<%BANNER%>

Differentiation of Carbohydrate Isomers by Tunable Infrared Multiple Photon Dissociation and Fourier Transform Ion Cyclo...

Permanent Link: http://ufdc.ufl.edu/UFE0024320/00001

Material Information

Title: Differentiation of Carbohydrate Isomers by Tunable Infrared Multiple Photon Dissociation and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
Physical Description: 1 online resource (139 p.)
Language: english
Creator: Stefan, Sarah
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anomers, carbohydrates, disaccharides, fticr, irmpd, isomers, monosaccharides, ms
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Carbohydrates and their derivatives play a crucial role in many biological processes including fertilization, cell growth, inflammation and post-translational protein modification. The function of carbohydrates in these systems is closely related to their structure, including monosaccharide sequence, glycosidic linkage and stereochemistry. Unfortunately, the number of anomeric configurations and possible linkages between monosaccharide units makes analysis of carbohydrate structures complex. In order to shed light on these larger oligosaccharides, the fragmentation patterns and infrared multiple photon dissociation (IRMPD) spectra of various mono- and disaccharides were obtained and compared. For this work, various tunable infrared sources including a line-tunable continuous-wave carbon dioxide laser and a free electron laser (FEL) were used in conjunction with Fourier transform ion cyclotron resonance mass spectrometry (FTICR-MS). The first three projects used a line-tunable carbon dioxide laser to fragment various mono- and disaccharides in both the positive and negative ion modes. In the first project, anomers of lithium-cation attached O-methyl-gluco- and galactopyranosides were fragmented. The identity and anomeric configuration of each monosaccharide was accurately determined by comparing fragmentation patterns and ratios of certain fragments. A second project explored the fragmentation pattern of lithiated glucose-containing disaccharides having various linkages (1-2, 1-3, 1-4 and 1-6) and anomeric configurations (alpha and beta). Both the linkage and anomeric configuration of the various disaccharides were successfully identified based on their fragmentation patterns at several wavelengths. Next, irradiation of deprotonated and chlorinated glucose-containing disaccharides produced fragmentation patterns in which cleavage of the glycosidic bond resulted in major abundances of m/z 161 and 179 fragment ions. Along with differentiating the anomeric configuration for the chlorinated disaccharides, comparison of the abundances for major fragment ions also resulted in the positive identification of the linkages for both sets of disaccharides. Lastly, several deprotonated (negatively charged) mono- and disaccharides were fragmented with a FEL. The IRMPD spectra of the monosaccharide anions (m/z 179) from both the deprotonated monosaccharides and those isolated by fragmentation of various disaccharides were taken. A C-O stretching band characteristic of aldehydes was present in all spectra at ~1720 wavenumbers and gave spectroscopic evidence of the monosaccharide ring opening and therefore loss of anomericity.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sarah Stefan.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Eyler, John R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-11-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024320:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024320/00001

Material Information

Title: Differentiation of Carbohydrate Isomers by Tunable Infrared Multiple Photon Dissociation and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
Physical Description: 1 online resource (139 p.)
Language: english
Creator: Stefan, Sarah
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anomers, carbohydrates, disaccharides, fticr, irmpd, isomers, monosaccharides, ms
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Carbohydrates and their derivatives play a crucial role in many biological processes including fertilization, cell growth, inflammation and post-translational protein modification. The function of carbohydrates in these systems is closely related to their structure, including monosaccharide sequence, glycosidic linkage and stereochemistry. Unfortunately, the number of anomeric configurations and possible linkages between monosaccharide units makes analysis of carbohydrate structures complex. In order to shed light on these larger oligosaccharides, the fragmentation patterns and infrared multiple photon dissociation (IRMPD) spectra of various mono- and disaccharides were obtained and compared. For this work, various tunable infrared sources including a line-tunable continuous-wave carbon dioxide laser and a free electron laser (FEL) were used in conjunction with Fourier transform ion cyclotron resonance mass spectrometry (FTICR-MS). The first three projects used a line-tunable carbon dioxide laser to fragment various mono- and disaccharides in both the positive and negative ion modes. In the first project, anomers of lithium-cation attached O-methyl-gluco- and galactopyranosides were fragmented. The identity and anomeric configuration of each monosaccharide was accurately determined by comparing fragmentation patterns and ratios of certain fragments. A second project explored the fragmentation pattern of lithiated glucose-containing disaccharides having various linkages (1-2, 1-3, 1-4 and 1-6) and anomeric configurations (alpha and beta). Both the linkage and anomeric configuration of the various disaccharides were successfully identified based on their fragmentation patterns at several wavelengths. Next, irradiation of deprotonated and chlorinated glucose-containing disaccharides produced fragmentation patterns in which cleavage of the glycosidic bond resulted in major abundances of m/z 161 and 179 fragment ions. Along with differentiating the anomeric configuration for the chlorinated disaccharides, comparison of the abundances for major fragment ions also resulted in the positive identification of the linkages for both sets of disaccharides. Lastly, several deprotonated (negatively charged) mono- and disaccharides were fragmented with a FEL. The IRMPD spectra of the monosaccharide anions (m/z 179) from both the deprotonated monosaccharides and those isolated by fragmentation of various disaccharides were taken. A C-O stretching band characteristic of aldehydes was present in all spectra at ~1720 wavenumbers and gave spectroscopic evidence of the monosaccharide ring opening and therefore loss of anomericity.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sarah Stefan.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Eyler, John R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-11-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024320:00001


This item has the following downloads:


Full Text

PAGE 1

1 DIFFERENTIATION OF CARBOHYDRATE ISOMERS BY TUNABLE INFRARED MULTIPLE PHOTON DISSOCIATION AND FOURIER TRAN SFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY By SARAH ELIZABETH STEFAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Sarah Elizabeth Stefan

PAGE 3

3 To my Mom and Dad

PAGE 4

4 ACKNOWLEDGMENTS Many people have supported and helped m e thr oughout my graduate career. First, I would like to thank my parents. Their support and unco nditional love have made my studies possible. I want to thank them for their understanding and encouragement when tough times occurred; I appreciate all their help and love more than they will ever know. I also thank my family for their support and for always making life interesting. I am grateful for my friends, old and new, w ho gave a helping hand and an ear for listening when I needed them. All the laughs and conversati ons over these past four years have lifted my spirits and helped me to keep going. I want to acknowledge my lab mates, past and present, for their help, knowledge and conversations ha ve been instrumental in my work. I would also like to thank all my professors at Wheaton College, specifically Drs. Elita Pastra-Landis and Laura Muller, whose support and investment in me opened my eyes and mind to the potential of graduate schoo l. Their enthusiasm and support have made all the difference. I have the deepest gratitude to all the people with whom I collaborated; they have made my project possible. First, I wish to thank my advisor, Dr. John Eyle r, for his guidance, patience and support during my graduate career. I want to tha nk Dr. Brad Bendiak for all the samples, advice and support that he has provided throughout this project. His guidance and suggestions were well needed and helped tremendously. I would also like to thank Dr. David Powell for use of his instrument for the negative disaccharide wor k. Next, I want to thank my other committee members, Drs. Nicolo Omenetto, Nicolas Polf er and Carrie Haskell-Luevano, whose questions and conversations have helped me along the wa y. Finally, I would like to thank Drs. Jos Oomens and Jeffrey Steill for their help and effo rt with the work perfor med at the Free Electron Laser for Infrared eXperiments (FELIX) facility. Without all of these pe ople, this dissertation would not be possible.

PAGE 5

5 Finally I need to thank the one person who has had to listen to me late at night and early in the morning, whose patience and loving shoulder made it easier to continue when I wanted to give up, Mr. Brad House. His immense comput er knowledge and lack of chemistry knowledge helped me survive the past four years.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................10 ABSTRACT...................................................................................................................................13 CHAP TER 1 INTRODUCTION..................................................................................................................15 Carbohydrates.........................................................................................................................15 Monosaccharides............................................................................................................. 15 Disaccharides.................................................................................................................. .18 Oligoand Polysaccharides............................................................................................. 19 Differentiation of Monoand Disaccharides..........................................................................21 Separation of Oligosaccharides....................................................................................... 22 Analysis Methods............................................................................................................24 Mass Spectrometry: Ionization Techniques.................................................................... 26 Fragmentation Methods................................................................................................... 27 Charged Ions....................................................................................................................29 Objective of This Research.....................................................................................................31 Overview....................................................................................................................... ..........32 2 FOURIER TRANSFORM ION CYCLOTRON RES ONANCE MASS SPECTROMETRY.................................................................................................................39 History....................................................................................................................................39 Apparatus................................................................................................................................40 Magnet.............................................................................................................................40 Vacuum System............................................................................................................... 40 Analyzer Cell...................................................................................................................41 Data System.................................................................................................................... .41 Theory.....................................................................................................................................42 Cyclotron Motion............................................................................................................ 42 Trapping Motion..............................................................................................................43 Magnetron Motion........................................................................................................... 44 Basic FTICR-MS Operation and Data Acquisition................................................................ 46 Mass Resolution......................................................................................................................49 Tandem Mass Spectrometry...................................................................................................51 Dissociation Techniques.........................................................................................................51 Conclusions.............................................................................................................................53

PAGE 7

7 3 INFRARED MULTIPLE P HOTON DISSOCIATION .........................................................58 Introduction................................................................................................................... ..........58 Mechanism of Infrared Multiple Photon Dissociation........................................................... 59 Lasers Used for IRMPD.........................................................................................................60 4 DIFFERENTIATION OF MONOSACCHARIDES IN THE POSITIVE ION MODE BY IRMPD W ITH A TUNABLE CO2 LASER..................................................................... 66 Introduction................................................................................................................... ..........66 Procedure................................................................................................................................67 Reproducibility................................................................................................................ .......68 Results and Discussion......................................................................................................... ..69 Methyl-glucopyranosides................................................................................................69 Unknown Study of Methyl-glucopyranosides.................................................................71 Methyl-galactopyranosides..............................................................................................71 Unknown Study of both Methyl-glucoand galactopyranosides....................................72 Conclusions.............................................................................................................................73 5 DIFFERENTIATION OF DISACCHARIDES IN THE POSITIVE ION MODE WITH A TUNABLE CO2 LASER....................................................................................................83 Introduction................................................................................................................... ..........83 Procedure................................................................................................................................84 Fragmentation Study.......................................................................................................84 Anomeric Configuration Study....................................................................................... 85 Results and Discussion......................................................................................................... ..85 Differentiation of Disaccharides...................................................................................... 85 Determination of the Anomeric Configurations.............................................................. 86 Differentiation of Unknowns...........................................................................................87 Conclusions.............................................................................................................................88 6 IRMPD STUDIES OF NEGATIVELY CHAR GED DISACCHARIDES WITH A TUNABLE CO2 LASER........................................................................................................93 Introduction................................................................................................................... ..........93 Procedure................................................................................................................................94 Deprotonated Disaccharides............................................................................................94 Chlorinated Disaccharides............................................................................................... 95 Reproducibility: Deprotonated Disaccharides................................................................. 96 Reproducibility: Chlorinated Disaccharides.................................................................... 96 Results and Discussion......................................................................................................... ..97 Deprotonated Disaccharides............................................................................................97 Chlorinated Disaccharides............................................................................................... 98 Identification of Fragment Ions..................................................................................... 102 Conclusions...........................................................................................................................102

PAGE 8

8 7 DIFFERENTIATION OF DISACCHARIDES IN THE NEGATIVE ION MODE W ITH FREE ELECTRON LASER INFRARED MULTIPLE PHOTON DISSOCIATION..................................................................................................................115 Introduction................................................................................................................... ........115 Procedure..............................................................................................................................115 Results and Discussion......................................................................................................... 116 Disaccharides................................................................................................................. 116 Monosaccharide Anion Produced from Disaccharides................................................. 118 Conclusions...........................................................................................................................119 8 CONCLUSIONS AND FUTURE WORK ........................................................................... 127 LIST OF REFERENCES.............................................................................................................131 BIOGRAPHICAL SKETCH.......................................................................................................139

PAGE 9

9 LIST OF TABLES Table page 5-1 Table of ratios used to determine th e laser power used for fragm entation........................ 91 6-1 Major fragment ions observed for the chlorinated disaccharides when the precursor ion ( m/z 377 ) was almost depleted by infrared mulitple photon dissociation (IRMPD) at 9.588 m......................................................................................................................108 6-2 Comparison of the fragments produced by collision induced di ssociation (CID) and IRMPD for the chlorinated disaccharides. ....................................................................... 108

PAGE 10

10 LIST OF FIGURES Figure page 1-1 Fischer projection for Dand L-glucose............................................................................ 34 1-2 Example of the numbering system fo r the carbons of m onosaccharides........................... 34 1-3 Fischer projections for the D-hexoses of the aldose family............................................... 34 1-4 Anomers of D-glucose. ...................................................................................................... 35 1-5 Inter-conversion of the ring structures for the 6-m embered ring, pyranose, and the 5-membered ring, furanose, of D-glucose......................................................................... 35 1-6 Examples of disaccharides composed of two glucose (G lc) monosaccharides. ................ 36 1-7 Structures of two common oligosaccharide derivatives. ................................................... 36 1-8 Typical steps for analysis of glycans................................................................................. 37 1-9 Fragmentation nomenclature for oligosaccharides............................................................ 38 1-10 Possible fragmentation pathways for fragm entation by infrared multiple photon dissociation (IRMPD)........................................................................................................ 38 2-1 Ion cyclotron motion....................................................................................................... ...54 2-2 Schematic diagram of the components of a Bruker 4.7 T FTICR (Fourier transform ion cyclotron) mass spectrometer...................................................................................... 54 2-3 Figures of merit for FTICR-MS as a function of m agnetic field strength......................... 55 2-4 Two of the typical analyzer cells used for in FTICR m ass spectrometers......................... 55 2-5 General schematic of a t ypical experim ental sequence..................................................... 56 2-6 Various domains and spectra obtained from an FTICR-MS experiment.......................... 56 2-7 Effect of number of data points acqui red and Fourier transform on m ass resolution........57 3-1 Energy potential well...................................................................................................... ...64 3-2 Depiction of the IRMPD mechanism in polyatomic molecules........................................ 64 3-3 Schematic of an undulator used for free elctrom lasers (FELs)......................................... 65 3-4 Layout schematic of Free Electrom La ser for Infrared eXperim ents (FELIX)................. 65

PAGE 11

11 4-1 Structures of the O-methylated monos accharides discussed in this ch apter...................... 74 4-2 Experimental set up of th e 4.7 T FTICR m ass spectrometer............................................. 74 4-3 Wavelength-dependent fragmentation patterns for the lithiated Omethyl-glucopyranosides for wavelength from 9.2 to 10.8 m..................................... 75 4-4 Infrared mulitple photon dissociation de pletion sp ectra of the precursor ions ( m/z 201) for both and -O-methyl-glucopyranoside lithium cation complexes....... 76 4-5 Comparison of the fragmentation of -m ethyl-glucopyranoside at wavelengths 9.588 and 10.611 m...................................................................................................................77 4-6 Relative percent abundance of fr agment ions for both lithiated and -O-m ethyl-glucopyranosides over the wavelength range from 9.201 to 9.675 m.........78 4-7 Spectra of unknowns in si ngle blind study of m ethyl-glu copyranosides at wavelength 9.588 m............................................................................................................................79 4-8 Fragmentation patterns over the wavelengths from 9.2 to 10.6 m ..................................80 4-9 Ratio of m/z 169 to m/z 151 for and -O-m ethyl-galactopyranoside............................ 81 4-10 Decision flowchart used to identif y the different monosaccharide anom ers..................... 81 4-11 Spectra of unknowns identified as galactopyranosides in single blind study obtained at wavelength 9.588 m .....................................................................................................82 5-1 Wavelength-dependent fragmentation for the various linked lithiated disaccharides ....... 89 5-2 Flow-chart depicting how linkage of the disacch arides was determined.......................... 90 5-3 Flow-chart showing ratios of peak heights and values used to determ ine anomeric configurations....................................................................................................................91 5-4 Bar graphs comparing ratios from kn owns and unknown lithiated gluco se-containing disaccharides at the wavelengths 9.342, 9.472 and 9.588 m...........................................92 6-1 Schematic drawing of the laser/mass spectrom eter set-up used for the analysis of deprotonated disaccharides.............................................................................................. 104 6-2 Relative percent abundan ce of the precursor ion ( m/z 341) of isom altose at selected wavelengths......................................................................................................................104 6-3 Wavelength-dependent fragmentation patte rns for the va rious deprotonated disaccharides.................................................................................................................. ..105 6-4 Ratio of m/z 161/179 for 1-3 and 1-6 linked disacchari des, showing that this ratio is not optim al for distinguishing the different anomers....................................................... 106

PAGE 12

12 6-5 Comparison of the fragmentation patte rns of deprotonated isom altose on two separate days....................................................................................................................107 6-6 Fragmentation spectra for the nearly depleted precursor ion ( m/z 377) for the chlorinated disacch arides at 9.588 m............................................................................. 109 6-7 Infrared multiple photon dissociation spectra for chlorin ated isomaltose obtained at three wavelengths on two different days......................................................................... 110 6-8 Average fragmentation spectra fo r the disaccharides at 9.342, 9.473 and 9.588 m ......111 6-9 Decision flow chart used to identify disacch aride samples with unknown identities in a single-blind study..........................................................................................................112 6-10 Comparison of various ratios used to determ ine the anomeric configurations of the chlorinated disaccharides................................................................................................. 113 6-11 Identification of some of the fragment ions for the various linked disaccharides. .......... 114 7-1 Schematic of the FTICR set-up at FELIX....................................................................... 121 7-2 Infrared multiple photon dissociation fr agm entation patterns over the wavelength range of 5.5 to 11 m for the deprotonated 18O-labeled disaccharides........................... 122 7-3 Fragmentation pattern of ch lorinated unlabeled sophorose. ............................................ 123 7-4 Comparison of the IRMPD spectra for O18-labeled sophorose and O16-chlorinated sophorose.........................................................................................................................123 7-5 Comparison of the IRMPD spectra of the monosaccharide anions (m/z 179) produced by deprotonation of glucose a nd by fragmentation of a disaccharide by sustained off-resonance irradiation collision-induced dissociation (SORI-CID) and CO2 laser irradiation........................................................................................................ 124 7-6 Schematic of the possible mechanism leading to the opening of the monosaccharide anion ring. ........................................................................................................................124 7-7 Infrared multiple photon dissociati on spectra of various dep rotonated monosaccharides..............................................................................................................125 7-8 Comparison of the IRMPD spectra for a nom ers of O-methyl-glucopyranoside to the spectrum of deprotonated glucose................................................................................... 125 7-9 Comparison of the fragmentation patterns of the deprotonated m onosaccharides over the wavelength range of 5.5 to 11 m.............................................................................. 126

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DIFFERENTIATION OF CARBOHYDRATE ISOMERS BY TUNABLE INFRARED MULTIPLE PHOTON DISSOCIATION AND FOURIER TRAN SFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY By Sarah Elizabeth Stefan May 2009 Chair: John R. Eyler Major: Chemistry Carbohydrates and their derivatives play a crucial role in many biological processes including fertilization, cell grow th, inflammation and post-transl ational protein modification. The function of carbohydrates in th ese systems is closely related to their structure, including monosaccharide sequence, glycosidic linkage and st ereochemistry. Unfortunately, the number of anomeric configurations and possible linkages between monosaccharide units makes analysis of carbohydrate structures complex. In order to sh ed light on these larger oligosaccharides, the fragmentation patterns and infrared multiple phot on dissociation (IRMPD) spectra of various monoand disaccharides were obtained and compar ed. For this work, various tunable infrared sources including a line-tunable continuous-wave carbon dioxide laser and a free electron laser (FEL) were used in conjunction with Four ier transform ion cycl otron resonance mass spectrometry (FTICR-MS). The first three projects used a line-tunable ca rbon dioxide laser to fragment various monoand disaccharides in both the positive and negative ion modes. In the first project, anomers of lithium-cation attached O-methyl-glucoand gal actopyranosides were fragmented. The identity and anomeric configuration of each monosacchar ide was accurately determined by comparing

PAGE 14

14 fragmentation patterns and ratios of certain fragments. A second project explored the fragmentation pattern of lithiated glucose-containing disaccharides having various linkages (1-2, 1-3, 1-4 and 1-6) and anomeric c onfigurations (alpha and beta). Both the linkage and anomeric configuration of the various disaccharides were successfully identified based on their fragmentation patterns at several wavelengths. Ne xt, irradiation of deprotonated and chlorinated glucose-containing disaccharides produced fragmentation patterns in which cleavage of the glycosidic bond resulted in major abundances of m/z 161 and 179 fragment ions. Along with differentiating the anomeric conf iguration for the chlorinated disaccharides, comparison of the abundances for major fragment ions also resulted in the positive identification of the linkages for both sets of disaccharides. Lastly, several deprotonated (negatively charged) monoand disaccharides were fragmented with a FEL. The IRMPD spectra of the monosaccharide anions (m/z 179) from both the deprotonated monosaccharides and those isolat ed by fragmentation of various disaccharides were taken. A C-O stretching ba nd characteristic of aldehydes was present in all spectra at ~1720 wavenumbers and gave spectroscopic ev idence of the monosaccharide ring opening and therefore loss of anomericity.

PAGE 15

15 CHAPTER 1 INTRODUCTION Carbohydrates and their derivativ es are biologically im portant. They participate in cellcell interactions and also act as target struct ures for microorganisms, toxins and antibodies. 1-3 Carbohydrates also interact with pr oteins and play a critical role in fertilization, cell growth, inflammation and post-transla tional protein modifications.1,3-5 The simplest unit within these larger carbohydrates is that of the monosacch aride. When two monosaccharides are joined together, the result is a disaccharide. The disaccharide is the smallest saccharide unit which contains the glycosidic bond. Depending on the anomeric configur ations of the monosaccharides that react, a disaccharide can either be or -linked. The role of carbohydrates depends not only on the subunits of sugars which compose them, but also how these units are linked together.6 Therefore, characterization of the both th e anomeric configuration and the linkage of the different types of monoand disaccharides is important. Carbohydrates Carbohydrates can be categorized based on thei r degree of polym erization. The smallest group is that of monosacch arides and their derivatives, all of which are not polymerized. The next category includes oligosaccharides, that have 2 to 10 degrees of polymerization. The last category is that of polysaccharides, that have gr eater than 10 degrees of polymerization. This chapter will discuss all the possible types of carbohydrates as well as give an overview of the methods used for carbohydrate analysis. Monosaccharides Monosaccharides are the sm allest units that compose larger oligosaccharides. There are several types of monosacchar ides and they all have th e general formula of (CH2O)n. Typically the more biologically common isomer of monos accharides in nature is the D-isomer, but

PAGE 16

16 L-isomers are also found. The monosaccharide isomer can be determined by drawing the Fischer projection. In the Fisc her projection when the hydroxyl group on the highest numbered stereocenter is on the right, it is the D-isomer and when the hydroxyl group is on the left it is the L-isomer, Figure 1-1.7 Since D-isomers of sugars are found mu ch more frequently in nature, this dissertation will deal only with D-isomers. The carbons in monosaccharides are numbered sequentially starting with the end of the chain nearest to the carbonyl gr oup, as seen in Figure 1-2. Carbon number 1, also known as the anomeric carbon, is where two monosaccharides can be joined together, through a glycosidic linkage or bond, to form larger oligosaccharides. The smallest possible monosaccharide has a backbone composed of only 3 carbon atoms, but 4, 5 and 6 carbons are other possible backbon es. The names of these monosaccharides are trioses, tetroses, pentoses, hexoses, and heptoses respectively. Monosacc harides that contain a keto group are called ketose whereas monosaccharides containing an aldehyde are called aldoses. Typically the names of the family and number of carbons are combined into one systematic name. For example, a monosaccharide containi ng both a 4 carbon backbone and an aldehyde group would be named an aldotetrose (aldo fo r the aldehyde group and tetrose for the 4 carbon backbone). For the aldose family, each of the eight D-aldohexoses differs in stereochemistry at carbon 2, 3 or 4 and has its own unique, common name such as D-glucose, D-galactose, etc., as shown in Figure 1-3. When two monosaccharides only differ at one carbon position, they are epimers. Since they only differ in the position of the hydroxyl group on carbon number 4, D-glucose and D-galactose are an example of epimers from the aldose family. Monosaccharides can be found in either the open chain or ring form, but typically the ring form is more common. In solution, monosaccharid es with a 5 or 6 carbon backbone can undergo

PAGE 17

17 nucleophilic attack of the carbonyl carbon by one of the hydroxyl gr oups along the chain, resulting in a ring. Six-membered rings are ca lled pyranoses and 5-membered rings are called furanoses.8,9 At least four carbons and one oxygen are ne eded to form a furanose. Therefore, aldotetroses and higher aldoses a nd 2-pentuloses and higher ketose s can be found in the furanose ring. While monosaccharide rings can be either 5or 6-membered, pyranosides are the most common form. When cyclic monosaccharides only diffe r by the position of the hydroxyl group on the anomeric carbon, they are anomers. If the hydroxyl group is axial relative to the plane of the ring then it is said to be in the -position and if it is equato rial then it is in the -position, Figure 1-4. The cyclic monosacchar ides can interconvert between and -anomers through a process known as mutarotation, Figure 1-5. Du ring mutarotation, the ring opens into the chain form. Once in the chain form, a nucleophilic attack results in the formation of the -anomer. Therefore, in solution there is an equilibrium mixture of all possible isomers including the furanose, pyranose, -, and open chain forms of the monosaccharides. This equilibrium mixture is different for each monosaccharide, bu t for D-glucose it is approximately one-third -anomer, two-thirds -anomer and less than 1% of both the open and five-membered ring forms.7 On the other hand, D-mannose has approximately 69% -anomer and 31% -anomer in solution, thus showing that the equilibrium doesnt always contain more of the -anomer than the -anomer. The two cyclic forms of D-glucose are known as hemi-acetals, which are formed by the reaction of the hydroxyl group on carbon number 5 and the aldehyde group. Typically any monosaccharide that contains a hemiacetal group is a reducing sugar and can react further. A reducing sugar is one that reacts with Tollens (Ag(NH3)2OH) or Benedicts reagents (solution of

PAGE 18

18 copper (II) sulfate, sodium carbonate, sodium citrate dihydrate and 2,5-difluorotoluene) to reduce eithe r Ag+2 or Cu+2. If a sugar contains an acetal group th en it cannot react w ith the Tollens or Benedicts reagents and it is called a non-reducing sugar. While hexoses are the most abundant sugars, there are a number of monosaccharide sugar derivatives that are naturally abundant and important. Some of these derivatives are N-acetylneuraminic acid (sialic acid), -D-acetylgalactosamine and -D-acetylglucosamine. These derivatives are found primarily in animals as the major components of glycoproteins and glycolipids. Disaccharides Disaccharid es, the next largest saccharid e are formed when a hydroxyl group of one monosaccharide reacts with the a nomeric carbon of the other, Fi gure 1-6. The resulting bond is known as an O-glycosidic linkage. When two cyc lic hexoses come together, a glycosidic linkage can occur at one of the five hydroxyl positions. This leads to numerous possible isomers with various linkages. Disaccharides are composed of a non-reducing monosaccharide that is fixed in the ring conformation and a reducing-monosacch aride that can inte rconvert between the and the -configuration. Therefore, in solu tion, there will be a mixture of the and -configurations of the reducing sugar of the disaccharide. While most sugars have a common, non-systematic name, there is a systematic nomenclature scheme for disaccharides. In it, the name of the first monosaccharide unit, its anomeric configuration and then the linkage followed by the second monosaccharide unit is given. For example, two glucose (Glc) units that are connected at the 1 and 6 carbon will be named glucose 1-6 glucose (Glc 1-6Glc), for which the common name is isomaltose. For larger oligosaccharides the nomenclature pro cess is the same, but for each monosaccharide attachment the linkage and anomeric configur ation followed by the monosaccharide is given.

PAGE 19

19 For example a trisaccharide that has a glucose -linked to carbon number 2 of a mannose (Man) monosaccharide which is -linked to carbon number 4 of anot her glucose unit would be named Glc 1-2Man 1-4Glc. When the anomeric carbons of both monosaccharide units are linked, the anomeric configuration of each saccharide is given. For example, sucrose is a disaccharide when the anomeric carbon of both the glucose and fruc tose (Fru) monosaccharide units are linked. For this, the systematic name would be Glc 1-2 Fru. Since both anomeric carbons are linked in sucrose, it is a non-reducing sugar, unlike kojibiose (Glc 1-2Glc) and sophorose (Glc 1-2Glc) that are examples of reducing sugars. Oligoand Polysaccharides Oligosacch arides are the next largest sacch aride chains that consist of 3 to 10 monosaccharide units linked together. Sugars th at contain more than ten monosaccharide units are called polysaccharides. Oligoand polysaccharides can be either homoor heter-oligosaccharides. Homo-oligosaccharides contain the same monosaccharide unit that repeats, whereas heter-oligosaccharides contain different monosaccharide units linked together. One homo-polysaccharide is starch, which can be found in foods such as potatoes. Starches characteristically have 1-4 linkage between two glucose units.10 Other polysaccharides that do not have this linkage, also known as non-starch polysaccharides, can be found in foods such as bran, bananas and hazelnuts. Other common polysaccharides are cellulose and glycogen. Cellulose is a polysacchar ide that contains severa l hundreds to thousands of 1-4 linked glucose units. It is the main component of the primar y cell walls of plants and can be found in some algae. Glycogen is a glucose-polysaccharide th at has a lot of branch ing and most commonly functions as short-term en ergy storage in animals. Common oligosaccharide derivatives are those of N-acetyl hexosamines, primarily N-acetylglucosamine (GlcNAc) and N-acetyl galactosamine (GalNAc), Figure 1-7.11 The

PAGE 20

20 GlcNAc reducing end is linked to serine or threonine residues whereas the GalNAc reducing end is linked to asparagines N-acetylglucosamine is a component of chitin and GalNAc is the terminal carbohydrate that form s the antigen of blood group A. N-acetylgalactosamine is also the first monosaccharide unit that connects to se rine and threonine in glycosylation and is necessary for intercellular communication. Polysaccharides and oligosaccharides are also known as glycans. Glycosylation is a post-translational modification where oligoa nd polysaccharides are linked to proteins and lipids, forming glycoconjugates. Glycosylati on is one of the most common post-translational modifications for proteins and it is approximate d that more than 50% of all proteins are glycosylated.12 Linkages between a glycan and a prot ein form glycoproteins and those with lipids form glycolipids. The type of glycoprotein is determined by th e linkage between the carbohydrates and the protein. Glycoproteins can be Oor N-linked. While N-lin ked are linked by a chitobiose ( dimer of 1-4-linked glucosamine units) unit to an amide nitroge n of an asparagine residue, O-linked are linked to the oxygen of a side chain of an am ino acid.13,14 Typically the linkage is through a serine or threonine residue. N-glycosidic bonds are found in all nucleotides (the resulting sugar and nucleotide structures are called nucleosi des, such as ribonucleic acid (RNA) and deoxyribonucleic acid (DNA)). Un like other oligosaccharides that are linked by oxygen bridges, RNA and DNA are polyesters that are linked by phosphate bridges. DNA is the largest known polymer with more than 1012 units found in human genes and the number of units found decreases as one goes down the evolutionary chain.8 Another example of a polysaccharide with N-linkages is chitin. Chitin is a natura lly occurring polysaccharide, composed of 1-4-linked N-acetyl-D-glucosamines, which is found in places like fungi and exoskeletons of arthropods

PAGE 21

21 such as crustaceans. The three classes of glyc oproteins are: N-glycos yl protein, O-glycosyl protein and N,O-glycosyl protein. Since multip le types of linkages (Oor N-linked) and anomeric configurations are possi ble, it is no surprise that many different isomers are possible. When attached to proteins (glycoproteins), oligosaccharides have b een found to aid in a plethora of functions in the human body including cellular recognition, signaling, receptor binding and immune responses.15-17 They also serve to influen ce folding, biological lifetime and recognition of binding pa rtners for proteins.17 Carbohydrates are also involved in the glycosyl phosphatidyl-inositol (GPI) anchor, by which proteins are attached to the plasma membrane and the oligosaccharides are linked to lipid s which are attached to cell membranes.18 In this process, a glycolipid can be connected to the C-terminus of a protein during post-tr anslation modification. Since the biological role of oligosaccharides depends on the linkage, branching, configuration and saccharide units, being able to distinguish and differentiate the smaller monoand disaccharides that compose larger olig osaccharides is very important. Due to the various linkages (carbons 1-6 of each monosacch aride unit), anomeric configurations ( or -) and monosaccharide units (any of the eight D-hexos es) there is a plethora of possible isomers, which makes analysis of carbohydrates a very difficult task. Differentiation of Monoand Disaccharides Glycans m ust be isolated and prepared for analysis. The preparation method can include releasing the glycans, separating them and then finally analyzing them. Once separated common methods for analysis have included nucl ear magnetic resonan ce (NMR) and/or mass spectrometry (MS). Figure 1-8 shows a schematic of the different methods used for separating and analyzing saccharides.

PAGE 22

22 Separation of Oligosaccharides Typically oligosacch arides can be released by several methods, either chemical or enzymatic. Enzymatic methods use a specific enzy me to pick out a particular substrate from mixtures.14 Enzymes, for example glycosidase and ga lactosidase, are used to remove specific sugar residues sequentially from the non-reducing end. A chemical method for releasing glycans is an alkali-borohydride treatment which then can be followed by hydrolysis, with the resulting species then separated by high performance liquid chromatography (HPLC) and/or gas chromatography (GC).19,20 Once released, the oligosaccharides can then be separated. Methods for determining and separating mixtures of carbohydrates include thin layer chromatogr aphy (TLC), column chromatography methods (including gas chroma tography, liquid chromatography, gas-liquid chromatography and high performance liquid chromatography) and capillary electrophoresis (CE).13,14,21 Thin layer chromatography is a relatively cheap and inexpensive method for separating analytes. Microcrystalline cellulose and silica gel are two typical solid supports. Cellulose separation occurs by a liquid-liquid partition where the s ugar of interest is distributed between the mobile phase and the cellulose-bound complex in water. The separation occurs based on the solubility of sugar in the elue nt and how easily it can enter the solid support. Cellulose TLC has the same chromatographic characteristics as pape r TLC, but allows for shorter elution time and increased sensitivity. Silica gel separation is similar to cellulose, but requires an additional adsorption component, typically an inorganic salt (phosphate, bisulfate). Numerous solvents are used to separate the various monosaccharides. High performance liquid chromatography, gas-chromatography (GC) and gas-liquid chromatography (GLC) can also be used to separate components of mixtures. All of

PAGE 23

23 these methods require somewhat expensiv e equipment. High performance liquid chromatography is typically preferred for monos accharide mixtures, oligosaccharide analysis and purification. Whereas GLC is limited to monosa ccharide mixtures only, HPLC requires different columns and various solvents are used to elute the mixture through the co lumn. Typical columns include sulfonated polymeric or amino-bonded s ilica columns. Typical solvents include acetonitrile/water mobile phase. Gas liquid chromatography is a sensitive technique and allows the analysis of sub-nanomol ar amounts of carbohydrates.14 Capillary electrophoresis is a newer technique th at yields results in relatively short times and with high efficiency. To achieve electrophore tic separation, the two ends of the capillary are submerged into two separate electrolyte reservoi rs that contain a high vo ltage electrode. The separation is due to the variation of molecular size and electric charge ratios of the sugars within the mixture. This method does not require deri vatization of the oligos accharides and cannot be used to identify and separate oligosaccharides that have the same degree of polymerization, i.e. isomers. Derivatization of oligosaccharides allows for th em to be more volatile and therefore more compatible with analysis methods such as mass spectrometry. One common derivative method is hydrolysis followed by chromatographic separation.22,23 Besides hydrolysis, other common methods used to derivative oli gosaccharides are permethylation 24 and peracetylation.25 Permethylation has been shown to easily determine branching and interglycosidic linkages. It also helps stabilize sialic aci d residues in acidic oligosaccharides and in conjunction with matrix-assisted laser desorption ionization (MALDI) has been show n to give more predictable ionization than non-permethylated oligosaccharides.26 Two common methods for permethylation are the use of dimethyl sulfoxide anion (DMSO-) to remove protons from the

PAGE 24

24 analyte and replace th em with methyl groups27 and the addition of methyl iodide to DMSOwhich contains powdered sodium hydroxide. This second method effectively replaces protons with a methyl group at both oxygen and nitrogen sites in oligosaccharides.24 Analysis Methods Once releas ed and separated, the oligosaccharides can then be analyzed individually. One past method for differentiation of isolated and separated carbohydrat es is NMR spectroscopy.28-30 Over the past 25 years advances in NMR have allowed it to become suitable for structural analysis of carbohydrates.31 Such advances include improve ments in instrumentation, pulse sequences, ability to interpret spectra, isot opic labeling of compounds and improvement in molecular modeling. With the advances of tech nology, the ability and accessibility of these techniques have become faster, better and more accessible. The improved coupling of molecular modeling with NMR has provided the ability to determine primary structure and three-dimensional structures of different biological molecules.31 While NMR has been used to study carbohydrat e structures, including glycosidic linkages of saccharide units, and has developed considerably in recent years, it still has several drawbacks and areas in need of improvement. First, the sa mple size required for NMR analysis is relatively large. Another major drawback is that data analysis can be complicated and time consuming. Typical 1H NMR spectra can be used to give par tial spatial arrangeme nt, but due to the incomplete separation of the proton resonance si gnals they cannot provide a lot of structural information. Other types of NMR have been us ed in the past to analyze carbohydrates and include 13C, 15N, 17O, 19F and 31P. The resolution and sensitiv ity of each method varies and therefore different information can be asce rtained by using each method. For example, 13C-NMR can give the information of the anomer ic configuration of th e carbohydrate residues. It can also provide sequence information of th e composite monosaccharides, their sequence and

PAGE 25

25 the overall conformation of the carbohydrates. Another NMR method that improves the results, but increases the complexity, of data analysis uses 2Dhomonuclear correlation types of spectra (2D-COSY) to assign resonances and give furthe r structural informati on. Although these spectra give more information, they do not provide monosaccharide sequence information because there is an absence of coupling over the glycosidic lin kage. For this, nuclear overhauser enhancement spectroscopy (NOESY) or rota ting-frame overhauser enhancement spectroscopy (ROESY) may be used. While there is some success with thes e methods, the linkage is not always identified.31 Since carbohydrates are inhe rently flexible, in solu tion carbohydrates may undergo alternations. Estimation of the solution structur e required knowledge of th e configuration of the composing monosaccharides. Flexible motions of the whole molecule on a short time scale involve fast vibrations at bonds and angles and on a longer tim e scale involve changes in the dihedral angles. Therefore chan ging the relaxation time can help de duce the internal flexibilities of carbohydrates in solution. As one can see, the data required for this type of analysis are extensive and analysis can be extremely time-consuming. Mass spectrometry is another ve ry popular analytical technique that is used for gas-phase analysis of carbohydrates. Several types of mass spectrometers have been used for analysis, including Fourier transform ion cyclotron res onance mass spectrometry (FTICR-MS), which will be discussed further in chapter 2. Mass spectrometry has been shown to have 3 to 4 orders of magnitude higher sensitivity than NMR.32 Mass spectrometry is highly sensitivity and can be used in multi-step approaches to determine struct ural information. In order for analysis with mass spectrometry to be done, one of several ioni zation methods can be used to introduce the analyte of interest into the mass spectrometer.

PAGE 26

26 Mass Spectrometry: Ionization Techniques Both hard and soft ionization m ethods exist. Hard ionization methods are ones that result in fragmentation and degradation of the sample during the ionization process, whereas soft methods produce little or no fragmentation during the ionization process. One previous hard ionization method widely applied is electron ionization (EI). In EI a beam of electrons is used to excite and ionize a volatile analyte. A main drawback of EI is fragmentation of the sample before detection.33 Soft ionization methods are currently preferred since they result in the ionization with molecules of the sample remaining intact. Electrospray ionization (ESI) is the most popular of the soft ionization methods. Seve ral soft ionization tech niques have been used in the past for carbohydrate analysis and include fast atom bombardment (FAB),34-36 MALDI37 and ESI.38,39 In FAB, the analyte is mixed with a liquid matrix and is bombarded under vacuum with a high energy beam of atoms. Fast atom bo mbardment results in the release of [M+H]+ or [M-H]ions which can then be analyzed.40 Analysis with FAB had seve ral constraints including poor ionization of neutral and basic o ligosaccharides and restriction of analysis to relatively smaller molecules. While basic oligosaccharides were ionized poorly with FAB, acidic oligosaccharides produced stronger signals in the negative ion mode.35 When FAB was coupled with FTICR-MS, extensive fragmentation, including cross-ring cleavages was seen.41,42 While FAB uses a liquid matrix, MALDI uses a crystalline matrix where the analyte of interest is co-crystallized with the solid matrix molecules. A laser is focused onto the matrix and its photon energy is absorbed by the matrix and the analyte of in terest is released as charged ions.43 While analysis with traditional MALDI is po ssible, analysis of smaller saccharide units is a challenge because most peaks of the t ypical matrix are present in the range m/z <500, where peaks due to smaller saccharides such as mono-,d iand trisaccharides are also found. Recently,

PAGE 27

27 use of an acid fullerene matrix instead of the tr aditional matrix has allowed for disaccharides to be successfully st udied with MALDI.44 This approach needs to be developed more fully and applied to other types of carbohydrates. Electrospray ionization is the least energetic of these three gentle i onization techniques. A primary benefit of ESI ionization is the absence of matrix peaks and therefore ease of analysis for smaller mono-, diand trisaccharides.33,45,46 In ESI, a solution of the analyte and solvent is passed through a capillary with a high voltage (2 to 5 kV) applied to it.47 This process allows for charged droplets to be formed. Once formed these charged droplets can then be transferred (through differential pumping and ion optics) in to a mass spectrometer and analyzed by mass spectrometry. Electrospray ionizat ion is versatile when it comes to carbohydrates since it can be used to ionize both basic and acidic oligosaccharides. Since multiply charged ions are formed, and mass spectrometers typically separate base d on mass-to-charge ratio, ESI has virtually no limit to the size of the ion that can be analyzed. This dissertation will concentrate on ESI since it was used exclusively in the research to be reported. Fragmentation Methods Since isom ers have the same mass and th erefore cannot be differentiated by mass spectrometry alone, differences in ion fragmentati on can be used to distinguish isomers. To obtain structural information, se veral fragmentation me thods have been used. These methods include electron capture dissociat ion (ECD), collision induced di ssociation (CID) and infrared multiple photon dissociation (IRMPD). Electron capture dissociation uses low energy electrons to induce fragmentation of the saccharide.6 It results in multiply, positively charged i ons that can then be analyzed with a mass spectrometer. Past research has included using ECD to do top-down analysis where a whole protein is sequenced simultaneousl y. Also, O-glycosylation sites on proteins have been explored

PAGE 28

28 using ECD.48 While ECD can be used for proteins and peptides, it has limited application to oligosaccharides. Another dissociation technique that is more a pplicable to oligosaccharides is CID. In on-resonance or traditional CID, a neutral backgr ound gas is pulsed into the cell, analyte ions are accelerated to higher kinetic energies, and collid e with the introduced gas. These collisions result in fragmentation of the analyte of interest.49 Another commonly used CID method in FTICR-MS is sustained off-resonance irradia tion collision induced dissociation (SORI-CID).50 In SORI-CID, ions are excited by an off-resonance frequency, cau sing their kinetic energies to increase and decrease repeatedly with time, resu lting in less-energetic collisions with background molecules over a longer time period than with conventional CID. These collisions can nonetheless result in fragmentati on of an isolated ion of intere st. Collision induced dissociation of oligosaccharides results in fragments that can be used to determine stereochemistries, linkage position and branching information.34,51,52 A disadvantage to SOR I-CID with respect to identification of oligosaccharide is that since SORI-CID is low energy, cross-ring fragmentations are less likely than the fragmentation of the glycosidic linkage. Also, the ability to control the energies of collisions is limited with CID. Due to the collisions and variance of energy, CID can give different fragmentation th an other dissociation methods. One fragmentation method that gives similar and complementary fragments to CID, but allows for finer control of the ener gy imparted to the system is IRMPD.53 IRMPD relies on absorption of photons by one vibrational normal mode of trapped ions and the subsequent redistribution of photon energy into other vibrational modes of th e ions. This redistribution occurs via intramolecula r vibrational relaxation.54,55 If sufficient photons are absorbed without excessive collisional or radiative relaxation, then the internal energy of the ion increases to a

PAGE 29

29 level above the dissociation threshold, resulti ng in fragmentation. One advantage of IRMPD over CID is that the power is only limited by the laser being used. Therefore, use of a tunable laser gives finer control over the power imparted into the system. The theory and history of IRMPD will be discussed in more detail in chapter 3. A systematic nomenclature method has b een developed for naming fragments of carbohydrate ions. In this method, the fragment s which contain a non-reducing end sugar are labeled with uppercase letters sequen tially starting with A, Figure 1-9.17 Those fragments that contain the reducing sugar are labeled sequentially with letters from the end of the alphabet (X, Y, Z). Ions formed by cleavage across a ring are A and X ions. The subscripts for these fragments are given by assigning each ring bond a number and then counting clockwise. Charged Ions Since m ass spectrometry only detects charged particles, metal ions have become a common way to ionize neutrals and then detect the complexes formed with mass spectrometry. Adduction of an alkali metal ion has been used with FAB, MALDI and ES I in both the positive and negative ion mode.56-61 For fragmentation of a metal-attached ions two pathways predominate. The first type of fragmentation is loss of the metal ion and the sec ond type is fragmentation of the molecule into smaller charged parts which often retain the metal ion. The fragmentation pathway that occurs depends on the strength of the bonds of the adduc tion of the metal to the molecule, Figure 1-10. If the binding energy of the metal ion is less than the dissociation thre shold, then loss of the metal will occur. This type of fragmentation is seen when large alkali metals are adducted to molecules. This is because the binding energy of th e larger alkali metals ions is lower than that of smaller alkali metal ions.58 The opposite has been seen with the smaller alkali metal ions. Since their binding energies are larger and thus me tal ion dissociation is less likely, the result is

PAGE 30

30 greater fragmentation of the molecules with the sm aller alkali metal ions remaining attached to the fragments. Cancilla et al. found that the relative binding energy for alkali metal ions is Li+>Na+>K+>Rb+>Cs+.58 The stronger the binding energy, th e more fragmentation that will be seen with IRMPD since it is more likely the mo lecule will fragment before losing the metal.6 Xie et al. have compared the ability of CID and IR MPD to fragment alkali-adducted molecules and showed that for smaller ions such as Li+ and Na+ both dissociation method yielded similar fragments.62 Specifically, adduction of lithium to saccharides has been studied by Hofmeister et al.60 In this research they determined that the lithium cat ion interacts with disacch arides through several oxygen sites, including the glycosidic bond. This tr iple interaction leads to stronger binding and therefore greater fragmentation is seen with IRMP D. The research performed in this dissertation primarily used adduction of lithium ions and analysis in the positive ion mode. In the negative ion mode, Cole & Zhu have s hown that chlorinated species can be studied conveniently.61 Formation of the chlorine adduct ha s proven successful for species that are polar, neutral molecules or slightly acidic mole cules that do not genera te negative ions through deprotonation. Therefore, chlorina tion has been shown to be one easy method for exploring ions in the negative mode when addition of a strong base does not promote deprotonation. While the addition of an appropriate salt can help facilitate th e ESI process through producing charged adducts, excessively high salt concentrations can cause background interferences; therefore caution needs to be taken when usi ng salts for the creation of ions. These interferences can lead to si gnal suppression and the subsequent inability to detect the ions of interest. The ease of the adduc tion of metals to create ions with oligosaccharides makes their

PAGE 31

31 use with IRMPD a promising method to differentia te the sugars in both positive and negative ion modes. Objective of This Research Since carbohydrates are biologi cally im portant, being able to differentiate both their linkages and anomeric configurations can give valuable information. For this research, FTICR-MS was used in conjunction with IRMPD to distinguish various monoand disaccharide ions in both the positive and ne gative ion mode. Fourier transf orm ion cyclotron resonance mass spectrometry not only gives superior mass resolution and mass accuracy when compared to other types of mass spectrometry, but it also allows for tandem mass spectrometric experiments to be done in the same region of space (within the analyzer cell), thereby eliminating extra instrumentation that is often need ed with other mass spectrometers.63,64 Since IRMPD uses lasers to introduce photons, va rious lasers have been used in the past including fixed frequency and wavelength-tunable CO2 lasers65-69 and free electron lasers (FELs).55,70-72 Fixed frequency CO2 lasers produce photons at one wavelength (10.6 m), thus the information that can be obtained with them is limited. Free electron lasers, on the other hand, have a large output wave length range (5 to 250 m) but these lasers are very expensive and access to beam time is limited. Therefore, a less expensive alternative w ith at least a (limited) range of wavelengths (9.2 to 10.6 m) is the tunable CO2 laser that will be emphasized in this research. The objective of this research was to produce a method for discriminating between various linked and anomeric configurations of monoand disaccharides. While previous research done by Polfer et al. with irradiation produced by a FEL ha d shown that the linkages and anomeric configurations could be disti nguished by wavelength-dependent ion fragmentation patterns, a

PAGE 32

32 method to do so in more conventional ( i.e. non-FEL equipped) labora tories had not been demonstrated.73,74 In this research the anomeric configuration of monoand disaccharid es was determined by examining the fragmentation patterns produced by IRMPD with a tunable CO2 laser in both the positive and negative ion modes using FTICR-MS While past methods have studied the lithiated disaccharides in the positive ion mode with FEL irradiation, the negative mode of monoand disaccharides has not been expl ored. Therefore the fragmentation of glucose-containing disaccharides, some of thei r specific fragment ions and some selected monosaccharides was also examined in the negative ion mode at the Free Electron Laser for Infrared eXperiments (FELIX) facility. Overview The next ch apter will give a description of FTICR-MS. This description will include a history as well as theoretical and practical as pects of FTICR-MS. Chapter 3 will discuss the mechanism and theory of IRMPD. The types of la sers used for IRMPD will also be described in this chapter. Chapter 4 is a detailed descri ption of the procedure and apparatus used to differentiate lithiated monosaccharides with a tunable CO2 laser at the University of Florida in Dr. John Eylers laboratory. The results of this study will also be discussed. Chapter 5 will discuss a method to determine both the linkage and anomeric configuration of lithiated glucosecontaining disaccharides in the positive ion mode with a CO2 laser. Chapter 6 will next describe IRMPD fragmentation of deprotonated and chlorinated disaccharides in th e negative ion mode by wavelength-tunable CO2 laser. A description of the pr ocedure and apparatus used for the fragmentation of deprotonated disaccharides done at the Universi ty of Florida in Dr. David Powells laboratory will also be given. Chapter 7 will give a detailed account of negative monoand disaccharides ions and some of their fragment ions explored at the FELIX facility. Finally, a

PAGE 33

33 conclusion including a summary of the strengths and weaknesse s of this work along with proposed future work will be presented.

PAGE 34

34 CHO OH H H HO OH H OH H CH2OH D-glucose CHO H HO OH H H HO H HO CH2OH L-glucose Figure 1-1. Fischer projection for Dand L-glucose. O H HO H HO H OH OH H H OH Alpha-D-glucose 1 2 3 4 5 6 Figure 1-2. Example of the numbering system for the carbons of monosacch arides. The carbons are numbered sequentially beginning with the anomeric (chiral) carbon. CHO OH H OH H OH H OH H CH2OH CHO H HO OH H OH H OH H CH2OH CHO OH H H HO OH H OH H CH2OH CHO H HO H HO OH H OH H CH2OH CHO OH H OH H H HO OH H CH2OH CHO H HO OH H H HO OH H CH2OH CHO OH H H HO H HO OH H CH2OH CHO H HO H HO H HO OH H CH2OH D-Allose D-Altrose D-Glucose D-MannoseD-GuloseD-Idose D-Galactose D-Talose Figure 1-3. Fischer projections for the D-hexoses of the aldose family. Isomers that vary in only one position are called epimers.

PAGE 35

35 O H HO H HO H OH OH H H OH O H HO H HO H H OH H O H OH A B Figure 1-4. Anomers of D-gluc ose. A) Structure of the -anomer of glucose, where the hydroxyl group on the anomeric carbon is in th e axial position. B) Structure of the -anomer of glucose, where the hydroxyl group on the anomeric carbon is in the equatorial position. O H HO H HO H OH OH H H OH O C H H HO H HO H OH H OH H O O H HO H HO H H OH H O H OH OH H H H OH HO H O H HO HO H OH H H OH HO H O H HO HO O H H OH HO H C H O H HO HO H Figure 1-5. Inter-conversion of the ring structures for the 6-membered ring, pyranose, and the 5-membered ring, furanose, of D-glucose. Once the cyclic ring of the -glucose opens, a nucleophilic attack results in the closing of the ring in the -position.

PAGE 36

36 O H HO H HO H O OH H H OH O H H HO H OH H OH O H HO H HO H H OH H OH O H O H HO H OH H OH 1 4H,OH H,OH 1 4A BNon-reducing end Reducing end Non-reducing end Reducing end Figure 1-6. Examples of disaccharides compos ed of two glucose (Glc) monosaccharides. A) Structure of maltose (Glc 1-4Glc) with an -link between carbon 1 of the nonreducing sugar and carbon 4 of the reduci ng sugar. B) Structure of cellobiose (Glc 1-4Glc) with a -link between carbon 1 of the non-reducing sugar and carbon 4 of the reducing sugar. N H OH OH OH O H O O N-acetyl glucosamineN H OH OH OH OH O O n-acetyl galactosamineA B Figure 1-7. Structures of two common oligosaccharide deri vatives. A) Structure of N-acetylglucosamine. B) Structure of N-acet yl galactosamine. These derivatives are found linked to proteins and are biologically important.

PAGE 37

37 Glycoconjugate Chemical or enzymatic methodReleased glycans Separated glycans Purification methods: -Gel filtration -Chromatography -Capillary electrophoresis Derivatization: Methylationwith CH3I and a strong base Derivatization: Hydrolysis with a strong acid Derivatization Hydrolysis with enzymes Direct analysis: NMR and MS Methylatedsaccharides Monosaccharides Smaller glycans -Sequence -Glycosidic bond conformation -Enzymes -Methylation -Sequence -Position -Glycosidic bond conformation -Type -Amount -Glycosidic bond conformation -CE -TLC -HPLCGlycoconjugate Chemical or enzymatic methodReleased glycans Separated glycans Purification methods: -Gel filtration -Chromatography -Capillary electrophoresis Derivatization: Methylationwith CH3I and a strong base Derivatization: Hydrolysis with a strong acid Derivatization Hydrolysis with enzymes Direct analysis: NMR and MS Methylatedsaccharides Monosaccharides Smaller glycans -Sequence -Glycosidic bond conformation -Enzymes -Methylation -Sequence -Position -Glycosidic bond conformation -Type -Amount -Glycosidic bond conformation -CE -TLC -HPLC Figure 1-8. Typical steps for analysis of glycans. Figure adap ted from Valle, J. J. Ph.D., University of Florida, Gainesville, 2005.74

PAGE 38

38 O OH O OH OH CH2OH O O OH OH CH2OH O O OH OH CH2OH R Y2B1 Z2C1Non-reducing end Reducing-end Y1B2 Z1C2 Y0B3 Z0C3 0,2A1 1,5X1 Figure 1-9. Fragmentation nomenclature for ol igosaccharides. Figure adapted from Zaia, J. Mass Spectrom. Rev. 2004, 23, 161-227.17 E2 E E1A+ M+A1+ [A2+M]+ E= E1-E2M=Li, Na where E1>E2A. E1 E E2A+ M+A1+ [A2+M]+ E= E2-E1B.M= K, Rb,Cswhere E2>E1 E2 E E1A+ M+A1+ [A2+M]+ E= E1-E2M=Li, Na where E1>E2A. E1 E E2A+ M+A1+ [A2+M]+ E= E2-E1B.M= K, Rb,Cswhere E2>E1 Figure 1-10. Possible fragmentation pathways fo r fragmentation by IRMPD. A) Fragmentation pathway for smaller alkali ions. B) Fragme ntation pathway for larger alkali ions. Figure adapted from Park, Y.; Lebrilla, C. B. Mass Spectrom. Rev. 2005, 24, 232264.6

PAGE 39

39 CHAPTER 2 FOURIER TRANSFORM ION CYCLOTRO N RES ONANCE MASS SPECTROMETRY Fourier transform ion cyclot ron resonance mass spectrometr y (FTICR-MS) is a powerful analytical technique with a plethora of resear ch applications. This chapter will discuss the history of the technique, the apparatus used for it and exam ples of research done with FTICR-MS. History Todays current research with FTICR-MS firs t b ecame possible with the invention of E.O. Lawrences cyclotron in the 1930s.75 Lawrences cyclotron accelerator was used to bombard target compounds with ions of various masses. In 1932, Lawrence et al demonstrated that an ion moving perpendicular to an uniform magnetic fi eld is restricted to circular, cyclotron motion with an angular frequency given by the following equation:76 m qBc. (2-1) In Equation 2-1, c is the cyclotron frequency, q is the ions charge, B (in Tesla) is the magnetic field strength and m is the mass of the ion. This motion is independent of the particles orbital radius. The direction of this motion depends on the charge of the ion, with positive ions rotating in one direction and negative ions in the opposite direction, Figure 2-1. This theory was incorporated into Sommer, Thomas and Hipples Omegatron in the 1950s,77 which was later developed into other instru ments that were used to study ion-molecule reactions.78 Then in the 1970s, Comisarow and Mars hall introduced Fourier transform methods into ion cyclotron resonance (ICR) mass spectrom etry to build the first Fourier transform mass spectrometer (FTMS).79,80 The number of Fourier transform mass spectrometers and

PAGE 40

40 applications using them has been increasing sinc e the initial demonstrati on of the technique in the 1970s. Apparatus While several types of FTICR-MS instrum ents are available, all have the same general components.78 These include the magnet, vacuum system analyzer cell and a data system. A schematic diagram of the components of a 4.7 T FT ICR-MS system (minus the data system) is shown in Figure 2-2. Magnet The first component is the m agnet, most commonly either an electromagnet or a superconducting magnet. Magnetic field strengths of electromagnets are below 3.0 T, normally around 1.5 T. Superconducting magnets are generally available in field strengths of 3.0 to 9.4 T, but higher field strengths such as 20 T have been used in FTMS instruments.78 As magnetic field strength increases, both the mass resolving power and highest non-coalesced mass increase (Figure 2-3). Therefore, as magnetic strength increases the ability to study higher masses with more resolving power is possible. Also, with stronger magnetic fi elds, longer ion trapping times are possible. Since the capabilities of the mass spectrometer increase with magnetic field strength, typical mass spectrometers are desi gned using the strongest magnet available or affordable. Vacuum System To avoid collisions of the analyte with ot her m olecules in the cell, low pressures are needed for optimal ion excitation and detection. For best results, background pressures in the 10-9 to 10-10 Torr range are typically used. In order to achieve these low pressures, a pumping system is needed. Generally such a system will use mechanical pumps for rough pumping and turbo-molecular pumps to achieve the low pre ssures needed for FTICR-MS. To allow the

PAGE 41

41 coupling of ambient ionization t echniques, there is normally a re gion where higher pressures are pumped down by differential or rough pumps near the source. Normally a gate valve separates the source region from the high vacuum region. Optics are used to guide ions from a high pressure region (10-5-10-6 Torr) to a lower pressure region, as seen in Figure 2-2. Fourier transform ion cyclotron resonance mass spectromet ers often use valves to permit pulses of gas (or air) to be leaked into th e cell, allowing fragmentation methods like collision induced dissociation (CID) to be performed. Analyzer Cell An analyzer cell is the next pa rt of the instrum entation. Ions are stored, mass analyzed and detected in the cell. The analyzer cell is wher e ions can also be isolated and fragmented in tandem mass spectrometry (MSn). Since the cell is the heart of the mass spectrometer, having the most efficient design is desired. While a number of designs have been proposed over the years,64 two typically used cells are of cubic and cylindrical geomet ry (Figure 2-4). These cells are both composed of si x electrode plates, which perform on e of three functions when voltages are applied to them. The first type of plate, the trapping plate, holds ions in the cell in the direction parallel to the magnetic fi eld. The second type of plate, the excitation plate, excites the trapped ions to larger radii. Th e last type of plate, the detecti on plate, detects the excited ions. In cubic cells, the trapping plates sometime have small openings that allow externally produced ions to enter the cell, where they can then be ex cited and detected. Cylindr ical cells are typically preferred since they are larger, conform more closely to the geometry of superconducting magnets (with cylindrical bores) and therefore can hold more ions Data System The next component is the data system. Th e data system takes the signal induced by excited ions on the detection plates and tr ansforms it into useable information. The

PAGE 42

42 instrumentation in this component includes a frequency synthesizer that generates the frequencies used for exciting the ions, a dela y pulse generator, broa dband radio frequency amplifier, fast transient di gitizer and computer system.78 A computer coordinates all of the electrical devices needed for the experimental pro cess. User-friendly interf aces allow for ease in use of the system by operators. With these interfaces, tandem mass spectrometry experiments can be performed simply by changing and/or addi ng events in the experimental sequence. It goes without saying that as technology improves the capabi lity and ease of FTICR-MS instruments will also improve. Theory Cyclotron Motion The force of the m agnetic field on an ion causes it to move in a circular orbit. As an ion with a charge (q) moves in a magnetic field (B) and electric field (E), the Lorentz force causes the ion to move in a circular orbit in a plane perpe ndicular to direction of the magnetic field (Equation 2-2). It should be noted that the Lo rentz force is dependent on the mass and velocity of the ion. BvqEq dt vd monaccelerati mass Force (2-2) The Lorentz force must be equal to the centrif ugal force for the ion to experience circular motion. The velocity of the ion in the x-y plane, a plane that is perpendicular to the magnetic field (B), is denoted by vxy and the angular acceleration (dvxy/dt) is expressed as v2 xy/r. In the absence of an electric field, Equa tion 2-2 then becomes the following: r v mBqvxy xy 2. (2-3) Rearranging Equation 2-3 and solving for r gives the radius of the cyclotron motion as:

PAGE 43

43 qB mv rxy c. (2-4) Substituting in the angular velocity (in rad/s) as = vxy/r, Equation 2-4 becomes: m 2r = qB r. (2-5) Rearranging Equation 2-5 gives the cy clotron frequency, Equation 2-1: m qBc. (2-1) Since =2 /t=2 f, the linear cyclotron frequenc y can therefore be given by: m qB fc c 22 (2-6) For example, at 4.7 T a singly charged ion of m/q 349 will have a cyclotron frequency of 209.8 kHz: kHz 8.209 10673.13492 7.410602.1127 19 kgu u TC. Equation 2-6 shows that the cyclotron freque ncy is dictated only by the magnetic field strength, the charge of the ion and the mass of the ion. This m eans that the cyclotron frequency is independent of the ions velo city and therefore independent of the ions kinetic energy. Since the frequency is independent of the velocity and kinetic energy, all ions with the same m/q ratio will have the same cyclotron frequency. Usi ng Equation 2-1 and Equations 2-3 to 2-6, the typical frequencies calculated are from the kilohe rtz (kHz) to the megahertz (MHz) range. These frequencies are detectable by most commerc ially available instrument electronics.64,78 Trapping Motion The presence of a uniform m agnetic field in the z-direction allows for unrestricted motion along the z-axis and confines the motion of ion in the x-y plane. To prevent ions from escaping along the z-axis, a trapping voltage, Vtrap, can be applied to the end-cap electrodes of the cell.

PAGE 44

44 This trapping voltage leads to a three-dimens ional quadrupolar potential, in the cell, in the form:64,81,82 22 22 2 rz a Vtrap (r,z) (2-8) In Equation 2-8, Vtrap is the trapping voltage, r is the radial position of the ion in the x y plane and equals 22yx a is a measure of the trap size and and are trap shape dependent constants. Equation 2-8 can used to solve in terms of the z-motion of the ion, giving: ),,(2 2zyxq dt zd mFaxial (2-9) Solving Equation 2-9, gives: )2cos()0()( tvztzz (2-10) An ion at a particular z-position will oscillate with a given frequency that be found by Equation 2-11: 2 3 10 x 2.21088 ma zV vtrap z (2-11) In Equation 2-10, vz is in Hz, Vtrap is in volts, a is in meters, m is in atomic mass units and z is in multiples of elementary charge. Magnetron Motion Combination of the electric and magnetic fields creates a three-dimensional trapping potential that allows ions to be stored and analyzed for extens ive intervals of time (seconds). Although the cyclotron and trapping motions are not coupled, their motions combine to induce a third type of motion: magnetron motion. The tr apping potential of Equa tion 2-8 also produces a radial force with the equation:

PAGE 45

45 r a qV qEFtrap r radial 2 )( (2-12) This radial force acts upon the ions in an outwa rd direction that opposes the inward Lorentz force of the magnetic field. An equation related to th e motion of an ion that is subjected to a static magnetic field and three-dimensional axial quadr upolar potential is give n when Equation 2-1 and 2-11 are combined to give Equation 2-13: r a qV rqBrmFtrap 2 2 (2-13) Solving this quadratic eq uation for zero gives: 02 2 ma qV m qBtrap (2-14) The absence of the radius, r in Equation 2-14, indicates that is independent of the radius. Therefore, each ion motion frequency is independent of the ion position within the ion trap. Solving Equation 2-14 for yields two natural rotational fre quencies. The first frequency, +, is given in Equation 2-15. This is the perturbed cyclotron freq uency that is observed in the presence of a direct current (d.c.) tr apping potential. The second frequency, -, is shown in Equation 2-16. This is the magne tron frequency which is a circular motion that is superimposed onto the cyclotron motion. 2222 2 z c c (2-15) 2222 2 z c c (2-16) The cyclotron frequency is far greater than both the magnetron and trapping frequencies. Therefore, only the cyclotron fre quency is used for ion detection.64,78

PAGE 46

46 Basic FTICR-MS Operation and Data Acquisition Due to the design of an FTICR mass spectrome ter, the various experi mental events occur in the same region of space, namely the analyzer cell. A typical event sequence can be seen in Figure 2-5. The basic events of a typical experime nt are: ionization, dela ys, excitation, detection and quenching ions from the cell. The first step of the experimental sequence is quenching. Quenching empties the analyzer cell of any ions that may have been present from previous experiments. These ions are typically ejected along the z-axis of the cell by changing or rem oving voltages on the trapping plate. Usually a quench pulse of about 1 millisecond gives ample time to empty the cell of all unwanted ions. The next step in the experimental sequence is ionization in whic h gas-phase ions are produced. Ions can either be formed internally in or externally to the cell. Externally made ions have to be transferred into th e cell for analysis through the use of ion optics. Once inside the cell, the ions are constrained to motion in the x-y plane by the magnetic field and are trapped along the z-axis by a voltage (typically 0.5 to 5 V) that is applied to the trapping electrodes. Both positive and negative ions can be trappe d and analyzed within the ICR cell by simply changing the polarity of the voltages applied to th e trapping electrodes. Also, ions of a large m/z range can be trapped in the cell, all of which os cillate at their own particular frequencies as determined by Equation 2-6. After ionization, a series of delays usually follow in the experimental sequence. Such delays allow time for ion injection, ejection of un wanted ions and reaction of trapped ions with neutral species or irradiation by laser sources Thus, during these delays the ions can be subjected to tandem mass spectrometry techniques such as introducing col lision gases (for CID),

PAGE 47

47 laser pulses (for IRMPD) or el ectrons (for electron capture di ssociation) into the analyzer cell.64,83,84 Excitation of the ions into larger, detectable radii is the next event in the experimental sequence. In order to detect a wide m/z range of ions, a swept frequency approach can be used in excitation. For this, a wide range of freque ncies is applied sequentially to the excitation electrodes. These frequencies create a short, high intensity, broadband radio frequency signal also know as a chirp. When the frequency appl ied matches the cyclotron resonance frequency of an ion, the ion absorbs energy and this results in the acceleration of the ion into a larger orbit. Ions of the same mass to charge, once excited, move together in ion p ackets. The excitation event is brief since if the ions are excited too much, their radii will become too large, causing them to impact the analyzer cell walls and thus resulting in their loss. Use of stored waveform inve rse Fourier transform (SWIFT)85 or chirp excitation allows for the undesired ions to be ejected from the cell while permitting the desired ions to remain in the cell for detection. While a SWIFT uses a calcul ated and then synthesized waveform and a chirp uses a high voltage swept r.f. signa l, they both excite unwanted ions into an orbital radius that is larger than the cell radius, causing ionwall collision of these undesired ions.64,78 This effectively eliminates the unwanted ions. Since a chirp is a high voltage, short durati on event, all the ions are both excited and detected almost simultaneously. For exampl e, a frequency synthesizer can sweep over frequencies from 100 kHz to 10 MHz in roughly 1 millisecond. This sweep excites all the ions with cyclotron frequencies in that range. The resulting time domain spectrum (Figure 2-6 A) is very complex. To produce the mass spectrum, the time domain signal is mathematically analyzed using a Fourier transform algorithm. This generates a frequency domain spectrum

PAGE 48

48 (Figure 2-6 B), with all the i ndividual ion frequencies being pr esent. A calibration formula derived from the cyclotron frequency equati on allows the frequency domain to be easily transformed into a mass spectrum (Figure 2-6 C). Since this approach excites and detects all frequencies simultaneously, the acquisition time needed is far less than that of classical ICR in which only one frequency at a time could be detected and analyzed The time it takes to perform an FT on the data is only hindered by the tec hnology available; as the speed of computers increases, so does the ability to do FT. Once excited, the ion packets create an alterna ting current (image current) in the detection plates, where the amplitude is related to the numbe r of charges in the cell. This image current gives FTICR-MS the unique ability to detect ions without destroying them. While FTICR-MS uses the image current, all ot her mass spectrometers, excluding orbitraps, detect ions in a destructive manner. Since the ions are not destroyed during detection in FTICR-MS, they remain in the cell and can be re-measured and reacted further without having to produce more ions. Also, since multiple frequencies are appl ied during the excitation step, FTICR-MS can be used to detect ions of many different masses si multaneously. This also allows FTICR-MS to have increased sensitiv ity and resolution. The entire sequence can then be repeated as many times as wanted or needed. The scans that are collected can be signal averaged. Signal averaging leads to spectr a with better signal to noise (S/N) ratios and improves the quality of th e collected spectra. Other events can be added into the sequence to allow for tandem methods such as CID or IRMPD to be performed on the ions within the cell. The actual experimental sequence and length can vary depending on the experiment.

PAGE 49

49 Mass Resolution One major advantage of FTICR-MS instrument s is their superior mass resolution when compared to other mass spectrometers. Mass resolution (m2-m1 m50% where m1 and m2 are the closest masses that can be resolved) is defined as the point where one va lley begins to appear between peaks of equal shape and height and is separated by m50%.64 Both high mass resolving power and high mass resolution can significantly im prove the quality of the experimental data obtained. As mass resolving power increases, the maximum number of components in a mixture that can be resolved also increases. Therefore, it may be possible to dist inguish and differentiate different chemical components in a mixture without prior separation. Another advantage is that high resolution can decrease peak width, there by giving a more accurate mass determination. Fourier transform ion cyclotron resonance mass spectrometry is capable of giving the highest mass resolving power and highest mass accuracy (for all ions up to m/z 5000) of all mass spectrometry methods. High mass resolution requires that a long time domain signal, also known as the transient response signal, be acquired. The mass resolution increases in direct propo rtion to the length of the transient recorded. The number of data points collected during the experiment is set by the user before the transient is collected. Figur e 2-7 shows that as the number of data points increases, the peak widths decrease and the resolution of the mass spectrum increases. However, the number of data points that can be processe d from a transient is limited. Thus far, approximately 106 data points can be processed using commercially available data analysis programs. The higher the number of data point s, the more computer memory is needed. Therefore as technology advances, larger numbers of data points can be taken. The number of data points required for a desired transient length can be calculated by Equation 2-16,

PAGE 50

50 S N Tacq (2-16) In Equation 2-16, Tacq is the transient duration, S is the sampling rate and N is the number of data points collected. The transient collection rate is based on the sampling frequency used. According to the Nyquist theorem, the sampling frequency must be at least twice the highest frequency (determined by the lowest m/z) being recorded. Based on the number of points collected, the maximum resolution that can be achieved is determined by: 2acqcTf R (2-17) In Equation 2-17, R is the resolving power, fc is the cyclotron frequency and Tacq is the duration of the transient. As seen in Figures 2-6 and 2-7, the transient signal decays over time. This occurs as the collisions between ions and neutra ls destroy the coherent ion packet within the analyzer cell. Therefore, to reduce the possibili ty of collisions within the cell, all FTICR-MS experiments are carried out in ultra-high vacuum. Another aspect that can affect the resoluti on is space charging. Space charging is due to repulsions between ions having simi lar charge. It is a consequence of Coulombs law and can be described by the following equation: )(2 r qq kF (2-18) In Equation 2-18, F is the force between the two ions, k is a proportionality constant, q and q' are the ion charges and r is the distance between the two ions. Space charging can affect mass measurement accuracy and sensitivity.78,86-88 The greater the force, the more space charging, thereby resulting in a decrease in resolution. Reducing the number of ions held within the cell or introducing an internal calibrant ca n reduce the effect of space charging.89-91

PAGE 51

51 Tandem Mass Spectrometry One distinct advantage of FTICR-MS is the ability to perform multiple (tandem) mass spectrometry (MS) experiments. For tandem MS, the precursor ion is excited and then dissociated. The resulting product ions are then detected and analyzed. Tandem MS allows more information than just the precursor mass spect rum of an ion to be obtained. For example, isomers with the same mass can be identified by ratios of the relative percent abundances of product ions.92-94 Since more steps are involved, tande m MS experiments are inherently more complex than regular mass spectrometric experiments. Current software allows tandem experiments to be performed by simply adding additional steps into the experimental sequence. Unlike tandem MS performed on magnetic sect ors or quadruple mass spectrometers, where additional mass analyzers are needed for each ad ditional step, FTICR-MS only needs additional steps added to the experimental sequence. The experimental sequence can also be altered to include isolation steps for the pr oduct ions. Once isolated, both the precursor and/or the product ions, can be dissociated by either collision induced dissociation (CID ), irradiation with a laser or by electron impact (EI).64,78,95,96 Dissociation Techniques Several dissociation techniques are employe d in tandem mass spectrometry. These methods include CID, surface induced dissociation (SID),97 electron capture dissociation (ECD)98,99, ultraviolet photodissociation (UVPD)100, blackbody infrared dissociation (BIRD)101 and infrared multiple-photon dissociation (IRMPD).95,102 One of the most popular dissociation techniques for biological molecules is dissociation by collision. This method involves the trapping and reaction of ions in the analyzer cell prior to dissociation. Applicatio n of an excitation pulse ejects all the ions of higher and lower masses than the previously selected and isolated precursor ion from the analyzer cell. Ejection of the

PAGE 52

52 unwanted ions can also be confi gured to involve exciting the precu rsor ion into a larger radius orbit and thus increasing th e kinetic energy of the ion.78,103 The relationship between the kinetic energy and the radius is s hown in Equation 2-19. m rBq E2222 (2-19) In Equation 2-19, E is the kinetic energy of the ion, q is the charge of the ion, B is the magnetic field strength, r is the radius of the ions orbit and m is the mass of the ion. This mass selected and kinetically energized ion undergoes collis ions with a background gas or a neutral gas (typically Ar) that is pulsed into the cell by a pulsed valve.49,104 As long as the pulsed gas does not increase the pressure in the cell too much the ions are retain ed and detected. One disadvantage of traditional CID is that the product ions are formed away from the center of the analyzer cell. The farther the ions are from the center of the cell, the more likely it is that there will be a decrease in detection effici ent and resolution. An alternative to traditional CID is sustained off-resonance irradiation (SOR I)-CID which does not have this disadvantage and is less energetic than traditional CID.50,105,106 Another tandem mass spectrometric method whic h was used for the research reported in this dissertation is IRMPD. Traditional IR MPD dissociation uses a fixed wavelength CO2 laser (10.6 m) to introduce photons and slowly heat the ions by increasing their vibrational energies, thus resulting in dissociation of the ions within the analyzer cell. In IRMPD, the photons are absorbed and their energy is redistributed intern ally until the dissociati on threshold is met or exceeded, resulting in the fragmentation of the precu rsor ion. Recently, tunable lasers, including free electron lasers, have been used to frag ment oligosaccharides and other biological samples.70-73,107

PAGE 53

53 Infrared multiple photon dissociation results in similar and/or complementary fragments to those produced by CID. The benef it of IRMPD over CID is that a gas pulse is not required for fragmentation. With no gas pulsing, there is no need for extra experimental time to reduce the pressure in the cell before detection. The ab ility to manipulate ions and the convenience with which photons can be delivered into the cel l makes coupling IRMPD with FTICR-MS an advantageous method. Conclusions Fourier transform ion cyclotr on resonance mass spectrometry has become a very valuable tool for bioanalytical studies in cluding proteomics and glycobiol ogy. The increased ability and popularity of FTICR-MS is mainly due to the in creased efficiency of and developments in hardware and software technology. It offers higher mass resolution and mass accuracy that any other type of mass spectrometry, thereby allowing superior mass assignment. Along with these benefits, the ability to do tandem mass spectrome try in time rather than space makes FTICR-MS superior over many other mass spectrometric met hods. With improvements in data acquisition and analysis technology, the power and ease of use of FTICR-MS also increases, making it an even more valuable mass spectrometric tool for future research.

PAGE 54

54 B B + B B + + Figure 2-1. Ion cyclotron motion. The ions move perpendicular to the magnetic field and the cyclotron motion is opposite for opposite charges. Figure adapted from Marshall, A. G.; Hendrickson, C. L. Int. J. Mass Spectrom. 2002, 215, 59-75.84 ZnSe Window Infinity Cell FOCL2 PL9 FOCL1Gate ValveHVO YDFL XDFL PL4 PL2 PL1 Extract/Trap Plate Hexapole Skimmer Modified Heated Metal Capillary Electrospray Tip Ion Optic Lenses Turbopump1 Turbopump2 Atmosphere Turbopump3 4.7T Superconducting Magnet ZnSe Window Infinity Cell FOCL2 PL9 FOCL1Gate ValveHVO YDFL XDFL PL4 PL2 PL1 Extract/Trap Plate Hexapole Skimmer Modified Heated Metal Capillary Electrospray Tip Ion Optic Lenses Turbopump1 Turbopump2 Atmosphere Turbopump3 4.7T Superconducting Magnet Figure 2-2. Schematic diagram of the component s of a Bruker 4.7 T FTICR mass spectrometer. Shown are the different vacuum pumping re gions, the ion optics, the cell and an electrospray source. Figure courtesy of Dr. Michelle Sweeney.

PAGE 55

55 7 T 9.4T 14.5T 21 T 0 21-Mass resolving power -Highest non-coalesced mass difference 21 0 7 T 9.4T 14.5T 21 TMagnet strength, B (tesla) Magnet strength, B (tesla) -Upper mass limit -Number of ions capable of being trapped -Ion trapping period 7 T 9.4T 14.5T 21 T 0 21-Mass resolving power -Highest non-coalesced mass difference 21 0 7 T 9.4T 14.5T 21 TMagnet strength, B (tesla) Magnet strength, B (tesla) -Upper mass limit -Number of ions capable of being trapped -Ion trapping period Figure 2-3. Figures of merit for FTICR-MS as a function of magnetic field strength. Adapted from Marshall, A. G.; Hendrickson, C. L.; Emmett, M. R.; Rodgers, R. P.; Blakney, G. T.; Nilsson, C. L. Eur. J. Mass Spectrom. 2007, 13, 57-59.108 Detection Trapping Excitation B Y X Z Trapping Detection Excitation A B Detection Trapping Excitation B Y X Z Trapping Detection Excitation A B Figure 2-4. Two of the typi cal analyzer cells used for in FTICR mass sp ectrometers. A) Schematic of a cubic cell. B) Schematic of a cylindrical cell. Both types of cells have three sets of plates that trap, excite or detect the ions within the cell. Figure adapted from Marshall, A. G.; Hend rickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. 1998, 17, 1-35.64

PAGE 56

56 QuenchIonization Excitation DetectionTime delays: Injection Ejection Reaction Repeat Time QuenchIonization Excitation DetectionTime delays: Injection Ejection Reaction Repeat Time Figure 2-5. General schematic of a typical experimental sequence. This sequence can be repeated as many times as needed. Frequency (kHz) 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 Time 150 100 50 0 Abundance 0.009 0.008 0.006 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 -0.008 -0.009 -0.01 A m/z 1,00 0 900 800 700 600 500 400 300 200 Abundance 400 350 300 250 200 150 100 50 B C Frequency (kHz) 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 Time 150 100 50 0 Abundance 0.009 0.008 0.006 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 -0.008 -0.009 -0.01 A Time 150 100 50 0 Abundance 0.009 0.008 0.006 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 -0.008 -0.009 -0.01 Time 150 100 50 0 Abundance 0.009 0.008 0.006 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 -0.008 -0.009 -0.01 A m/z 1,00 0 900 800 700 600 500 400 300 200 Abundance 400 350 300 250 200 150 100 50 B C Figure 2-6. Various domains a nd spectra obtained from an FTICR-MS experiment. A) The time domain transient response signal. B ) The frequency domain. C) The m/z domain.

PAGE 57

57 m/z 350 349 348 Abundance 650 600 550 500 450 400 350 300 250 200 150 100 50 Frequency (kHz) 350 349 348 Abundance 450 400 350 300 250 200 150 100 50 128 K Frequency (kHz) 350 349 348 Abundance 250 200 150 100 50 256 K m/z m/z Time 90 80 70 60 50 40 30 20 10 0 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 Time 45 40 35 30 25 20 15 10 5 0 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 Time 150 100 50 0 0.009 0.008 0.006 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 -0.008 -0.009 -0.01 m/z512 K m/z 350 349 348 Abundance 650 600 550 500 450 400 350 300 250 200 150 100 50 Frequency (kHz) 350 349 348 Abundance 450 400 350 300 250 200 150 100 50 128 K Frequency (kHz) 350 349 348 Abundance 250 200 150 100 50 256 K m/z m/z Time 90 80 70 60 50 40 30 20 10 0 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 Time 45 40 35 30 25 20 15 10 5 0 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 Time 150 100 50 0 0.009 0.008 0.006 0.005 0.004 0.003 0.001 0 -0.001 -0.003 -0.004 -0.005 -0.006 -0.008 -0.009 -0.01 m/z512 K Figure 2-7. Effect of number of data points acquired and Fourier transform on mass resolution. As the number of points increases, the peak width decreases and the resolution increases.

PAGE 58

58 CHAPTER 3 INFRARED MULTIPLE PHOTON DISSOCIATION Introduction Several methods have been used to fragment ions in mass spectrometers. One of them is the absorption of light, which can result in differ ent reactions and fragmentation of ions. Two types of light that have been used in the past for ion irradiation involve wavelengths in the ultraviolet/visible (UV/vis) and in the infrared (IR) region of the el ectromagnetic spectrum. Absorption of UV/vis light by the io ns usually involves promotion to excited electronic states of the ions, which either by direct dissociation or internal conversi on to high vibrational levels of the ground state gives rise to inte rnal energies above the dissociat ion threshold of the molecule, thereby resulting in fragmentation. Another source of light is infrared (IR) radiation, which can result in fragmentation by induci ng step-wise vibrational excitation of the molecule. Infrared photons are less energetic (typical ly 0.001 to 1.7 eV/photon) than UV/vis photons (typically 2 to approximately 8 eV/photon). Therefore, the number of IR photons needed for dissociation of a molecule is greater than that ne eded with UV/vis radiation. The invention of high-power IR lasers enabled IR light to be utilized for ion and neutral dissociation. In the 1970s, IR lasers were used to explor e the dissociation of trapped ions.53,109-111 These experiments and the others th at shortly followed demonstrated the phenomenon of infrared multiple photon dissociation (IRMPD) and have been reviewed in detail by Eyler and Polfer.95,96 Such experiments include dissociat ion of large biomolecules that were ionized by electrospray (ESI)45,112 and matrix assisted de sorption ionization (MALDI).43,113 This chapter will discuss the mechanism by which IRMPD occurs and the various lasers that can be used as IRMPD radiation sources.

PAGE 59

59 Mechanism of Infrared Multiple Photon Dissociation A polyatomic molecule or ion that is irradiated with infrared radiation can absorb the light energy as photons. The energy of the photon is converted to vibrati onal energy within the molecule or ion. Since energy is quantized, th e energy absorption leads to a transition between energy levels if the energy of the photon matches the difference between the energy level of the molecule or ion. Figure 3-1 depicts a typical potential-energy well of a diatomic molecule. Initially the mechanism of IRMPD was thought to proceed via a ladder-climbing mechanism in which rotational levels compensated for the unequ al spacing of the vibrational levels as one ascends the energy well. In this mechanism, the addition of each photon sequentially between energy levels in the potential-energy well caused the total vibrational energy of the ion to increase as pictured in Figure 3-1. This is no t the case and therefore, th is step-ladder mechanism cannot be the mechanism for IRMPD. The actual mechanism of IRMPD involves th e slow, sequential ab sorption of multiple infrared photons. Once absorbed, the energy of each photon is internally redistributed until the dissociation threshold is either met or exceeded, resulting in fragmentation of the molecule of interest. A schematic view of this process can be seen in Figure 32. The first photon is absorbed by the fundamental of th e vibrational mode in question (vi =0 vi =1). Redistribution of the photon energy that is absorbed by a normal mode occurs by rapid intramolecular vibrational relaxation (IVR), which distributes the photons energy to an assembly of other vibrational modes within the molecule.54,114-116 If the IVR is rapid enoug h (ps to ns time scale), the absorbing vibrationa l mode can be de-excited with enough time to allow the absorption of a sequential photon. This process can be repe ated many times until enough internal energy has been gained by the molecule or ion so that the dissociation threshold is either met or exceeded and the molecule or ion fragments. As the number of photons absorbed increases, so does the

PAGE 60

60 internal energy of the molecule. After a few phot ons have been absorbed, the internal energy of the ion is such that it enters a quasi-continuum. In the quasi-con tinuum, the vibrational levels are closer together and are indirec tly coupled to the absorbing mode, allowing for the absorption efficiency of the molecule to increase. Once th is point is reached, if there is both favorable absorption strength and laser power then multiple absorptions (10 to 100+ photons) can occur, leading to higher internal temper atures, thus allowing dissociation to occur more rapidly. This mechanism has been used to explain the abso rption of up to hundreds of photons for polycyclic aromatic hydrocarbons (PAHs)117,118 and fullerenes.119,120 Lasers Used for IRMPD R.C. Dunbar first performed an experiment demonstrating the photodissociation of gaseous ions in an ICR cell in the early 1970s.121 While these experiments were performed using a slide projector as the light source and a coarse cuto ff filter as the wavelength selector, many technological advances since then have been in corporated into more complicated experiments and, in particular, allow IRMPD to be performed quite routinely. One specific advancement is the development of sophisticated lasers. The first experiments with IRMPD used high-power continuous wave (cw)-tunable CO2 lasers, but there are several other types of lasers that can be used for irradiation. All lasers consist of a gain medium that is contained within a highly reflective optical cavity.122 Amplification of the light occurs when phot ons are passed through the gain medium and stimulated emission of radiation take s place. It is this amplification from which the laser gets its power. At one end of the cavity there is a sm all opening or partially reflective mirror through which a low percentage of the light can pass creat ing the output beam of the laser. While the general idea of a laser is simp le, there are several types of lasers each having its own methodology for creating light. Such lase r systems include optical parametric

PAGE 61

61 oscillators/amplifiers (OPO/OPA), tunable CO2 lasers and free electron lasers (FELs). The research reported in this dissertation used both a cw-tunable CO2 laser and a FEL. The first type of laser used in th is research was a continuous wave CO2 laser. Carbon dioxide lasers use a gas gain medium which contains carbon dioxide, helium, nitrogen and sometimes a small amount of hydrogen, water vapor and/or xenon.123 Typically, CO2 lasers are electrically pumped by a gas discharge. Nitrogen molecules are excited by this discharge into a higher vibrational level. Once ex cited, the nitrogen molecules transfer their energy to the CO2 molecules that collide with th em. The helium in the mixture serves as a depopulator, lowering the laser power and removing the heat. Non-tunable CO2 lasers typically emit at a wavelength of 10.6 m and tunable-CO2 lasers have outputs in the region of 9.2 to 10.8 m, which corresponds to the stretching frequency of C-O b onds. The laser power available for CO2 lasers can be from a few watts to several hundred watts. The benefit of CO2 lasers, especially tunable lasers, is that they are affordable bench-top lasers. The major drawback to tunable-CO2 lasers is that only a limited wavelength range can be explored. Free electron lasers are laser systems that no t only are large and complex, but also are expensive.124-126 For this reason, FELs are generally located in national laboratories. The amplification and wavelength rang e of FELs is achieved through th e use of an undulator. In the undulator, the placement of magnets with alternating polarities, as seen in Figure 3-3, allows free electrons to be accelerated, re sulting in the releas e of photons. The spacing of the magnets within the undulator and the energies of the electrons dictates the wave length of the photons being released. The released photon s results in coherent light that can then be used in various ways. The irradiation is composed of 5 to 20 macropulses (composed of hundreds of high-power micropulses spaced 1 ns apart).127 The macropulses are delivered to the user station at a rate of 5

PAGE 62

62 to 10 Hz. While the power of each micropulse is in the MWatt range, each macropulse has an average energy of ~30 to 50 mJ. The major be nefit of FELs is that they allow access to wavelengths (~5 to 250 m), that correspond to most of the chemically interesting infrared wavelength range, much wider than a CO2 laser can achieve. The major drawback of FELs is that they are very costly and access to beam time is limited since they are generally housed in national facilities with many user s applying for beam time. Several FELs have been coupled with Fourie r transform ion cyclotron resonance (FTICR) mass spectrometers. Examples include the Free Electron Laser for Infrared eXperiments (FELIX)127 at the FOM-Institute for Plasma P hysics Rijnhuizen in The Netherlands, the Centre Laser Infrarouge Orsay (CLIO)71 facility in Orsay, France and the FEL at the Science University of Tokyo (SUT).70 The research done in this diss ertation with IR photons from a FEL was performed at FELIX. Figure 3-4 shows a schematic of the laser instrumentation of the FELIX facility. Two lasers, FEL-1 and FEL2 give FELIX its wide wavelength range capability.127 Accelerator 1 allows FEL-1 to access wavelengths from 25 to 250 m. When the two accelerators are used in conjunction with each other, FEL-2 can access wavelengths from 5 to 30 m. The free electrons are accel erated to either 15 to 25 or 25 to 45 MeV by one or two radio-frequency linear accelerator s. An undulator is used, wher e the positioning of samariumcobalt permanent magnets tunes the wavelength of th e laser beam. The resonator of the laser is defined by two gold-plated copper mirrors. The F ELIX is a pulsed laser composed of micro and macro-pulses. The micropulses are spaced 1 ns apart and have a duration of 3 to 6 ps. Macropulses are possible for a duration up to 20 s at a rate of 5 Hz or 10 Hz. All lasers have their benefits and drawb acks. For this dissertation the relatively inexpensive and technically simple wavelength-tunable CO2 laser was used for a majority of the

PAGE 63

63 work. Although access to FELIX beam time is limite d, some shifts were available, so some of the research reported here was also done in The Netherlands. Use of both a FEL and CO2 laser gave a plethora of information that neit her laser alone would have provided.

PAGE 64

64 v =0 v =1 v =2 v =3 v =4 v =5 j =1 j =0 v =0 v =1 v =2 v =3 v =4 v =5 j =1 j =0 Figure 3-1. Energy potential well. As one move s up the well, the spacing of the vibrational levels decreases. v =0 v =0 v =1 v =2 v =3 v =4 v =5 v =0 v =1 v =2 v =3 v =4 v =5 IVR IVR Dissociation threshold IVR v =2 v =3 v =4 v =5 v =1 v =0 v =0 v =1 v =2 v =3 v =4 v =5 v =0 v =1 v =2 v =3 v =4 v =5 IVR IVR Dissociation threshold IVR v =2 v =3 v =4 v =5 v =1 Figure 3-2. Depiction of the IRMPD mechanis m in polyatomic molecules. One photon of IR radiation is absorbed, its energy is then dist ributed into an array of vibrational modes through IVR. This process is repeated until the dissociation threshold is met or exceeded, resulting in the fragmentation of the molecule.

PAGE 65

65 N N N N N N N S S S S S S S Electron beam Released photons N N N N N N N S S S S S S S Electron beam Released photons Figure 3-3. Schematic of an undulator used for FELs. Here a beam of free electrons enters the undulator and magnets of altern ating polarities forces the el ectrons to travel in an oscillating path, resulting in the release of photons. The photons combine coherently to give the final beam of light. Electron Injector Accelerator 1 Accelerator 2 Undulator1 Undulator2 FEL 1 FEL 2 Electron Injector Accelerator 1 Accelerator 2 Undulator1 Undulator2 FEL 1 FEL 2 Figure 3-4. Layout schematic of FELIX. Two accelerators and FELs are used to give FELIX its continuous wavelength range. Figure adapte d from Oepts, D.; van der Meer, A. F. G.; van Amersfoort, P. W. Infrared Phys. Technol. 1995, 36, 297-308.127

PAGE 66

66 CHAPTER 4 DIFFERENTIATION OF MONOSACCHARIDES IN THE POSITIVE ION MODE BY IRMPD W ITH A TUNABLE CO2 LASER Introduction As described in chapter 1, monosaccharides are the smallest of all the sugar units and play an important role in biological systems. Past mass spectrome tric methods have used soft ionization techniques such as fa st atom bombardment (FAB), el ectrospray ionization (ESI) and matrix assisted laser desorption ionization (MAL DI) to analyze monosaccharides in the positive ion mode.128-132 Gaucher and Leary showed that metal ions complexed with various monosaccharides can be used to differentiate the anomeric configuration of monosaccharides.128 They used electrospray ionizati on (ESI) and collision induced di ssociation (CID) to identify and differentiate hexoses (glucose and ma nnose) that were derivatized with zinc (diethylenetriamine). Other adducts that ha ve been used for the di fferentiation of anomers and include copper (II)131, ammonium133, lead134 and sodium135. All these past research methods have utilized fragments produced by CID to dist inguish the identity of the monosaccharides and derivatives being studied. Infrared multiple photon dissociation (IRMPD) is another fragmentation method that gives different, but complementary fragme nts to those produced by CID. Jose Valle used irradiation by a free electron laser (FEL) to fragment rubi diumand potassium-attached monosaccharides. Monitoring the dissociation of the various ions over a range of wavelengths showed differences in the IRMPD spectra which were used to diff erentiate positively charged saccharide isomers and anomers.74 As described in chapter 3, line-tunable CO2 lasers, when compared to FELs, have a limited wavelength range. Although the wavelength is limited, the cost of a line tunable CO2 laser is far less than that of a FEL. Therefore, use of a tunable CO2 laser in this work makes the method

PAGE 67

67 developed here more affordable and accessible to other laboratories. This chapter will discuss IRMPD research done at the University of Florida on lithiated O-methyl-glucoand galactopyranoside monosaccharides anomer s in the positive ion mode using a CO2 laser. Procedure Each of the monosaccharides was prepared at a concentration of 0.1 mM in 80:20 general-use grade methanol to MilliQ ultra-pure H2O solution containing 0.1 mM LiCl. The monosaccharides used in these studies were obtai ned from Dr. Brad Bendiak at the Department of Cellular and Structural Biology, University of Colorado Health Sciences Center. Their structures are seen in Figure 4-1 and include -O-methyl-glucopyranoside, -O-methyl-glucopyranoside, -O-methyl-galactopyranoside and -O-methyl-galactopyranoside. A schematic of the instrumental set-up is shown in Figure 4-2. The lithiated monosaccharides were ionized with a commercial ESI source (Analytica of Branford, Branford, CT, USA). The capillary of this source has been user modified136,137 with a conical capillary138 inlet to increase ion introduction into the mass sp ectrometer. For these experiments, the capillary was set to temperatures between 120 and 125C. The flow rate for all experiments was 15 L/hr. All experiments were performed on a Bruke r 47e Fourier transf orm ion cyclotron resonance (FTICR) mass spectrometer (Bruker Da ltonics; Billerica, MA, USA) with a 4.7 T superconducting magnet (Magnex Scientific Ltd.; Abington, UK) and an InfinityTM cell (Figure 4-2).139 Precursor ions were isolated using a stored waveform inverse Fourier transform (SWIFT)85 and irradiated for 1 second with a Lasy-20G tunable CO2 laser (Access Laser Co.; Marysville, WA, USA). This lase r has a power range of 0-20 W w ith a wavelength range of 9.2 to 10.8 m. Typical laser powers for the experi ments described in this chapter were approximately 0.7 W, but were occasionally as high as 3 W to overcome the effects of slight laser misalignment.

PAGE 68

68 To facilitate fragmentation and increase signa l intensity, ions were accumulated in the hexapole for a period of 1.0-1.5 seconds. Th is accumulation period was kept constant throughout all experiments when comparing monosacch aride anomeric pairs. Irradiation with the CO2 laser was facilitated by a mechanical mirror. When the mechanical mirror was in one position, the laser beam was blocke d and directed into a power meter. When the mirror was in the other position, the beam wa s passed into the back of the FTICR cell through a ZnSe window. No internal mirrors were used, therefore ther e was only one pass of the laser beam through the ion cloud within the cell. The wavelengths used were determined based on the stability and power stated in the laser manua l, where wavelengths with exce llent stability and high power were chosen. Each day, two sets of twenty-f ive scans of 512 K data points were collected and averaged at each wavelength. All experiments were repeated at least once, several days apart. Significance of results reported here is base d on the 95% confidence interval of the mean.140 Reproducibility Since variations in the alignment of the lase r and the abundance of i ons within the cell change the amount of fragmentation observed, a method for calibrating th e overall power of the laser beam actually irradiating the was needed. To ensure reproducibility, the total number of ions and irradiation power were kept constant throughout the day. The laser power used daily was found by determining the power needed to keep a ratio of 1.26 0.04 for the m/z 127 to m/z 201 fragment ions of lithiated -O-methyl-glucopyranoside at an irradiation wavelength of 9.588 m. Once the power needed for this ratio wa s found, it was used throughout the day. The power was monitored for each experiment with a power meter and adjustments were made to the CO2 control electronics as needed to keep the power constant. Due to slight variations in laser alignment as well as the cell heating during the course of the day, even with a daily calibration, some variance in the fragmentation was seen. Also, the

PAGE 69

69 inherent fluctuation of ion production by the ESI process could have added to the observed variances. The method of calib ration described in the last paragraph was settled upon after several other methods were tried. Preliminary attempts kept the percent dissociation of the precursor ion (m/z 201) constant for -O-methyl-glucopyranoside at 9.588 m and then used that laser power for the day. Whil e both methods of calibration resu lted in similar ratios for the abundances of key fragment ions (mainly m/z 109 and m/z 127), using the ratio of m/z 127 to m/z 201 daily gave a smaller relative error, better reproducibility and finer control over the energy imparted to the system. The fragment ion at m/z 127 was used for the calibration because its appearance and disappearance changed significantly when compared to that of the precursor ion (m/z 201) as a function of laser power. Results and Discussion Methyl-glucopyranosides Anomers of O-methyl-glucopyranoside were first studied. Since monosaccharides can open up into the chain form and then close again, permitting interconversion between the different anomers, use of O-methylated monos accharides insured that the anomer under study was locked in its closed conformation and ther efore could not interconvert. The fragmentation patterns were obtained by using a CO2 laser to irradiate both the and anomers of O-methyl-glucopyranoside. At each wavelengt h the percent abundance of each fragment was determined by the following formula: 100 Abundance Precursor Abundance Fragment Abundance Fragment abundance percent Relative (4-1) This percent abundance was plotted for each fragment over the range of 9.2 to 10.8 m as a function of both mass and wavelength. The major fragments for both anomers were m/z 67, 81, 91, 97, 109, 127, 141, 151 and 169, as seen in Figu re 4-3. The fragmentation patterns of the

PAGE 70

70 and -anomers proved to be different as seen in Figure 4-3. Specifically, the relative percent abundance of key fragments such as m/z 91, 109 and 127 appeared to be significantly different for the anomers. The abundance of m/z 109 was always higher for -O-methyl-glucopyranoside than for -O-methyl-glucopyranoside. Also, the relative percent abundance for m/z 127 was always significantly higher for the -anomer than the -anomer. These fragments are key identifying fragments and allow for easy discrimination between the two anomers. The IRMPD spectra of these ions showed that using disappearance of the precursor ion to distinguish between the and -anomer is impossible. As seen in Figure 4-4, there are several wavelengths at which the percen t abundances of the remaining pr ecursor ion signal overlap for these anomers. These wavelengths include 9.230, 9.250, 9.305, 9.473, 9.448, 9.520, 9.675 and 9.7 to 10.8 m. The power needed to fragment -O-methyl-glucopyranoside to obtain a ratio of 1.26 0.04 for m/z 127 to m/z 201 at 9.588 m, was far less than the power that was needed for fragmentation of either anom er at wavelengths 9.7-10.8 m. Figure 4-5 demonstrates the fragmentation efficiency at th e two wavelengths 9.588 and 10.611 m. While fragmentation at the higher wavelength is possible, it requires longer irradiation tim es and/or higher laser power. Typical laser power needed to fragment at the lower wavelengths was approximately 0.70 W, but to obtain any fragmentation at th e higher wavelengths (on the same day with the same alignment and number of ions in the cel l) required more than 3 W. Since the IRMPD depletion spectra of the precu rsor ion was found to be of little help in distinguishing the anomers, comp aring the percent abundances of m/z 109 and m/z 127 fragments for O-methyl-glucopyranosides allowed for the differe nt anomers to be distinguished, Figure 4-6. The ratio of m/z 109 to m/z 127 was always higher for the -anomer than for the -anomer. While the values changed slightly at the different wavelengths, the -anomer always produced a

PAGE 71

71 value that was greater than one and the -anomer gave values that were less than one. The greater abundance of the smaller fragments and th e larger depletion of the parent ion indicate that the -anomer of O-methyl-glucopyranoside requi res less energy for fragmentation than -O-methyl-glucopyranoside. Unknown Study of Methyl-glucopyranosides To demonstrate the ability to differentiate isomers of glucopyranosides by the developed method, a single-blind test was performed for two samples each of and -O-methyl-glucopyranoside. Repr esentative spectra taken at 9.588 m are shown in Figure 4-7 A and B. These unknowns were identif ied based on their fragmentation patterns. Figure 4-7 A has a greater m/z 109 to m/z 127 and a relatively small amount of m/z 169 in comparison to the unknown seen in Figure 4-7 B. Both methyl-glucopyran osides were positively identified based solely on their spectra. Methyl-galactopyranosides This fragmentation procedure was then a pplied to another hexoses anomer pair, and -O-methyl-galactopyranoside. To compare the re sults of the O-methyl-glucopyranosides with the O-methyl-galactopyranosides, a daily calibration with the -O-methyl-glucopyranoside was performed at 9.588 m. This ensured that the same amount of power was used for each anomer. For these studies, experiments were performed tw ice about a week and a half apart. Figure 4-8 shows the resulting fragmentation pattern of the O-methyl-galactopyranosides as a function of wavelength and mass. One key difference in the fragmentation patterns of and -O-methyl-galactopyranoside was the significantly gr eater abundance of the m/z 121 fragment ion for the -anomer as compared to that seen for the -anomer. Also, the ratio of m/z 169 to m/z 151 proved to be useful in the identification of the and -anomers, as shown in Figure 4-9 for wavelengths 9.2

PAGE 72

72 to 9.7 m. The abundance ratio of m/z 169 to m/z 151 was always less than 5 for -O-methyl-galactopyranoside and was always greater than 5 for the -anomer. When comparing Figure 4-3 to Figure 4-8, several differences in the fragmentation patterns of the methyl-glucoand galact opyranosides are apparent. For example, m/z 109, 127, 151 and 169 are key fragment ions whose relative percent abundances vary for the glucoand galactopyranosides. Comparing Figure 4-3 to Fi gure 4-8, one can see that the abundances of m/z 169 are greater for both methyl-galactopyranos ides than for the me thyl-glucopyranoside anomers, while the abundances of m/z 109 and m/z 127 are lower for the methyl-galactopyranosides than for the methyl-g lucopyranosides. Using these differences in fragmentation patterns the glucoand galactopyranos ides can be distinguished from each other. Unknown Study of both Methyl-glucoand galactopyranosides To test the method described above for differentiating both and anomers of both the gluco-and galactopyranosides, a si ngle-blind study was performed. For this study, two samples each of -, -O-methyl-glucopyranoside and -, -O-methyl-galactopyranoside were randomized and their identity concealed. The unknowns were then analyzed individually at wavelengths of 9.230, 9.473 and 9.588 m. The identities of the unknown samples were then determined using the flowchart shown in Figure 4-10. Similar spectra to those show n in Figure 4-7 A and B were obtained for and -O-methyl-glucopyranoside. Spectra obtaine d for the galactopyranosides are shown in Figure 4-11 A and B. As displaye d in Figure 4-11 A, there was a relatively small abundance of m/z 109 and m/z 127 and a high abundance of m/z 169, identifying this unknown as one of the galactopyranoside anomers. Comparing the sp ectrum in Figure 4-11 A to the spectrum in Figure 4-11 B, there is a higher ratio of m/z 127 to m/z 109 and very little m/z 151. The lack of m/z 121 and the ratio of 8.9 for m/z 169 to m/z 151 positively identify this unknown as

PAGE 73

73 -methyl-galactopyranoside. This identification was confirmed with the fragmentation seen when the unknown was irradiated with a highe r laser power. For the next unknown, whose spectrum is seen in Figure 4-11 B, the ratio of m/z 169 to m/z 151 of 3 and the appearance of m/z 121 made it clear that the id entity of this unknown was -O-methyl-galactopyranoside. All of the eight unknown samples were correc tly identified in a similar way. In order to simulate a real-life laboratory environment, no calibration of laser power was done before the unknown studies reported in the last paragraph. Although various laser powers were used for the unknown studies, the precursor ion was never depleted. In general, higher powers were needed for O-methyl-galactopyranos ides since they fragmented less than the O-methyl-glucopyranosides. For example, th e methyl-glucopyranosides that produced the spectra seen in Figure 4-7 A and B were both irradiated with 2.61 W and the methyl-galactopyranosides that produced the spectra seen in Figure 4-11 A and B were irradiated with 3.98 and 4.33 W, respectively. Conclusions Use of a tunable CO2 laser produced unique fragmenta tion patterns for anomers of O-methyl-glucopyranoside and O-methyl-galactopyr anoside over the wavelength range of 9.2 to 9.7 m. Various fragment ions and their ratios were used to differentiate between the two sets of monosaccharides (O-methyl-galactopyranoside and O-methyl-glucopyranoside) anomers. Since only two monosaccharides were studied, future work should include other hexoses.

PAGE 74

74 O H HO H HO H H OH H OCH3 OH O OH H H HO H OCH3 OH H H OH O OH H H HO H H OH H OCH3 OH O H HO H HO H OCH3 OH H H OH Alpha-methyl-glucopyranoside Beta-methyl-glucopyranoside Alpha-methyl-galactopyranoside Beta-methyl-galactopyranoside Figure 4-1. Structures of the O-methylated monosaccharides discussed in this chapter. CO2Laser M1 M2 Salt window Power meter WavemeterMirror gateM3 Laser table M4M5 Iris ZnSe window Source4.7 T Bruker Mass SpectrometerInfinity cell CO2Laser M1 M2 Salt window Power meter WavemeterMirror gateM3 Laser table M4M5 Iris ZnSe window Source4.7 T Bruker Mass SpectrometerInfinity cell Figure 4-2. Experimental set up of the 4.7 T FTICR mass spectrometer. When the mechanical mirror M3 is down the laser beam passes through the ZnSe wi ndow into the cell. If the mechanical mirror up the laser beam is blocked from entering the cell and reflected into the power meter.

PAGE 75

75 Figure 4-3. Wavelength-dependent fragmentation patterns for the lithiated O-methyl-glucopyranosides for wavelength from 9.2 to 10.8 m. A) Lithiated -O-methyl-glucopyranoside. B) Lithiated -O-methyl-glucopyranoside.

PAGE 76

76 Figure 4-4. Infrared multiple photon dissociation depletion spectra of the precursor ions (m/z 201) for both and -O-methyl-glucopyranoside lithium cation complexes.

PAGE 77

77 m/z 200 190 180 170 160 150 140 130 120 110 Abundance 20 15 10 5 201 169 151 141 127 109A111 m/z 200 190 180 170 160 150 140 130 120 110 Abundance 20 15 10 5 201 169 151 141 127 109A m/z 200 190 180 170 160 150 140 130 120 110 Abundance 20 15 10 5 201 169 151 141 127 109 m/z 200 190 180 170 160 150 140 130 120 110 Abundance 20 15 10 5 201 169 151 141 127 109A111 m/z 200 190 180 170 160 150 140 130 120 110 Abundance 45 40 35 30 25 20 15 10 5 201 169 141 127 109B m/z 200 190 180 170 160 150 140 130 120 110 Abundance 45 40 35 30 25 20 15 10 5 201 169 141 127 109B Figure 4-5. Comparison of the fragmentation of -methyl-glucopyranoside at wavelengths 9.588 and 10.611 m. A) Spectrum obtained at 9.588 m. B) Spectrum obtained at 10.611 m. The laser power used was 2.21 W for 9.588 m and 5.86 W for 10.611 m. This demonstrates the difference in absorbance and fragmentation for the different disaccharides. Also it s hows that conventional non-tunable CO2 lasers, with an output wavelength of 10.6 m, are not optimal for fragmentation.

PAGE 78

78 Figure 4-6. Relative percent abundance of fragment ions for both lithiated and -O-methyl-glucopyranosides over the wavelength range from 9.201 to 9.675 m. A) Relative percent ab undance of product ion m/z 109. B) Relative percent abundance of product ion m/z 127. C) The ratio of m/z 109 to m/z 127.

PAGE 79

79 m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201A79 81 m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201A m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201A79 81 m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201B79 81 m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201B m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201 m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 m/z 200 150 100 Abundance 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 67 91 97 109 127 141 151 169 201 67 91 97 109 127 141 151 169 201B79 81 Figure 4-7. Spectra of unknowns in single blind study of methyl-g lucopyranosides at wavelength 9.588 m. A) Spectrum of -O-methyl-glucopyranoside. B) Spectrum of -O-methyl-glucopyranoside.

PAGE 80

80 Figure 4-8. Fragmentation patterns ove r the wavelengths from 9.2 to 10.6 m. A) Lithiated -O-methyl-galactopyranos ide. B) Lithiated -O-methyl-galactopyranoside.

PAGE 81

81 Figure 4-9. Ratio of m/z 169 to m/z 151 for and -O-methyl-galactopyranoside. m/z 169 > m/z 91,109,127,151 yes no -O-methyl-galactopyranoside -O-methyl-galactopyranoside -O-methyl-glucopyranoside -O-methyl-glucopyranoside > 5 < 5 > 1 < 1 m/z 169/151 m/z 109/127 -O-methyl-galactopyranoside -O-methyl-galactopyranoside -O-methyl-glucopyranoside -O-methyl-glucopyranoside m/z 169 > m/z 91,109,127,151 yes no -O-methyl-galactopyranoside -O-methyl-galactopyranoside -O-methyl-glucopyranoside -O-methyl-glucopyranoside > 5 < 5 > 1 < 1 > 1 < 1 m/z 169/151 m/z 109/127 -O-methyl-galactopyranoside -O-methyl-galactopyranoside -O-methyl-glucopyranoside -O-methyl-glucopyranoside Figure 4-10. Decision flowchart used to id entify the different monosaccharide anomers.

PAGE 82

82 m/z 200 150 100 Abundance 50 45 40 35 30 25 20 15 10 5 67 91 109 111 127 141 151 169 201 97A79 m/z 200 150 100 Abundance 50 45 40 35 30 25 20 15 10 5 67 91 109 111 127 141 151 169 201 97A79 m/z 200 150 100 Abundance 150 100 50 67 91 109 111 127 141 151 169 201 97 121B81 79 m/z 200 150 100 Abundance 150 100 50 67 91 109 111 127 141 151 169 201 97 121B81 79 Figure 4-11. Spectra of unknowns identified as galactopyranosides in single blind study obtained at wavelength 9.588 m. A) Spectrum of -O-methyl-galactopyranoside. B) Spectrum of -O-methyl-galactopyranoside.

PAGE 83

83 CHAPTER 5 DIFFERENTIATION OF DISACCHARIDES IN THE P OSITIVE ION MODE WITH A TUNABLE CO2 LASER Introduction The structure of carbohydrates dictates th eir biological function. This includes the monosaccharide sequence, the anomeric configuration and the linkage between the monosaccharides within the oligosaccharides, wh ich are the carbohydrate building blocks. Since various linkages and anomeric configuration are possible, being able to differentiate the smaller oligosaccharides that co mpose the larger carbohydrates is a complicated task. As described in chapter 1, disaccharides are the smallest sacchari de units that contain the glycosidic bond. The anomeric configuration of this bond can play an important role in the functi on of the saccharide. In the past, collision induced dissociation (CID)60,141 and infrared multiple photon dissociation (IRMPD)62,73,142 have been used for fragmentation of saccharides, in particular disaccharides. Past experiments by Polfer et al. examined fragmentation patterns of lithiated disaccharides with the Free Electron Laser for Infrared eXperiments (FELIX) at the FOM-Institute for Plasma Physic s Rijnhuizen in the Netherlands.73 They found that isomeric ions with various linkages had different frag mentation patterns. They also found that the intensity ratio of specific fragments (m/z 169/187) was higher for -anomers than for -anomers and may be used to differentiate the anomeric configuration of the disaccharides. Although the fragmentation and some ratios we re explored, there was no attempt to quantitatively ascertain the anomeric configuration. This chapter describes attempts to develop a method for the differentiation of lithiated disaccharides. Wavelength-selective fragmentati on of glucose-containing disaccharide anomers was performed by IRMPD with a tunable CO2 laser, and differentiation of the disaccharides was

PAGE 84

84 based on their fragmentation patterns and ratios of the relative abundances of specific fragment ions. Procedure All work described in this chapter was performed with the instrumental set-up described in chapter 4. Each of the disaccharides was prep ared as a 0.10 mM solution which also contained 0.10 mM LiCl. To aid in ionization, the solven t was composed of an 80:20 ratio of general-use grade methanol to MilliQ ultra-pure H2O. The disaccharides used in these studies were obtained from Dr. Brad Bendiak at the Department of Ce llular and Structural Biology, University of Colorado Health Sciences Center, and were composed of two glucose units having varying linkages and anomeric configurations: kojibiose ( 1-2), sophorose ( 1-2), nigerose ( 1-3), laminaribiose ( 1-3), maltose ( 1-4), cellobiose ( 1-4), isomaltose ( 1-6) and gentiobiose ( 1-6). Fragmentation Study For each lithiated disaccharide, the precursor ion (C12O22H11Li+, m/z 349) was produced by electrospray ionization (ESI) and isolated via a stored waveform inverse Fourier transform (SWIFT). The isolated precursor ion was irradiated for 1 second using a laser power determined by a daily calibration involving the precursor ion of sophorose ( 1-2). This ion was fragmented to obtain an m/z 349:229 intensity ratio of 1.04 0. 16. The laser power required to achieve this ratio was then used for the remai nder of the experiments over the wavelength range from 9.2 to 9.7 m. For each disaccharide, three data sets composed of 15 scans of 512 K points were collected and averaged at each wavelength. To test reprod ucibility, the entire fragmentation pattern of sophorose was obtained once on two separate days. Once fragmented, all the relative percent abundances of the fragments were calculated and correlated w ith the disaccharide linkage.

PAGE 85

85 Anomeric Configuration Study After the fragmentation patte rns corresponding to different linkages were determined, experiments to differentiate the anomeric confi guration for the pairs of anomers were performed at 9.342, 9.473 and 9.588 m. For these experiments, the laser power was adjusted to obtain a peak height ratio of 1:2 for the isolated precursor ion (m/z 349) to a specific fragment ion (m/z 169, 187 or 229) for different anomers, see Table 5-1. Abundance ratios for other peak pairs were measured and correlated w ith the anomeric configuration. Results and Discussion Differentiation of Disaccharides The wavelength-dependent fragme ntation patterns for all of the disaccharides are shown in Figure 5-1. These fragmentati on patterns from 9.2 to 9.7 m are similar to the patterns found by Polfer et al. with a free electron laser.73 For example, fragmentation of 1-2 linked disaccharides produces a major fragment of m/z 229, while the spectra of disaccharides with linkage 1-3 have major fragments of m/z 169 and 331 and linkage 1-4 spect ra have similar amounts of m/z 169 and 187, with very little m/z 229. Lastly, the spectra of disaccharides with linkage 1-6 have fragments m/z 169, 187, 229, 259, 289, but very little 331. While the fragments produced with the CO2 laser are identical to those produced with FELIX, the relative percent abundances of each fr agment are slightly different. Since the relative abundance of the fragment s is dependent on the amount of dissociation of the precursor ion (m/z 349), variations in the laser power experienced by th e ion clouds and the energy imparted to the ions during electrospray could be causes of these differences. Also, differences in the nature of the laser ir radiation (continuous wave CO2 laser vs. several macropulses composed of multiple high-power micropulses) could be th e cause of some of the differences seen.127

PAGE 86

86 As seen in Figure 5-1, fragmentation of the pr ecursor ion over a range of wavelengths from 9.2 to 9.7 m gave fragmentation patterns unique for each of the disaccharides. The wavelength of fixed-frequency CO2 lasers (10.6 m) was also explored, but very little (if any) fragmentation was seen for the eight disaccharides. Either l onger irradiation times and/or much higher laser power were needed to see the little fragmentation that was observed. This result is consistent with those found by Polfer et al. in which the fragmentation of the lithiated disaccharides dramatically declined around 10.6 m.73 Furthermore, this research shows that the typical wavelength of non-tunable CO2 lasers is not optimal for differentiation of the various disaccharides. The flowchart in Figure 5-2 shows that the presence and absence of a certain fragment makes determination of the different linkages possible. For example, if the dissociation spectrum contains the fragments m/z 331 and m/z 289, but not the m/z 259 fragment, the linkage can be identified as 1-4. Similar patterns can be specified for each of the different disaccharides that were studied. While the linkage can be determined by the presence or absence of certain fragments alone, more information is needed to identify the anomeric configuration. Determination of the Anomeric Configurations To determine the anomeric configurations, a dditional experiments were performed at three wavelengths (9.342, 9.473 and 9.588 m) in which the m/z 349 precursor ion was dissociated with sufficient laser power to de crease its abundance to one half th at of a fragment ion specified in Table 5-1. Using this laser power, the ratio s of the relative percen t abundances of other fragment ions were measured and used to iden tify the anomeric configuration, as shown in Figure 5-3. For example, if the fragmentation pattern indicated that the linkage was 1-4, the precursor ion was then irradiated to give a ratio of 1:2 for m/z 349 to m/z 187. The ratio of two

PAGE 87

87 product ions (m/z 229 and 289) was then calculated and, based on the flowchart in Figure 5-3, the anomeric configuration was determined. The resulting ratios from the relative percent abunda nces of the specific ions provide a method to differentiate all of the anomers at the various wavelengths. While only one wavelength is needed to differentiate the anomer s from each other, use of multiple wavelengths confirms the results. For the 1-2 linkage disaccharides the ratio of m/z 187 to 229 was always higher for kojibiose ( -linked) than for sophorose ( -linked). The ratio of m/z 169 to 187 was always lower for nigerose ( -linked) than for laminaribiose ( -linked). For 1-4 linked disaccharides, the ratio of m/z 229 to 289 is greater than 0.25 for cellobiose ( -linked) and less than 0.25 for maltose ( -linked). Lastly, the ratio of m/z 169 to 187 is greater for gentiobiose ( -linked) than isomaltose ( -linked) at all of the three wavelengths. Differentiation of Unknowns To make this work applicable to other labor atories, a method to determine both the linkage and the anomeric configuration of the differen t disaccharides was explored. To test this experimental procedure, one sample of each di saccharide was transferre d to a coded vial to conceal its identity. Each unknown was then tested on two separate days and the identity was predicted based on the methods described above a nd in Figures 5-2 and 5-3. Figure 5-4 shows the results of the unknowns in comparison to the known samples analyzed previously. All error bars shown are the 95% confidence interval of the mean for the experimental scans for a given day. In all cases, the identity of the unknown was positively identified based on the ratios and flow charts. Some of the fluctuation and discrepancy between the ratios of the unknown and known samples could be due to variation in the electrospray source. Since the laser power setting is based on the ratio of the precursor ion to a fr agment, slight changes in the abundance of the

PAGE 88

88 precursor ion can cause the ratios of fragments to be affected. Even with the variation, the results are such that the disaccharid es were distinguished unambiguously. Conclusions This research demonstrated the use of a tunable CO2 laser to identify both the linkage and the anomeric configuration of various lithium-attached disaccharides. When compared to FELs, tunable CO2 lasers have a limited wavelength range, but this research showed that results comparable to those from the broad output wave length range of the FEL can be achieved in the CO2 wavelength range of 9.2 to 9.7 m. These results provide a method for differentiation of isomers that is accessible and economical ly feasible for other laboratories.

PAGE 89

89 Figure 5-1. Wavelengthdependent fragmentation for the various linked lithiated disaccharides. A) Kojibiose ( 1-2). B) Sophorose ( 1-2). C) Nigerose ( 1-3). D) Laminaribiose ( 1-3). E) Maltose ( 1-4). F) Cellobiose ( 1-4). G) Isomaltose ( 1-6). H) Gentiobiose ( 1-6).

PAGE 90

90 Nigerose ( 1-3) Laminaribiose ( 1-3) Maltose ( 1-4) Cellobiose ( 1-4) Isomaltose ( 1-6 ) Gentiobiose ( 1-6) Kojibiose ( 1-2) Sophorose ( 1-2) Isomaltose ( 1-6) Gentiobiose ( 1-6) m/z 289 observed? yes no Maltose ( 1-4) Cellobiose ( 1-4) Isomaltose ( 1-6) Gentiobiose ( 1-6) Nigerose ( 1-3) Laminaribiose ( 1-3) m/z 259 observed? Maltose ( 1-4) Cellobiose ( 1-4) m/z 331 observed? yes no yes no Nigerose ( 1-3) Laminaribiose ( 1-3) Maltose ( 1-4) Cellobiose ( 1-4) Isomaltose ( 1-6 ) Gentiobiose ( 1-6) Kojibiose ( 1-2) Sophorose ( 1-2) Isomaltose ( 1-6) Gentiobiose ( 1-6) m/z 289 observed? yes no yes no yes no Maltose ( 1-4) Cellobiose ( 1-4) Isomaltose ( 1-6) Gentiobiose ( 1-6) Nigerose ( 1-3) Laminaribiose ( 1-3) m/z 259 observed? Maltose ( 1-4) Cellobiose ( 1-4) m/z 331 observed? yes no yes no yes no yes no yes no yes no Figure 5-2. Flow-chart depi cting how linkage of the disaccharides was determined.

PAGE 91

91 Table 5-1. Table of ratios used to determin e the laser power used for fragmentation. Disaccharide Ratio used for standardizing irradiation power Kojibiose ( 1-2) Sophorose ( 1-2) 1:2 m/z 349:229 Nigerose ( 1-3) Laminaribiose ( 1-3) 1:2 m/z 349:169 Maltose ( 1-4) Cellobiose ( 1-4) 1:2 m/z 349:187 Isomaltose ( 1-6) Gentiobiose ( 1-6) 1:2 m/z 349:169 Nigerose ( 1-3) Laminaribiose ( 1-3) Kojibiose ( 1-2) Sophorose ( 1-2) Isomaltose ( 1-6) Gentiobiose ( 1-6) Maltose ( 1-4) Cellobiose ( 1-4) m/z 187/229 > 0.1 < 0.1 Kojibiose Sophorose m/z 169/187 > 5.0 < 5.0 m/z 229/289 > 0.25 < 0.25 m/z 169/187 > 1.0 < 1.0 Laminaribiose Nigerose Cellobiose Maltose Gentiobiose Isomaltose Nigerose ( 1-3) Laminaribiose ( 1-3) Kojibiose ( 1-2) Sophorose ( 1-2) Isomaltose ( 1-6) Gentiobiose ( 1-6) Maltose ( 1-4) Cellobiose ( 1-4) m/z 187/229 > 0.1 < 0.1 m/z 187/229 > 0.1 < 0.1 Kojibiose Sophorose m/z 169/187 > 5.0 < 5.0 m/z 169/187 > 5.0 < 5.0 m/z 229/289 > 0.25 < 0.25 m/z 229/289 > 0.25 < 0.25 m/z 169/187 > 1.0 < 1.0 m/z 169/187 > 1.0 < 1.0 Laminaribiose Nigerose Cellobiose Maltose Gentiobiose Isomaltose Figure 5-3. Flow-chart showing ratios of peak heights and values used to determine anomeric configurations.

PAGE 92

92 Kojibiose and Sophorose m/z 187/2290 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4Kojibiose 9.342 Kojibiose 9.473 Kojibiose 9.588 Sophorose 9.342 Sophorose 9.473 Sophorose 9.588SampleRati o Knowns Unknown trial 1 Unknown trial 2 RatioA Kojibiose and Sophorose m/z 187/2290 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4Kojibiose 9.342 Kojibiose 9.473 Kojibiose 9.588 Sophorose 9.342 Sophorose 9.473 Sophorose 9.588SampleRati o Knowns Unknown trial 1 Unknown trial 2 RatioA Nigerose and Laminaribiose m/z 169/1870 2 4 6 8 10 12Nigerose 9.342 Nigerose 9.473 Nigerose 9.588 Laminaribiose 9.342 Laminaribiose 9.473 Laminaribiose 9.588SampleRati o Known Unknown trial 1 Unknown trial 2 RatioB Nigerose and Laminaribiose m/z 169/1870 2 4 6 8 10 12Nigerose 9.342 Nigerose 9.473 Nigerose 9.588 Laminaribiose 9.342 Laminaribiose 9.473 Laminaribiose 9.588SampleRati o Known Unknown trial 1 Unknown trial 2 RatioB Maltose and Cellobiose m/z 229/2890 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9Maltose 9.342 Maltose 9.473 Maltose 9.588 Cellobiose 9.342 Cellobiose 9.473 Cellobiose 9.588SampleRati o Known Unknown trial 1 Unknown trial 2 RatioC Maltose and Cellobiose m/z 229/2890 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9Maltose 9.342 Maltose 9.473 Maltose 9.588 Cellobiose 9.342 Cellobiose 9.473 Cellobiose 9.588SampleRati o Known Unknown trial 1 Unknown trial 2 RatioC Isomaltose and Gentiobiose m/z 169/1870 0.5 1 1.5 2 2.5 3 3.5Isomaltose 9.342 Isomaltose 9.473 Isomaltose 9.588 Gentiobiose 9.342 Gentiobiose 9.473 Gentiobiose 9.588SampleRati o Known Unknown trial 1 Unknown trial 2 RatioD Isomaltose and Gentiobiose m/z 169/1870 0.5 1 1.5 2 2.5 3 3.5Isomaltose 9.342 Isomaltose 9.473 Isomaltose 9.588 Gentiobiose 9.342 Gentiobiose 9.473 Gentiobiose 9.588SampleRati o Known Unknown trial 1 Unknown trial 2 RatioD Figure 5-4. Bar graphs comp aring ratios from knowns and unknown lithiated glucose-containing disaccharides at the wavelengths 9.342, 9.472 and 9.588 m. A) Ratio of m/z 187/229 for kojibiose and sophorose. B) Ratio of m/z 169/187 for nigerose and laminaribiose. C) Ratio of m/z 229/289 for maltose and cellobiose. D) Ratio of m/z 169/187 for isomaltose and gentiobiose.

PAGE 93

93 CHAPTER 6 IRMPD STUDIES OF NEGATIVELY CHAR GED DISAC CHARIDES WITH A TUNABLE CO2 LASER Introduction The fragmentation of alkali-attached disaccharides73 and polysaccharides formed by electrospray ionization (ESI) has been previously studied.57 The results showed that disaccharides and polysaccharides with different linkages give different fragments and that the wavelength-dependent fragmentation patterns can be used to identify the linkages in these positively charged species. Both co llision induced dissociation (CID),59,60,141,143,144 and infrared multiple photon dissociation (IRMPD) using a free electron laser (FEL)62,70,73,74,142 have been used to fragment saccharides. While adduction of a metal in both positive and negative ion modes makes ionization of car bohydrates easier, when develo ping a general approach for saccharide isomeric determination it would be si mplest to analyze the saccharide without any metal ions attached. Past studies done in the negativ e ion mode by Jiang and Cole showed that the addition of chloride yields a higher abundance of the precur sor ion than that seen for the deprotonated disaccharide, thus making it easier to isolate and fragment the precursor.145 Early studies of chlorinated disaccharides such as sucrose were performed with fast atom bombardment (FAB) and CID by Prome et al.,146 who found that the saccharides give characteristic fragment ions that can be used for identification in the negative mode. Along with other fragments, they observed that the fragmentation of the chloride-attached species resulted in loss of HCl to yield a high abundance of deprotonated sucrose. More re cent studies in the negative mode on the fragmentation of 1-3, 1-4 and 1-6 linked gl ucose-containing disaccharides were conducted by Zhu and Cole,147 who showed that it is possible to id entify the linkage of the chlorinated disaccharides by CID. They also found that th e spectra of the deprotonated species are very

PAGE 94

94 similar to the spectra obtained for the chlorina ted species, further suppor ting the theory of HCl loss prior to dissociation.147 Even more recent studies by Jiang and Cole have shown that CID fragmentation of chlorinated disaccharides not only gives structural linkage information, but also has the potential for anomeric discrimination,145 using the relative abunda nces of chlorinated products vs. the non-chlorinated products. Of the 1-4 linked disaccharides explored by Jiang and Cole, the ratio of chlorinated to non-chlor inated products was always higher for the -anomer than the -anomer. They observed the opposite trend for 1-6 linked disaccharides. This chapter reports studies of the fragme ntation patterns of several deprotonated glucose-containing disaccharides with various linkages and anomer ic configurations obtained by irradiation with a CO2 laser in the negative ion mode. The fragmentation patterns of chlorinated disaccharides at selective wavelengths are also described. Procedure Deprotonated Disaccharides All the fragmentation results for the deprot onated disaccharides discussed here were obtained at the University of Fl orida in Dr. David Powells labor atory. A Bruker Bio-Apex II Fourier transform ion cyclotr on resonance (FTICR) mass spectrometer with a 4.7 T magnet (Bruker Daltonics, Billerica, MA) and an InfinityTM cell were used to analyze several glucosecontaining disaccharides including: kojibiose ( 1-2), nigerose ( 1-3), laminaribiose ( 1-3), maltose ( 1-4), cellobiose ( 1-4), isomaltose ( 1-6) and gentiobiose ( 1-6). All disaccharide samples were provided by Dr. Brad Bendiak of the Department of Cellular and Structural Biology, University of Colorado Health Sciences Center. The disaccharides were ionized with a pneumatically-assisted Apollo exte rnal electrospray source (Bruker Daltonics, Billerica, MA). For irradiation, the beam from a Lasy-20G tunable CO2 laser (Access Laser Co.; Marysville, WA) was passed into the ICR cell through a KBr window located on the end of the FTICR mass

PAGE 95

95 spectrometer opposite the end from which externally generated ions were admitted. To control the irradiation time, a mechanical mirror was used to direct the laser beam. When not directed into the cell, the laser beam was directed into a power meter head, allowi ng the laser power to be monitored throughout the day. Figure 6-1 shows a general schematic of the FTICR-MS and laser instrumentation used for these experiments. Solutions of each disaccharide were prepar ed at a concentration of 0.1 mM. For deprotonation of the disacch arides, several bases were tried, including sodium hydroxide (NaOH), tri-ethylamine (N(CH2CH3)3) and ammonium hydroxide (NH4OH). Of the bases used, NaOH gave the most stable signal with largest abundance of the deprotonated disaccharide. Therefore, NaOH was used for all of the deprotonated disaccharide experiments. Several solutions with concentration ratios fr om 0.1 to 1.0 mM NaOH to 0.1 mM disaccharide were tried, and a concentration of 1 mM NaOH was found to give the best signal. For improved ionization, all solutions were composed of 80:20 methanol:water made with general-use grade methanol and MilliQ water. For all e xperiments, flow rates between 3 and 7 L/min were used. The flow rate was adjusted for each disaccharide to obtain the most stable and abundant signal. All of the isolated disaccharide precursor ions (m/z 341) were irradiated with the CO2 laser for one second. The wavelengths used were chos en based on specificatio ns given in the laser manual, so that only wavelengths with good or excellent power and stability were used. For each wavelength, three experimental sets of 10 scan s of 128 K points each were collected. Once the precursor ions were fragmented, the relative percent abundance of each fragment was determined. Chlorinated Disaccharides Experiments which studied the fragmentation of chlorinated disaccharides were performed at the University of Florida in Dr. Eylers laboratory with the same instrumental set-up as

PAGE 96

96 described in chapter 4. All samples were 0.1 mM solutions made with 1:1 LiCl:saccharide in 80:20 methanol:water. To obtain higher signal intensities, ions were accumulated for 1.5 seconds before being transferred to the anal yzer cell followed by isolation, irradiation and detection. All disaccharides were ir radiated for 1 second with a Lasy-20 CO2 laser. The laser power used was adjusted to create a peak heig ht ratio of 1:1 for the isolated precursor ion (m/z 377) to a specific fragment ion (m/z 161 or 179). For disaccharides kojibiose ( 1 -2), sophorose ( 1-2), isomaltose ( 1-6) and gentiobiose ( 1-6), a ratio of 1:1 for m/z 377 to m/z 179 was used. For nigerose ( 1-3), laminaribiose ( 1-3), maltose ( 1-4) and cellobiose ( 1-4), a ratio of 1:1 for the peak heights of m/z 377 to 161 was used. For each linkage, the choice of fragment was based on greatest abundance and highest sensitivity relative to the disappearance of the precursor ion peak. Reproducibility: Deprotonated Disaccharides To aid in reproducibility, a daily calibration was performed using isomaltose, which was the first successfully detected deprotonated disaccharide. In this calibration, the laser power needed to produce an m/z 179:341 abundance ratio of 1.19 0.17 was determined daily. The laser power was kept constant for all experiments performed on the same day. Reproducibility: Chlorinated Disaccharides To increase the reproducibility of these experime nts, the laser power was adjusted to give a 1:1 ratio of the precursor ion to a specific fragment ion. The laser power was determined by monitoring the fragment ion (either m/z 179 or 161) for each wavelength in the experiment. Most of the variation seen in these experiments can be attr ibuted to electrospray source ionization and/or laser power fluctuations. Specif ic wavelengths were chosen to give a general coverage of the full ra nge between 9.2 to 9.7 m. For time considerations and for simplicity of the experiments, spectra at three wavelengths (9.342, 9.473 and 9.588 m) were obtained.

PAGE 97

97 Results and Discussion Deprotonated Disaccharides Figure 6-2 shows the percent abundance of th e precursor ion of isomaltose following IRMPD for the various wavelengths studied (i.e. a depletion spectrum). While the calibration wavelength of 9.588 m has a 95% confidence interval of th e mean from experiments performed on two separate days of approximately 2% the maximum variation (at wavelength 9.657 m) was 26%. Figure 6-2 shows that th e day-to-day variation of this method is larger than desired. Some of the fluctuations causing daily variations could result from instab ility in the electrospray source or slight variation of the laser power. Based on the variance of relative percent abundance for the fragments, this proved to be a qualitative rather than quantitative method for determining the linkage position of the disaccharides. The relative percent abundances for the fragme nts are plotted as a function of wavelength and mass in Figure 6-3, which shows that th e major fragments produced for all of the disaccharides, except for kojibiose, were m/z 179 and m/z 161. For kojibiose, the major fragments seen were m/z 263 and 323, as well as minor contributions from fragments m/z 331, 281, 113 and 101. For both isomaltose and gentiobiose, the appearance of m/z 281 was unique for and thus characteristic of the 1-6 linkage. Thus, 1-2 and 1-6 linked disaccharides can be distinguished based on the presence of the m/z 281 fragment ion. Disacc harides with linkages of 1-3 and 1-4 gave only fragments at m/z 161 and 179. To distinguish the two linkages, the abundance of m/z 161 and 179 for each linkage was explored further. For the 1-3 linked, the ratio of m/z 161:179 was approximately 1:2, whereas the ratio was approximately 6:1 for the 1-4 linked disaccharides. Thus, within the wavelength range 9.2 to 9.7 m, the linkage of disaccharides can be identified from the fragmentation pattern.

PAGE 98

98 Although the linkage position of the disaccharides could easily be determined based on the appearance of certain key fragments, determinin g the anomeric configuration was problematic. Since fragmentation of 1-3 and 1-4 linked disaccharides produces only m/z 161 and 179 ions, the ratio of these two was calculated for each enanti omeric pair. The results for the 1-3 and 1-6 linked disaccharides are shown in Figure 6-4. Unfortunately, the ratios for each anomeric pair were very similar and the identities of the a nomers cannot be distinguished by this means. Therefore more research in this area is needed including greater fragmentation of the precursor ion to determine if lower mass ions can be used to distinguish the anomers. Due to time constraints and chlorine contamination, the spectrum of deprotonated sophorose was not obtained in this study. While the relative percent abundances of the frag ment ions appear to vary from day to day, the identities of the fragment ions from each di saccharide remained the same. Figure 6-5 shows an example of the percent dissociation of the prec ursor ion from isomaltose and relative percent abundances of the fragment ions at two different wavelengths (9.201 and 9.657 m) on two separate days. These differences, for example the relative percent abundance of the precursor ion (m/z 341), could be due to the inherent fluctu ation of the electrospra y source and the laser power. Since the degree of depl etion of the precursor ion dict ates the abundances of the fragments produced, an increase in laser power results in greater abunda nces of the smaller m/z ions due to multiple fragmentation. Chlorinated Disaccharides During the experiments on deprotonated disacchar ides, the presence of chlorine hindered detection of the deprotonated spec ies. If chlorine was present in the cell and/or solutions, the sugar would preferentially bind to Clrather than lose H+. To take advantage of this, several

PAGE 99

99 experiments with the chlorinated disaccharides we re performed in Dr. Eyle rs laboratory at the University of Florida. First, the fragmentation indu ced by the irradiation of the chlorinated disaccharides was studied. Table 6-1 shows the fragments for each chlorinated disaccharide obtained at 9.588 m by varying the laser power to decrease the signa l abundance of the precursor ion to a relative percent abundance of 2.6 1.6%, Figure 6-6. Almost depleting the precursor ion allowed for more of the lower mass ions (for example m/z 101) to be observed. By monitoring the fragme nt ions produced, it is possible to identify the linkage. The presence of m/z 263 and 323 without m/z 281 indicates a 1-2 linkage. The absence of m/z 263, 281 and 323 indicates a 1-3 linkage, wher eas the presence of these three ions indicates a 1-6 linkage. La stly, the presence of only m/z 143, 161, 179 and 341 indicates a 1-4 linkage. This study showed that the fragments produced by IRMPD are in fact different from those seen in the CID studi es of Zhu and Cole.147 The fragments produced for each fragmentation method are compared in Table 6-2. First, fo r 1-3 linked disaccharides, the IRMPD spectra contained lower mass fragment ions m/z 101, 119, 131 and 143 that were not present in the CID spectra. For the 1-4 linked disaccharides, the higher mass ions of m/z 263 and 281 present in the CID spectra were not present in the IRMPD sp ectra. For the 1-6 linked disaccharides, the IRMPD produced the lower mass ions m/z 119, 131 and 143 that were not found in the CID spectra and the higher mass ions with m/z 221 and 251 present in the CID spectra were not present in the IRMPD spectra. These results indi cate that, in general, the higher laser powers used for IRMPD allow production of some lower ma ss ions that are not seen with CID, perhaps by photodissociation of higher mass fragments.

PAGE 100

100 The ratios of the relative pe rcent abundances of specific fragment ions were obtained to differentiate the anomeric conf iguration of disaccharides with the same linkage. For these experiments, three wavelengths (9.342, 9.473 and 9.588 m) were chosen for fragmentation. Selecting only three wavelengths allowed for a (l imited) range of wavelengths to be explored while decreasing the experiment time. For thes e wavelengths, the laser power was adjusted to obtain a peak height ratio of approximately 1:1 for m/z 161 to 377 (for nigerose, laminaribiose, maltose and cellobiose) and m/z 179 to 377 (for kojibiose, sophorose, isomaltose and gentiobiose). As has been discussed in earlier chapters, using such a ratio to standardize the laser power allowed for improved day to day repr oducibility compared to solely monitoring the depletion of the precursor ion. Figure 6-7 show s the day to day reprodu cibility at the three wavelengths for the isomaltose precursor and fragment ions. In comparison to the average uncertainty of the precursor ion abundance found for the deprotonated isomaltose ( 15%) the variation for chlorinated isomaltose precursor ion abundances over the three was only 2%, indicating much better dayto-day reproducibility. Similar to the results found with IR MPD of deprotonated disaccharides, m/z 161 and 179 were the major fragments produced by irradiation of the chlorinated disaccharides. The spectra obtained for each disaccharide at each of the thr ee wavelengths are shown in Figure 6-8. Except for the 1-3 linked disaccharides, both the linkage and anomeric configuration can be determined for all the disaccharides. For the 1-3 linked disacc harides, only the linkage could be determined. The appearance of m/z 263 as the primary peak for both kojibiose and sophorose at 9.342 m makes this a useful wavelength for distinguishing the linkage and anomeric configuration of the 1-2 linked disaccharides. For nigerose and lami naribiose, the fragment ions produced were mainly m/z 161 and 179, with the abundance of m/z 161 approximately 1.3 times that of m/z 179.

PAGE 101

101 In contrast, the spectra of maltose and cellobiose have major fragments at m/z 161 and 179, but the abundance of m/z 161 is approximately 6 (for maltose) and 8 (for cellobiose) times greater than the abundance of m/z 179. While m/z 161 and 179 are the major peaks in the spectra for gentiobiose and isomaltose, the presence of the fragment at m/z 281 in the spectra facilitates identification of 1-6 linked. The fragmentation patterns and specific ratio s were used in a singl e-blind study where one sample of each disaccharide was analy zed at wavelengths of 9.342, 9.473 and 9.588 m. The decision flow-chart used to identify the unknowns in the blind study is shown in Figure 6-9. Analysis of the spectra shows that specific rati os of fragments can be used to differentiate the anomeric pairs of the disaccharides, as shown in Figure 6-10. The ratio of m/z 263 to m/z 179 fragment ion abundances of the 1-2 linked disaccharides shows differences at 9.342 and 9.473 m, with this ratio greater for kojibiose than for sophorose at 9.342 m and smaller for kojibiose at wavelength 9.473 m. To differentiate the 1-4 linke d anomers, the data show that the ratio of m/z 161/179 can be used. While all three wavelengths gave similar results, with the ratio of cellobiose always being gr eater than that for maltose, 9.342 m gave the greatest difference with the smallest error bars (ratio of 6.5 0.04 for maltose versus 8.1 0.3 for cellobiose). Lastly, the ratio of m/z 161 to 143 is always higher for gentiobiose than for isomaltose. The average value for gentiobiose is 9.6 1.1 vs. 3.1 0.4 for isomaltose. Unfortunately, more work is needed to differentia te the anomers of the 1-3 linked disaccharides. Since the major fragments produced fo r the 1-3 linked disaccharides were m/z 161 and 179 and the error bars for the ratios ove rlap at the three wavelengths, further fragmentation with more power may be needed to distinguish the and -anomers.

PAGE 102

102 Identification of Fragment Ions Because, the disaccharides used in these studies are all composed of two glucose units, it is impossible to distinguish from which of the two glucose monosaccharides, reducing or non-reducing, a fragment was produced without 18O labeling of the sugar at the reducing-end. Figure 6-11 shows the possible frag ment identities for fragments m/z 119,161, 179, 221, 251 and 289 based on 18O-labeling studies done by Hofmeister et al. in the positive ion mode60 and Fang et al. in the negative ion mode.93 For the chlorinated disaccharides, the resulting fragments are produced by the loss of HCl. Fang et al. found that fragments m/z 73, 89, 97, 119, 179, 251, 263 and 281 are produced from the non-reducing e nd. Since the reducing-end monosaccharide can interconvert in solution, there is a mixture of both the and -anomers for the reducing-end monosaccharide. Therefore, being able to use the non-reducing fragments to identify the anomers allows one to be certain of the anomeric configuration being identified. Conclusions These studies of the deprotonated and chlo rinated disaccharides revealed that the fragments produced by IRMPD over the wavelength range of 9.2 to 9.7 m can be used to identify the linkage positions of the glucose m onosaccharides which comprise the disaccharides. Identification of the anomeric conf iguration of the disaccharides is a more difficult task, but it can be achieved for most of the a nomeric pairs by calcula ting ratios of certain fragment ions for the various linkages. The similarity of results for the deprotonate d and chlorinated disaccharides supports the conclusions of Cole and co-w orkers that loss of HCl occurs before fragmentation.145,147 This study also showed that IRMPD can be used in the negative ion mode to determine both the anomeric configuration and linkage of ch lorinated glucose-containing disaccharides.

PAGE 103

103 Future research should include multiple frag mentations (hopefully produced using higher laser powers) to see if depleti on of the higher mass fragments re sults in a greater abundance of the lower masses for the 1-3 linked disaccharides. It may be possible that ratios of these lower masses can be used in the differen tiation of the anomeric configurati on. Therefore, future studies should use an abundance ratio of the precursor ion to a fragment ion at all wavelengths to see how the abundances of the fragment ions change. Also, since all the disa ccharides studied here were composed of two glucose units, the fragme ntation of disaccharides composed of other monosaccharide units such as mannose or allose should be tried.

PAGE 104

104 Power-meter CO2laser Ion optics Pneumaticallyassisted ESI Infinity cell KBrwindow Mirror Mechanical MirrorBruker Apex II FTICR4.7T magnet Power-meter CO2laser Ion optics Pneumaticallyassisted ESI Infinity cell KBrwindow Mirror Mechanical MirrorBruker Apex II FTICR4.7T magnet Figure 6-1. Schematic drawing of the laser/ma ss spectrometer set-up used for the analysis of deprotonated disaccharides. 0 10 20 30 40 50 60 70 9.2019.3299.4739.5889.657 RPW Wavelength ( m)Relative Percent Abundance 0 10 20 30 40 50 60 70 9.2019.3299.4739.5889.657 RPW Wavelength ( m)Relative Percent Abundance Figure 6-2. Relative percent a bundance of the precursor ion (m/z 341) of isomaltose at selected wavelengths. Error bars show the 95% confidence interval of the mean based on data acquired on two separate days.

PAGE 105

105 Figure 6-3. Wavelength-depe ndent fragmentation patterns for the various deprotonated disaccharides. A) Kojibiose. B) Nigerose. C) Laminaribiose. D) Maltose. E) Cellobiose. F) Isomal tose. G) Gentiobiose.

PAGE 106

106 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 9.2019.239.2829.3059.3299.3429.4739.4889.529.5889.657 WavelengthRati o Nigerose Laminaribiose Isomaltose Gentiobiose Wavelength (m)Ratio 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 9.2019.239.2829.3059.3299.3429.4739.4889.529.5889.657 WavelengthRati o Nigerose Laminaribiose Isomaltose Gentiobiose Wavelength (m)Ratio Figure 6-4. Ratio of m/z 161/179 for 1-3 and 1-6 linked disacchari des, showing that this ratio is not optimal for distinguishing the differ ent anomers. Error bars are the 95% confidence interval of the mean for each ratio. The observed overlap of many of the error bars indicates that this ratio alone cannot be used to positively identify the anomers.

PAGE 107

107 9.201-10 0 10 20 30 40 50 60 70 101113115119125143161179221251281309311323341 m/z Day 1 Day 2 9.657-10 0 10 20 30 40 50 101113115119125143161179221251281309311323341 m/z Day 1 Day 2 Am/z m/zRelative Percent Abundance Relative Percent Abundance9.657 m 9.201 m B 9.201-10 0 10 20 30 40 50 60 70 101113115119125143161179221251281309311323341 m/z Day 1 Day 2 9.201-10 0 10 20 30 40 50 60 70 101113115119125143161179221251281309311323341 m/z Day 1 Day 2 9.657-10 0 10 20 30 40 50 101113115119125143161179221251281309311323341 m/z Day 1 Day 2 9.657-10 0 10 20 30 40 50 101113115119125143161179221251281309311323341 m/z Day 1 Day 2 Am/z m/zRelative Percent Abundance Relative Percent Abundance9.657 m 9.201 m B Figure 6-5. Comparison of the fragmentati on patterns of deprotonated isomaltose on two separate days. A) The IRMPD spectru m of isomaltose at wavelength 9.201 m. B) The IRMPD spectrum of isomaltose at wavelength 9.657 m. Notice that the percent abundances of the pr ecursor and fragment ions ar e different, but in general the same fragments are produced day to day.

PAGE 108

108 Table 6-1. Major fragment ions observed for th e chlorinated disaccharides when the precursor ion (m/z 377) was almost depleted by IRMPD at 9.588 m. The solid boxes indicate the presence of a fragment ion of that m/z, whereas the stripped boxes indicate the absence of the fragment. 323 Gentiobiose ( 1-6) Isomaltose ( 1-6) Cellobiose ( 1-4) Maltose ( 1-4) Laminaribiose ( 1-3) Nigerose ( 1-3) Sophorose ( 1-2) Kojibiose( 1-2) 341 281 263 179161 143131119113 101 Disaccharide m/z 323 Gentiobiose ( 1-6) Isomaltose ( 1-6) Cellobiose ( 1-4) Maltose ( 1-4) Laminaribiose ( 1-3) Nigerose ( 1-3) Sophorose ( 1-2) Kojibiose( 1-2) 341 281 263 179161 143131119113 101 Disaccharide m/z Table 6-2. Comparison of the fragments pr oduced by CID and IRMPD for the chlorinated disaccharides. Fragments produced by both IRMPD and CID are coded in orange, by CID only in yellow, by IRMPD only in blue, and those not produced by either CID or IRMPD are coded in white. 1-6 linked 1-4 linked 1-3 linked341 323281263 251221179161143131119113101m/z Linkage 1-6 linked 1-4 linked 1-3 linked341 323281263 251221179161143131119113101m/z Linkage

PAGE 109

109 m/z 350 300 250 200 150 100 Abundance 200 150 100 50 377 341 323 263 179 161 143 131 119 113A m/z 350 300 250 200 150 100 Abundance 200 150 100 50 377 341 323 263 179 161 143 131 119 113 m/z 350 300 250 200 150 100 Abundance 200 150 100 50 377 341 323 263 179 161 143 131 119 113A m/z 350 300 250 200 150 100 Abundance 300 250 200 150 100 50 377 341 323 263 179 161 143 131 119 113 101B m/z 350 300 250 200 150 100 Abundance 300 250 200 150 100 50 377 341 323 263 179 161 143 131 119 113 101 m/z 350 300 250 200 150 100 Abundance 300 250 200 150 100 50 377 341 323 263 179 161 143 131 119 113 101B m/z 350 300 250 200 150 100 Abundance 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 377 341 179 161 143 131 119 113 101 251 *C m/z 350 300 250 200 150 100 Abundance 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 377 341 179 161 143 131 119 113 101 251 * m/z 350 300 250 200 150 100 Abundance 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 377 341 179 161 143 131 119 113 101 251 *C m/z 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 341 179 161 143 131 119 113 101 251 377D m/z 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 341 179 161 143 131 119 113 101 251 377 m/z 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 341 179 161 143 131 119 113 101 251 377D m/z 350 300 250 200 150 100 Abundance 350 300 250 200 150 100 50 377 341 179 161 143E m/z 350 300 250 200 150 100 Abundance 350 300 250 200 150 100 50 377 341 179 161 143 m/z 350 300 250 200 150 100 Abundance 350 300 250 200 150 100 50 377 341 179 161 143 377 341 179 161 143E m/z 350 300 250 200 150 100 Abundance 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 377 179 161 143F m/z 350 300 250 200 150 100 Abundance 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 377 179 161 143 m/z 350 300 250 200 150 100 Abundance 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 377 179 161 143F m/z 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 281 377 341 323 263 179 161 143 131 119 113 101 *G m/z 350 300 250 200 150 100 Abundance 400 350 300 250 200 150 100 50 281 377 341 323 263 179 161 143 131 119 113 101 *G281 377 341 323 263 179 161 143 131 119 113 101 *G m/z 350 300 250 200 150 100 Abundance 80 70 60 50 40 30 20 10 377 341 281 323 263 179 161 143 131 119 113 101 *H m/z 350 300 250 200 150 100 Abundance 80 70 60 50 40 30 20 10 377 341 281 323 263 179 161 143 131 119 113 101 m/z 350 300 250 200 150 100 Abundance 80 70 60 50 40 30 20 10 377 341 281 323 263 179 161 143 131 119 113 101 377 341 281 323 263 179 161 143 131 119 113 101 *H Figure 6-6. Fragmentation spectra for the nearly depleted precursor ion (m/z 377) for the chlorinated disaccharides at 9.588 m. A) Kojibiose. B) Sophorose. C) Maltose. D) Cellobiose. E) Nigerose. F) Laminaribiose. G) Isom altose. H) Gentiobiose.

PAGE 110

110 9.3420 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379WavR P Day 1 Day 2 9.342 m m/z Relative Percent Abundance 9.4730 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379m/zRP A Day 1 Day 2 9.473 m m/z Relative Percent Abundance 9.5880 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379mR P Day 1 Day 2 9.588 m m/z Relative Percent Abundance 9.3420 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379WavR P Day 1 Day 2 9.342 m m/z Relative Percent Abundance 9.3420 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379WavR P Day 1 Day 2 9.342 m m/z Relative Percent Abundance 9.342 m m/z Relative Percent Abundance 9.4730 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379m/zRP A Day 1 Day 2 9.473 m m/z Relative Percent Abundance 9.4730 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379m/zRP A Day 1 Day 2 9.473 m m/z Relative Percent Abundance 9.473 m m/z Relative Percent Abundance 9.5880 5 10 15 20 25 30 357173818389101113119125131143161179251263281323341377379mR P Day 1 Day 2 9.588 m m/z Relative Percent Abundance 9.588 m m/z Relative Percent Abundance Figure 6-7. Infrared multiple phot on dissociation spectra for chlori nated isomaltose obtained at three wavelengths on two different days. Si milar reproducibilities were observed for the other chlorinated disaccharides.

PAGE 111

111 0 10 20 30 40 50 717381838789101113119125131143161179251263281323341377379Sophorose 9.342 9.473 9.588 0 10 20 30 40 717381838789101113119125131143161179251263281323341377379Laminaribiose 9.342 9.473 9.588 0 20 40 60 717381838789101113119125131143161179251263281323341377379Kojibiose 9.342 9.473 9.588 0 10 20 30 40 717381838789101113119125131143161179251263281323341377379Nigerose 9.342 9.473 9.588 0 10 20 30 40 50 717381838789101113119125131143161179251263281323341377379Maltose 9.342 9.473 9.588 0 10 20 30 40 50 717381838789101113119125131143161179251263281323341377379Cellobiose 9.342 9.473 9.588 0 10 20 30 717381838789101113119125131143161179251263281323341377379Isomaltose 9.342 9.473 9.588 0 10 20 30 40 717381838789101113119125131143161179251263281323341377379Gentiobiose 9.342 9.473 9.588 m/z m/zRelative Percent Abundance Kojibiose Sophorose Nigerose Laminaribiose Maltose Cellobiose Isomaltose Gentiobiose 0 10 20 30 40 50 717381838789101113119125131143161179251263281323341377379Sophorose 9.342 9.473 9.588 0 10 20 30 40 717381838789101113119125131143161179251263281323341377379Laminaribiose 9.342 9.473 9.588 0 20 40 60 717381838789101113119125131143161179251263281323341377379Kojibiose 9.342 9.473 9.588 0 10 20 30 40 717381838789101113119125131143161179251263281323341377379Nigerose 9.342 9.473 9.588 0 10 20 30 40 50 717381838789101113119125131143161179251263281323341377379Maltose 9.342 9.473 9.588 0 10 20 30 40 50 717381838789101113119125131143161179251263281323341377379Cellobiose 9.342 9.473 9.588 0 10 20 30 717381838789101113119125131143161179251263281323341377379Isomaltose 9.342 9.473 9.588 0 10 20 30 40 717381838789101113119125131143161179251263281323341377379Gentiobiose 9.342 9.473 9.588 m/z m/zRelative Percent Abundance Kojibiose Sophorose Nigerose Laminaribiose Maltose Cellobiose Isomaltose Gentiobiose Figure 6-8. Average fragment ation spectra for the disaccha rides at 9.342, 9.473 and 9.588 m. The major fragments observed for kojibiose and sophorose are m/z 179, 263 and 323, whereas the major fragments for nigerose, laminaribiose, maltose and cellobiose are m/z 161 and 179. The major fragments observed for isomaltose and gentiobiose are m/z 161, 179 and 281.

PAGE 112

112 1-3 linked 1-4 linked 1-6 linked 1-2 linked 1-3 linked Nigerose Laminaribiose m/z 323? yes no 1-6 linked 1-3 linked 1-4 linked m/z 161 >>179? 1-4 linked m/z 281? yes no yes no > 2 < 2 > 4.35 < 4.35 > 6.75 < 6.75 m/z 263/179 at 9.473 m 1-2 Kojibiose 1-2 Sophorose 1-6 Gentiobiose 1-6 Isomaltosem/z 161/143 at 9.342, 9.473 and 9.588 m m/z 161/179 at 9.473 and 9.588 m 1-4 Maltose 1-4 Cellobiose 1-3 linked 1-4 linked 1-6 linked 1-2 linked 1-3 linked Nigerose Laminaribiose m/z 323? yes no yes no yes no 1-6 linked 1-3 linked 1-4 linked m/z 161 >>179? 1-4 linked m/z 281? yes no yes no yes no yes no yes no > 2 < 2 > 4.35 < 4.35 > 6.75 < 6.75 m/z 263/179 at 9.473 m 1-2 Kojibiose 1-2 Sophorose 1-6 Gentiobiose 1-6 Isomaltosem/z 161/143 at 9.342, 9.473 and 9.588 m m/z 161/179 at 9.473 and 9.588 m 1-4 Maltose 1-4 Cellobiose Figure 6-9. Decision flow chart used to identify disaccharide samples with unknown identities in a single-blind study. Once the linkage is de termined, the anomeric configuration of all the disaccharides (except 1-3 linked) can then be determined from ratios of specific fragment ions.

PAGE 113

113 m/z 263/1790 0.5 1 1.5 2 2.5 3 3.5 9.3429.4739.588 WavelengthRati o Kojibiose Day 1 Kojibiose Day 2 Kojibiose Unknown Sophorose Day 1 Sophorose Day 2 Sophorose Unknown m/z 161/1790 1 2 3 4 5 6 7 8 9 9.342 9.473 9.588 WavlengthRati o Maltose Day 1 Maltose Day 2 Maltose Unknown Cellobiose Day 1 Cellobiose Day 2 Cellobiose Unknown m/z 161/1430 2 4 6 8 10 12 14 16 9.3429.4739.588 Wavelength Rati o Isomaltose Day 1 Isomaltose Day 2 Isomaltose Unknown Gentiobiose Day 1 Gentiobiose Day 2 Gentiobiose Unknown Wavelength ( m)Ratio m/z 263/179A Wavelength ( m)Ratio m/z 161/143C Wavelength ( m)Ratio m/z 161/179B m/z 263/1790 0.5 1 1.5 2 2.5 3 3.5 9.3429.4739.588 WavelengthRati o Kojibiose Day 1 Kojibiose Day 2 Kojibiose Unknown Sophorose Day 1 Sophorose Day 2 Sophorose Unknown m/z 161/1790 1 2 3 4 5 6 7 8 9 9.3429.4739.588 WavlengthRati o Maltose Day 1 Maltose Day 2 Maltose Unknown Cellobiose Day 1 Cellobiose Day 2 Cellobiose Unknown m/z 161/1430 2 4 6 8 10 12 14 16 9.3429.4739.588 Wavelength Rati o Isomaltose Day 1 Isomaltose Day 2 Isomaltose Unknown Gentiobiose Day 1 Gentiobiose Day 2 Gentiobiose Unknown Wavelength ( m)Ratio m/z 263/179A Wavelength ( m)RatioWavelength ( m)Ratio m/z 161/143C Wavelength ( m)RatioWavelength ( m)Ratio m/z 161/179B Figure 6-10. Comparison of various ratios used to determine the anomeric configurations of the chlorinated disaccharides. The unknowns that were identified in the single-blind study by the decision flow chart (Figure 6-9) are also shown. A) The ratio of m/z 263/179 for 1-2 linked kojibiose a nd sophorose. B) The ratio of m/z 161/179 for 1-4 linked maltose and cellobiose. C) The ratio of m/z 161/143 for 1-6 linked isomaltose and gentiobiose.

PAGE 114

114 O OH OH OH O OH OH O HO OH H, OH O OH OH OH O HO O OH OH OH O HO O OH OH OH O OH OH HO H, OH H,OH A B C -HCl -HCl -HCl 179 161 221 119 251 161 179 281O OH OH OH O OH OH O HO OH H, OH O OH OH OH O HO O OH OH OH O HO O OH OH OH O OH OH HO H, OH H,OH A B C -HCl -HCl -HCl 179 161 221 119 251 161 179 281 Figure 6-11. Identification of so me of the fragment ions for th e various linked disaccharides. A) Fragments produced by 1-2 linked disacch arides. B) Fragments produced by 1-4 linked disaccharides. C) Fragments produced by 1-6 linked disaccharides. Based on results from Hofmeister, G. E.; Zhou, Z.; Leary, J. A. J. Am. Chem. Soc. 1991, 113, 5964-5970 and Fang, T. T.; Zi rrolli, J.; Bendiak, B. Carbohydr. Res. 2007, 342, 217-235.60,93

PAGE 115

115 CHAPTER 7 DIFFERENTIATION OF DISACCHARIDES IN THE NEGATIVE ION MODE WITH FREE ELECTRON LASER INFRARED MULTIPLE PHOTON DISSOCIATION Introduction While tunable CO2 lasers have a limited wavelength range of 9.2 to 10.8 m, the output wavelengths produced by free electron lasers (FELs ) span a much wider range. Past infrared multiple photon dissociation (IRMPD) studies with a FEL on both monoand disaccharides involved positive ions formed by the adduction of metal ions.73,74 Polfer et al. found that fragmentation of various glucose-containing lithium-attached disaccharides yielded unique fragmentation patterns that were a function of both mass and wavelength. The work of Jose Valle showed that the IRMPD spectra of various rubidium-attached O-methylated pyranosides were unique for the various monosaccharides.74 The Free Electron Laser for Infrared eXperiments (FELIX) of the FOM-Institute for Plasma Physics Rijnhuizen, The Netherlands is a user facility that began operation in the 1990s.74,127,148 The components of FELIX, discussed in ch apter 3, are housed in the basement of the facility, and the laser beam is directed through evacuated beam tubes into different user stations by pneumatically controlled mirrors. For the studies reported in this chapter, a laboratory-constructed Fourier transform ion cy clotron resonance (FTICR) mass spectrometer, Figure 7-1, was used.148 The FELIX laser beam passes into the back of the FTICR mass spectrometer through a ZnSe window, where a mirror system creates multiple passes resulting in higher laser fluence over the waveleng ths of approximately 5.5 to 11.0 m used for most studies. Procedure All the experiments reported in this chapter we re performed at the F ELIX facility with the help of Drs. Jos Oomens and Jeffrey Steill. First the fragmentation and IRMPD spectra of the

PAGE 116

116 deprotonated precursor ions of several monoand disaccharides were obtained. Next, the IRMPD spectra and fragme ntation patterns of the m/z 179 monosaccharide anion isolated from the fragmentation of deprotonated disaccharides in the negative ion mode were obtained. Also, the fragmentation and IRMPD spectra of various deprotonated monosacchar ides were studied. All ions were irradiated with FELIX for 2.5 to 3.5 seconds with a macropulse repetition rate of 5 Hz (for the disaccharide studies) or 10 Hz (f or the monosaccharide studies). For the dual laser experiments, a fixed-frequency CO2 laser was used to fragment a disaccharide to yield the monosaccharide anion (m/z 179). The disaccharides were irradiated with the CO2 laser with a power of 0.8 W for 0.35 seconds to produce m/z 179 as the predominant fragment ion. All deprotonated disaccharide samples of O18-labeled kojibiose, sophorose, nigerose and laminaribiose were prepared in 80:20 methanol (MeOH):H2O solutions at 1.0 mM disaccharide and 1.0 mM NaOH concentrations. The solution of chlorinated sophorose was prepared in 8:2 MeOH:H2O with 1 mM concentrations of both sophorose and LiCl. All monosaccharide samples were prepared with 1 mM s accharide and 1 mM NaOH in 80:20 MeOH:H2O, with the exception of glucose, which was made at 2 mM glucose to 1.5 mM NaOH (deprotonated studies) or 1.5 mM LiCl (chlorinated studies). Several concentrations of NaOH were tried, but the concentrations that gave the best results were ei ther 1.0 or 1.5 mM. Also, triethylamine was tried as an alternative base, but deprotonation with NaOH provided the best signal for both the disaccharides and the monosaccharides. Results and Discussion Disaccharides The deprotonated disaccharides and their frag mentation patterns were studied first. Fang et al. showed that fragmentation of a cross-ring cleavage product (m/z 221), itself produced by CID from the deprotonated parent ion, was us eful in the differentia tion of disaccharides.93

PAGE 117

117 The research here explored the possibility of using another fragment ion (m/z 179) to differentiate the disaccharides. First studies of the fragmentation of the precursor ion of the 18O-labeled deprotonated disaccharides (m/z 343) were performed to confirm that the anion monosaccharide fragment (m/z 179) is produced solely from th e non-reducing end. This step was necessary to determine if the monosaccharid e anion could be used to differentiate the disaccharides and anomeric configurations For these experiments, two sets of 18O-labeled glucose-containing disaccharide anomer s, kojibiose/sophorose (1-2 linked) and nigerose/laminaribiose (1-3 linked), were used The spectrum of chlorinated disaccharide of sophorose was also obtained to explore the effect of chlorine ion adduction on the fragmentation of disaccharides. The fragmentation patterns of deprotonated 18O-labeled kojibiose, sophorose, nigerose and laminaribiose are shown in Figure 7-2. The fragments produced for the two anomer pairs vary based on the linkage. For example, fragments m/z 101, 119, 143, 223, 265 and 325 are present for the 1-2 linked (kojibiose and sophorose) but are not present for the 1-3 linked disaccharides (nigerose and laminaribiose). Higher mass fragments are present in IRMPD spectra of the 1-2 linked disaccharides, but only the lower masses of m/z 59 and 97 are produced by the IRMPD fragmentation of 1-3 linked disaccharides. The fragmentation patterns show that the relative percent abundances of specific fragments are different for each anomer. For example, the relative percent abundances of m/z 223 and 265 were higher for kojibiose than for sophorose. The fragmentation patterns for both nigerose and laminaribiose contain m/z 59, 62, 69, 71, 89, 97, 113, 115, 161, 163 and 179, but fragmentation of nigerose produces a higher abundance of m/z 97 while fragmentation of laminari biose produced a higher abundance of m/z 59 and 62 over the wavelength range of approximately 9.0 to 11.0 m.

PAGE 118

118 The IRMPD fragmentation spectrum of chlori nated unlabeled sophorose was also obtained over the wavelength range of 5.5 to 11.0 m. The fragmentation of the chlorinated disaccharide, Figure 7-3, shows a pattern very similar to that of the deprotonated disaccharide, Figure 7-2 B, in which m/z 323, 263 (265 for 18O-labeled sophorose) and 179 are the major fragments produced. One difference is that the ion m/z 223 (221 for 16O-sophorose) fragment is produced by irradiation of the deprotonated sophorose, but not the chlorinated sophorose. Comparing the IRMPD spectra of the deprotonated and chlorina ted sophorose parent ions, Figure 7-4, shows that adduction of chlorine produ ces a similar spectrum, but that the overall spectral peaks are sharpened. Monosaccharide Anion Produced from Disaccharides The absence of m/z 181 in the fragmentation patterns fo r both sets of disaccharide anomer pairs confirmed that the fragment ion m/z 179 comes solely from the non-reducing end. Next, the IRMPD spectra of the monosaccharide anion (m/z 179) produced from fragmentation of the various disaccharides were obtained. For this, m/z 179 was produced by sustained off-resonance irradiation collision induced dissociation (SORICID) and then by laser irradiation with a non-tunable CO2 laser. The results of this study we re compared to those obtained from the irradiation of deprotonated gluc ose. As seen in Figure 7-5, the presence of a peak around ~1720 cm-1 for all of the IRMPD spectra is consistent with a characteristic C=O stretch of an aldehyde. Since the IRMPD spectra of the m/z 179 fragment ion produced by CID and laser irradiation both contain the aldehyde stretch, th ese results indicate that the monosaccharide anion opens upon fragmentation. A possibl e schematic for this process is shown in Figure 7-6. To confirm the ring opening, several depr otonated monosaccharides including allose, galactose, glucose and mannose were irradiated w ith FELIX and their spectra were obtained. As seen in Figure 7-7, all monosaccharides produced very broad spectra over the range of 1000 to

PAGE 119

119 1800 cm-1. More importantly, the IRMPD spectra for all the deprotonated monosaccharides contained a peak around 1720 cm-1, thereby indicating the ring ope ning and loss of anomericity. To give further spectroscopic evidence for the ring opening, the IRMPD spectra of O-methylated anomers of glucose were obtaine d. Methylation of the C-1 oxyge n locks the conformation of the anomer and thereby eliminates mutarotation. Wh en comparing these spectra, Figure 7-8, to the spectrum of deprotonated glucose (also seen in Figures 7-5 or 7-7), th e O-methylated compounds lacked the aldehyde peak at 1720 cm-1, thereby confirming the suspicion of the opening of the ring. Since the IRMPD spectra alone could not differentiate the monosaccharides, more information was needed. Irradiation of the monosaccharides over the wavelengths 5.5 to approximately 11 m produced fragment ions with m/z 59, 71, 89, 101, 113, 119,143 and 161. While all of the monosaccharides produced the same fragment ions, the relative percent abundances of the fragments varied depending on the monosaccharide. For example, the percent abundances of m/z 89, 101, 131 and 161 were the largest fr agments of glucose, while fragments with m/z 59, 71 and 101 were the most abundant for allose, galactose and mannose. Since the monosaccharides used for this study were not methyl ated, they existed as a mixture of anomeric configurations, and therefore anomeric configurat ions were not studied. Also, due to limited time, the spectra of these monosaccharides were obtained only once. Conclusions The studies performed with FELIX on monoa nd disaccharides confirmed that at least some of the monosaccharide anions open upon depr otonation resulting in loss of the anomeric configuration. These findings also indicate that the fragment ion m/z 179 may not be the best to use when differentiating disaccharides. These results also demonstrated that the fragmentation

PAGE 120

120 patterns for the various monoand disaccharides in the negative ion mode are unique and depend on both the linkage and anomeric co nfiguration of the saccharide.

PAGE 121

121 Figure 7-1. Schematic of the FTICR set-up at FELIX. Drawing courte sy of Dr. Jos Oomens.

PAGE 122

122 O OH OH OH O HO O OH OH OH 18OH,H -H+ 325 265 223 179 143 119 113 101 89 179 223m/z 179 m/z Wavelength (m)Relative Percent AbundanceA B325 265 223 179 143 119 113 101 89 O OH OH OH O HO O OH OH OH 18OH,H -H+223O OH OH OH O HO O OH OH OH 18OH,H -H+ 325 265 223 179 143 119 113 101 89 179 223m/z 179 m/z Wavelength (m)Relative Percent AbundanceA B325 265 223 179 143 119 113 101 89 O OH OH OH O HO O OH OH OH 18OH,H -H+223O OH OH OH O HO O OH OH HO 18OH,H -H+ Wavelength (m) 161 179 113 115 97 89 71 59 161 179 113 115 97 89 71 59 m/z m/zRelative Percent Abundance 163 163 179 163 179 163C DO OH OH OH O HO O OH OH HO 18OH, H -H+O OH OH OH O HO O OH OH HO 18OH,H -H+ Wavelength (m) 161 179 113 115 97 89 71 59 161 179 113 115 97 89 71 59 m/z m/zRelative Percent Abundance 163 163 179 163 179 163C DO OH OH OH O HO O OH OH HO 18OH, H -H+ Figure 7-2. Infrared multiple photon dissociati on fragmentation patterns over the wavelength range of 5.5 to 11 m for the deprotonated 18O-labeled disaccharides. A) Kojibiose. B) Sophorose. C) Nigerose. D) Laminaribiose.

PAGE 123

123 341 323 263 179 161 143 131 119 101 89m/z Wavelength (m)Relative Percent AbundanceO OH OH OH O OH OH O HO OH H, OH -HCl 341 323 263 179 161 143 131 119 101 89 341 323 263 179 161 143 131 119 101 89m/z Wavelength (m)Relative Percent AbundanceO OH OH OH O OH OH O HO OH H, OH -HCl Figure 7-3. Fragmentation pattern of chlorinated unlabeled sophorose. 0 0.1 0.2 0.3 0.4 0.5 0.6 700800900100011001200130014001500160017001800 cmDissociation y e O-18 Deprotonated Sophorose Chlorinated Sophorose cm-1Dissociation Yield 0 0.1 0.2 0.3 0.4 0.5 0.6 700800900100011001200130014001500160017001800 cmDissociation y e O-18 Deprotonated Sophorose Chlorinated Sophorose 0 0.1 0.2 0.3 0.4 0.5 0.6 700800900100011001200130014001500160017001800 cmDissociation y e O-18 Deprotonated Sophorose Chlorinated Sophorose cm-1Dissociation Yield Figure 7-4. Comparison of the IRMPD spectra for O18-labeled sophorose and O16-chlorinated sophorose.

PAGE 124

124 Spectra of m/z 179 0 0.1 0.2 0.3 0.4 0.5 0.6 100011001200130014001500160017001800 WavenumberDissociation yi e Deprotonated glucose from kojibiose produced by SORI-CID from kojibiose produced by CO2 laser from sophorose produced by CO2 laser cm-1Dissociation yield Spectra of m/z 179 0 0.1 0.2 0.3 0.4 0.5 0.6 100011001200130014001500160017001800 WavenumberDissociation yi e Deprotonated glucose from kojibiose produced by SORI-CID from kojibiose produced by CO2 laser from sophorose produced by CO2 laser cm-1Dissociation yield Figure 7-5. Comparison of the IRMPD spect ra of the monosaccharide anions (m/z 179) produced by deprotonation of glucose a nd by fragmentation of a disaccharide by SORI-CID and CO2 laser irradiation. O H HO H HO H OOH H H OH OH HO H HO H O OH H OH Figure 7-6. Schematic of the possible mechanis m leading to the opening of the monosaccharide anion ring.

PAGE 125

125 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 10001050110011501200125013001350140014501500155016001650170017501800 cmdis s Allose Galactose Glucose Mannose Dissociation yieldcm-1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 10001050110011501200125013001350140014501500155016001650170017501800 cmdis s Allose Galactose Glucose Mannose 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 10001050110011501200125013001350140014501500155016001650170017501800 cmdis s Allose Galactose Glucose Mannose Dissociation yieldcm-1 Dissociation yieldcm-1 Figure 7-7. Infrared multiple photon dissociation spectra of various deprotonated monosaccharides. All the spectra are very broad with a peak around ~ 1720 cm-1, indicating the opening of the ring. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8Dissociati o Alpha-O-methyl-glucopyranoside Beta-O-methyl-glucopyranoside 0 0.1 0.2 0.3 0.4 0.5 0.6 820920102011201220132014201520162017201820 Dissowav e Deprotonated glucose -O-methyl-glucopyranoside -O-methyl-glucopyranoside Deprotonated glucose cm-1 Dissociation yield Dissociation yield 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8Dissociati o Alpha-O-methyl-glucopyranoside Beta-O-methyl-glucopyranoside 0 0.1 0.2 0.3 0.4 0.5 0.6 820920102011201220132014201520162017201820 Dissowav e Deprotonated glucose -O-methyl-glucopyranoside -O-methyl-glucopyranoside Deprotonated glucose cm-1 Dissociation yield Dissociation yield 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8Dissociati o Alpha-O-methyl-glucopyranoside Beta-O-methyl-glucopyranoside 0 0.1 0.2 0.3 0.4 0.5 0.6 820920102011201220132014201520162017201820 Dissowav e Deprotonated glucose -O-methyl-glucopyranoside -O-methyl-glucopyranoside Deprotonated glucose cm-1 Dissociation yield Dissociation yield Figure 7-8. Comparison of the IRMPD spectra fo r anomers of O-methyl-glucopyranoside to the spectrum of deprotonated glucose.

PAGE 126

126 DWavelength (m) m/zRelative Percent Abundance59 71 89 101 143 161 Wavelength (m)C Relative Percent Abundance89 101 113 119 131 161 m/zAWavelength (m) m/zRelative Percent Abundance59 71 89 101 119 143 161 m/zBWavelength (m)Relative Percent Abundance59 71 89 101 113 143 131 113 59 71 143 131 113 119 119 131 161 O H HO OH H H OH H OH OH,H -H+O OH H H HO H OH H OH OH,H -H+O H HO H HO H OH H OH OH,H -H+O H HO H HO OH H H OH OH,H -H+ DWavelength (m) m/zRelative Percent Abundance59 71 89 101 143 161 Wavelength (m)C Relative Percent Abundance89 101 113 119 131 161 m/zAWavelength (m) m/zRelative Percent Abundance59 71 89 101 119 143 161 m/zBWavelength (m)Relative Percent Abundance59 71 89 101 113 143 131 113 59 71 143 131 113 119 119 131 161 O H HO OH H H OH H OH OH,H -H+O OH H H HO H OH H OH OH,H -H+O H HO H HO H OH H OH OH,H -H+O H HO H HO OH H H OH OH,H -H+ Figure 7-9. Comparison of the fragmentation patterns of the de protonated monosaccharides over the wavelength range of 5.5 to 11 m. A) Allose. B) Galactose. C) Glucose. D) Mannose.

PAGE 127

127 CHAPTER 8 CONCLUSIONS AND FUTURE WORK Infrared m ultiple photon dissociation (IRMPD) was used in conjunction with Fourier transform ion cyclotron resonance mass spectr ometry (FTICR-MS) to obtain fragmentation patterns of monoand disaccharide isomers in both the positive and negative ion mode. The fragmentation patterns and IRMPD spectra produc ed with tunable irradiation from both a continuous wave, line-tunable CO2 and free electron laser (FEL) were used to differentiate various lithiated, deprotonated and ch lorinated monoand disaccharides. The major benefit demonstrated by this res earch was that an af fordable and accessible line-tunable CO2 laser can be used for the differentiati on of isomers. The fragmentation patterns produced over the wavelength range from 9.2 to 9.7 m for lithiated monoand disaccharides were used to identify and differentiate both th eir linkages and anomeric configurations. Along with showing that the output wavelength of fixed frequency CO2 lasers (10.6 m) is not at all optimal for differentiation of isomers, this rese arch also demonstrated the benefits of using multiple wavelengths from a tunable CO2 laser. The first project of this dissertation showed that CO2 laser irradiation of O-methyl-glucoand galactopyranosides produced unique fragmentation patterns over the wavelength range of 9.2 to 9.7 m. The relative percent abundance of fragment m/z 169 could be used to distinguish the glucopyranosides from the galactopyranosides. Furthermore, ratios of the relative percent abundance of specific fragments (m/z 109/127 for the glucopyranosides and m/z 169/151 for the galactopyranosides) were used to differentiate the anom eric configuration of monosaccharide isomers. In both cases, the rati os of the specified fragment ions for the -anomers were larger than for the -anomers. A single-blind study confirmed that the isomers could be identified based on fragment abundances in conjunction with ra tios of the relative

PAGE 128

128 percent abundances of specific key ions. In this project, a method for differentiation that could be useful to other researchers was developed. In a second project, irradiati on of lithiated disaccharide isomers with a line-tunable CO2 laser over the wavelengt h range of 9.2 to 9.7 m produced fragmentation si milar to that obtained with a more expensive and complex FEL. The fragmentation patterns seen from 9.2 to 9.7 m could be used to differentiate the linkage, while ratios of specific ions were used to determine the anomeric configurations. Fragmenting the precursor ion (m/z 349) with a laser to produce a 1:2 peak height of precurs or ion to fragment ion (m/z 229 for 1-2 linked, m/z 169 for 1-3 and 1-6 linked and m/z 187 for 1-4 linked disaccharides) at wavelengths 9.342, 9.473 and 9.588 m allowed the eight isomers to be differentiated. Comparing ratios of the re lative abundances of other key fragments (m/z 187/229 for 1-2 linked, m/z 169/187 for 1-3 and 1-6 linked and m/z 229/289 for 1-4 linked disaccharides) gave a method to differentiate the anomeric configuration. Specifically, the ratios calculated for fragments from the -anomers were larger than the ratios obtained for the -anomers for all except the 1-2 linked disaccharides. The study of deprotonated and chlorinated dis accharides fragmented with a line-tunable CO2 laser demonstrated that differentiation of th e disaccharides in the negative mode is more difficult than that involving lithiated disacchar ides in the positive ion mode. The fragments obtained from the dissociation of the de protonated disaccharides were primarily m/z 161 and m/z 179 and were similar to those obtained for th e chlorinated species. While the linkage for each deprotonated disaccharide could be determined based on the relative percent abundances of the fragment ions, the anomeric c onfigurations of the deprotonated ions were not determined in this study. Fragmentation patterns were used to determine the linkage of the eight chlorinated disaccharides studied. Also, comparison of specific ratios of the relative percent abundances of

PAGE 129

129 specific fragment ions (m/z 263/179 for 1-2 linked, m/z 161/179 for 1-4 linked and m/z 161/143 for 1-6 linked) for the chlorinated disaccharides gave a method of discriminating the various anomers. Lastly, study of deprotonated disaccharides with a FEL gave spectroscopic evidence for opening of the monosaccharide anion. An 18O-labeling study of the fr agmentation of 1-2 and 1-3 linked deprotonated disaccharides conf irmed that the monosaccharide anion (m/z 179) produced over wavelengths 5.5 to 11.0 m contained solely the non-reducing monosaccharide. Multiple fragmentation methods, including sustaine d off-resonance irradiation collision-induced dissociation (SORI-CID) and laser irradiation by a fixed-frequency CO2 laser were used to fragment various disaccharides and isolate the m/z 179 anion. The IRMPD spectra for the isolated m/z 179 fragment ion revealed a band co rresponding to C=O al dehyde stretch around 1720 cm-1. This gave strong evidence for opening of the non-reducing monosaccharide ring and subsequent loss of anomericity of the monosacchar ide anion produced from the fragmentation of the glucose-containing disaccharides. Furthermore, the IRMPD spectra of several other deprotonated monosaccharides also contained th is peak. Opening of the monosaccharide ring and thereby loss of the anomeric configuration confirms the need for more information, such as fragmentation patterns, to differentiate th e monosaccharide anomers that compose larger oligosaccharides when using the deprotonate d forms of these compounds for analysis. Only glucose-containing disaccharides were used in the research discussed in this dissertation, therefore future studies should examine the fragmentation patterns of other hexosecontaining disaccharides. It may be possible, since the monosaccharides studied in this dissertation gave unique fragment ation patterns, that disaccharides composed of different monosaccharide units could also give unique frag mentation patterns that could be used for

PAGE 130

130 isomeric differentiation. The patterns of these smaller saccharides could then be used to differentiate the linkage and monosaccharide units within larger oligoand polysaccharides. Since the largest saccharide units studied here were the disaccharides, larger saccharide units such as trisaccharides composed of various monosaccharide units should be studied. A major limitation of this research was that only pure samples of each disaccharide were used. Since in nature mixtures of anomers are often present simultaneously in solution, a method that can determine the percentage of each anomer within a mixture of saccharides should be developed. Lastly, an instrumental set-up that utilizes an optical parametric oscillator (OPO) in conjunction with a FTICR mass spectrometer may be useful for studying various hexoses. The use of an OPO allows access to the 2.28-4.67 m wavelength range, which corresponds to the O-H stretch region. The various O-H stretches could be helpfu l in differentiating anomers of monoand disaccharide

PAGE 131

131 LIST OF REFERENCES (1) Clowers, B. H.; Dwivedi, P.; Steine r, W. E.; Hill, H. H.; Bendiak, B. J. Am. Soc. Mass Spectrom. 2005, 16, 660-669. (2) Dwek, R. A. Chem. Rev. 1996, 96, 683-720. (3) Lis, H.; Sharon, N. Chem. Rev. 1998, 98, 637-674. (4) Dyer, J. W., I.S.; Palejwala, A. ; Ellis, A.; Shirazi-Beechey, S.P. Am. J. Physiol Gastrointest. Liver Physiol. 2002, 282, G214-G248. (5) Varki, A. Glycobiology 1993, 3, 97-130. (6) Park, Y.; Lebrilla, C. B. Mass Spectrom. Rev. 2005, 24, 232-264. (7) Solomons, G.; Fryhle, C. Organic Chemistry, 7 ed.; John Wiley & Soncs, Inc.: New York, NY, 2002. (8) El Khadem, H. S. Carbohydrate Chemistry: Monosaccharides and Their Oligomers; Academic Press, Inc.: San Deigo, 1988. (9) Lindhorst, T. K. Essentials of Carbohydrate Chemistry and Biochemistry, 2nd ed.; WileyVCH: Weinheim, 2003. (10) Englyst, K. N.; Liu, S.; Englyst, H. N. Eur. J. Clin. Nut. 2007, 61, S19-S39. (11) Voet, D.; Voet, J. G. Biochemistry, 2 ed.; John Wiley & Sons: New York, 1995. (12) Apweiler, R.; Hermjakob, H.; Sharon, N. Biochim. Biophys. Acta. 1999, 1473, 4-8. (13) Rassi, Z. E., Ed. Carbohydrate analysis: high perfor mance liquid chromatography and capillary electrophoresis; Elsevier Science B. V.: Amsterdam, 1995. (14) Chaplin, M. F.; Kennedy, J. F., Eds. Carbohydrate Analysis: A practical approach; Oxford University Press Inc.: New York, 1994. (15) Hakomori, S. Curr. Opin. Hematol. Rev. 2003, 10, 16-24. (16) Ramasamy, R.; Reese, R. T. J. Immunol. 1985, 134, 1952-1955. (17) Zaia, J. Mass Spectrom Rev 2004, 23, 161-227. (18) Amon, S.; Zamfir, A. D.; Rizzi, A. Electrophoresis 2008, 29, 2485-2507. (19) Coppin, A.; Maes, E.; Flahaut, C.; Coddeville, B.; Strecker, G. Eur. J. Biochem. 1999, 266, 370-382.

PAGE 132

132 (20) Martensson, S.; Levery, S. B.; Fang, T. T.; Bendiak, B. Eur. J. Biochem. 1998, 258, 603622. (21) Rassi, Z. E., Ed. Carbohydrate analysis by modern chromatography and electrophoresis; Elsevier Science B.V.: Amsterdam, 2002. (22) Bergstrom, N.; Nair, G. B.; Weintraub, A.; Jansson, P.-E. Carbohydr Res. 2002, 337, 813817. (23) Zhang, Y.-H. P.; Lynd, L. R. Anal. Biochem. 2003, 322, 225-232. (24) Ciucanu, I.; Kerek, F. Carbohydr. Res. 1984, 131, 209-217. (25) Bourne, E. J.; Stacey, M.; Tatlow, J. C.; Tedder, J. M. J. Chem. Soc. 1949, 2976-2979. (26) Kang, P.; Mechref, Y.; Klouckova, I.; Novotny, M. V. Rapid Commun. Mass Spectrom. 2005, 19, 3421-3428. (27) Hakomori, S. I. J. Biochem. 1964, 55, 205. (28) Armstrong, G. S.; Bendiak, B. J. Magn. Reson. 2006, 181, 79-88. (29) Duus, J. O.; Gotfredsen, C. H.; Bock, K. Chem. Rev. 2000, 100, 4589-4614. (30) Bendiak, B. Carbohydr Res. 1999, 315, 206-221. (31) Vliegenthart, J. F. G.; Woods, R. J., Eds. NMR Spectroscopy and Computer Modeling of Carbohydrates: Recent Advances; Oxford University Press: Washington, DC, 2006. (32) Fang, T. T.; Bendiak, B. J. Am. Chem. Soc. 2007, 129, 9721-9736. (33) Yamashita, M.; Fenn, J. B. J. Phys Chem. 1984, 88, 4451-4459. (34) Carr, S. A.; Reinhold, V. N.; Green, B. N.; Hass, J. R. Biomed Mass Spectrom 1985, 12, 288-295. (35) Egge, H.; Peter-Katalinic, J. Mass Spectrom Rev. 1987, 6, 331-393. (36) Dell, A. Adv. Carbohydr. Chem. Biochem. 1987, 45, 19-72. (37) Stahl, B.; Steup, M.; Karas, M.; Hillenkamp, F. Anal. Chem. 1991, 63, 1463-1466. (38) Duffin, K.; Welply, J. K.; Huang, E.; Henion, J. D. Anal. Chem. 1992, 64, 1440-1448. (39) Fura, A.; Leary, J. A. Anal. Chem. 1993, 65, 2805-2811. (40) Barber, M.; Bordoli, R. S.; Sedgwick, R. D.; Tyler, A. N. J. Chem. Soc. Chem. Comm. 1981, 325-327.

PAGE 133

133 (41) Carroll, J. A.; Lebrilla, C. B. Org. Mass Spectrom. 1992, 27, 639-643. (42) Carroll, J. A.; Ngoka, L.; Beggs, C. G.; Lebrilla, C. B. Anal. Chem. 1993, 65, 1582-1587. (43) Karas, M.; Bachmann, D.; Hillenkamp, F. Anal. Chem. 1985, 57, 2935-2939. (44) Zhang, H.; Brokman, S. M.; Fang, N.; Pohl, N.; Nicola, L.; Yeung, E. S. Rapid Commun. Mass Spectrom. 2008, 22, 1579-1586. (45) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (46) Cech, N. B.;Enke, C.G. Mass Spectrom. Rev. 2001, 20, 362-387. (47) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Mass Spectrom. Rev. 1990, 9, 37-70. (48) Mirgorodskaya, E.; Roepstorff, P.; Zubarev, R. A. Anal. Chem. 1999, 71, 4431-4436. (49) Cody, R. B.; Burnier, R. C.; Freiser, B. S. Anal. Chem. 1982, 54, 96-101. (50) Gauthier, J. W.; Trautman, T. R.; Jacobson, D. B. Anal. Chim. Acta. 1991, 246, 211-225. (51) Mueller, D. R.; Domon, B. M.; Bl um, W.; Raschdorf, F.; Richter, W. J. Biomed. Mass Spectrom 1988, 15, 441-446. (52) Laine, R. A.; Yoon, E.; Mahier, T. J.; Abbas, S. A.; de Lappe, B.; Jain, R.; Matta, K. Biol. Mass Spectrom. 1991, 20, 505-514. (53) Woodin, R. L.; Bomse, D. S.; Beauchamp, J. L. J. Am. Chem. Soc. 1978, 100, 3248-3250. (54) Black, J. G.; Yablonovitch, E.; Bloembergen, N.; Mukamel, S. Physical Review Letters 1977, 38, 1131-1134. (55) Oomens, J.; Tielens, A. G. G. M.; Sartakov, B. G.; von Helden, G.; Meijer, G. Astrophys J. 2003, 591, 968-985. (56) Adamson, J. T.; Hakansson, K. Anal. Chem. 2007, 79, 2901-2910. (57) Asam, M. R.; Glish, G. L. J. Am. Soc. Mass Spectrom. 1997, 8, 987-995. (58) Cancilla, M. T.; Penn, S. G.; Carroll, J. A.; Lebrilla, C. B. J. Am. Chem. Soc. 1996, 118, 6736-6745. (59) Firdoussi, A. E.; Lafitte, M.; Tortajada, J.; Kone, O.; Salpin, J.Y. J. Mass Spectrom. 2007, 42, 999-1011. (60) Hofmeister, G. E.; Zhou, Z.; Leary, J. A. J. Am. Chem. Soc. 1991, 113, 5964-5970.

PAGE 134

134 (61) Cole, R. B.; Zhu., J. H. Rapid Commun. Mass Spectrom. 1999, 13, 607-611. (62) Xie, Y.; Lebrilla, C. B. Anal. Chem. 2003, 75, 1590-1598. (63) March, R. E. J. Mass Spectrom. 1997, 32, 351-369. (64) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. 1998, 17, 1-35. (65) Jasinski, J. M.; Rosenfeld, R. N.; Meyer, F. K.; Brauman, J. I. J. Am. Chem. Soc. 1982, 104, 652-658. (66) Gaumann, T.; Riveros, J. M.; Zhu, Z. Q. Helv. Chim. Acta 1990, 73, 1215-1218. (67) Shin, S. K.; Beauchamp, J. L. J. Am. Chem. Soc. 1990, 112, 2057-2066. (68) Shin, S. K.; Beauchamp, J. L. J. Molec. Struct. 1991, 249, 1-9. (69) Peiris, D. M.; Riveros, J. M.; Eyler, J. R. Int. J. Mass Spectrom. Ion Process 1996, 159, 169-183. (70) Fukui, K.; Takada, Y.; Sumiyoshi, T.; Imai, T.; Takahashi, K. J. Phys Chem. B. 2006, 110, 16111-16116. (71) Prazeres, R.; Glotin, F.; Insa, C.; Jarosynski, D. A.; Ortega, J. M. Eur. Phys J. D. 1998, 3, 87-93. (72) Prazeres, R.; Glotin, F.; Orte ga, J. M., Tsukuba, Japan 2003; 83-87. (73) Polfer, N. C.; Valle, J. J.; Moore, D. T.; Oomens, J.; Eyler, J. R.; Bendiak, B. Anal. Chem. 2006, 78, 670-679. (74) Valle, J. J. Ph.D., University of Florida, Gainesville, 2005. (75) Lawrence, E. O.; Edlefsen, N. E. Science 1930, 72, 376. (76) Lawrence, E. O.; Livingston, M. S. Phys. Rev. 1932, 40, 19. (77) Sommer, H.; Thomas, H. A.; Hipple, J. A. Phys. Rev. 1949, 76, 1877. (78) Amster, I. J. J. Mass Spectrom. 1996, 31, 1325-1337. (79) Comisarow, M. B.; Marshall, A. G. Chem. Phys. Lett. 1974, 25, 282. (80) Marshall, A. G.; Comisarow, M. B.; Parisod, G. J. Chem. Phys. 1979, 71, 4434. (81) Jackson, G. S.; Hendrickson, C. L.; Reinhold, B. B.; Marshall, A. G. Int. J. Mass Spectrom. Ion Process. 1997, 165/166, 327-338. (82) Dehmelt, H. Rev. Mod. Phys. 1990, 1990, 525-530.

PAGE 135

135 (83) Marshall, A. G. Int. J. Mass Spectrom. 2000, 200, 331-356. (84) Marshall, A. G.; Hendrickson, C. L. Int. J. Mass Spectrom. 2002, 215, 59-75. (85) Guan, S; Marshall, A. Int. J. Mass Spectrom. Ion Processes 1996, 157/158, 5-37. (86) Busch, K. Spectrosc. 2004, 19, 35-38. (87) Ledford, E. B.; Rempel, D.; Gross, M. L. Anal. Chem. 2005, 1984, 2744-2748. (88) Zhang, L.; Rempel, D.; Pramanik, B. N.; Gross, M. L. Mass Spectrom. Rev. 2005, 24, 286309. (89) Burton, R. D.; Matuszak, K. P.; Watson, C. H.; Eyler, J. R. J. Am. Soc. Mass Spectrom. 1999, 10, 1291-1297. (90) Hannis, J. C.; Muddlman, D. C. J. Am. Soc. Mass Spectrom 2000, 11, 876-883. (91) Taylor, P. K.; Amster, I. J. Int. J. Mass Spectrom. 2003, 222, 351-361. (92) Garozzo, D.; Impallomeni, G.; Spina, E.; Green, B. N.; Hutton, T. Carbohydr. Res. 1991, 221, 253-257. (93) Fang, T. T.; Zirrolli, J.; Bendiak, B. Carbohydr. Res. 2007, 342, 217-235. (94) Stefan, S. E.; Eyler, J. R. Anal. Chem. 2009, 81, 1224-1227. (95) Eyler, J. R. Mass Spectrom. Rev. 2009, 28, 448-467. (96) Polfer, N. C.; Oomens, J. Mass Spectrom. Rev. 2009, 28, 468-494. (97) Ijames, C. F.; Wilkins, C. L. Anal. Chem. 1990, 62, 1295-1299. (98) Zubarev, R. A. Curr. Opin. Chem. Biotechnol. 2004, 15, 12-16. (99) Zubarev, R. A.; Haselmann, K. F.; Budnik, B.; Kjeldsen, F.; Jensen, F. Eur. J. Mass Spectrom. 2002, 8, 337-349. (100) Gorshkov, M. V.; Guan, S. H.; Marshall, A. G. Int. J. Mass Spectrom. Ion Process. 1993, 128, 47-60. (101) Dunbar, R. C.; McMahon, T. B. Science 1998, 279, 194-197. (102) Peiris, D. M.; Cheeseman, M. A.; Ramanathan, R.; Eyler, J. R. J. Phys Chem. 1993, 97, 7839-7843. (103) Senko, M. W.; Speir, J. P. Anal. Chem. 1994, 66, 2801-2808. (104) Carlin, T. J.; Freiser, B. S. Anal. Chem. 1983, 55, 571-574.

PAGE 136

136 (105) Laskin, J.; Futrell, J. H. Mass Spectrom. Rev. 2003, 22, 158-181. (106) Laskin, J.; Futrell, J. H. Mass Spectrom. Rev. 2005, 24, 135-167. (107) Oomens, J.; Sartakov, B.G.; Meijer, G.; von Helden, G. Int. J. Mass Spectrom. 2006, 254, 1-19. (108) Marshall, A. G.; Hendrickson, C. L.; Emme tt, M. R.; Rodgers, R. P.; Blakney, G. T.; Nilsson, C. L. Eur. J. Mass Spectrom. 2007, 13, 57-59. (109) Bomse, D. S.; Beauchamp, J. L., Eds. Multiphoton dissociation of molecules with low power CW infrared lasers.; Springer-Verlag: Berlin-Heidelberg, 1978. (110) Woodin, R. L.; Bomse, D. S.; Beauchamp, J. L. Chem. Biochem. Appl. Lasers 1979, 4, 355-388. (111) Woodin, R. L.; Bomse, D. S.; Beauchamp, J. L. Chem. Phys. Lett. 1979, 63, 630-636. (112) Dole, M.; Mack, L. L.; Hines, R. I.; Mobley, R. C.; Ferguson, L. D.; Alice, M. B. J. Chem. Phys. 1968, 49, 2240-2249. (113) Tanaka, K.; Ido, Y.; Akita, S.; Yoshida, T. Iyo Masu Kenkyuki Koenshu 1987, 12, 219-222. (114) Lehmann, K. K.; Scoles, G.; Pate, B. H. Annu. Rev. Phys. Chem. 1994, 45, 241-274. (115) Grant, E. R.; Schulz, P. A.; Sudbo, A. S.; Shen, Y. R.; Lee, Y. T. Phys. Rev. Lett. 1978, 40, 115-118. (116) Bagratashvili, V. N.; Letokhov, V. S.; Makarov, A. A.; Ryabov, E. A. Mulitple Photon Infrared Laser Photophysics and Photochemistry; Hardwood Academic Publishers: Amsterdam, 1985. (117) Oomens, J.; Meijer, G.; von Helden, G. J. Phys. Chem. A 2001, 105, 8302-8309. (118) Oomens, J.; van Roji, A. J. A.; Meijer, G.; Von Helden, G. Astrophys. J. 2000, 542, 404410. (119) von Helden, G.; Holleman, I.; Knippels, G. M. H.; van der Meer, A. F. G.; Meijer, G. Phys. Rev. Lett. 1997, 79. (120) von Helden, G.; Holleman, I.; Meijer, G.; Sartakov, B. G. Opt. Express 1999, 4, 46-52. (121) Dunbar, R. C. J. Am. Chem. Soc. 1971, 93, 4354-4358. (122) Atkins, P.; de Paula, J. Physical Chemistry, 7th ed.; W. H. Freem an and Company: New York, 2002. (123) Patel, C. K. N. Phys. Rev. 1964, 136, A1187.

PAGE 137

137 (124) Colson, W. B.; Sessler, A. M. Annu. Rev. Nucl. Part. Sci. 1985, 35, 25-54. (125) Colson, W. B.; Johnson, E. D.; Kelly, M. J.; Schwettman, H. A. Phys. Today 2002, 55, 3541. (126) O'Shea, P. G.; Freuden, H. P. Science 2001, 292, 1853-1858. (127) Oepts, D.; van der Meer, A. F. G.; van Amersfoort, P. W. Infrared Phys. Technol. 1995, 36, 297-308. (128) Gaucher, S. P.; Leary, J. A. Anal. Chem. 1998, 70, 3009-3014. (129) Gaucher, S. P.; Leary, J. A. J. Am. Soc. Mass Spectrom. 1999, 10, 269. (130) Smith, G.; Leary, J. A. J. Am. Chem. Soc. 1998, 120, 13046-13056. (131) Desaire, H.; Leary, J. A. Anal. Chem. 1999, 71, 1997-2002. (132) Smith, G.; Leary, J. A. J. Am. Soc. Mass Spectrom. 1996, 7, 953-957. (133) Madhusudanan, K. P. J. Mass Spectrom. 2006, 41, 1096-1104. (134) Salpin, J.-Y.; Tortajada, J. J. Phys. Chem. A 2003, 107, 2943-2953. (135) March, R. E.; Stadey, C. J. Rapid Commun Mass Spectrom. 2005, 19, 805-812. (136) Ikonomou, M. G.; Kebarle, P. J. J. Am. Soc. Mass Spectrom. 1994, 5, 791-799. (137) Wigger, M.; Nawrocki, J. P.; Watson, C. H.; Eyler, J. R.; Benner, S. A. Rapid Commun. Mass Spectrom. 1997, 11, 1749-1752. (138) Wu, S.; Zhang, K.; Kaiser, N. K.; Bruce, J. E.; Prior, D. C.; Anderson, G. A. J. Am. Soc. Mass Spectrom. 2006, 17, 772-779. (139) Caravatti, P.; Allemann, M. Org. Mass Spectrom. 1991, 26, 514-518. (140) Urdan, T. C. Statistics in Plain English, 2 ed.; Lawrence Erlbaum Associates, Inc. : Mahwah, 2005. (141) Zhou, Z.; Ogden, S.; Leary, J. A. J. Org. Chem. 1990, 55, 5444-5446. (142) Zhang, J.; Schubothe, K.; Li, B.; Russell, S.; Lebrilla, C. B. Anal. Chem. 2005, 77, 208214. (143) Daikoku, S.; Ako, T.; Kurimoto, A.; Kanie, O. J. Mass Spectrom. 2007, 42, 417-423. (144) Konig, S.; Leary, J. A. J. Am. Soc. Mass Spectrom. 1998, 9, 1125-1134. (145) Jiang, Y.; Cole, R. B. J. Am. Soc. Mass Spectrom. 2005, 16, 60-70.

PAGE 138

138 (146) Prome, D.; Prome, J.-C.; Puze, G.; Aurelle, H. Carbohydr. Res. 1985, 140, 121-129. (147) Zhu, J.; Cole, R. B. J. Am. Soc. Mass Spectrom. 2001, 12, 1193-1204. (148) Valle, J. J.; Eyler, J. R.; Oomens, J.; Moor e, D. T.; van der Meer, A. F. G.; Von Helden, G.; Meijer, G.; Hendrickson, C. L.; Marshall, A. G.; Blakney, G. T. Rev. Sci. Instrum. 2005, 76, 023103.

PAGE 139

139 BIOGRAPHICAL SKETCH Sarah Elizabeth Stefan was born in 1983 to R obert and Elizabeth Stefan. She grew up in Plym outh, Massachusetts, with he r parents and four siblings. Sarah attended Plymouth public schools for elementary through high school. She gr aduated in the top five of her high school class in May 2001. She then attended Wheaton Coll ege, a small liberal arts college in Norton, Massachusetts. Under the dire ction of Dr. Laura Muller, she worked on her undergraduate honors thesis entitled Analysis of Fingerprint Residue via Infrared Microscopy. In May 2005, she graduated summa cume laude and with chemis try departmental honors, receiving a Bachelor of Arts degree in chemistry with a minor in Amer ican politics. After gr aduation, Sarah moved to Gainesville, Florida, to begin her graduate studies in analytical chemistry at the University of Florida. She then joined the group of Dr. John Eyler and began her work using infrared multiple photon dissociation and Fourier tr ansform ion cyclotron resona nce mass spectrometry in the differentiation of carbohydrates. She received her Doctor of Philosophy from the University of Florida in the spring of 2009.