogen atom from

both stands of double-stranded DNA, thus causing irreversible DNA cleavage.35-38

However, the clinical use of these anticancer reagents is hindered by their high toxicity.34

The unwanted side effects of these radical intermediates can also be attributable to their

reactions with proteins.12-14 Compared to ortho- and para-benzynes, meta-benzynes (like

2) have not received the same degree of attention.

Figure 6.1 Structures of ortho- (1), meta- (2), and para-benzyne (3).

A key parameter that affects the reactivity of benzynes is the degree of interaction

between the two biradical electrons.39 This interaction can be described by the magnitude

of singlet-triplet (S-T) splitting, which is defined as the energy difference between the

129

singlet ground state and the lowest energy triplet state. The S-T splittings of the three

benzynes have been measured by negative ion photoelectron spectroscopy.41 The ortho-

benzyne has a large S-T splitting (-37.5 kcal/mol) due to strong through-space coupling,

therefore, a large amount of energy is required to uncouple the biradical electrons.41 This

explains why nonradical reactivity was observed for ortho-benzynes.42 The para-benzyne

has a much smaller S-T splitting (-3.8 kcal/mol) due to weak through-bond interaction

between the biradical electrons.41 Hence, it displays radical reactivity.43,44 It should be

noted that the radical reactivity of para-benzyne analogues is substantially lower than

that of related monoradicals because of the through-bond coupling of the biradical

electrons.39 The meta-benzyne has an intermediate S-T splitting (-21.0 kcal/mol),41 which

is thought to hinder radical reactivity and can potentially make a more selective warhead

for antitumor reagents than para-benzynes.39,40

Gas-phase studies have indicated that the reactivity of meta-benzynes is affected

by both EA and the extent of S-T splitting, as expected based on above discussion.

Additionally, another important reactivity controlling factor for meta-benzynes is

distortion energy (△E2.30), which is the energy required to distort the minimum energy

dehydrocarbon atom separation (DAS) to the separation of the transition state, which is

approximately at 2.30 Å.45-48 A small △E2.30 means that it is energetically easier to

uncouple the singlet biradical electrons in the transition state for radical reactions. This

explains why some meta-benzynes with very small △E2.30 undergo radical reactions

despite of their large S-T splitting.48 Moreover, the reactivity of meta-benzynes can be

"tuned" from electrophilic to radical-like by changing the substituent groups, which

influence S-T splitting, EA and distortion energy of the benzynes.47,48 Therefore, it is

130

critical to understand the factors that control the reactivity of meta-benzynes toward

organic substrates, and more importantly, biomolecules, including DNA and proteins.

Such knowledge could facilitate rational design of better antitumor drugs. This study

focuses on the understanding of meta-benzynes' reactivity toward proteins. Due to the

large size and complexity of proteins, the reactions of selected meta-benzynes with free

amino acids were examined. So far, only one report has been published on the reactions

of a meta-benzyne analogue (N-methyl-6,8-didehydroquinolinium cation) with free

amino acids and dipeptides.49 In this work, the reactivity of four meta-benzyne analogues

with varying EA, S-T splitting, and distortion energy was studied towards amino acids in

an FT-ICR mass spectrometer. To the best of our knowledge, this is the first study on the

effects of the above parameters on the gas-phase reactivity of meta-benzynes toward

different amino acids.


All the experiments were carried out in a Finnigan FTMS 2001 dual-cell Fourier-

transform ion cyclotron resonance (FT-ICR) mass spectrometer. The amino acids were

introduced into the mass spectrometer by using a manual solids probe. The probe was

heated to 140 oC for all amino acids except lysine for which the probe was heated to

200°C. Observation of an abundant protonated amino acid and only minor amounts of

other ions upon reaction with protonated acetone (proton affinity = 194 kcal/mol50)

confirmed that the amino acids were introduced into the instrument without much thermal

decomposition. The proton affinities of the amino acids are 211.9 kcal/mol for glycine,

218.6 kcal/mol for leucine, 238 kcal/mol for lysine, 223.6 kcal/mol for methionine, 215.9

131

kcal/mol for cysteine, 220 kcal/mol for proline, and 220.6 kcal/mol for phenylalanine.50

All amino acids (purity ≥ 98.5 %)) except DL-lysine-ε-15N were obtained from Fluka

Biochemika. DL-Lysine-ε-15N (purity ≥ 98.5%) was obtained from Cambridge Isotope

Laboratories, Inc. All amino acids were used without further purification. The other

reagents, tetrahydrofuran, allyl iodide, dimethyl disulfide, and tert-butyl isocyanide, were

obtained from Sigma Aldrich Co. and were used as received.

The radicals’ precursor ions were generated by CH3I chemical ionization in one

side of the dual-cell mass spectrometer as described previously.27,49,51 For example, N-

methyl-6,8-didehydroquinolinium cation (radical b) was formed by introducing 6,8-

dinitroquinoline and methyl iodide into the same cell of the instrument through a solids

probe and a pulsed valve, respectively. An electron beam of 20-25 eV kinetic energy was

used; the filament current was 7 μA and the ionization time 1 s. Methyl iodide undergoes

electron ionization and self-chemical ionization to form dimethyl iodide cation, which

then transfers a methyl cation to the neutral radical precursor, subsequently generating N-

methyl-6,8-dinitroquinoline cation. The N-methylated precursor ion was transferred into

the other cell by grounding the conductance limit plate for about 154 μs. Quadrupolar

axialization (QA) was employed to increase ion transfer efficiency.52 Next, the radical

sites were generated by sustained off-resonance irradiated collision-activated dissociation

(SORI-CAD).53 This involved applying an off-resonance RF voltage to the excitation

plates of the cell and pulsing argon (at a nominal pressure about 10-5 torr) into the cell.

Collisions with argon for 0.5-1 s at an RF frequency 1000 Hz higher or lower than the

ions' cyclotron frequency resulted in homolytic cleavages of the two carbon-nitrogen

bonds. The other radicals were generated using a similar procedure except that different

132

precursors were used: 3,5-diiodopyridine for a, 5,7-dinitroisoquinoline for c, 5,7-

dinitroquinoline for d, 3-iodopyridine for e, 5-nitroisoquinoline for f, and 6-

nitroquinoline for g.

After the radicals were generated, they were isolated by ejecting all other ions

from the cell by applying a series of stored-waveform inverse Fourier transform (SWIFT)

excitation pulses to the plates of the cell.54 The isolated radical ions were allowed to react

with an amino acid for a variable period of time (typically 0.5-1000 s). Detection was

performed by using “chirp” excitation of 124 V amplitude, 2.7 MHz bandwidth, and 3.2

kHz/μs sweep rate to kinetically excite the ions so that they move coherently as ion

packets and closer to the detection plates which is required for their detection. All mass

spectra presented here are the average of five transients, which were recorded as 64k data

points and subjected to one zero fill prior to Fourier transformation. Each reaction

spectrum was background corrected by using a procedure described previously.55

All reactions were found to follow pseudo-first order kinetics, which allows for

the determination of the second-order reaction rate constant (kexp) from a semilogarithmic

plot of the relative abundance of the reactant ion versus reaction time and the

concentration of the amino acid. In the FT-ICR, the concentration of ions (charged

radicals) inside the cell is much smaller than the concentration of neutral molecules

(amino acids). Therefore, the concentration of the amino acid can be assumed to be

constant. The concentration of the amino acids was determined by measuring the pressure

inside the cell by an ionization gauge that is located on each side of the dual cell. The ion

gauge pressure readings were corrected for the sensitivity of the ion gauge toward each

amino acid and for the pressure gradient between the ion gauge and the cell. The

133

correction factors were obtained by measuring the reaction rate of an exothermic proton-

transfer reaction from protonated acetone or protonated methanol to the given amino acid.

Such reactions can be expected to occur at collision rate.56 The accuracy of the measured

rate constants is estimated to be around 50%, and the precision is estimated to be better

than 20%. The theoretical collision rate constants (kcoll) were obtained using a

parameterized trajectory theory.57 The efficiency of each reaction (the fraction of

collisions that leads to reaction) is given by kexp/kcoll. The primary products’ relative

abundances (branching ratios) are given as the ratio of a given primary product ions’

abundance divided by the sum of all primary product ions’ abundances.

Quantum chemical calculations were performed with the Gaussian 03 and Molpro

electronic structure program suites. Molecular geometries for the (bi)radicals were

calculated as described previously.48 For meta-benzyne analogues a-d, the S–T splitting

at the dehydrocarbon atom separation of the transition state (∆ES-T), the electron affinity

at the dehydrocarbon atom separation of the transition state (EA2.30), and the distortion

energy (∆E2.30) were calculated at the RHF-UCCSD(T)/cc-pVTZ//B3LYP/cc-pVTZ level

of theory. The potential energy surfaces for meta-benzynes a-d were calculated at the

UBLYP/cc-pVDZ//UBLYP/cc-pVDZ level of theory. The charge densities were

calculated at the UB3LYP/cc-pVTZ//UB3LYP/cc-pVTZ level of theory. The activation

enthalpies for hydrogen atom abstraction from methane and addition of water and

ammonia to different radical sites of the meta-benzyne analogues were calculated at the

MPW1K/6-31+G(d,p)//MPW1K/6-31+G(d,p) level of theory. For σ-monoradicals e-g,

the electron affinities were calculated at the RHF-UCCSD(T)/cc-pVTZ//B3LYP/cc-

pVTZ level of theory.

134

6.3 Results and Discussion

The four positively charged meta-benzyne analogues and three related σ-

monoradicals selected for this study are N-methyl-3,5-didehydropyridinium (a), N-

methyl-6,8-didehydroquinolinium (b), N-methyl-5,7-didehydroisoquinolinium (c), N-

methyl-5,7-didehydroquinolinium (d), N-methyl-3-dehydropyridinium (e), N-methyl-5-

dehydroisoquinolinium (f), and N-methyl-6-dehydroquinolinium (g) cations (Figure 6.2).

Before the reactivities of the meta-benzynes (a-d) toward amino acids were examined,

their reactivities towards simple organic molecules, such as tetrahydrofuran, allyl iodide,

dimethyl disulfide, and tert-butyl isocyanide, were investigated (Table 6.1). The reactions

of amino acids with related monoradicals were also examined (Table 6.2). Finally, meta-

benzyne analogues (a-d) were allowed to react with glycine, leucine, lysine, isotopically

labeled lysine (lysine-ε-15N), methionine, cysteine, proline and phenylalanine for variable

periods of time. Their reaction efficiencies and product branching ratios were determined,

as summarized in Tables 6.3-6.5. Three important reactivity-controlling parameters for

meta-benzynes were listed in the tables. They are S-T splitting (△ES-T), EA at the

transition state geometry (EA2.30), and distortion energy (△E2.30). The reason of

computing EA2.30 for meta-benzynes is because EA at the 2.3 Å transition state geometry

is a much more important reactivity controlling factor than the EA at the minimum

energy (ground state) geometry.48

135

Figure 6.2 Structures of the meta-benzyne analogues (a-d) and related monoradicals (e-g) studied.

In order to better understand the role of distortion energy as a reactivity

controlling parameter for meta-benzynes, potential energy surfaces for biradicals a-d are

shown in Figure 6.3. The distortion energies for N-methyl-3,5-didehydropyridinium (a)

and N-methyl-5,7-didehydroquinolinium (d) are similar, which are 7.6 and 8.6 kcal/mol

respectively. In contrast, the distortion energies for N-methyl-6,8-didehydroquinolinium

(b) and N-methyl-5,7-didehydroisoquinolinium (c) are much smaller, which are 4.2 and

3.5 kcal/mol respectively. Therefore, a small amount of energy is required to distort the

minimum energy dehydrocarbon atom separation (DAS) to the separation of the

transition state for biradical b and c. Hence, they are more likely to display radical

reactivity compared to biradical a and d. This finding was demonstrated in the present

study, as discussed below.

B

m

te

is

on

st

si

A

Fig

Biradicals’ r

The r

molecules ha

etrahydrofur

socyanide (tB

n the reactiv

tudied.

Biradi

imple organi

AI, a CN gro

gure 6.3 Rela

reactions wi

eactivity of

ave not be

an (THF),

BuNC) were

vity of the fo

ical b demo

ic substrates

up from tBu

ative energy meta-b

th simple or

the four N-

en examine

allyl iodide

e investigate

our meta-ben

onstrates mo

s. It abstracts

uNC (Table 6

versus dehy

benzyne anal

rganic mole

-methylated

ed previous

(AI), dime

ed. More im

nzynes befo

ostly radical

s an hydroge

6.1), just as

ydrocarbon alogues a-d.

ecules

biradicals a

sly. Therefo

ethyl disulfi

mportantly, th

ore their reac

l reactivity

en atom from

monoradica

atom separat

a – d towar

ore their re

fide (DMDS

his knowledg

ctions with a

when allow

m THF, an i

als do.58-60 A

tion for

d simple or

eactivity tow

S), and tert-

ge can shed

amino acids

wed to react

iodine atom

At the first gl

136

ganic

wards

-butyl

d light

were

with

from

lance,

137

this behavior appears to be in conflict with general expectations. A large S-T splitting,

such as that of b (-18.7 kcal/mol), is generally thought to hinder radical reactivity, as is

the case for many other meta-benzyne type biradicals.39,40 However, the key to the

behavior of biradical b is its small distortion energy (4.2 kcal/mol), an important

reactivity controlling parameter for meta-benzynes.48 Only a small amount of energy is

required to partially uncouple the biradical electrons in the transition state for radical

reactions. Some other reactions products, including 2H-atom abstraction, SCH3

abstraction, HCN abstraction, could be the results of nucleophilic addition reactions.

These results indicated that biradical b could react via both radical and nonradical

mechanisms.

For biradical c, fewer characteristic radical reactions were observed upon its

interaction with simple organic molecules (Table 6.1). Biradical c reacted with AI and

tBuNC by mainly I atom and minor CN group abstraction respectively, which are

characterized as radical reactions. However, no characteristic H-atom abstraction (radical

reaction) was observed for c upon reaction with THF. The other reactions products, such

as SCH3 abstraction from DMDS, HCN abstraction from tBuNC, could arise from

nucleophilic addition reactions. Generally, c was found to react with AI, DMDS, and

tBuNC slower than b did. These results indicated that biradical c could react with organic

substrates via both radical and nonradical mechanisms, but its radical reactivity is lower

than that of b. This can be rationalized as lower EA2.30 of c than that of b, although both

radicals have similar S-T splitting and distortion energy values.

Different from b and c, the only definite radical reaction observed for biradical a

was an iodide atom abstraction from AI (Table 6.1). It reacted with THF by 2H-atom

138

abstraction and C4H6 abstraction, while no H atom abstraction was observed. Addition

reactions were observed for a upon reaction with AI, and an exclusive HCN abstraction

was observed upon reaction with tBuNC. The predominant nonradical reactivity of a can

be rationalized by its relatively large distortion energy, i.e. an energy of 7.6 kcal/mol is

required to distort the biradical's minimum energy geometry to the separation of the

transition state. To avoid this energetically costly uncoupling of the biradical electrons,

biradical a tends to undergo nucleophilic addition reactions as opposed to radical

reactions. In terms of total reaction efficiencies, a reacted with most organic substrates

the fastest, except for THF. This is due to the fact that it has the largest EA2.30, which is

thought to enhance both radical and nonradical reactions. However, compared to

monoradicals, biradical a is less reactive.58-60 This is because of the biradical's relatively

large S-T splitting (-23.3 kcal/mol), which has been demonstrated to hinder radical

reactions.

Biradical d was found to be unreactive toward THF, AI and DMDS. It only

reacted with tBuNC by HCN abstraction at an efficiency 13-times lower than that of b.

The nonradical reactivity of d can be rationalized by its largest distortion energy (8.6

kcal/mol) among the four biradicals. The low reaction efficiency could be attributable to

its relatively small EA2.30 (5.14 eV). In agreement, a previous study reported that d is

unreactive toward dinucleoside phosphates whereas biradical b is not.51 In summary, the

radical reactivity of the four biradicals can be ranked as below based on their reactions

with organic substrates, b > c > a > d (none).

139

Monoradicals’ reactions with amino acids

The electorn affinity (EA) has been demonstrated as a major reactivity controlling

parameter for reactions of charged monoradicals with small organic substrates as well as

amino acids.21,22,25 The monoradicals studied here have decreasing EA values as follows:

radical e (5.75 eV) > radical f (4.74 eV) > radical g (4.57 eV). The ordering of EA's was

found to correlate with the monoradicals’ reaction efficiencies towards the amino acids

studied. In general, the greater the EA of the radical, the greater the total reaction

efficiency. Additionally, more NH2 abstraction and less hydrogen atom abstraction were

observed for radical e with larger EA than radical f and g with smaller EA's. This is in

agreement with the previous finding that, as the EA of a radical increases, its addition

reactions become faster than hydrogen atom abstraction.19,20 The details of the reactions

are discussed below.

Glycine and leucine react via two reaction pathways with radicals e – g, hydrogen

atom abstraction and NH2 abstraction (Table 6.2). The more electrophilic radical e reacts

by NH2 abstraction faster than the less electrophilic radicals f and g. The greater the EA

of the monoradical, the more the NH2 abstraction is favored. NH2 abstraction has been

suggested to occur via a nucleophilic addition-elimination reaction pathway that

progresses through the formation of an addition intermediate.21 In the case of glycine and

leucine, the formation of this addition intermediate is highly exothermic and fragments

due to the inability for gas-phase systems to transfer energy to the surroundings.

Additionally, leucine yields a higher branching ratio for hydrogen atom abstraction than

glycine (Table 6.2), suggesting that hydrogen atom abstraction preferentially occurs from

the side-chain of the amino acids. In fact, a previous study on the reactivity of partially

140

isotope labeled amino acids towards charged phenyl monoradicals confirms that most

hydrogen atoms are abstracted from the alkyl side chain and a small amount of hydrogen

atoms are abstracted from the α-carbon.24

Proline reacts with radicals e – g mainly via hydrogen atom abstraction. Proline

does not react by NH2 abstraction because its nitrogen atom is part of a five-membered

ring. However, a ring-opening product, formed by C2H4N abstraction, was also observed

for e – g. The C2H4N abstraction product was also observed for reactions of a charged

phenyl radical (N-phenyl-3-dehydropyridinium) with proline in an earlier study.24

C2H4N abstraction from proline likely starts by nucleophilic addition of the proline

nitrogen to the radical site, followed by elimination of CO2 as well as ethylene from the

ring in proline.24 Several additional reaction pathways were observed including

abstraction of OH group, addition and addition – OH, which could also be initiated by

similar nucleophilic addition of the proline to the radical.

For lysine and lysine-ε-15N, hydrogen atom abstraction and NH2 abstraction were

observed for reactions of radicals e – g. Both NH2 groups of L-lysine-ε-15N can react with

the radical sites, which agrees with the finding from a previous study.24 Additional

reaction pathways were observed including addition, addition – COOH, and abstraction

of CH2NH group. A possible mechanism for the formation of a stable adduct has been

proposed earlier for a charged phenyl radical (N-phenyl-3-dehydropyridinium) when it

reacts with lysine. 24

141

Biradicals’ reactions with amino acids

Glycine, Leucine, Lysine, and 15NH2-Lysine

The results for reactions of biradicals a – d toward the simple aliphatic amino

acids, glycine, leucine, lysine, and 15NH2-lysine are summarized in Table 6.3. A previous

study has demonstrated that the N-methyl-6,8-didehydroquinolinium cation b reacts with

amino acids via both radical reaction pathways and nucleophilic addition-elimination

pathways.49 The reaction products observed in this study are similar to those in the

literature report, including one (or two) hydrogen atoms abstraction, H2O abstraction,

addition, addition – CO2, addition – HCOOH and addition – COOH (Table 6.3). The

abstraction of two hydrogen atoms can be rationalized by radical reaction mechanisms,

which is likely initiated by H-atom abstraction from α-carbon or from the alkyl chain of

the amino acids.49 The other reactions observed are likely initiated by nucleophilic

addition of NH2 or OH group to the more electrophilic radical site at carbon 6.49

For biradical c which has similar S-T splitting (-17.6 kcal/mol) and distortion

energy (3.5 kcal/mol) compared to that of biradical b, it is reasonable to predict that it

may show some radical-type reactivity just as biradical b does. However, one hydrogen

atom abstraction (radical reaction) was not observed for biradical c upon interaction with

aliphatic amino acids. Only trace amount and small amount of 2H-atom abstraction were

observed when biradical c reacts with glycine and lysine respectively. Unlike H-atom

abstraction, 2H-atom abstraction can be either radical (consecutive hydrogen atom

abstraction) or nonradical reactions (hydride abstraction followed by proton transfer).

The majority of other reactions pathways observed for biradical c are addition, NH3

abstraction, H2O abstraction, addition – COOH and addition – H2O (Table 6.3), which

142

are likely initiated by nucleophilic addition-elimination reactions (nonradical reactions).

The absence of H-atom abstraction pathway for biradical c is likely due to its low EA2.30,

yet the radical reactivity of biradical c could not be ruled out based on these results.

For biradical a with a relatively large S-T splitting (-19.4 kcal/mol) and large

distortion energy (7.6 kcal/mol), only trace amount of 2H-atom abstraction was observed

when it reacted with glycine and leucine. Most of the other reactions were nucleophilic

addition reactions, which were similar to those observed for biradical b with the same

group of amino acids, such as H2O abstraction, addition, addition – CO2, addition –

HCOOH and addition – COOH (Table 6.3). This suggests that nonradical reaction

pathway dominates for the reactions of biradical a with aliphatic amino acids. It should

be noted that the branching ratio of 2H-atom abstraction increased for biradical a upon

interaction with lysine, which has a larger alkyl side chain. This finding agrees with what

has been observed for positively charged phenyl monoradicals, i.e. the larger the alkyl

side chain of an amino acid, the higher the branching ratio of H-atom abstraction.21 This

is due to the fact that hydrogen atoms can be abstracted from both α-carbon and the alkyl

side chain.24 Different from biradical b, biradical a displays an additional reaction

pathway, NH3 abstraction. Since both NH3 and 15NH3 groups were abstracted from lysine

labeled with 15N on the side chain, NH3 abstraction is likely initiated by NH2 abstraction

from either the amino terminus or the side chain of lysine followed by a hydrogen atom

abstraction by another unquenched radical site.

For biradical d with a relatively large S-T splitting (-24.6 kcal/mol) and large

distortion energy (8.6 kcal/mol), it is predicted to react mostly via nonradical instead of

radical pathways. Indeed, the only reaction pathways observed for d are H2O abstraction

143

and addition – COOH upon interaction with lysine, whereas no reaction products were

observed upon interaction with glycine or leucine. Moreover, d has the lowest reaction

efficiency, if any, among the four biradicals, albeit it has a slightly higher EA2.30 (5.14 eV)

than that of c (4.93 eV). This can be explained by the largest distortion energy of d

among the four biradicals. A possible mechanism for the formation of adduct – COOH

for biradical d upon reaction with lysine is shown in Scheme 6.1. Based on atomic charge

calculations, the radical site 7 of biradical d is the more electrophilic site, which is most

likely the radical site that initiates the reactions. Moreover, the calculated activation

enthalpies for the nucleophilic addition of ammonia and water to the radical site 7 of d

are 12.2 kcal/mol and 22.4 kcal/mol respectively, which indicates that the nucelophilic

addition of the amino acid's NH2 group to the radical site is kinetically favored.

N

CH3

N

CH3

N

CH3

HO

O

NH2

NH2

NHO

O

NN

CH3

N

NHCOOHLoss of

Addition-COOH

H

H

H H H

H

Scheme 6.1 Proposed mechanism for the formation of adduct – COOH for biradical d upon reaction with lysine

C

su

m

C

H

ac

co

sh

H

le

du

co

S

P

w

Cysteine and

The p

usceptible to

monoradicals

CH3).21 In co

HSR abstract

cids just lik

omplex diss

hown in Sch

H2O abstracti

ess prevalent

ue to the hi

ompeting rea

Scheme 6.2 P

Proline and P

The r

with biradica

d Methionin

presence of

o biradical a

s with this gr

ntrast, birad

tion (Table 6

ke for mono

sociates.49 A

heme 6.2. Co

ion, addition

t for reactio

igh reactivity

actions.

Proposed me

Phenylalani

eaction path

ls b are sim

ne

f the S-C

attack. Previ

roup of amin

dicals’ reactio

6.4), which

oradicals, fo

A possible m

ompared to H

n, addition –

ons of a – d

y of the SR

echanism for

ine

hways obser

ilar to those

bond in cy

ious studies

no acids are

ons with this

is likely init

llowed by H

mechanism f

HSR abstrac

CO2, additio

with this gr

R group in th

r HSCH3 abs

rved for the

e observed fo

ysteine and

have demon

e dominated

s group of a

tiated by SR

H atom abs

for HSCH3

ction, the oth

on – COOH

roup of ami

he amino ac

straction fro

reactions o

or glycine an

methionine

nstrated that

by SR abstr

amino acids a

R abstraction

straction bef

abstraction

her reaction p

H, and additio

ino acids. Th

cids, which h

m methionin

of proline an

nd leucine. T

e is particu

t the reactio

raction (R =

are dominate

n from the a

fore the coll

by biradical

pathways su

on – HCOOH

his is most l

hinders the

ne by biradic

nd phenylal

They include

144

ularly

ons of

H or

ed by

amino

lision

l c is

uch as

H are

likely

other

cal c.

anine

e one

(o

–

v

pr

ph

n

b

ra

b

S

or two) hydr

HCOOH an

ia both radi

roline mostl

henylalanine

onradical pa

e ruled out a

adical and n

iradical c, as

Scheme 6.3 P

rogen atoms

nd addition –

cal and non

ly by H2O a

e exclusively

athways. Ho

as discussed

nonradical m

s shown in S

Proposed rad

abstraction,

– COOH (Ta

nradical path

abstraction,

y by additio

wever, the r

d above for t

mechanisms w

Schemes 6.3

dical mechan

, H2O abstra

able 6.5), wh

hways, as di

addition and

on. These re

radical react

the reaction

were propos

and 6.4 resp

nism for H2O

action, additi

hich indicate

iscussed abo

d addition –

esults sugge

tion mechan

with organi

sed for H2O

pectively.

O abstraction

ion, addition

ed that biradi

ove. Biradica

– CO2, whil

ested that c

nism for bira

ic substrates

O abstraction

n from prolin

n – CO2, add

ical b could

al c reacted

le it reacted

tend to reac

adical c coul

s. Therefore,

n from prolin

ne by biradic

145

dition

react

d with

with

ct via

ld not

, both

ne by

cal c.

ph

to

re

p

ph

at

sm

d

ar

Scheme

Biradi

henylalanine

o react via n

eacted with

athways wer

henylalanine

t low efficie

mall EA2.30.

d upon reacti

re shown in

e 6.4 Propos

ical a reacte

e mainly by

nucleophilic

proline mo

re also obse

e. As discus

encies due t

Possible me

on with prol

Schemes 6.5

ed nonradica

ed with prol

H2O abstrac

addition-eli

ostly by add

erved. No re

sed above, d

to its large d

echanisms fo

line, i.e. H2O

5 and 6.6 res

al mechanismby biradica

line exclusiv

ction and ad

imination pa

dition, while

action produ

d is either u

distortion en

or all the no

O abstraction

spectively.

m for H2O aal c.

vely by H2O

ddition. Thes

athways as d

e H2O abstr

ucts were ob

unreactive to

nergy, large

onradical rea

n as well as

abstraction fr

O abstractio

se results ind

discussed ab

raction and

bserved whe

oward amino

e S-T splittin

actions obser

addition and

rom proline

on, and it re

dicate that a

bove. Biradi

addition –

en d reacted

o acids or re

ng and relat

rved for bira

d addition –

146

eacted

a tend

ical d

CO2

d with

eacted

tively

adical

CO2,

147

Scheme 6.5 Proposed mechanism for H2O abstraction from proline by biradical d

Scheme 6.6 Proposed mechanism for formation of adduct and adduct-CO2 for biradical d upon reaction of with proline.

Tab

le 6

.1

ES

-T,a k

cE

A2.

30,a

E2.

30,a k

c

MW

7 M

W 1

MW

9

Rea

ctio

n ef

fici

endi

met

hyl d

isul

fi

cal/m

ol

a eV

ca

l/mol

72

2 C E

168

A

94

S SS nc

ies

(Eff

.) a

nd p

de, t

ert-

buty

l iso

c

m

/z 9

2 a

-23.

3 6.

17

7.6

x H

abs

e 57%

C

4H6

abs

43%

E

ff. =

0.2

1%

I ab

s 5

5%

(2°)

I a

bs

Add

ition

45%

E

ff. =

3%

CH

3 ab

s 8

7%

(2°)

SC

H3

abs

(2°)

SS

CH

3 ab

s(2

°) C

H2

abs

SC

H3

abs

13%

(2

°) S

CH

3 ab

s E

ff. =

39%

N CH

3

rodu

ct b

ranc

hing

cyan

ide,

and

cyc

lth

e pr

imar

y

m/z

b-1

85.

2 4.H

abs

2 x

H a

bE

ff. =

Unr

eact

ive

I ab

s

Ally

l ab

All

yl-H

E

ff. =

U

nrea

ctiv

e i

SC

H3

ab (

2°)

SC

HS

CH

3 a

SS

CH

3 a

Eff

. =

Unr

eact

ive g

rati

os f

or b

irad

iulo

hexa

ne; s

econ

dpr

oduc

ts th

at p

ro

14

2 b 8.

7 25

2 7

0%

bs 3

0%

= 1

%

isom

er 7

%

83%

bs

15%

ab

s 2

%

2.5%

is

omer

40%

bs 9

6%

CH

3 ab

s ab

s 2

%

abs

2%

30

%

isom

er 7

%

N CH

3

ucal

s a

– d

upo

n r

dary

pro

duct

s ar

e od

uce

them

. m

/z 1

42

c-1

7.6

4.93

3.

5

No

Rea

ctio

n

I ab

s 8

3%A

ddit

ion

17 %

Eff

. = 0

.3%

SC

H3

abs 1

00(2

°) S

CH

3 a

E

ff. =

6%

reac

tion

wit

h te

trno

ted

as (

2o ) an

d

n %

% 0%

ab

s

rahy

drof

uran

, ally

d ar

e li

sted

aft

er

m/z

142

d

-2

4.6

5.14

8.

6

No

Rea

ctio

n

No

Rea

ctio

n

No

Rea

ctio

n

148

yl io

dide

,

148

ES

-T,a k

cE

A2.

30,a

E2.

30,a k

c M

W 8

a Cal

cula

ted

at th

cal/m

ol

a eV

ca

l/mol

83

H

he R

HF

-UC

CS

D(T

)/c

m

/z 9

2 a

-23.

3 6.

17

7.6

CN

abs

100

%

Eff

. = 6

9%

cc-p

VT

Z//B

3LY

P/cc

N CH

3

Ta

m/z

b-1

85.

2 4.

HC

N a

b (

2°)

C4

(2°

) H

C (

2°)

Ad

CN

ab

(

2°)

CE

ff. =

U

nrea

ctiv

e -p

VT

Z le

vel o

f th

eorab

le 6

.1, c

onti

nue

14

2 b 8.

7 25

2 bs

93%

4H

8 ab

s C

N a

bs

dditi

on

bs 7

%

H2+

. tran

sfer

52

%

isom

er 5

%

ry. N C

H3

ed

m

/z 1

42

c-1

7.6

4.93

3.

5

HC

N a

bs 8

5% (

2°)

HC

N a

b (

2°)

Add

itio

CN

abs

15 %

(2°)

CN

abs

Eff

. = 3

9%

%

bs

on

%

s %

m/z

142

d

-2

4.6

5.14

8.

6

HC

N a

bs 1

00%

(2

°) H

CN

abs

(2

°) A

dditi

on

Eff

. = 4

%

149

149

150

Table 6.2 Reaction efficiencies (Eff.) and product branching ratios for monoradicals upon reaction with L-glycine, L-leucine, L-proline, L-lysine, DL-lysine-ε-15N.

m/z 93 e

m/z 143

f

m/z 143

g

EA (eV)a 5.75 4.74 4.57

Glycine MW 75

H absb 35% NH2 abs 65%

Eff. = 19%

H abs 82% NH2 abs 18% Eff. = 2.5%

H abs 68% NH2 abs 32% Eff. = 2.6%

Leucine MW 131

H abs 62% NH2 abs 38%

Eff. = 71%

H abs 97% NH2 abs 3% Eff. = 12%


Eff. = 6%

Proline MW 115

H abs 62% C2H4N abs 16%

OH abs 16% Addition 3%

Addition-OH 3% Eff. = 26%

H abs 84% C2H4N abs 4% Addition 8%

Addition – OH 4% Eff. = 17%

H abs 34% C2H4N abs 8% Addition 37%

Addition– OH 21% Eff. = 6%

Lysine MW 146

H abs 43% NH2 abs 41% Addition 14%

Addition-COOH 2%

Eff. = 58%


CH2NH abs 3% Addition 2% Eff. = 31%

H abs 68% NH2 abs 20% Addition 12%

Eff. = 15%

DL-lysine-ε-15N MW 147


15NH2 abs 18% Addition 10%

Addition-COOH 1%

Eff. =65%


15NH2 abs 2% CH2NH abs 5%

Addition 1% Eff. =33%


15NH2 abs 16% Addition 12%

Eff. = 13%

aCalculated at the RHF-UCCSD(T)/cc-pVTZ//B3LYP/cc-pVTZ level of theory.

NH2

CH

C

CH2

OH

O

CH

CH3

CH3

N

CH3

151

Tab

le 6

.3 R

eact

ion

effi

cien

cies

(E

ff.)

and

pro

duct

bra

nchi

ng r

atio

s fo

r bi

radi

ucal

s a

– d

upo

n re

acti

on w

ith

gl

ycin

e, L

-leu

cine

, L-l

ysin

e, a

nd D

L-l

ysin

e-ε-

15N

.

m

/z 9

2 a

m

/z 1

42

b

m/z

142

c

m

/z 1

42

d

ES

-T,a k

cal/

mol

E

A2.

30,a e

V

E

2.30

,a kca

l/m

ol

-23.

3 6.

17

7.6

-18.

7 5.

25

4.2

-17.

6 4.

93

3.5

-24.

6 5.

14

8.6

G

lyci

ne

MW

75

Tra

ce 2

H a

bs

H2O

abs

57%

A

ddit

ion-

CO

2 4

%

Add

itio

n -

HC

OO

H 3

9%

Eff

. = 3

2%

H a

bs 7

%

2 H

abs

9%

H

2O a

bs 1

3%

Add

itio

n 2

4%

Add

itio

n-C

O2

38%

A

ddit

ion-

HC

OO

H 9

%

Unr

eact

ive

isom

er 2

4%

Eff

. = 7

.5%

Tra

ce 2

H a

bs

Add

itio

n 1

00%

E

ff. =

0.1

%

No

Rea

ctio

n

L

-leu

cine

M

W 1

31

Tra

ce 2

H a

bs

H2O

abs

42%

A

ddit

ion

32%

A

ddit

ion-

CO

2 1

%

Add

itio

n- H

CO

OH

20%

A

ddit

ion-

C4H

8 5

%

Eff

. = 7

0%

H a

bs 8

%

2 H

abs

9%

H

2O a

bs 2

7%

Add

itio

n 2

0%

Add

itio

n-C

O2

18%

A

ddit

ion-

HC

OO

H 1

8%

Eff

. = 2

3%

NH

3 ab

s 9

%

Add

itio

n 9

1%

Eff

. = 2

.2%

No

Rea

ctio

n

N CH

3

N CH

3

151

152

Tab

le 6

.3, c

onti

nued

m/z

92

a

m

/z 1

42

b

m/z

142

c

m

/z 1

42

d

ES

-T,a k

cal/

mol

E

A2.

30,a e

V

E

2.30

,a kca

l/m

ol

-23.

3 6.

17

7.6

-18.

7 5.

25

4.2

-17.

6 4.

93

3.5

-24.

6 5.

14

8.6

L

-lys

ine

M

W 1

46

2 H

abs

6%

N

H3

abs

23%

H

2O a

bs 5

%

Add

itio

n 1

%

Add

itio

n-C

O 3

%

Add

itio

n-C

OO

H 6

2%

Eff

. = 6

7%

2 H

abs

12%

H

2O a

bs 1

5%

Add

itio

n-C

OO

H 7

3%

Eff

. = 2

1%

2 H

abs

9%

H

2O a

bs 6

%

Add

itio

n-C

OO

H 7

5%

Add

itio

n-H

2O 1

0%

Eff

. = 1

9%

H2O

abs

12%

A

ddit

ion-

CO

OH

88%

E

ff. =

4%

D

L-l

ysin

e-ε-

15N

M

W 1

47

2 H

abs

10%

N

H3

abs

22%

15

NH

3 ab

s 1

2%

Add

itio

n-C

OO

H 5

6%

Eff

. = 8

1%

Not

rel

evan

t N

ot r

elev

ant

Not

rel

evan

t

a Cal

cula

ted

at th

e R

HF-

UC

CSD

(T)/

cc-p

VT

Z//B

3LY

P/c

c-pV

TZ

leve

l of

theo

ry.

N CH

3

N CH

3

152

153

Tab

le 6

.4 R

eact

ion

effi

cien

cies

(E

ff.)

and

pro

duct

bra

nchi

ng r

atio

s fo

r bi

radi

ucal

s a

– d

upo

n re

acti

on w

ith

L-m

ethi

onin

e an

d L

-cys

tein

e.

m

/z 9

2 a

m

/z 1

42

b

m/z

142

c

m

/z 1

42

d

ES

-T,a k

cal/

mol

E

A2.

30,a e

V

E

2.30

,a kca

l/m

ol

-23.

3 6.

17

7.6

-18.

7 5.

25

4.2

-17.

6 4.

93

3.5

-24.

6 5.

14

8.6

L

-met

hion

ine

MW

149

2 H

abs

d 33%

H

2O a

bs 3

0%

OH

abs

4%

SH

2 ab

s 6

%

HS

CH

3 ab

s 2

2%

Add

itio

n –

CO

2 5%

E

ff. =

74%

H a

bs 4

%

2 H

abs

4%

H

2O a

bs 3

9%

SH2

abs

1%

, S

CH

3 ab

s 4

%

HS

CH

3 ab

s 3

9%

SC2H

4 ab

s 4

%,

Add

itio

n 5

%

Unr

eact

ive

isom

er 1

2%

Eff

. = 5

5%

H2O

abs

12%

SH

2 ab

s 6

%

HS

CH

3 ab

s 8

2%

Eff

. = 4

7%

H2O

abs

88%

H

SC

H3

abs

12%

E

ff. =

5%

L

-cys

tein

e M

W 1

21

H2O

abs

22%

SH

2 ab

s 7

2%

Add

ition

– C

OO

H 6

%

Eff

. = 4

3%

H a

bs 4

%

2 H

abs

4%

N

H2

abs

2%

N

H3

abs

8%

H

2O a

bs 2

0%

SH

abs

1%

SH

2 ab

s 5

2%

Add

itio

n 5

%

Add

ition

– C

OO

H 3

%

Add

itio

n –

HC

OO

H 1

%

Unr

eact

ive

isom

er 2

5%

Eff

. = 2

2%

SH

abs

19%

SH

2 ab

s 7

1%

Add

itio

n 1

0%

Unr

eact

ive

isom

er 2

0%

Eff

. = 3

%

No

Rea

ctio

n

a Cal

cula

ted

at th

e R

HF-

UC

CSD

(T)/

cc-p

VT

Z//B

3LY

P/c

c-pV

TZ

leve

l of

theo

ry.

N CH

3

NH

2

CHC

CH

2

OH

O

CH

2S

CH

3

153

154

Tab

le 6

.5 R

eact

ion

effi

cien

cies

(E

ff.)

and

pro

duct

bra

nchi

ng r

atio

s fo

r bi

radi

ucal

s a

– d

upo

n re

acti

on w

ith

L-p

rolin

e an

d L

-phe

nyla

lani

ne.

m/z

92

a

m

/z 1

42

b

m/z

142

c

m

/z 1

42

d

ES

-T,a k

cal/

mol

E

A2.

30,a e

V

E

2.30

,a kca

l/m

ol

-23.

3 6.

17

7.6

-18.

7 5.

25

4.2

-17.

6 4.

93

3.5

-24.

6 5.

14

8.6

L

-pro

line

M

W 1

15

H2O

abs

d 100

%

Eff

. = 9

5%

H a

bs 2

%

2 H

abs

4%

H

2O a

bs 6

6%

Add

itio

n 6

%

Add

itio

n –

CO

2 7

%

Add

itio

n –

HC

OO

H 1

5%

Unr

eact

ive

isom

er 2

2%

Eff

. = 4

2%

2 H

abs

6%

H

2O a

bs 3

7%

Add

itio

n 3

8%

Add

itio

n –

CO

2 1

9%

Eff

. = 2

%

H2O

abs

17%

A

ddit

ion

70%

A

ddit

ion

– C

O2

13%

E

ff. =

0.7

%

L-p

heny

lala

nine

M

W 1

65

H2O

abs

50

%

OH

abs

13%

A

ddit

ion

30%

A

ddit

ion

– H

CO

OH

7%

E

ff. =

72

%

2 H

abs

5%

H

2O a

bs 3

4%

Add

itio

n 2

4%

Add

itio

n –

CO

2 1

2%

Add

ition

– C

OO

H 7

%

Add

itio

n –

HC

OO

H 1

3%

Add

itio

n –

OH

5%

U

nrea

ctiv

e is

omer

12%

E

ff. =

28%

Add

itio

n 1

00%

E

ff. =

3%

No

Rea

ctio

n

a Cal

cula

ted

at th

e R

HF-

UC

CSD

(T)/

cc-p

VT

Z//B

3LY

P/c

c-pV

TZ

leve

l of

theo

ry.

N CH

3

154

155

6.4 Conclusions

The reactivities of four meta-benzyne analogues, a – d, toward four organic

substrates and eight amino acids were examined in a dual-cell Fourier-transform ion

cyclotron resonance (FT-ICR) mass spectrometer. Based on the four biradicals' reactivity

toward THF, AI, DMDS, and tBuNC, their radical reactivity follows the order of b > c >

a > d (none). Similarly, biradicals b and c display more radical reactivity toward amino

acids than biradical a and d do. The radical or nonradical reactivity toward organic

substrates and amino acids is largely affected by the distortion energies. EA at the

transition state geometry affects the total reaction efficiency of meta-benzynes, with the

three biradicals following the order of a > b > c that is generally consistent with the

EA2.30 ordering.

Overall, three important parameters were found to affect the reactivity of meta-

benzynes toward amino acids as well as organic substrates. They are (1) the S-T splitting

at the separation of the transition state, ∆ES-T; (2) the electron affinity at the separation of

the transition state, EA2.30; (3) the energy required to distort the minimum energy

dehydrocarbon atom separation to the separation of the transition state, ∆E2.30. This is the

first study on how these distinct chemical properties can affect the gas-phase reactivities

of the selected meta-benzynes towards amino acids. Different from related monoradicals,

which reacted with amino acids mainly by H atom abstraction and NH2 group abstraction,

the biradicals reacted via multiple pathways, including one (or two) hydrogen atoms

abstraction, H2O abstraction, addition, addition – CO2, addition – HCOOH and addition –

COOH. The extent to which of these reaction pathways (radical or nonradical) dominate

156

is highly influenced by the three reactivity-controlling parameters of the meta-benzyne

analogues.

157

6.5 References

1. Davies, M. J.; Dea, R. T. Radical-Mediated Protein Oxidation: From Chemistry to Medicine; Oxford University Press: Oxford, 1997; pp 1–443.

2. Davies, J. M.; Fu, S.; Wang, H.; Dean, R. T. Free Radic. Biol. Med. 1999, 27, 1151–1163.

3. Dean, T. R.; Fu, S.; Stocker, R.; Davis, J. M. Biochem. J. 1997, 324, 1–18.

4. Hawkins, C. L.; Davies, M. J. Biochim. Biophys. Acta. 2001, 1504,196–219.

5. Davies, J. K. Protein Damage and Degradation by Oxygen Radicals. IV. Degradation of Denatured Protein. J. Biol. Chem. 1987, 262, 9914–9920.

6. Goshe, B. M.; Chen, H. Y.; Anderson, E. V. Biochemistry 2000, 39, 1761–1770.

7. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data 1988, 17, 513–887.

8. Bonifacic, M.; Stefanic, I.; Hug, G. L.; Armstrong, D.A.; Ausmus, K. D. J. Am. Chem. Soc. 1998, 120, 9930-9940.

9. Nicolaou, K. C.; Dai, W.-M. Angew. Chem. Int. Ed. Engl. 1991, 30, 1387–1416

10. Grissom, J. W.; Gunawardena, G. U.; Klingberg, D.; Huang, D. H. Tetrahedron 1996, 52, 6453–6518.

11. Wenk, H. H.; Winkler, M.; Sander, W. Angew. Chem. Int. Ed. Engl. 2003, 42, 502–528.

12. Zein, N.; Casazza, A. M.; Doyle, T. W.; Nadler, S. G. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 8009–8012.

13. Zein, N.; Solomon, W.; Casazza, A. M.; Kadow, J. F.; Krishnan, B. S.; Tun, M. M.; Vyas, D. M.; Doyle, T. W. Bioorg. Med. Chem. Lett. 1993, 3, 1351–1356.

14. Jones, G. B.; Plourde, G. W.; Wright, J. M. Org. Lett. 2000, 2, 811–813

15. Braslau, R.; Anderson, M. O. Tet. Lett. 1998, 39, 4227-4230.

16. Kenttämaa, H. I. Org. Mass Spectrom. 1994, 29, 1–10.

17. Chyall, L. J.; Kenttämaa, H. I. J. Am. Chem. Soc. 1994, 116, 3135.

18. Smith, R. L.; Kenttämaa, H. I. J. Am. Chem. Soc. 1995, 117, 1393.

158

19. Heidbrink, J. L.; Ramirez-Arizmendi, L. E.; Thoen, K. K.; Guler, L.; Kenttämaa, H. I. J. Phys. Chem. A 2001, 105, 7875-7884.

20. Ramirez-Arizmendi, L. E.; Guler, L.; Ferra, J. J.; Thoen, K. K.; Kenttämaa, H. I. Int. J. Mass Spectrom. 2001, 210/211, 511-520.

21. Huang, Y.; Guler L.; Heidbrink, J.; Kenttämaa, H. I. J. Am. Chem. Soc. 2005, 127, 3973–3978.

22. Huang, Y.; Kenttämaa, H. I. J. Am. Chem. Soc. 2005, 127, 7952–7960.

23. Li, S.; Fu, M.; Habicht, S. C.; Pates, G. O.; Nash, J. J.; Kenttämaa, H. I. J. Org. Chem. 2009, 74, 7724–7732.

24. Pates, G. O.; Guler, L.; Nash, J. J.; Kenttämaa, H. I. J. Am. Chem. Soc. 2011, 133, 9331–9342.

25. Jing, L.; Nash, J. J.; Kenttämaa, H. I. J. Am. Chem. Soc. 2008, 130, 17697–17709.

26. Donahue, N. M.; Clarke, J. S.; Anderson, J. G. J. Phys. Chem. A. 1998, 102, 3923–3933.

27. Adeuya, A.; Yang, L.; Amegayibor, F. S.; Nash, J. J.; Kenttämaa, H. I. J. Am. Soc. Mass Spectrom. 2006, 17, 1325.

28. Wittig, G. Naturwissenschaften. 1942, 30, 696–703.

29. Roberts, J. D.; Simmons, H. E.; Carlsmith, L. A.; Vaughan, C. W. J. Am. Chem. Soc. 1953, 75, 3290–3291.

30. Chapman, O. L.; Mattes, K.; McIntosh, C. L.; Pacansky, J.; Calder, G. V.; Orr, G. J. Am. Chem. Soc. 1973, 95, 6134-6135.

31. Radziszewski, J. G.; Hess, B. A. J.; Zahradnik, R. J. Am. Chem. Soc. 1992, 114, 52-57.

32. Schweig, A.; Muenzel, N.; Meyer, H.; Heidenreich, A. Struct. Chem. 1990, 1, 89-100.

33. Warmuth, R. Angew. Chem., Int. Ed. 1997, 36, 1347–1350.

34. Batchelder, R. M.; Wilson, W. R.; Hay, M. P.; Denny, W. A. Br. J. Cancer, Suppl. 1996, 27, S52–S56

35. Bergman, R. G. Acc. Chem. Res. 1973, 6, 25–31.

36. Myers, A. G.; Cohen, S. B.; Kwon, B. M. J. Am. Chem. Soc. 1994, 116, 1255–1271.

159

37. Jones, G. B.; Warner, P. M. J. Am. Chem. Soc. 2001, 123, 2134–2145.

38. Baroudi, A.; Mauldin, J.; Alabugin, I. V. J. Am. Chem. Soc. 2010, 132, 967–979.

39. Chen, P. Angew. Chem. Int. Ed. Engl. 1996, 35, 1478-1480.

40. Price, J. M.; Kenttämaa, H. I. J. Phys. Chem. A. 2003, 107, 8985-8995.

41. Wenthold, P. G.; Squires, R. R.; Lineberger, W. C. J. Am. Chem. Soc. 1998, 120, 5279-5290.

42. Thoen, K. K.; Kenttämaa, H. I. J. Am. Chem. Soc. 1997, 119, 3832–3833.

43. Logan, C. F.; Chen, P. J. Am. Chem. Soc. 1996, 118, 2113–2114.

44. Hoffner, J.; Schottelius, M. J.; Feichtinger, D.; Chen. P. J. Am. Chem. Soc. 1998, 120, 376–385.

45. Johnson, W. T. G.; Cramer, C. J. J. Am. Chem. Soc. 2001, 123, 923-

46. Johnson, W. T. G.; Cramer, C. J. J. Phys. Org. Chem. 2001, 14, 597-

47. Nash, J. J.; Nizzi, K.; Adeuya, A.; Yurkovich, M. J.; Cramer, C. J.; Kenttämaa, H. I. J. Am. Chem. Soc. 2005, 127, 5760-5761.

48. Gao, J.; Jankiewicz, B. J.; Reece, J.; Sheng, H.; Cramer, C. J.; Nash, J. J.; Kenttämaa, H. I. Chem. Sci. 2014, 5, 2205–2215.

49. Fu, M.; Li, S.; Archibold, E.; Yurkovich, M. J.; Nash, J. J.; Kenttämaa, H. I. J. Am. Soc. Mass. Spectrom. 2010, 21, 1737-1752.

50. Linstrom, P. J.; Mallard, W. G., Eds. NIST Chemistry WebBook, NIST Standard Reference Database Number 69, March 2003; National Institute of Standards and Technology: Gaithersburg, MD, 20899 (http://webbook.nist.gov).

51. Yang, L.; Nash, J. J.; Yurkovich, M. J.; Jin, Z.; Vinueza, N. R.; Kenttämaa, H. I. Org. Lett. 2008, 10, 1889-1892.

52. Guan, S.; Kim, H. S.; Marshall, A. G.; Wahl, M. C.; Wood, T. D.; Xiang, X. Chem. Rev. 1994, 94, 2161–2182.

53. Gauthier, J. W.; Trautman, T. R.; Jacbson, D. B. Anal. Chim. Acta 1991, 246, 211–225.

54. Chen, L.; Wang, T. C. L.; Ricca, T. L.; Marshall, A. G. Anal. Chem. 1987, 59, 449–454.

160

55. Leeck, D. T.; Stirk, K. M.; Zeller, L. C.; Kiminkinen, L. K. M.; Castro, L. M.; Vainiotalo, P.; Kenttämaa, H. I. J. Am. Chem. Soc. 1994, 116, 3028-3030.

56. Thoen, K. K.; Smith, R. L.; Nousiainen, J. J.; Nelson, E. D.; Kenttämaa, H. I. J. Am. Chem. Soc. 1996, 118, 8669–8676.

57. Su, T.; Chesnavich, W. J. J. Chem. Phys. 1982, 76, 5183–5185.

58. Nelson, E. D.; Artau, A.; Price, J. M.; Kenttämaa, H. I. J. Am. Chem. Soc. 2000, 122, 8781 –8782.

59. Amegayibor, F. S.; Nash, J. J.; Kenttämaa, H. I. J. Am. Chem. Soc. 2003, 125, 14256 –14257.

60. Vinueza, N. R.; Archibold, E. F.; Jankiewicz, B. J.; Gallardo, V. A.; Habicht, S. C.; Aqueel, M. S.; Nash, J. J.; Kenttämaa, H. I. Chem. Eur. J. 2012, 18, 8692 – 8698.

VITA

161

VITA

Weijuan Tang was born in Guiyang, Guizhou Province, China on April 9, 1988.

She graduated from Guiyang No.1 High School in the June 2006. Afterwards, she entered

China Pharmaceutical University in Nanjing, China, where she graduated with a bachelor

degree in pharmaceutical science in June 2010. Her undergraduate research focused on

synthesis and biological evaluation of promising drug molecules. In the summer of 2010,

she went to Purdue University for her Ph.D. degree. She joined Professor Hilkka I.

Kenttämaa’s group where she worked on a variety of analytical chemistry projects,

including structural characterization of petroleum asphaltenes and organosulfur model

compounds, functional group selective ion-molecule reactions for drug metabolite

identification and fundamental studies on the gas-phase reactivity of meta-benzyne

towards amino acids. She was granted the Doctor of Philosophy degree in Chemistry

from Purdue in May, 2015.

MASS SPECTROMETRIC STUDIES ON PETROLEUM ASPHALTENES …

Documents