The Use of Infrared Multiple Photon Dissociation and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry to Study the Thermochemical Dynamics of Dephosphorylation and the Spectral and Density Functional Theory Determined Structural Characteristics of Monosaccharide Isomers

Material Information

The Use of Infrared Multiple Photon Dissociation and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry to Study the Thermochemical Dynamics of Dephosphorylation and the Spectral and Density Functional Theory Determined Structural Characteristics of Monosaccharide Isomers
Pearson III, Wright
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (191 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Committee Chair:
Eyler, John R.
Committee Members:
Ohrn, Nils Y.
Yost, Richard A.
Hirata, So
Denslow, Nancy D.
Graduation Date:


Subjects / Keywords:
Ions ( jstor )
Kinetics ( jstor )
Laser beams ( jstor )
Laser power ( jstor )
Lasers ( jstor )
Mass spectroscopy ( jstor )
Molecules ( jstor )
Photons ( jstor )
Signals ( jstor )
Wavelengths ( jstor )
Chemistry -- Dissertations, Academic -- UF
arrhenius, co2, dft, energy, esi, fticrms, irmpd, kinetics, modeling, opo, phosphopeptide, rex, saccharide, spectroscopy, thermochemical
City of Miami ( local )
Electronic Thesis or Dissertation
born-digital ( sobekcm )
Chemistry thesis, Ph.D.


In recent years, there has been an explosion of interest in the fields of saccharide and protein chemistry. Though much of the initial work has been conducted in solution, the desire to avoid complicated liquid-phase interactions has resulted in the promotion of gas-phase experiments. Capable of multiple stages of experimentation (MSn) and long ion retention times while minimizing extraneous ion-ion/ion-molecule reactions, Fourier transform ion cyclotron resonance mass spectrometry (FTICR MS) has proven to be a powerful instrument in these studies. Of particular importance is the addition of techniques, like infrared multiple photon dissociation (IRMPD), that promote ion dissociation/fragmentation for structural and thermochemical information. The current research employed two 4.7T FTICR mass spectrometers with different types of IR laser systems to examine the possibility of and the reasons for differentiation of saccharide isomers and the kinetic and energetic picture of dephosphorylation through multiple photon dissociation processes. A LINOS OS-4000 continuous wave optical parametric oscillator (OPO) laser was purchased and the laser beam transfer optics were designed, positioned (within a purge box), and focused on the ions in an ICR cell. Photon (vibrationally resonant) induced dissociation or action spectra, spanning the C-H-and O-H stretch regions, were then acquired for four rubidium cation-bound glycosides. Comparison of the experimental and Density Functional Theory calculated spectra revealed a network of intermolecular hydrogen bonding and rubidium attachments that provided the means to differentiate D-glucoside and D-galactoside anomers in the O-H stretching region of the infrared spectrum. In the second project, two cw-CO2 lasers were also set up for gas-phase experiments. Dephosphorylation rate constants (at varying laser powers) were obtained for both positively and negatively charged phosphopeptide ions at three different wavelengths. In these experiments, the dissociation rate constants favored the negatively charged species at all wavelengths (a result of lower activation energies) as the overall Arrhenius activation energies were independent of vibrational mode intensities. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
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.
Thesis (Ph.D.)--University of Florida, 2009.
Adviser: Eyler, John R.
Electronic Access:
Statement of Responsibility:
by Wright Pearson III.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Pearson III, Wright. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
Resource Identifier:
489094586 ( OCLC )
LD1780 2009 ( lcc )


This item has the following downloads:

Full Text




2 2009 Wright L Pearson III


3 To God my wife, family, colleagues, and friends who encouraged me to stay on the path


4 A CKNOWLEDGEMENTS I would like to thank my wife for her constant encourageme nt, belief, and love through my graduate school years and m y mom who has aided me financially and morally with love and concern. I would like to thank my research advisor Prof essor John Eyler who with insight, patience, and kindness ha s encouraged me to continue and provide d the mentoring and the skills that enabled me to become a competent scientist I would also like to thank Dr. Kath ryn Williams for her friendship and ability to see and develop in me the qualities of a scientist trusting not only my insight in to the fields of physics and chemistry but the humanities as well. I would like to acknowledge my committee members Professor Richard Yost, Professor Nils Ohrn, Professor So Hirata, Professor Nancy Denslow, Professor Adrian Roitberg, and Pro fessor Christopher Batich for their support and Dr. Brad Bendiak and Dr David Powell for the saccharide samples and for the use of the FTICR MS in the mass spectrometry ser v i ces lab ( even to early hours of the morning ), respectively Thanks also go out to t he Eyler group and many dear friends in the d epartment an d the greater Gainesville area: Dr. Alex Aksenov, Boris Kr y stal, Dr. Erkan Kose, Dr. Nick Polfer, Powell Wheeler, Joe Diaz, Aaron Beatty, Mark Griffi th e t al who have encouraged me greatly an d have my sincerest wi shes, with a special word of thanks to Dr. Cesar Contreras who aided me in putting together the current laser systems as well as a few experiments these last years And to Jesus Christ through all my doubt, questions, and unfaithful ness he has proven to be faithful and true.


5 T ABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................ ................................ ................................ ............. 4 TABLE OF CONTENTS ................................ ................................ ................................ ................. 5 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 13 CHPATER 1 INTRODUCTIO N ................................ ................................ ................................ .................. 15 Saccharides ................................ ................................ ................................ ............................. 16 Monosaccharides ................................ ................................ ................................ ............. 17 Disaccharides ................................ ................................ ................................ ................... 20 Saccharide Structure & Differentiation ................................ ................................ ........... 21 Peptides ................................ ................................ ................................ ................................ ... 22 Amino Acids ................................ ................................ ................................ .................... 23 Polypeptides ................................ ................................ ................................ .................... 25 Unimolecular Dissociation of Phosphate Groups ................................ ............................ 27 Overview ................................ ................................ ................................ ................................ 28 2 FOURIER TRANSFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY (FTICR MS) ................................ ................................ ........................... 39 History and Fundamentals ................................ ................................ ................................ ...... 39 Ion Cyclotron Radius ................................ ................................ ................................ ....... 39 Ion Cyclotron Resonance Mass Spectrometer ................................ ................................ 41 Fourier Transform Ion Cyclotron Resonance Mass Spec trometer ................................ .. 41 Infinity Cell ................................ ................................ ................................ ............................. 42 Oscillation and Magnetron Motion and Trapping ................................ ........................... 43 Cyclotron Motion and Excitation ................................ ................................ .................... 45 Detection ................................ ................................ ................................ .......................... 4 8 Ion Sources ................................ ................................ ................................ ............................. 50 Inter nal Ion Sources ................................ ................................ ................................ ......... 51 External Ion Sources ................................ ................................ ................................ ........ 51 Ion Accumulation and Ion Optics ................................ ................................ .................... 55


6 3 I N FRARED MULTIPLE PHOTON DISSOCIATION (IRMPD) THEORY ........................ 62 Vibrational Motion ................................ ................................ ................................ ................. 62 Diatomic Vibrational Motion ................................ ................................ .......................... 63 Polyatomic Vibrations ................................ ................................ ................................ ..... 68 Interaction of Light and Matter ................................ ................................ ............................... 69 Diatomic Electric Dipole Mome nt ................................ ................................ .................. 70 Polyatomic Electric Dipole Moment ................................ ................................ ............... 71 Infrared Photodissocation Spectroscopy ................................ ................................ ................. 72 Multiple Photon Dissociation ................................ ................................ .......................... 73 Action Spectroscopy ................................ ................................ ................................ ........ 75 U nimolecular Dissociation Kinetics ................................ ................................ ....................... 76 Laser Intensity and Temperature ................................ ................................ ..................... 77 Molecular Categories ................................ ................................ ................................ ....... 83 4 CARBON DIOXIDE (CO 2 ) AND OPTICAL PARAMETR IC OSCILLATOR (OPO) LASERS ................................ ................................ ................................ ................................ 94 Carbon Dioxide Laser ................................ ................................ ................................ ............. 94 Four Level System ................................ ................................ ................................ ........... 96 Vibrational Rotational Picture ................................ ................................ ......................... 97 Experimental C arbon Dioxide Lasers ................................ ................................ ................... 100 Apollo C arbon Dioxide Laser ................................ ................................ ....................... 100 Lasy C arbon Dioxide Laser ................................ ................................ ........................... 103 O ptical P arametric O scillator Laser ................................ ................................ ...................... 104 O ptical P arametric O scillator Theory ................................ ................................ ............ 104 Nonlinear Optics ................................ ................................ ................................ ............ 105 Quasi Phase Matching ................................ ................................ ................................ ... 109 Pound, Drever an d Hall Technique ................................ ................................ ............... 110 O ptical P arametric Oscillator Operation and Set Up ................................ .................... 113 O ptical P arametric O scillator Operation ................................ ................................ ....... 114 O ptical P arametric O scillator Set Up ................................ ................................ ............ 116 5 SPECTRAL AND MOLECULAR DIFFERENCES IN O METHYLATED GLYCOSIDE ISOMERS ................................ ................................ ................................ ..... 127 Experimental Methods ................................ ................................ ................................ .......... 129 Materials ................................ ................................ ................................ ........................ 129 Instrumentation ................................ ................................ ................................ .............. 129 O ptical P arametric O scillator Laser ................................ ................................ .............. 130 Experimental Procedure ................................ ................................ ................................ 131 Computation Methods ................................ ................................ ................................ ... 133 Experimental Results and Discussion ................................ ................................ ................... 134 Glucosides ................................ ................................ ................................ ..................... 134 Galactosides ................................ ................................ ................................ ................... 137 Conclusion ................................ ................................ ................................ ............................ 140


7 6 THERMOCHEMICAL DYNAMICS OF DEPHOSPHORYLATION ............................... 151 Experimental Methods ................................ ................................ ................................ .......... 153 Materials ................................ ................................ ................................ ........................ 153 Instrumentation ................................ ................................ ................................ .............. 154 Experimental Procedure ................................ ................................ ................................ 155 Experimental Results and Discussion ................................ ................................ ................... 156 Phosphopeptide Mass Spectra ................................ ................................ ....................... 156 Phosphopeptide Kinetics ................................ ................................ ............................... 157 Relative Activation Energies and IRMPD Spectra ................................ ....................... 159 Conclusion ................................ ................................ ................................ ............................ 161 7 C ONCLUSION S AND FUTURE WORK ................................ ................................ ........... 173 LIST OF REFERENCES ................................ ................................ ................................ ............. 177 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 191


8 L IS T OF TABLES Table Page 3 1 Comparison of Arrh enius activation energies from infrared multiple photon dissociation (I RMPD ) and blackbody infrared radiative dissociation (BIRD) experiment s ................................ ................................ ................................ ........................ 93 6 1 C onversion of the to a point on the kinetic plots ................................ ................................ .............................. 171 6 2 Unim olecular rate constants for dephosphorylation of both negatively and positively charged phosphopetide ions at the given wavelengths and corresponding powers. ........ 171 6 3 Comparison of relative activa tion energies taken at different wavelengths for positive and negative ions ................................ ................................ ................................ .............. 172


9 L IST OF FIGURES Figure Page 1 1 D glucose and alpha D glucopyranose .............. 29 1 2 Glyceraldehyde L and D isomers. ................................ ................................ ...................... 29 1 3 E ight stereoisomers of D aldohexose. ................................ ................................ ............... 30 1 4 Mechanism for mutarotation ................................ ................................ .............................. 31 1 5 Stereoisomers of alpha D aldohexose rings ................................ ................................ ..... 32 1 6 Mechanism for O methylation ................................ ................................ .......................... 33 1 7 Chair and boat conformations ................................ ................................ ............................ 34 1 8 Disaccharide examples. ................................ ................................ ................................ ...... 34 1 9 Structures of the four methylglycosides ................................ ................................ ............ 35 1 10 Amino acids by polarity and charge. ................................ ................................ ................ 36 1 11 Alanin e L and D isomers as zwitterions. ................................ ................................ ........... 37 1 12 Electron resonance stabilization of the amide linkage ................................ ....................... 37 1 13 Trans and cis form of the am ide linkages ................................ ................................ .......... 38 1 14 ................................ ................................ ........... 38 2 1 Ion trap configurations. ................................ ................................ ................................ ...... 56 2 2 Quadrupolar cylindrical trap ................................ ................................ .............................. 57 2 3 Ion cyclotron motion of positive and negative ions ................................ ........................... 57 2 4 Ion cyclot ron, magnetron and oscillation motions ................................ ............................. 58 2 5 Ion excitation and detection ................................ ................................ ............................... 58 2 6 General description of data acquisition for a Fourie r transform ion cyclotron resonance (FTICR) mass spectrometer including i on detection, time domain signal (FID) and the mass spectrum ................................ ................................ ............................. 59 2 7 Ultraviolet m atrix assisted laser desorption/ionizatio n ( MALDI ) source ......................... 59 2 8 Electrospray ionization as an electrochemical cell ................................ ............................ 60


10 2 9 Droplet jet fusion model ................................ ................................ ................................ .... 60 2 10 Fourier t ransform i on c yclotron r esonance m ass s pectrometry (FTICR MS) ion optics ................................ ................................ ................................ ................................ 61 3 1 Harmonic and Morse corrected anharmonic potential energy curve s ............................... 88 3 2 Four carbon dioxide (C O 2 ) normal modes ................................ ................................ ......... 89 3 3 Anharmonic bottleneck ................................ ................................ ................................ ...... 89 3 4 Infrared multiple photon dissociation (IRMPD) process ................................ ................... 90 3 5 Interplay of degrees of freedom ( DOF ) rate constants, and the nature of the transition states in determining REX ................................ ................................ ................. 91 3 6 Effects of DOF, rate constants, and transition states on molecular kinetic categories ...... 91 3 7 Boltzmann distribution for modified Tolman modeling ................................ .................... 92 3 8 Boltzmann distribution for medium molecule kinetics ................................ ...................... 92 4 1 C arbon dioxide ................................ ................................ ...... 118 4 2 Four stage CO 2 laser ................................ ................................ ................................ ........ 118 4 3 General iIlustration of rotational vibrational transitions of P and R branches ............... 119 4 4 ................................ .......................... 119 4 5 Initial set up : t he Apollo CO 2 laser part I ................................ ................................ ........ 120 4 6 Initial set up: t he Apollo CO 2 laser part II ................................ ................................ ....... 120 4 7 Introduction of the l aser gate and associated control electronics. ................................ ... 121 4 8 Implementation of a second Lasy CO 2 laser ................................ ................................ ... 122 4 9 General d escription of the o ptical p arametric o scillator (O PO ) operation .......... 122 4 10 Features in the OPO laser as related to Pound Drever and Hall method for frequency stabilization ................................ ................................ ................................ ..................... 123 4 11 Derivative of the intensity with respect to frequency of a Gaussian like be am profile 123 4 12 LINOS OPO laser with labels ................................ ................................ .......................... 124 4 13 Wavelength tuning: adjusting the periodically poled lithium niobate ( PPLN ) crystal and the etalon ................................ ................................ ................................ .................. 124


11 4 14 Temperature tuning: temperture verses wavelength plots for each poling period ......... 125 4 15 Experime ntal set up of the OPO laser with the Infinity cell ................................ ........... 126 5 1 Ba sic structure of the glucoside/ galactoside conformations ................................ ........... 141 5 2 Actio n spectra of the four glycoside isomers from 600 1700cm 1 ................................ 142 5 3 O ptical parametric oscillator laser general set up ................................ ............................ 143 5 4 Differenc es in the C ................................ ............... 143 5 5 Theoretical and experimental in the 900 to1400 cm 1 region ................................ ................................ .............................. 144 5 6 Action spectra of O methyl D in the 2750 to 3750 cm 1 region .... 144 5 7 Comparison of the experimental a nd theoretical spectra and struc tures of the four conformers of O methyl D glucoside ......................... 145 5 8 Overview of the experimental and theoretical spectra and of O methyl D gl ucoside ................................ ................................ .................. 146 5 9 I nfrared multiple photon dissociation (Action) ..................... 146 5 10 Comparison of the experimental and theoretical s pectra of O H stretching region for ................................ ................................ ............ 147 5 11 Experimental and theoretical s .. 148 5 12 Experimental and theoretical spectra and structure ................................ ........... 148 5 13 Experimental and theoretical spectra and structure ................................ ........... 149 5 14 Spectra of the O H stretch region for all four isomers ................................ ..................... 149 5 15 Spectra of the C H stretch region for all four isomers ................................ ..................... 150 6 1 I nfrared multiple photon dissociation phosphopeptide mass spectrum of the +2 charge state ................................ ................................ ................................ ...................... 162 6 2 I nfrared multiple photon dissociation ph osphopeptide mass spectrum of the 2 charge state ................................ ................................ ................................ ................................ 163 6 3 Diagram illustrating the elimination product for dephosphorylation of phosphoserine ................................ ................................ ................................ .................. 163 6 4 Pseudo first order kinetic plots of the dephosphorylation of positively charged (+2) ions ................................ ................................ ................................ ................................ ... 165


12 6 5 Pseudo first order kinetic plots of the dephosphorylation of negatively charge d ( 2) ions. ................................ ................................ ................................ ................................ .. 166 6 6 Cis and trans conformers of the the alpha carbons from serine proline residues ............ 167 6 7 Arrhenius relative activat ion energy plots for positively charged (+2) ions ................... 168 6 8 Arrhenius relative activation energy plots for negatively charged ( 2) ions ................... 169 6 9 I nfrared multiple photon dissociation ( A ction) s pectra for both positively (A) and negatively (B) charged phosphopeptides ................................ ................................ ......... 170


13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE USE OF INFRARED MULTIPLE PHOTON DISSOCIATION AND FOURIER TRANSFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY TO STUDY THE THERMOCHEMICAL DYNAMICS OF DEP HOSPHORYLATION AND THE SPECTRAL AND DENSITY FUNCTIONAL THEORY DETERMINED STRUCTURAL CHARACTERISTICS OF MONOSACCHARIDE ISOMERS By Wright Leroy Pearson III August 2009 Chair: John R Eyler Major: Chemistry In recent years there has been an explosion of interest in the fields of saccharide and protein chemistry. Though much of the initial work has been conducted in solution, the des ire to avoid complicated liquid phase interactions has resulted in the promotion of gas phase experiments. Capable of multip le stages of experimentation (MS n ) and long ion retention times while minimizing extraneous ion ion/ion molecule reactions, Fourier transform ion cyclotron resonance mass spectrometry ( FTICR MS) has proven to be a powerfu l instrument in these studies. Of p articular importance is the addition of techniques, like infrared multiple photon dissociation (IRMPD), that promote ion dissociation/fragmentation for structural and thermochemical information. The current research employed two 4.7T FTICR mass spectromete rs with different types of IR laser systems to examine the possibility of and the reasons for differentiation of saccharide isomers and the kinetic and energetic picture of dephosphorylation through multip le p hoton di ssociation processes. A LINOS OS 4000 continuous wave optical parametric oscillator (OPO) laser was purchased and the laser beam transfer optics were designed, positioned (within a purge box), and


14 focused on the ions in an ICR cell. Photon (v ibrationally resonant) induced dissociation or actio n spectra, spanning the C H and O H stretch regions, were then acquired for four rubidium cation bound glycosides. Comparison of the experimental and D ensity F unctional T heory calculated spectra revealed a network of intermolecular hydrogen bonding and rub idium attachments that provided the means to differentiate D glucoside and D galactoside anomers in the O H stretching region of the infrared spectrum In the second project, two cw CO 2 lasers were also set up for gas phase experiments. Dephosphorylation r ate constants (at varying laser powers) were obtained for both positively and negatively charged phosphopeptide ions at three different wavelengths. In the se experiment s, the dissociation rate constants favored the negatively charged species at all wavelen gths (a result of lower activation energies) as the overall Arrhenius activation energies were independent of vibrational mode intensities.


15 CHAPTER 1 INTRODUCTION I nfrared multi ple pho ton dissociation (IRMPD) of gas phase ions has been of importance to mass spectro metrists for a number of years. A large part of this interest has been advanced through the use of Fourier transform ion cyclotron resonance mass spectromet ry ( FTICR MS ) in Penning traps and the de velopment of tunable IR lasers. Penning tra ps make ideal mass spectrometers for these experiments due to two factors low pres sures and long ion storage times. Low pressure conditions (10 8 to10 10 Torr) minimize ion neutral interactions that lead t o collisional energy transfer. Long ion retention t imes allow the ions to be exposed to IR photons for longer periods to facili tate multiple photon processes. In IRMPD, infrared photons arise from two sources, black body radiation and an external IR laser (CO 2 OPO, etc.) beam centere d on the ions in the trap. B lackbody radiation can be adjusted by manip ulating the Keeping the temperature relatively low decreases the intensity and frequencies of black body radiation, thus leaving most of the photon production to the focused, relatively high powe r monochromatic IR laser beam. Infrared light sources produce single or multiple (tunable) wavelength(s) with a continuous (continuous wave, cw) and/or a pulsed beam depending o n the needs of the experiment. Multiple photon ab sorption is used to promote ion dissociation where the resonant vibrational mode absorbs IR photons and undergoes excitation. The energy is then quickly distributed throughout the vibrational modes raising the internal energy until th e weakest bond(s) dissociates. Spectroscopy, kinetics, thermodynamics, structures, and energetics of many systems have been explored using IR laser beams directed onto ions contained in Penning traps. In the present research, IRMPD has been applied to biol ogical molecules, in particul ar saccharides and peptides. In conjunction with Dr. Cesar Contreras, vibrational ly dependent


16 photodissociation spectra (action spectra) covering the C H and O H stretch region (2750 to 3750 cm 1 ) of four cyclic D aldohexose is omers have been measured with a n FTICR mass spectrometer utilizing an optical parametric oscill ator (OPO) laser light source. These spectra reveal areas of isomeric differentiation and help elucidate gas phase structures with the aid of computational model s. Kinetics and relative activation energies have also been determined for dephosphorylation of positively and negatively charged phosphopeptide ions, with a CO 2 laser light source. These experiments have been repeated at three distinct widely spaced wavel engths to determine the effect o f laser vibrational resonance. Prior to describing the experiments and results, the basics of saccharides and proteins are presented. Saccharides Saccharides have been studied for many years and have very important biologic al and economic functions, from energy storehouses in the human body and sweeteners in our beverages and foods to cellulose for plant str ucture and potential biofuels. Emil Fischer ( 1852 1919 ), the father of saccharide chemistry, is responsible for car bohydrate classification and the initial s tudies of reaction mechanisms. 1,2 His work has been carried forward through the tireless efforts of many in the scientific community, with experiments usually conducted in the solution phase. Given the complexity o f solution phase chemistry, where carbohydrate interactions form derivatives and polymers that vary with temperature, solvent, molecular structure, etc., gas phase chemistry is often utilized to help simplify the picture. Mass spectrometry is one of the mo re important instrumen ts in these gas phase studies. Saccharide ions are isolated in the gas phase to be examined through ion molecul e reactions, dissociation techniques (including IRMPD), and a host of other MS n experiments to obtain information. 3,4 Backg round information on saccharides is helpful for an appreciation of these complex interactions.


17 Monosaccharides The basic units of saccharides are the monosaccharides that comprise one of two families: aldoses and ketoses. Each monosaccharide is made of 3 t o 10 carbon atoms connected by C C sin gle bonds. When in the chain form, an alcohol is found on all but one carbon, which shares a double bond with an oxygen atom, resulting in the empirical formula (C n H 2n O n ). The most biologically significant saccharides and the focus of this section, are the aldose sugars, wh ic h have the carbonyl on one end of the ch ain forming an aldehyde group. In ketoses, the carbonyl is found inside the chain resulting in a ketone group. Carbonyls of open monosaccharides are reducing agents, oxidized by mild oxidizing agents such as Fe +3 or Cu +2 (found in positive Tests to monitor the presence and the quantity of monosaccharides (reducing sugars) are common today, e.g. in glucose monitoring for diabete s patients 5 10 Emil Fischer developed a carbon numbering system and simple nomenclature to identify each monosaccharide (Fischer projections) 11,12 Beginning with the carbonyl carbon designated as C 1 (in aldoses) or C 2 (in ketoses), each carbon is numbe red sequentially ending with the CH 2 OH group, which for D g lucose is C 6 (Figure 1 1 ). Monosaccharides with four, five, six, and seven carbon atoms are known as tetroses, pentoses, hexoses, and heptoses respectively, and are given the title prefix keto o r aldo depending on the family. The most naturally occurring are the hexose and pentose varieties in particular the aldohexose D glucose and ketohexose D fructose. Chiral c enters Most sugars have multiple chiral centers (carbons with four different func tional groups). As a general rule saccharides containing m chiral centers can have 2 m stereoisomers. In the case of aldohexoses which have 4 chiral centers, there are a total s ixteen possible stereoisomers. These isomers are split into L and D categories (This is not to be


18 confused with the terms dextrorotatory (d or + for cw rotation ) and levorotatory (l or for ccw rotation ) which specify the direction of rotation of plane p olarized light when interacting with the electron clouds of chiral molecules). As shown in Figure 1 2, w hen the configuration of the chiral carbon f a rthest from the carbonyl is the same as the Fischer projection of D gly ceraldehyde (with the O H to the right side of the chiral carbon) the sugar is designated as a D enantiomer Wit h the L enantiomer the OH is pointing to the left (or S and R in the Cahn Ingold Prelog convention respectively ) 1 3 For the aldohexoses, this results in two sets of eight uniquely named isomers shown in F igure 1 3 The number of stereo isomers is furthe r complicated by the formation of favored cyclic structures. Cyclic s tructures In solution s of monosaccharides containing five carbons or more, the combination of potentially reactive carbonyl and hydroxyl groups and a ring s relative stability create con ditions that promote cyclic structures in equilibrium with the linear forms Th is mechanism known as mutarotation, is show n in Figure 1 4 for D glucose. In D glucose t he carbonyl oxygen is protonated opening the carbonyl C 1 to nucleophilic attac k by the oxygen from the C 5 hydroxyl group to create a six membered ring. The hydrogen atom from the former C 5 oxygen is subsequently lost, creating the hemiacetal structure with an anomeric carbon, C 1 at the point of further mutarotations 14 The ring also creates a new chiral center at C 1 However, the D designation is conserved as the C 6 hydroxyl group remains relatively unperturbed The eight isomers of alpha D a ldohexose rings are illustrated in Figure 1 5 Depending on the approach of the nucleo philic attack, above or below the plane of the carbonyl, the cyclic molecule will be in one of two anomeric forms, alpha ( ) or beta ( ). An alpha sugar is the result of a nucleophilic attack above the plane causing the newly formed C 1 hydroxyl group to be axial to the plane of the molecule. A beta sugar forms when the nucleophile


19 approaches from the opposite end (below) cau sing the C 1 hydroxyl group to be equatorial or in the plane of the ring. In nature, these two forms mutarotate resulting in an observed specific (optical) rotation of + 52 .7 for D glucose at 20 o C. With the pure anomeric forms having = +112 ( anomer) and = + 1 8 7 ( anomer), at the D line of sodium, the equilibrium percent concentration is easily calculated to be 36% alpha and 64% beta in solution. 15 Depending on the molecular structure, solvent, etc., these percentages vary for different sugars. Mutarotation of the alpha and beta isomers can be halted through O methylation of the anomeric carbon, a p rocess known as glycos y lation. As shown in Figure 1 6 protonation of the C 1 hydroxyl group causes the release of water to creat e an oxon ium ion intermediate. The methanol oxygen in turn attacks C 1 and, depending on the point of attack, forms an O methylated alpha or beta isomer, which is a nonreducing sugar. Although the process is reversible in acidic medi a the conformati on is stable at moderate pH values 16 Cyclic c onformations Influenced by the size of the ring, functional group interactions, solvent effects, etc., single bonded rings can have many conformations. Five membered rings are typically planar due to the steri c conditions o f the tightly grouped carbons. They are found in pentagon or envelope (E) shapes where some puck ering and twisting does occur. The six membered rings favored by D aldo hexoses can take on more forms. The most common are the boat and chair (Fig ure 1 7 ). Boat formations are designated as either 1,4 B or B 1,4 depending on whether both C 1 and C 4 carbons are pointed above or below the plane of the ring, respectively. The chair conformations are similarly designated as 4 C 1 and 1 C 4 With the 4 C 1 stru cture, C 4 is above the plane of the ring and carbon C 1 is below. 16 The 1 C 4 chair is the opposite. In solution, the most abundant of these conformations is by far the 4 C 1 However, in the gas phase, al though


20 the 4 C 1 chair is favored, a fair amount of the 1 C 4 can exist 17 As with any flaccid ring these structures can twist (i.e. skewed (S)) and bend (i.e. half chair (H) ) into a number of variations. Disaccharides Most cyclic monosaccharides form larger mo lecules through glycosylation The mechanism for O glycoside linkage is similar to t hat described previously for O m ethylation (Figure 1 6) except that the nucleophilic oxygen c omes from a hydroxyl group located on one of the C 2 to C 6 carbons of a second cyclic monosaccharide ( the glycosyl donor ), not methanol When a glycosidic bond is created between these two monomers the hemiacetal of the first cyclic monosaccharide, the glycosyl acceptor is replaced by an acetal group. cetal carbon C 1 is locked and unable to isomerize (nonreducin g sugar) while the former glycosyl donor is now ready to act as acceptor on its reducing end The exception to this is when the nucleophilic oxygen comes from the hydroxyl group on C 1 which produces only a nonreducing sugar. 14 Isomers of these molecules are numerous. If the acceptor is in the alpha/beta conform ation the disaccharid e is alpha/beta respectively. Further, the structures multiply based on the sition/orientation on the ring ; e.g one of the oxygen atoms of a pentose sugar (the glycosyl donor) can attack the C 1 carbon of another pentose sugar (the glycosyl acceptor) from different hydroxyl group positions (( C 2 ) OH 1, (C 3 ) OH 1, (C 4 ) OH 6 ) OH 1) each with one of two orientations axial or equatorial (Figure 1 8). Dissacharides are named in the following order: the configuration ( or of the nonreducing sugar, the name of that monomer, the num bered designation of the two carbons joined (C n), and the name of the second monomer. Furano and pyrano are added to the names of both


21 units depending on whether the ring consists of fiv e or six carbons, respectively. 18 Structures of larger polysacch arides are largely based on these linkages. Polysaccharides Polysaccharides (glycans) differ in the quantity and types of monosaccha ride subunits, linkages (i.e. g lycoprotein s S glycoside link ages) and degree of branching. They are divided into two majo r groups: homopolysacchari des and heteropolysaccharides. Homopolysaccharides contain on ly one type of monosaccharide. Starch, glycogen in animals and cellulose are all hom opolysaccharides of D glucose. Heteropolysaccharides consist of many different monom ers and other subunits. They form the basis for proteoglycans, glycoproteins, glycolipids, etc., a s well as many complex sugars. They are important in cell communication, structure, transport, and a h ost of other bodily functions. 14 Carbohydrates form very complex structures and consist of many isomeric components making saccharides challenging for gas phase study. Saccharide Structure & Differentiation E arly experiments with limited IR laser sources joined to mass spectrometers illustrated ion dissociatio n through the absorption of vibrationally resonant IR photon s and set the stage for gas phase vibrational spectroscopy 19,20 By sequential scanning through the wavelengths of a tunable IR laser instruments like the FTICR mass spectrometer can monitor the relative abundance of the fragments and precursor ions at each wavelength to create a dissociation or action spectrum 21,22 With new developments in laser technology and availability the range of available wavelengths has expanded. Free electron laser fac ilities like the one at the FOM Instit ute for P lasma P hysics Rijn huizen, in the Netherlands, make availa ble IR wavelengths of 5 110m. Though impressive, the wavelength range does not include most of the near IR where the saccharide O H and C H stretches a re found The cw OPO laser, however, covers a large part of the region allowing the study of these important vibrational motions.


22 Infrared multiple photon dissociation in Penning traps has recently been applied to the problem of identification and struct ural elucidation of gas phase di and monosaccharides. In experiments at the FOM I nstitute with the free electron laser Dr. Nick Polfer et al. showed that specific fragmentation patterns for eight isomers of lithium tagged glucose based disaccharide m at a total irradiation energy of 340mJ) are capable of differentiating the isomers. 22 Comparisons of the vibrational spectra of these isomers however, prove less conclusive, due to the congested vib rationa region. The numerous saccharide hydroxyl groups led to the possibility that spectral differences may exist for the vibrationally isola ted near IR O H stretch region. In this research four Rb + tagged O methylated mon o saccharides were examin ed: Rb + [O methyl D glucopyranoside + [O methyl D + [O methyl D galactopyranoside + [O methyl D galactopyranoside] ( Gal)] (Figure 1 9) A cw OPO laser provide d the necessary wavelength range (1.38 2. 0 and 2.28 4.67 m ) and power (50 150mW) in the O H stretching region to dissociate the attached Rb + ion The results from these experiments discussed in Chapter 5, give a promising means to diffe rentiate each monomer complex. They also provide gas ph ase structural information when the observed spectra are compared with those calculated theoretically Peptides Proteins are the principle components of muscle s and organs, and play crucial roles in chemical reactions (enzymes), communication (phosphorylat ion, etc.), transport (albumin, hemoglobin, etc.) and structure (organelles and c ell membrane) within the body. 23 Even though proteins are very important and have been subject ed to intense experimental scrutiny, complex structures and their important doma ins ha ve proven difficult to ascertain. With the development


23 of the polymerase chain reaction (PCR) the blueprint of proteins, deoxyribonucleic acid (DNA), became readily available and led to the mapping of the genetic component of protein production prom oting t he study of proteins. 24 Methods such as electrophoresis and electrospray ionization (ESI) components, providin g a wealth of new information. 25 The present work contributes to this body of knowledge by demonstrating the use of IRMPD for phos phorylation elucidation in Chapter 6 To gain a better understanding of these findings, a brief introduction to the structure and nature of peptides is presented. Amino Acids O f the numerous amino acids that exist in nature, only twenty are considered standard to prot ein composition The others are produced after protein formation or are not involved. Figure 1 10 shows the 20 standard amino acids, all of which have a carboxyl gr oup, amine group, hydrogen atom, and a side chain (R group) bonded to the same carbon, called the alpha carbon. In solution at physiological pH (~ neutral) amino acids form zwitterions by proton transfer from the COOH to the NH 2 Side chain groups determi charge (solubility), and role in the protein. 24 With the exception of glycine whose R group is hydrogen, carbons are chiral centers. Similar to the monosaccharides, if the chiral group is in the same position as L (on the left side of the carbon) then it is a L isomer; w hen in the opposite configuration of D alanine, a D isomer ( F igure 1 11 ). Proteins are made up of L isomers, a result of asymmetric enzyme acti ve sites in peptide formation. T he carboxyl carbon is labeled 1, followed by the carbon 2, and so forth continuing with the carbons of the R group.


24 The twenty R groups are subcategorized into five classes according to the relative polarity and charge (Figure 1 10 ). 23 Nonpolar, aliphat ic groups consist of alanine (Ala, A), glycine (Gly, G), isoleucine (Ile, I), valine (Val, V), leucine (Le u, L), proline (Pro, P), and methionine (Met, M). Hydrophobic by nature, these amino acids are often found in the interior of globular pro teins and st abilize structure. Relatively nonpolar, aromatic R groups include phenylalanine (Phe, F), tryptophan (Tr p, W), and tyrosine (Tyr, Y). Aromatic side chains are important to enzyme formation and strongly absorb ultraviolet light at 280 nm, thus serving as im portant an alytical tools for researchers. Polar neutral side chains are found in asparagine (Asn, N), cysteine (Cys, C), glutamine (Gln, Q), proline (Pro, P), serine (Se r, S), and threonine (Thr, T). The S H groups of cysteine aid in the formation (through sulfide bonds) and maintenance of the t ertiary structure of proteins. Serine and threonine (and tyrosine) are potential phosphorylation sites important to protein regulation and energy transport. Lysine (Lys, K), arginine (Arg, R), and histidine (His, H) are bases (positiv e ly charge d at physiological pH). These hydrophilic molecules are often found on th e surface of proteins. Histidine is a key component in hemoglobin, aiding the exchange of O 2 and CO 2 and is impor tant in enzyme reactive sites. Lastly, as partate (Asp, D) and glutamate (Glu, E) are acidic (negativ e ly charge d at physiological pH). These serve in cell transport and have important f unctions in neurotransmitters. Peptide bond The simplest peptide, a dipeptide, is produced by the condensation o f the carboxyl group of one amino acid with the amino group of another, creating an amide linkage. Electron resonance stabilization of the amide linkage restricts its rotation (Figure 1 1 2 ) keeping the peptides in the predominant trans form in 99% of all l inkages (Figure 1 1 3 ). 26 The reaction however, is thermodynamically unfavorable and requires more sophisticated reactions to explain peptide formation and orientation.


25 Transcription and t ranslation A peptide bond is created as part of a longer chain of r eactions during the process of transcription and translation. 24 nucleus with the opening of a deoxyribonucleic acid (D NA) section, from which a comple mentary messenger ribonucleic acid (mRNA) strand is created (wit h the h elp of RNA Polymerase). cytoplasm and attaches to the ribosome exposing two groups of three cyclic bases (a codon). Meanwhile, tran sfer RNA (tRNA), with the comple mentary three base seque nce (anticodon) coded to a specific amino acid, binds to the amino acid and carries i t to the ribosome. The tRNA anticodon forms H bonds with complementary mRNA on the first of two exposed codons A second tRNA is then added to the other exposed codon, bri nging the two a mino acids in close proximity. The amino acids then form a peptide bond (with the help of peptidyl transferase). The ribosome moves down the mRNA exposing a new set of bases, a third tRNA comes into pla ce, and the first tRNA leaves. The proc ess of translation continues until the stop codon ceases the production and the peptide is released. Th is intricate process lowers the activation energy required for peptide bond formation and fixes L isomers in the trans state so that the bulky side chain s are located in the energetically favored position on opposite sides of the C N bond. Polypeptides Polypeptides consist of oligopeptides (~2 10 residues), polypeptides (~10 100 residues), and prote ins (~more than 100 residues). Their structures are writte n with the free carboxyl group ( COO ), on the far right, the C terminus, and the free amine group ( NH 2 ), N ter minus, located at the far left. Among their many roles, oligopeptides act as protease inhibitors, poisons, and means of drug transport, and they are often sy nthesized for peptide research. Polypeptides of 10 100 residues are primarily found in h ormones and cell m essengers particular to ribosomal activity as well as some enzymes 23 ,24 Proteins or large polypeptides, however, are described


26 accord in g to a structural hierarchy. Primary structure specifies the amino acid sequence and the positions of S S linkages. Alpha helixes and pleated sheets shape the primary strands into more elaborate secondary structures, as shown in Figure 1 1 4 The helixe s are typically right handed ( counterclockwise ) coil like spirals about ~ 5.4 wide with ~3. 7 residues per turn (100 o per amino acid) at a distance of ~1.5 along the axis. 27 Built with a single primary strand, the spiral frame is supported by hydrogen b onds between the N H of one residue and the C=O of ano ther placed four groups before. Multiple amino acids strands are also joined to one another by hydrogen bonds from the N H groups of one s These pleated backbone, with alternating R groups protr uding in and out of the sheet. Beta sheets are found as a t turns and loops in many proteins with prolin e, glycine, threonine and /or serine residues at each turn and twist. 28 Functional proteins are largely created from primary and secondary arrangements connected by hydrogen and di sulfide bonds and hydrophobic and electrostatic interactions These arrangem ents are known as tertiary (framing globular and other protein formations) structures In some proteins two or more tertiary structures group together and fashion even more complex quaternary structures. Supersecondary structures (motifs, folds, etc.) are found within these constructions, revealing specific protein function s in regions known as domains. From the quaternary structure of actin/tropomyosin/myosin/collagen in muscle tissue and hemoglobin for O 2 transport to ribosomes and enzymes key to protein synthesis, proteins are among the most studied molecules in nature. 23 Many instruments and techniques are applied to illuminate p eptide structure and function. X ray crystallographic evidence for three dimensional arrangements is the basis for the knowled ge about the aforementioned structural classes. 27,29 Nuclear m agnetic r esonance (NMR)


27 provides information on the molecular make up as well as solution phase dynamics including conformational changes and protein fold ing, among other interactions. 30 With th e advance of ionization sources, mass spectrometry is currently another emerging tool for gas phase experiments. Electrospray ionization (ESI), for instance, gently ionizes the protein(s) leaving multiple protonation (+ ions) or deprotonation ( ions) site s, depending on the conditions set by the experiment (discussed in Chapter 2 ). This multiply charged ionization capability allows large proteins to be mass detected within the limited mass to charge ( m/q ) ran ge of most mass spectrometers. Study of these io ns is carried out largely through fragmentation of the ions of interest, IRMPD providing a major means for dissociation. Fragmentation peaks (along with other tools) are used to determine peptide components and the kinetics of protein reactions under the p roper conditions. 25 ,31 Unimolecular D issociation of P hosphate G roups Dephosphorylation kinetics and relative activation energies have been determined for positive ly and negative ly charged phosphopeptide ions trapped in a FTICR MS in work reported in Chapte r 6 Phosphorylation and dephosphorylation regulate cellular processes that are important to metabo lism, growth, and reproduction. The activation energies associated with dephosphorylation can give insight into the nature of phosphopeptide interactions and provide a clearer picture of the effects of pho sphorylation. One method of determining Arrhenius activation energies uses temperature controlled blackbody radiation as an energy source 31,32 A nother approach is use of a low intensity cw CO 2 laser (in reso nance with the P O stretch) that models thermal properties of the blackbody radiator 33,34 In this picture a bsorption and emission of the low intensity infrared photons by the phosphopeptide molecule promotes thermal equilibrium with the surroundings, cre ating a


28 Boltzmann like distribution of energy similar to a blackbody radiator with an effective temperature. Dissociation of the lowest energy phosphate bond in the ion occurs when the internal energy of the ion exceeds a threshold energy (E t ), typically a ssociated with the high energy tail of the Boltzmann distribution. The result is loss of a water and a neutral phosphate group producing at a minimum two peptide fragment peaks (P f. w/o H20 & P f w/o H2O & H3 PO 4 ), and a phosphopeptide (P T ) peak of reduce d intensity Dissociation rate constants (k d ) are then obtained from the slopes of ln{[P T ]/([P T ]+[P f.s ])} or ln[relative ion a bundance] vs. time plots. When conditions for thermal equilibrium between the phosphopeptide ions and the environment exist, the r apid exchange limit (REX) is met. 32,33 An Arrhenius plot of the natural log of the rate constants vs. natural log of the laser p ower (watts) associated with each k d can then be constructed. The slope of the plot gives the activation energy multiplied by a constant Overview The objective of th e research reported in this dissertation is the use of IRMPD FTICR MS to give a picture of isomeric gas phase saccharide structure (with d ensity f unctional t heory ( DFT) calculations) and differentiation, as well as u nimolecular dis sociation from phosph opeptides C hapter 2 will examine FTICR MS features, such as the ion trapping cell and ion sources that play key roles The theoretical framework o f IRMPD will be the subject of C hapter 3 w hich will cover the nature of vibrational modes, the general IRMPD process, and rel ated statistic al /kinetic models. C arbon dioxide and OPO laser s are covered in C hapter 4 their relev an ce to the IRMPD experiments. Chapter 5 will present experimental and DFT calculated IR spectra of four O methylated D glucopyranoside ions with a unique look into the gas phase structures and the ability to differentiate between them. Unimolecular dissocia tion of


29 phosphate groups from both negative and positive phosphopeptide ions is the subject of C hapter 6 Vibrational, kinetic, and relative activation energy information and the relationship between photon absorption and the dissocia tion intensities are e xplored. Chapter 7 summarizes the research done to date and offers a look into future considerations for the continuation of these very promising experiments. Figure 1 1. gluco se and a lp ha D glucopyranose Figure 1 2. Glyceraldehyde L and D isomers


30 Figure 1 3. The eight stereoisomers of D aldohexose


31 Figure 1 4. The mechanism for mutarotation of the alpha and beta cyclic structures and their corresponding specific rotation s


32 Figure 1 5. The e ight stereoisomers of alpha D aldohexose rings


33 Figure 1 6 The mechanism for O methylation of the alpha and beta cyclic structures to prevent mutarotation is also known as glycos y lation. Glycos y lation is responsible f or attaching two monosaccharides together to form larger molecules, where the ROH group belongs to another monosaccharide.


34 Figure 1 7 T he chair and boat conformations of cyclic D glucose Figure 1 8 Examples of disa c charides : ( a) Glc 1 2Glc, ( b) Glc 1 3Glc, ( c) Glc 1 4Glc, and ( d) Glc 1 6Glc.


35 Figure 1 9 Structures of the four methylglycosides examined in the work reported in Chapter 5 : Rb + [O methyl D glucopyranoside + [O methyl D glucopyranoside + [O meth yl D + [O methyl D galactopyranoside] ( Gal)]


36 Figure 1 10 The amino acid groups separated by polarity and charge in their zwitterion forms.


37 Figure 1 1 1 Alanine L and D isomers as zwitterions Figure 1 12 Electron resonance stabilization of the amide linkag e


38 Figure 1 13 The trans and cis f orm of the amide linkages where 99% of all peptide linkages are in the trans configuration keeping R groups and alpha carbons on either side. Alp ha Helix Beta Pleated sheets ~3.7 residues per turn Figure 1 14 Three dimensional and general structure of h elix and pleated she e ts respectively; the alpha helix illustration gives the agreed upon secondary structure as pictured in the original paper by Pauling, Cory, and Branson. 27 [Reprinted with permission from the National Academy of Sciences. Pauling, L.; Corey, R.B.; Branson, H.R. 1951. The structure of proteins: Two hydrogen bonded helical configurations of the polypeptide chain. 28 Proc. N.A.S. (Volume 37, Page 207, Figure 2) and Elsevier. Nesloney, C.L.; Kelly, J.W. 1996. Progress Towards Understand ing sheet Structure. Bioorganic & Medicinal Chemistry. (Volume 4, Page 740, Figure 1). ]


39 CHAPTER 2 F OURIER TRANSFORM ION CYCLOTRON RESONANCE MASS SPECTROMETRY During the last 50 years i on c yclotron r esonance m ass s pectromet ry (ICR MS) and later Fourier t ra nsform ICR MS ( FTICR MS) have evolved into one of the mo re powerful tools in m ass s pectrometry. Beginning with the discovery of i on cyclotron motion the first ICR mass spectrometers were single frequency scanning instrument s where slow measurements and s imple ion sources ( chemical and electron impact ionization ) limited their potential With the advance of computers a fast Fourier transform algorithm was developed and put to use in mass spectromet ry Marshall and Comisarow appli cation of the technology with a new cell design led to the creat ion of the FTICR MS 35 39 The total ion signal taken virtually at once provide s relatively quick m/q determination of all trapped ions and has promoted real time gas phase experiments. Development of external ion s ources like electrospray ionization (ESI), a nondestructive multiple charge state platform, has allowed the study of very large molecules, greatly expand ing Because of the se and other advances and long ion retention times FTI CR MS has been very useful for the study of gas phase dissociation processes History and Fundamentals Ion Cyclotron R adius Lawrence and Livingston in 1 9 32 discovered that as an ion moves through a uniform magnetic field it encounters a force that prope ls the ion in a circular motion perpendicular (x,y plane) to the magnetic field (z axis) 40 This inward directed force is known as the Lorentz force and is given by ( 2 1)


40 where m is the mass of the ion is the velocity vector of the ion q is the ion charge and B is the magnetic field vector The Lorentz force acting as a c entripetal force is countered by an equivalent outward directed centrifugal force ( ) which creates a stable orbit for the ion(s) such that (2 2 ) where B o is the magnetic field strength along the z axis and the x,y components of are defined by Ion cyclotron motion is a circular orbit with a frequency obtained from a simple rearrange ment of E quation 2 2, where the cyclotron angular frequency is determined by the m/q of the ion(s) and the magnetic field strength, ( 2 3) The implications of this result for mass spectromet ry are clear. An ensemble of ions of varying m/q have different frequencies in the presence of a uniform magnetic field making it possible to discriminate ions with equal m/q without concern as to differences in velocity or translational energy ( k inetic e nergy focusing is unnecessary). A spectrum of these frequencies is easily converted into a mass spectrum by the implementation o f E quation 2 3. Lawrence further is excited to a larger coherent radius by applying a r elatively low energy r adio frequency (rf) voltage resonant with the angular frequency. Thus, the ions of interest can be discriminate d by either changing the magnetic field with a static applied resonant frequency or by having a static magnetic field with changing or concurrent resonant rf.


41 I on C yclotron R esonance M ass S pectrometer The first ICR mass spectrometer determined the mass of the ions by sequentially changing the magnetic field while using an applied static rf voltage to excite the ions as they c ame into resonance. Ions were then detected by electrometer plates placed in the magnetic field or by energy loss due to absor p tion by resonant frequencies. 41,42 These operating features are found in th pectrometer s respecti vely The process of sequential scanning made the task of taking a spectrum long and tedious, and severely limited the resolution. Double resonance techniques (the ability to excite/isolate one m/q ion separately from another), however made the ICR MS an instrument of choice for the study of gas phase ion molecule reactions F ourier T ransform I on C yclotron R esonance M ass S pectrometer In 1965 with the advance of computing, Cooley and Tu k e y developed a fa st Fourier transform algorithm. 43 The ability to conv ert large quantities of data into discrete frequency components in a short period of time opened up new possibilities. In NMR, discrete s ignals from an ensemble of varying frequencies of nuclear moment precession could be detect ed and used to acquire data in fractions of the time Marshall had seen the development of FT NMR a number approached Comisarow, who had recently built an ICR mass spectrometer with a proposition to create a Fourier trans form version of the i nstrument. Two main obstacles stood in the way : how to excite and detect the ions. The excitation of the ions is difficult due to the large frequency bandwidth associated with the m/ q range in ICR MS. To test the feasibility a n rf pulse of a na rrow frequency range (typically used in FT magnet. The resonant ions responded to give a detectable signal. Comisarow and Marshall reasoned that a shorter timed pulse could potentially incre ase the bandwidth for ICR excitation.


42 However, as the frequency amplitude is proportional to the area of the time of the pulse shortening the pulse duration calls for a much higher voltage high enough to make the project unfeasible. It had been long unde rstood that a slow frequency sweep of around 20 min utes at rf amplitude in millivolts could create a spectrum. So, Comisarow and Marshall decided to try a 1 ms rf sweep of ~10Volts for broadband detection, and they met success. 35 Detection of the ions req uired a different vision. I on motion had been simulated as a rotating monopole mimicking the cyclotron motion. The model predicted that by exciting the ICR motion to a sufficient radius, an image current would be produced on recording metal plates. This cu rrent is stated as (2 4) where I s is the image current, N is the number of ions in the cell, r is the radius of motion and d is the width of the ICR cell. 36 Detection of the signal was achieved by converting the image current to vo ltage t hrough a broadband RC circuit. However the key to making all this work lay in the basic design of the cell. Infinity Cell The ICR cell take s on many shapes and sizes ( F igure 2 1 ) but a common configuration is a cylinder in a quadru polar linear con figuration with the excitation and de tection plates on the left /right and top/bottom respectively, forming the four quarters of the cylinder and two trapping plates o n the front and back (Figure 2 2) 4 4 Centered inside the bore of a superconducting electr omagnet, the middle of the cell is aligned with the homogeneous region of the magnetic field. The ion s created by either an external or internal ion source enter through a small opening on one side of the trapping plate s (along the z coordinate) parallel t o the direction of the magnetic field. As mentioned the Lorentz force causes the ion s to move perpendicular ( in the


43 x,y plane ) to the axis x,y establishing a stable cyclic motion with a radius r P ositive ion s move in the counterclockwise direction ; n egative ion s move in a clockwise direction (Figure 2 3) according to Equation 2 1 4 5 In order to prohibit ejection along the z axis a small equivalent voltage is applied to each of the trapping plates, causing the ions to oscillat e harmonically within the small central region A rf voltage applied to the e xcitation plates then provide s the resonant ele ctric field necessary for the ions to spiral outward to a detectable orbit, where an induced ion image current is recorded through the detection plates. This configuration allows stable ion position ing as well as discrete excitation and detection at broad bandwidths. A closer look into this operation might prove helpful in understanding the FTICR mass spectrometer Oscillation and Magnetron Motion and Trapping When a positive (or negative) ion passes to the back of the cell it is repelled by a positive (or negative) voltage applied to the rear plate and travels in the opposite direction encountering an equal voltage now applied to the front entrance plate. Oscillating back and forth within the ped. This 3 D quadrupolar trapping potential ( is defined as, (2 5) where V trap is the trap voltage, a is a measure of trap dimensions, and are constants based on 4 6 T he electric field and force components of the surface help shed light onto the oscillating and magnetron motions. By taking the negative derivative of the 3 D quadrupolar trapping potential with respect to the direction of the ion, the equation for the electric field is


44 (2 6) corresponding to an axial directed force of (2 7) This produces an oscillating frequency on the z axis similar to simple harmonic motion and the frequency is given by, z = 2 v z = (2 8) Isolated, the axial force moves the ions in harmonic motion along the z axis but has no net effect on their x y motion. Radial forces are another matter. The radial force F radial = q E( r ) component acts in the x,y p lane and is calculated as 4 8 (2 9) motion and a term to Equation 2 2, such that (2 10) Simplification and re arrangement yield the quadratic relation, (2 11) Equation 2 11 indicates that is independent of the radius, meaning that the no e ffect on these complex motions. 4 7 The cyclotron motion is redefined by a solution to Equation 2 11, which gives two equations of motion: the reduced cyclotron ( and magnetron ( f requencies. These are


45 (2 12) and (2 13) respectively, where c is the unperturbed cyclotron frequency and z is the oscillation frequency. 4 9 The trapping oscillation, magnetron, and cyclotron rotations are il lustrated in Figure 2 4. 4 4 Cyclotron motion interacting with the trapping fields results in the additional magnetron motion. Both of these motions move the ion in the x, y plane as they simultaneously travel back potential well. Detection is hardly affected by the magnetron and oscillating trapping movements due to their insignificant role in defining the orbital radius after compensation for the reduced cyclotron motion. Only when the cell is misaligned and/or the radius of the ions approach es the limits of the trap are these effects even noticed Cyclotron Motion and Excitation The moment the ions enter the trap, the phase of their cyclotron motion is incoherent. Detection is impossible, as any induced charge on one plate is canceled by the 180 0 out of phase signal of another ion on the opposite plate. The radius is also too small for any notable detection. A simple rearrangement of Equation 2 2 gives the simplified (not taking into account magnetron motion, etc.) radius as, (2 14)


46 (2 15) with the radius redefined as, (2 16) in which r k B is the Boltzma nn constant, and T is the temperature, excitation is necessary to bring all but the largest m/ q ions into larger orbits. This is accomplished by applying a n oscillating electric field of sufficient amplitude between the excitation plates perpendicular to B o Ions in resonance with the electric field absorb energy causing the ions to spiral outward in the plane. The off axis movement of the spiraling motion also brings the ions into spatial coherence, as shown in Figure 2 5. 5 0 The final orbit radius is deter mined by the amplitude of the frequency. Instantaneous amplitude is given as (2 17) By integrating over time the absorbed energy is (2 18) As the absorbed energy of the ion is equal to its kinetic energy, setting terms in Equation 2 2 equal to Equation 2 18 and rearranging for r gives (2 19) Approximating the electric field as produced by the voltage placed on two infinitely extended plates,


47 (2 20) w ith the radius is redefined as (2 21) where A(t) is the amplitude, E 0 is the electric field, t excite is the excitation period, E abs is the absorbed energy, dip olar is the scaling factor for finite trap, and V p p is the peak to peak voltage a detectable ion radius is reached with the appropriate excitation period and voltage. 4 5 ,52 This post m ass Thus all masses of the same charge generate an equivalent image current and resulting signal. In other words, as long as an equivalent voltage and pulse duration are applied, the mass spectrum of similarly charged species should reflect the relative abundances of these ions Increasing the excitation energy much further can cause ion dissociation ( h ere a bsorbed resonant frequency energy ( t ranslational energy ) is converted into internal energy through collisions with background gas resulting in ion fragmentation i.e. sustain ed off resonance irradiation collisional induced dissociation, SORI CID) ion molecule reactions or ion ejection from the trap 4 4 As previously mentioned, a t excite pulse can excite ions at broader bandwidths than its specified fr equency ( o ) according to the relation, 2 (2 22) However, increasing the bandwidth by decreasing the excitation time requires an increase in the plate voltages to maintain the same amplitude and radius (Equations 2 4 through 2 7). Using this method, the ICR bandwidth (kHz to MHz) necessitates a pulse so short that the compensating


48 high voltage produces other intrusions. In order to minimize these problems, a modulated radio frequency (chirp) is applied to the plates This covers th e entire ba n dwidth while maintaining the amplitude at minimal voltages. In equation 2 21 t excite converts to set at the desired frequency range Though idealized, these equations are very similar to the conditions in the Bruker I nfi nity cell. 44, 51 As mentioned, the cell is placed where the magnetic field is most homogeneous with only small perturbations affecting B o and to the small ion the cell excitation plates appear as infinitely long. The applied electric field is also engineer ed to provide as consistent a field as possible, not withstanding trapping voltage interference. Even so, any necessary corrections are connected by a constant multiplied factor easily applied to these equations. For cylindrical traps this factor is dip o lar = 0.8018. 52 Detection Once the ions are excited and achieve a detectable coherent orbit, an image current is produced by a change in the image charge at both plates such that (2 2 3 ) The alternating current is then convert ed into voltage through an RC circuit where the solution to the RC circuit differential equation, (2 2 4 ) gives (2 2 5 )


49 In Equations 2 2 4 and 2 2 5 the resistance in the circuit is denoted by R the capacitanc e by C the instantaneous ICR voltage by V s and the phase difference between I c and I R by 53,54 The ion signal reflects not only the relative number of diverse m/q ions in the trap, but also the multiple advantage allows the FTICR MS to gather N data points simultaneously in amount of time of the older scanning ICR methods. 4 4 Signal to noise and resolution are improved as well. The signal to noise ratio is is described by the relationship The noise in the voltage ( V n ) is understood to be the product of the noise in the current (cell) and the impedance in the RC circuit such that (2 2 6 ) where the temperature of the resistor is T R and f is the detection bandwidth. By taking Equation 2 2 6 and dividing into from Equation 2 2 5 the S/N is (2 2 7 ) Superposition multiplies the effect of N i ons many times over causing the signal to increase ov er the noise. 53,54 (smaller m/q IR =V) Resolution is defined as the full width at half the height o f the peak (FWHM) m 1/2 and


50 the resolving power given by In terms of the corresponding angular frequency, the derivative of c with respect to mass yields, (2 2 8 ) where (2 29 ) The increased number of da ta points generates better frequency discrimination, which decreases 1/2 and improv es the mass resolving power and resolution Resol ving power also increases inversely with m/q and is enhanced by the magnetic field strength. 38 The signal is recorded in t he time domain as the free induction decay (FID) or transient response. This signal is often damped and broadened over time due to ion collisional interactions associated with cell pressure conditions as well as radial ejection from the trap (Figure 2 6). But the remaining signal is converted through the use of a fast Fourier transform algorithm into a c orresponding spectr al plot of the amplitude vs. frequency. Amplitude is equivalent to the number of ions of the same m/ q and the frequencies are directly re lated to their corresponding mass to charge. Prior to FT, the FID is apodized to correct broadened profiles of the peak bases. The cell is the heart of the mass spectrometer; however, the heart cannot beat without ions Ion Sources I on sources for the FTIC R MS create gas phase ions for detection. This is accomplished with a number of devices that are categorized in two groups: internal and external. When the molecules are introduced directly into the trap by a leak valve or probe, internal ion sources, such


51 as electron impact, chemical ionization and photo ionization, are often employed to ionize neutral species. External ionization sources like matrix assisted laser desorption ionization (MALDI) and electrospray ionization (ESI) form ions outside cell. The generated ions are injected through regions of differential pumping, where a series of ion optics and/or quadru/hexapole ion guides steer the charged molecules to the lower pressure region of the cell. Internal I on S ources Volatile molecules are introduce d by means of inserting a container into a sealed port from which a leak valve allows the neutrals to enter. Nonvolatile components are often inserted on heated probes to be vaporized. In both cases, the internal ion source interacts with the molecule in t he cell to produce ions. Electron impact (EI), one of the earliest ionization methods, uses a high energy electron beam (~70eV) centered on the molecules to remove a valence electron (positive charge) or a low energy electron beam that results in the captu re of an electron (negative charge). 55 57 Multiple ion fragmentation often occurs due to the excess energy absorbed by the molecule and can promote chemical ionization of uncharged species and ion molecule reactions. Chemical ionization (CI) is a softer te chnique, which ionizes the analyte through charge proton, or larger group transfer from a reagent gas in excess ( methane, ammonia, etc .) or via another radical/ionized molecule. 58 60 Photons (visible light, ultraviolet lasers, etc.) with sufficient energy can cause the ejection of an electron in the process known as photoionization. 61,62,63 A type of photoelectric effect, the probability of photoionization depends primarily on the threshold energy and the cross section of the interaction. 64 External I on S o urces The two most widely used external ion sources today are MALDI and ESI. Matrix assisted laser desorption ionization (MALDI) is a technique that marries the features of fast atom bombardment (FAB) and laser desorption ionization (LDI). Like FAB, the a nalyte is supported


52 in a matrix. This design enhances the transfer of energy from a laser beam aimed at the matrix surface (LDI) to the analyte, thus promoting desorption and ionization. Electrospray ionization (ESI) applies electrochemical and thermospray principles. A solution (sometimes acid or base modified) is passed through a charged capillary/needle (the working electrode) and exits as a nebulized spray (thermospray) into an electric field between the needle and a plate mounted on the opposite side ( counter electrode). In the electric field, the spray breaks down into very small droplets. Some of these remaining droplets (and already ionized components) then pass through a heated metal capillary (and/or heated drying gas) to emerge as gaseous ions. M a trix assisted laser desorption/ionization T hree crucial areas are necessary for an understanding of M atrix a ssisted l aser d esorption/ i onization ( MALDI ) : the makeup/preparation of the m atrix the function of the laser, and their interaction. As previously mentioned, matrices are designed to facilitate desorption and ionization. The substrate is an analyte specific acid ( 2,5 dihydroxybenzoic acid DHB), which is m 2,6 dihydroxyacetophenone (DHAP) are two common matrix substrates for saccharides and peptides, respectively. 65,66 These ingredients are mixed with the analyte in experimentally determined ratios (typical ly ~ 500 to 5000:1, substrate:analyte) and are often spiked with Na + K + Ag + Cu + or other monovalent cations to optimize and control ionization characteristics. 67 69 Relatively volatile solvents like acetone are used to dissolve the matrix, which is then spread out evenly in target spots on polished semiconducting plates. Small, even layered crystals, ideal for light absorption and matrix interactions, are produced through g entle heating of the plates


53 The plate is then mounted on a stage and a laser is focused on it to a 100 200 m spot size (Figure 2 7). 70 Pulsed at 3 10ns, the UV Nitrogen (UV MALDI) and, at 6 200ns, the IR YAG (IR MALDI) lasers are common and provide the energy absorbed by the sample. This energy is distributed from substrate to analy te causing desorption from the surface. Ionization occurs either 71 Internal ionization is promoted through analyte derivatives, matrix interactions (energy resonance transfer, metal additions, et c.), and changes in pH. 72 Post laser gas phase photoionization, electron capture, and charge exchange are additional means of external charging. 64,73 Power attenuated, the laser is tuned to supply an optimal amount of energy for desorption and ionization a bove the threshold laser irradiance of ~ 74 Each pulse is repeated until enough ions are collected for a clean low S/N spectrum. By raising the laser irradiance, examining the fragmentation yields, and piecing together the fragment p uzzle, MALDI has been be used for structural characterization of N linked glycan (Man) 6 (GlcNAc ) 2 in DHB amo ng other saccharide complexes. 75 Matrix assisted laser desorption ionization, though useful, does present problems, including preparation of the matr ix, laser alignment and the inclination to produce only single charge states. These are among the reasons why ESI has gained a more prominent role as an external ion source. E lectrospray ionization Electrospray ionization (ESI) occurs in three stages: the creation ions. A pump assisted syringe, or liquid chromatography (LC) instrument delivers a solution from the transfer line to the high voltage capillary. At t he capillary 3 5 kV is applied (working


54 electrode) creating an electric field between the capillary tip and a parallel plate (counter electrode) placed 0.5 to 2cm away. At a strength of ~10 6 the electric field is given by (2 3 0 ) where V c is the a pplied potential, r c is the outer radius of the capillary, and d is the distance between the working and counter electrodes. 76 In the positive ion mode, the capillary voltage is positive; the anions move toward the i nner surface of the capillary surface as cations continue to the tip. The voltage and electrophoretic movement are opposite for negative ions. Cations at the Tay lor cone (Figure 2 8), where further oxidation occurs (reduction for negative ions). 77 Once the tip of the cone builds up enough charge density, the force of the electric field overcomes the opposing solution surface tension and a charged droplet leaves in the direction of the electric field. Consistent spray comes from the controlled rate of charged droplet formation. This in turn is a product of proper solution chemistry, reasonable solution flow rates, and the appropriate voltage to oxidize/ionize the el ectroactive species. The rate is best characterized by the electrospray current given as (2 3 1 ) where the surface tension and dielectic constant are given a value H v f and specific conductivity is s 77,78 As the droplet travels in the field, the solvent begins to evaporate. surface tension or the Raleigh limit specified as (2 3 2 )


55 where the excess charge is q R e o is the is the surface tension, and R is the radius. 79 Coulomb fission then occurs, resulting in two droplets. According to droplet jet fission, this happens at the elongated end of the comet shaped droplet as it slowly breaks down the volume into smaller components (Figure 2 9). 77 A heated gas and/or metal capillary helps the process, although two competing processes explain the final ion construction. The charge residue model (CRM) states that the desolvation continues until the solvent disappears. 80,81 The ion evaporation model (IEM) maintains that the electric field provides enough energy for the ion to escape the confines of the droplet. 82,83 Many of these ions hit the opposing plate (counter electrode) and are reduced (or oxidized if they are negative ions) completing the electrochemical circuit maintaining the current. However, a significant number enter the mass spectrometer to be analyzed in the ICR cell. By adjusting the applied voltages, the pH (proton ating or deprotonating agent) the solvent polarity, and/or adding a cation(s) or anion(s) the charge state of the analyte can vary greatly. This helps place large m/q frame of reference and allows for interesting multiple charge state experiments. For example, electrospray ionization has been used to study molecules the size of myoglobin, a mass of 16,951 Daltons, with charge states of +8 to +23, to give a complete, although at times overwrought, spectrum. 84 The addition of microspray, varying gas assisted systems, and other designs further help s in spray efficiency and control. Ion Accumulation and Ion Optics O ften beginning at atmospher ic pressure, ions from an external ion sources are subject to various stages of pumping as they are guided through each region to enter the cell. Ions exiting the heated metal capillary pass through a set of skimmer s that help concentrate them by removing excess neutrals. These ions, all with different kinetic energies, are collected in a


56 hex apole where they are held by a tra pping voltage for 500 to 2000 ms. This allows the ions to assume uniform energy and to accumulate, maximizing ion concentration. Afterwards the ions are ejected from the hexapole trap as a group toward the cell. Ions are gu ided to the analyzer trapping cell in the Bruker FTICR mass spectrometer through cylindrically shaped electrostatic plates or einzel lenses. As seen in Figure 2 10, the ion optics designated PL 1, FOCL1, PL9, and FOCL2 are hollow cylinders whose primary pu rpose is to keep the electrostatically repulsive ions focused by application of like potential. The split cylinders DPL2, DPL4, XDFL, and YDFL, steer the ions into place through differentially controlled potentials with HVO. Octapole ion guides serve the s ame purpose in other FTICR MS instruments. The principles of IRMPD in an FTICR MS cell are explored in C hapter 3 Figure 2 1. Ion trap configurations are: ( a ) cubic, ( b ) cylindr ical, ( c ) cylindrical w/segmented endcaps, d&e opened ende d with change in capacitance coupl ing f dual, and g matrix where excitation is abbreviated E, detection D, and trapping T. [Reprinted with permission from John Wiley & Sons, Ltd. Marshall, A. G.; Hendrickson, C.L.; Jackson, G.S. 1998. Fourier Transform Ion Cyclotron Resonance Mass Spectrometry: A Primer Mass Spectrom. Rev. (Volume 17 Page 12 Figure 12 ).]


57 Figure 2 2 A quadru polar cylindrical trap consisting of excitation and detection plates Figure 2 3. The cyclotron motio n of positive and negative ions. [Reprinted with permission from Elsevier. Marshall, A.G.; Hendrickson, C.L. 2002. Fourier transform ion cyclotron resonance detection: principles and experimental configurations Int. J. Mass Spectrom. (Volum e 215, Page 60, Figure 1).] Detection Plates Excitation Plates R C


58 Figure 2 4. The pictures define the outside ion orbit where the ion cyclotron motion v c is coupled with the magnetron motion v m providing the variations in the motion as the ions oscillate v T between the two trapping plate s [ Re printed with permission from John Wiley & Sons, Ltd. Marshall, A. G.; Hendrickson, C.L.; Jackson, G.S. 1998. Fourier Transform Ion Cyclotron Resonance Mass Spectrometry: A Primer Mass Spectrom. Rev. (Volume 17 Page 14 Figure 13 ).] Figure 2 5. Ion cyclotron incoherent motion (left center) is excited coherently exciting the ions out to a higher radius (left) with the application of a resonant rf electric field The final radius (right) coupled with coherent motion leads to detection through an im age current induced on the detection plates which is then converted to voltage through an RC circuit. [Reprinted with permission from John Wiley & Sons, Ltd. Comisarow, M.B.; Marshall, A. G. 1996. The Early Development of Fourier Transform Ion Cyclotron Re sonance ( FTICR ) Spectroscopy J. Mass Spectrom. (Volume 31 Page 583 Figure 1 ).]


59 Figure 2 6. C omplete picture of the FTICR signal production, from detection, free induction decay (FID) of the time domain si gnal, Fourier transform of the signal (f requenc y spectrum not pictured) to m/q spectrum. [Reprinted with permission from the author. Hixson, K. K. 2000. FTICR Theory Environmental Molecular Sciences Laboratory. http://www.ems Vol ume 2008 (July 2009) .] Figure 2 7 The UV MALDI source depicting the power attenuated laser source, plate/matrix sample and plasma [Reprinted with permission from Elsevier Mamyrin, B.A. 1994. Laser assisted reflection time of flight mass spectrometry. Int. J. Mass Spectrom. (Volume 131, Page 5, Figure 3).]


60 Figure 2 8 The picture of electrospray as a n electrochemical cell with the Taylor cone formed due to the competing cohesive forces of the dropl et and the applied electric field. When charged droplets come close to the Raleigh limit they are separated from the Taylor cone and begin to break down further as they move toward the counter elect r ode and mass spectrometer entrance. [Reprinted with permi ssion from the American Chemical Society. Kebarle, P.; Tang, L. 1993. From ions in solution to ions in the gas phase the mechanism of electrospray mass spectrometry Anal. Chem. (Volume 63, Page 974A, Figure 1).] Figure 2 9 The droplet jet fusion mode l depicts the comet sh aped droplet as more likely to undergo co ulomb fission at the tail end. [Reprinted with permission from the American Chemical Society. Kebarle, P.; Tang, L. 1993. From ions in solution to ions in the gas phase the mechanism of electr ospray mass spectrometry Anal. Chem. (Volume 63 Page 977A, Figure 3).]


61 Figure 2 10. Focusing and guiding ion lenses in the Bruker FTICR mass spectrometer. [Reprinted with permission fro m Contreras, C.S. 2008. Carbohydrates and Amino Acids: Infrared Multiple Photon Dissociation Spectroscopy and Density Functional Theory Calculations. Ph.D. dissertation (Page 66, Figure 2 8) University of Florida, Gainesville, Florida.]


62 CHAPTER 3 I N FRARED MULTIPLE PHOTON DISSOCIATION THEORY Fifty years ago the scientific community believed infrared radiation from a blackbody or light source was incapable of causing molecular dissociation. The vibrational picture had been that of isolated diatomic b onds oscillating with minimal interactions within a larger molecular framework. From this perspective, it follows that even though a photon in resonance with a anharm onic vibrational energy ladder, it moves out of resonance and energy is no longer transferred. Under this paradigm, the only way a bond would break is through the application of consecutive frequencies resonant with the anharmonic vibrational levels until critical bond energy is reached, causing dissociation. This proved impossible for the instruments of the day. However, with the development of high power IR light sources, such as the CO 2 laser, infrared radiation at a fixed frequency proved sufficient to dissociation. 85 In order to understand these processes, the theories of vibrational motion, light matter interactions, and IRMPD spectro scopy/quasi BIRD will be reviewed. Vibrational M otion Molecular motion is expressed by three interacting components: translation, rotation, and motion around the molec within the molecule. Vibrational movement builds from diatomic motion, to linear combinations of diatomic vibrations (modes), to an interconnected collection of modes (that describes most polyatomic molecules). In this section, each of these concepts is considered as well as their harm onic and anharmonic behavior.


63 Diatomic V ibrational M otion Diatomic molecul ar vibration s are the building blocks for the complex motions of bonds. Startin g with the application of the classic harmonic picture to quantum and anharmoncity treatments an understanding of int ra molecular molecular motion is possible. Classical mechanics picture Molecular vibrational motion is often pictured as two atoms attache d by a spring that oscillates harmonically from an equilibrium position This classical viewpoint gives the diatomic oscillation frequency ( f r eq ) as (3 1) where k f is the bond force constant (a measurement of the bond strength), a nd is the reduced mass of two atoms: one with the mass m 1 the other with mass m 2 such that (3 2) The energy of a bond s harmonic motion is broken down into potential and kinetic energ y parts. P otential energy reaches a maximum a t the two extreme ranges of motion K inetic energy reaches its maximum at the equilibrium point In between, the energies are characterized by a potential loss and kinetic gain of energies as the bond moves from the pure potential energy to kinetic energy point and vice versa Continuation of this action is due to the elastic restoring force between the two atoms. The restoring force ( F restore ) is equal to the product of the spring s f orce constant k f and the displacement ( R ) from equilibrium ( R e ) such that (3 3) where x= R R e The vibrational (harmonic) potential energy ( U ) is given by (3 4)


64 and describes the motion of a parabola which is the characteristic movement of a harmonic os cillator (Figure 3 1) 86,87 To better describe this picture quantum considerations are necessary. Quantum c onsiderations A solution to the quantum picture begins with the time dependent Schrdinger equation, (3 5) In Equation 3 5, is the wavefunction described by the electronic e and nuclear coordinates N time is t and H is the Hamiltonian. The Hamiltonian is described by potential and kinetic terms such that H= K Tot + U Tot = K e + K N + U ee +U NN +U Ne (3 6) The total poten tial energy U tot consists of energy contributions from the repulsions between the electrons ( U ee ), the nuclei ( U NN ), and the attraction between nuclei and electrons ( U Ne ) K inetic energy K Tot contributions are the motions of the electrons ( K e ) and nuclei ( K N ) If the potential energy is considered independent of time Equation 3 5 is simplified by the separation of the time and spatial variables to give (3 7) The general form of the time in dependent Schrdinger equation becomes (3 8) where the operator H acts on the wave function to give the eigenvalue E and the original function ) Further simplification is found through the Born Oppenheimer approximation which states that the large slower mo ving nuclei are considered fixed with respect to the fast moving electrons, in order to separate and solve the electronic and nuclear components separately. The first part of the solution involves the use of the electron Hamiltonian ( H e )


65 H e =K e + U ee + U Ne (3 9) Solving the electronic time independent Schrdinger equation produces the electronic potential E e for the arrangement of the positioned nuclei. The solutions for a number of possible nuclei arrangements help define a potential energy surface whose m inimum identifies the molecule s ground state equilibrium conformation. The effective potential is then incorporated into nuclear Hamiltonian H N = K N +U NN (3 10) to give the nuclear component of the time in dependent Schrdinger equation (H N +E e (N) ) (N )= E N (N) (3 11) The solutions describe the translational, rotational, and vibratio nal molecular motions. In terms of the simple harmonic oscillator, the time dependent Schrdinger equation is given as (3 12) where and h is Plan c Separating the time ( ) and spatial ( ) components results in (3 13) D ividing Equation 3 13 gives (3 14) and the time and spatial terms are isolated. For a time ind ependent function the right side of E quation 3 14 is constant meaning the left side is as well. Furthermore, a s the left side of the


66 equation is an expression of the Hamiltonian operator with the dimensions of kinetic and potential energy it follows tha t their product, the eigenvalue, is the total energy such that (3 15) and (3 16) Solving the differential Equation 3 16 gives the time dependent wavefunction as (3 17) and a solutio n to differential E quation 3 15 gives the energy values (3 18) Equation 3 18 indicates that the energy values are quantized ( n ) and evenly spaced along the harmonic vibrational ladder (Figure 3 1). The ladder begins with the lowest possible energy or the zero point energy and moves from this position to the higher energy levels, whose difference is defined by 88,89 Anharmonicity The harmoni c picture adequately describes the behavior near the bottom of the potential energy well. However, as the vibrational motion is excited a nd the system move s up the energy ladder the restoring force and the displacement are no longer proportional. The resu lt is that the displacement from equilibrium becomes greater, and energy levels become more closely spaced redefin ing the harmonic potential energy curve as anharmonic ( Figure 3 1) At longer internuclear distances (i.e. high vibrational levels), the res toring force eventually becomes zero and the bond dissociates with a dissociation energy defined by the potential energy


67 from the ground state. A means of describing th e anharmonic potential energy curve is through the Morse potential, (3 19) w here D e marks the curve minimum, c is the speed of light, and a is the potential energy curve width, defined by (3 20) The potential energy function can be placed into Equation 3 8, the time independent Schrdinger equa tion, and solved to produce an equation for vibrational energy corrected by an anharmonic factor such that, (3 21) In Equation 3 e is given as (3 22) Equations 3 21 and 3 22 illustrate that the anharmonic correction is subtracted from the harmonic term n gets larger (overall energy increase) vib becomes smaller decreasing the gap between each successive vibrational energy level. Although the Morse potential is an approximation (additional corrective terms to Equation 3 21 are required when a better fit is necessary), it does provide a fairly good match for simpler systems It predicts the maximum vibrational level for dissociation to be (3 23)


68 and the dissociation energy ( D 0 ) as (3 24) where E 0 is the zero point energy. 90,91 P olyatomic Vibrations Di atomic molecules have only one stretching motion Larger molecules contain multiple vibrational motions with bond lengths and angles that change with respect to other vibrations. To represent each of these movements and their interactions precisely require s complicated kinetic and potential energy treatments. In order to simplify the problem, linear combinations ( Q V ) of vibrational movements are employed. For example, the linear combinations for carbon dioxide s two C O stretches (Figure 3 2) are (3 27) and (3 26) where m c or o is the mass of the carbon or oxygen, m is the total mass (=m c +2m o ), q n are the weighted coordinates, ( x n is the displacement of atom n ) n is equal to 1 (oxygen atom), 2 (carbon atom) and 3 (oxygen atom), and is the effective force constant. The dynamics of the a nti symmetric ( Q 1 ) and symmetric ( Q 2 ) stretches (Figure 3 2) are seen as the sum and difference of mutual and opposing directions of the C O stretching motions, respectively. A new effective force cons tant corrects for the dynamic combinations of these motions. Linear molecules like CO 2 have 3N 5 (N is equal to the number of atoms, i.e. 3(3) 5=4) degrees of freedom (DOF) and nonlinear molecules have 3N 6 normal vibrational DOF, each corresponding


69 to a linear combination of vibrational motions. Normal coordinates for the translational Q T and rotational Q R motions are subtracted from the total (3N) DOF. Linear and nonlinear molecules both have 3 translational DOF and their rotational DOF are 2 and 3, tota ling 5 and 6 nonvibrational DOF, respectively. Besides the symmetric and assymmetric stretches other molecular normal vibrational modes include rocking, wagging, twisting and scissoring motions or varying combinations of the six mo vements The Hamiltonia n for the harmonic normal modes is represented as a sum of terms, (3 27) The molecule s vibrational wavefunction is a product of each mode s wavefunctions giving the Schrdinger equation (3 28) The solution of Equation 3 28 gives total harmonic vibrational energy of the molecule as (3 29) Similar to the diatomic picture, an anharmonic correction to the Schrdinger equation can be performed However, further corrections are more involved because they account for complications, such as electrical anharmonicity from the interplay of dipole moments contributing to the corrective terms. 88 92,93 Interaction of Light and Matter Electromagneti c (e/m) resonance occurs when the oscillating electromagnetic (e/m) radiation of light of a given frequency comes into contact with a vibrational mode that has the


70 same frequency as the oscillating e/m radiation e/m radiation results from a change in its electric dipole moment ( ) induced and/ or amplified by the vibration. Once the light energy is absorbed, this energy results in a higher vibrational energy level. Normal modes that have the ability to absorb energy are infrared active. Other modes that do not affect a change in d ipole moment are infrared inactive. Diatomic E lectric D ipole M oment In a one dimensional oscillator, the electric dipole moment depends on the positions of the electrons and nuclei. An approximate relationship consider s the dipole moment as a product of t he two charges dq and their displacement from the equilibrium position as R = R e +x, such that (3 30) where o is the electric dipole moment at the equilibrium distance. From Equation 3 30 it follows that, (3 31) w here is the expression for the transition dipole moment. The equilibrium dipole moment cancels because are orthogonal ( when the nuclei are at equilibrium The addition of the relationship allows Equation 3 31 to be rewritten as (3 32) Equation 3 32 illustrates that if there is no change in the dipole with a change in displacement then the transition dipole moment is zero. Following from utilizing the


71 appropriate H ermite polyno mial wavefunction s the selection rules for harmonic vibrational transitions are An expansion of Equation 3 32 leads to an expression more appropriate for the anharmonic case, (3 33) As the vibrational energy is increased, the displacement x becomes greater and is no longer proportional to the dipole moment. When this occurs the second term in Equation 3 33 becomes more important and the associated electrical anharmonicities allow transitions of and so on. 90 94 These transitions contribute to the first overtones or second harmonics in vibrational spectra. Polyatomic E lectric D ipole M oment In a similar fashion the normal modes depend upon the displacement of charge over the linear combin ations of vibrational modes such that, (3 34) The transition dipole moment for each normal mode is (3 35) In terms of the CO 2 molecule s normal modes Q 1 and Q 2 the changes in x along the normal coordinate (the direction of motion) are active if they will produce a change in For Q 1 the asymmetric stretch, opposing and attracting motions causes a net change in the charge distribution which in turn indicates a change in the dipole moment ( dxdq = ) of the molecule. The symmetric stretch Q 2 however, does not produce a net ch ange in charge distributio n and, therefore, is inactive. 88 92 94


72 Similar to the diatomic case, anharmonicities are also taken into account by additional terms in E quation 3 35. Along with overtones and harmonics combination bands ( ) and Fermi resonances have to be taken into consideration. Combination bands arise due to the mixing of similar normal modes and produce peaks of lower intensity than fundamental transitions ( to ). The exceptions are Fermi resonances that occur due to interactions between two combination normal modes (i.e. fundamental and an overtone) of the same symmetry and similar energies. This interaction results in the mixing in to the overtone, producing overtone peaks with higher intensity than the fundamental and complicat ing the assignment of peaks for vibrational spectra. 92 94 Infrared Photodissocation Spectroscopy Absorption spectroscopy measures the energy of electromagnetic radi ation absorb ed or transmitted by a sample. Electromagnetic radiation in resonance with active vibrational modes of strong/weak transition moments is absorbed strongly/weakly resulting in an absorbance ( A ) of (3 36) where I o is the incident light intensity, I is the transmitted light intensity, is the molar absorptivity (molecular extinction coefficient), [C ] is the concentration, and L cell is the e/m radiations path length through the sample. Equation 3 36 is c ommonly known as the Beer Lambert law. A vibrational spectrum consist s of peaks (absorbance) or valleys (%T) corresponding to the vibrations resonant modes, with intensities varying according to the strength of the change in dipole moment. Absorption spec troscopy requires a minimum concentration of 10 10 molecules per cm 3 for proper detection. In the cell of the FTICR mass spectrometer, ion concentrations are


73 approximately10 6 ions per cm 3 (due in part to space charge effects, size of the trap etc.) produci ng ([~10 6 ions per cm 3 ], ~1 x10 4 M 1 cm 1 L cell ~12 cm) a n absorbance of 2 x10 10 a value far below the limits of detection. 95 Under these conditions, probing the vibrational modes requires a different approach. By monitoring the dissociation products a t varying infrared wavelengths, vibrational spectra are acquired, although the intensities are now products of light dipole interactions and the complex IRMPD mechanisms. 95 97 Multiple Photon Dissociation As mentioned previously, in the early years of spec troscopy it was thought that all resonant vibrational absorption occurred as part of a stepwise coherent mechanism. Under this scheme, a vibrational mode in resonance with an IR laser of a given frequency undergoes excitation but as the excited molecule cl imbs the vibrational energy ladder, the frequency of the IR laser becomes out of resonance and energy can no longer be absorbed (Figure 3 3). This process is still true for some molecules (with high dissociation energies and limited DOF), but it no longer applies to the extent once thought. Multiple photon dissociation is based on the generally in coherent mechanism that takes place in three stages, characterized by v ibrationally coherent excitation, internal vibrational redistribution into the quasicontin u u m, and photodissociation, as seen in Figure 3 4. 98 105 At lower vibrational levels the relatively harmonic behavior and rotational vibrational coupling allow for the resonant transfer of energy from the IR laser to the molecule at the fundamental frequenc y. As the excited molecule climbs the ladder, internal vibrational redistribution (IVR) of the energy occurs through anharmonic coupling of the vibrational mode with a number of background modes that in turn distribute the energy to the other states, etc. This process continues with an increasing number of interacting states which shift s and broaden s resonant absorption levels until a quasicontinuum of vibrational states is reached. In the quasicontinuum,


74 photons of all IR wavelengths are able to be absorbe d as the mutually interacting high density of states makes the molecule resonant. During this time, the IR laser, still focused at the fundamental vibrational mode, is continually producing photons that are being absorbed and redistributed until the molecu le reaches a critical energy for the dissociation of the weakest bond. This is physically allowed because the time frame for IVR (~ ps) is much shorter than the time required for absorption of a photon at the fundamental and the energy build s up at a fast er rate than stimulated or collisionally induced emissions (~ms) are able to counter. At room temperature, ions in equilibrium with their environment have multiple populations of differing energy, governed by their Boltzmann thermal distribution. Thus, the energy absorbed by the ions E abs ( product of the laser s fluence and the ion s absorption cross section) does not have the same effect on every ion in the distribution. Add this to the coupling of the vibrational rotational modes and the increasing anha rmonicity of excited ions, and the spectrum becomes broader and shifts to the red (i.e., to lower energies). This effect is decreased at lower fluences which tend to perturb the system less and also reduce complicating issues, such as Rabi broadening Be cause the efficiency of the multiple photon dissociation depends on the closely spaced ( high density ) states, larger molecules are generally better candidates for this process In fact, at room temperature some large molecules are already in the quasiconti nuum and dissociate readily. Conversely, small molecules with low density of vibrational states are more prone to stop absorbing laser energy at the anharmonic bottleneck. However, with higher laser fluence, even some of these reluctant molecules are capab le of IRMPD. The interaction of a vibrational level with high energy laser e/m field causes oscillation in the fixed vibrational motion, bringing other frequencies into resonance. 85


75 Actio n S pectroscopy In order to obt ain an action spectrum in FTICR MS instruments, the laser beam is first aligned with the ion cloud in the ICR cell. The ion signal is optimized, isolated, and the laser beam triggered to coincide with the ion s retention time. Mass spectra, at a series of consecutive wavelengths (at given powers and set times), are acquired and any IRMPD products are noted. The action spectrum is then created from a plot of the natural log of the power corrected relative abund ances of the total ion population divided by the relative abundance of the parent ion population versus wavelength. Positions of the spectral peaks are primarily a result of resonant laser/vibrational mode interactions for a given wavelength. Peak intensit ies are based on ion dissociation, which depends on the ion s initial internal energy the interplay of the resonant frequencies and the laser fluence with the vibrational transition dipole moment and the IRMPD process. A mong the first to implement these techniques in Fourier transform ion cyclotron resonance mass spectrometry were Dunbar, Eyler, and Beauchamp. 106 110 Of particular interest was the use of a two laser CO 2 system, which allowed Watson and Eyler to overcome the lack of power (causing low or l ack of ion dissociation) in the tunable laser, by following the tunable laser with a high powered pulsed laser. The first laser excites on resonance vibrational modes to quasicontinuum states. Now resonant with all infrared frequencies within the continuum the ions absorb photons from the fixed off resonance pulsed laser and dissociate. 111 The spectra from these experiments are a bit broader than spectra obtained without the assistance of a second laser but illustrate clearly defined peaks. Multiple fragm entations It is not uncommon for an ion to have a number of competing channels for dissociation, with multiple fragments created by similar dissociation energy profiles. However, most multiple fragmentation pathways are the result of excess internal energ y


76 carried over from IRMPD. Here, new low energy pathways are made available with each fragment, often creating a congested spectrum, especially at high laser fluencies Products from larger molecules are excellent candidates for these processes. Multiple f ragments can still be utilized to form an action spectrum, although sensitivity and peak resolution suffer due to the uncertainty in the final ion populations and the complicated energy picture. Messenger technique In order to correct for the above, lower laser fluences are used to probe the resonant vibrational frequencies of weakly bound ligand molecule complexes in what is known as the messenger technique. Designed to lower the number of photons required to dissociate the ligand without perturbing the v ibrations of the molecule, the messenger ligand increases the density of states to favor dissociation T he spectra also produce very sharp features, a result of increased sensitivity ( a one or two photon dissociation process ) due to the relatively low co upling of the van der Waals forces in the ligand attachment 95,96 Of interest in these processes are the kinetic and energetic pictu res for ion dissociation, which a re specific to the phosphopeptide dissociation experiments discussed later in Chapter 6 U n imolec u lar Dissociation Kinetics In the ICR cell, when the ions a bsorption of IR photons (from a low intensity CO 2 laser energy source) is equivalent to the stimulated and spontaneous emission, a thermal equilibrium with the surroundings is reached. This creat es a Boltzmann like distribution of energy with an effective temperature. Dissociation of an ion s lowest energy bond occurs when the ion s energy exceeds a threshold energy ( E t ) typically associated with the high energy tail of the Boltzmann distribu tion. The result is loss of a neutral, producing a daughter (D) peak (parent neutral) and a parent (P) peak of reduced intensity. The effect is carried out over a wide range of wavelengths to produce action spectra. Alternatively, if the laser is kept at a given wavelength and power while the irradiation times are varied, the outcome is used to obtain d issociation rate constants


77 ( k d ) from the slopes of ln{[P]/([P]+[D])}or ln[Relative Ion Abundance] versus time plots. Activation en e rgies E a are then deter mined by the slope of the line from a modified Arrhenius plot of the natural log of the rate constant ( k d ) versus the natural log of the p ower associated with each k d such that (3 38) The constant ( C ) represents the relationship between laser power and temperature and is determined using two different approaches, which will now be discussed. Laser Intensity and Temperature A CO 2 laser of sufficient fluence with an IR wavelength in resonance with an absorbing vibrational mode can i mitate a blackbody radiator with an effective temperature. However complementary temperature of a blackbody radiator is difficult. In order to accomplish this, two different models have been proposed independently by Robert Dunbar and by Kolja Paech and Evan Williams. The Dunbar m odel In the early 1990s, Dunbar suggested that an FTICR mass spectrometer, with long, relatively collision free retention times, provides an excel lent means to study slow kinetic information that evolves from ions irradiated in the cell. 112 114 Carbon dioxide lasers, with wavelengths in the congested vibrational region, served as source of choice, as many molecules are already in reso nance. Within this setup, an IR laser can be viewed as a thermal source leading to a thermal analysis (i.e., Arrhenius) of the affected systems, provided that certain conditions are met. The continuous wave (cw) laser power must be kept relatively low and serve as the principal energy source in the thermal equilibrium like processes (equivalent photon absorption and emission), with the kinetics governed accordingly. And the Einstein coefficients define the system, where spontaneous relaxation is more import ant than the


78 corresponding collisional processes. 113 Dunbar derived Equation 3 38 by approximating the relationship of the intensity and the blackbody temperature through Plan c equation, (3 39) where I ( h ) is the intensity of the laser (power) at a given frequency I o is the absolute light intensity (constant set by the data), and T is the effective temperature. The derivative of the natural log of Equation 3 39 with respect to resul ts in the relation, (3 40) In comparison, the derivative of the natural log of the Arrhenius equation with respect to gives (3 41) Solving Equations 3 40 and 3 41 for and equating these terms (with I ntensity = Power) provides the activation energy as (3 42) where q is estimated by the partition function at frequency and is often a value between 1.00 1.10 for temperature ranges associated with these laser dissociation experiments (often taken as


79 the average, 1.05). The constant expressed in Equation 3 38 is simply a multiple of q h and (i.e. at a wavelength of 9 43 cm 1 the constant is 8.14). Dunbar expressed two potential issues with this treatment. The first involved the interaction of strong radiative intensities near but not on resonance with the CO 2 laser. Dunbar dismissed this issue as, at most, a scalar mu ltiplier to blackbody results. Thus, as long as equilibrium conditions are met, the activation energy is not affected. The second problem involved the possibility of extensive unaccounted for IR radiative ( spontaneous ) emissions from vibrational intensitie the number of nondegenerate vibrational DOF, the greater the possibility of misrepresenting the distant modes and the number of emitted photons, resulting in low activation energies. A second treatment by Paech and Williams compensates for these emissions by using a steady state approximation, and they obtained a new constant that applies to peptides and proteins. Paech and Williams The P a ech and Williams model 33 begins with the assumption that the system is in thermal equilibrium ( k 1 1 ). Here the dissociation rate ( k d ) is small enough to assume that a relative steady state exists, according to the equilibrium condition, k 1 k 1 >>>k d ( t he rapid exchange limit, REX) such that ( 3 43 ) whe re and are the Einstein coefficients, ( ) is the energy density, and P ( )is the vibrational level occupation probability The Einstein and A coefficients are given by (3 44) and (3 45)


80 where I abs is the absorption intensity The left side of Equation 3 43 is related to stimulated absorption ( ) and the right side is related to both stimulated ( ) and spontaneous( A ) emission as they govern the movement of energy up and down multiple vibrational energy levels according to their respective probabilities, P ( T ) and ( T ). Further the occupation probabilit ies of an ion population with a Boltzmann distribution at an effective temperature T are characterized by (3 46) and ( 3 47 ) In Equations 3 46 and 3 47, n + 1 and n are quantum numbers that represent squares of the creation and annihilation operators. These terms correspond to the higher probability of absorption and emission at higher vibrational occup ation levels. Setting Equation 3 43 to zero gives (3 48) From Equations 3 46 and 3 47, [ ] =1. Using this simplification and removing the resonant laser frequency from the summation gives (3 49 ) Equation 3 49 is then rearranged to isolate ( laser ) (= I laser measured power) (3 50) Equation 3 50 was then computationally evaluated for four peptides (leucine enkephalin,


81 gramicidin S bradykinin, and meli ttin) at different temperatures, giving their related laser intensities In this procedure, the frequencies and transition dipole moments for each peptide were c omputed (AM 1) and the dipole moments (x1.8) were corrected. The corresponding wavelengths and the calculated transition probabilities, within a 100cm 1 region were added and averaged and the model based on Equation 3 50 executed. A plot of the natural log of the laser powers versus reciprocal of the temperatures was taken and the relationship esta blished through the slope( s ) such that, (3 51) Figure 3 5 illustrates that the slopes of all four peptides are very similar, deviating less than 9%, over a wide range of temperatures and laser powers Equation 3 51 is an expressio n of the relative power density (i.e., relative laser power). The relative power density removes the scalar multipliers, i.e., f rom Equation 3 50, as seen in the derivative of the natural log of the laser power (= Equation 3 50) wit h respect to such that, (3 52) Paech et al. believe th is relationship makes the proportionality constant s applicable to many different peptides above the REX limit. For instance, decreasing/increasing the pe ptide s size, though decreasing/increasing the degrees of freedom, does not significantly alter the relative intensities of absorption (peptides ~biological polymers) and since scalar multipliers are not a factor, a similar vibrational profile should be o bserved. Similarly, uncertainty between the


82 measured and actual laser power the peptides experience has no appreciable effect on the slope ( s ) Thus the proportionality factor s should be in close agreement with other similar peptides, if not peptides as a whole (within the REX), and applicable t o a range of power conditions. Solving Equations 3 41 ( except ) and 3 51 for and equating these terms the value for the Arrhenius activation energy is (3 5 3 ) Paech et al. compared Arrhenius activation energies from IRMPD simulated blackbody radiation utilizing Dunbar s and Paech and Williams constants with actual blackbody data (i.e., blackbody infrared radiative dissociation (BIRD) experiments, i n which the ICR cell is physically heated to a given temperature, producing black body photons for ion dissociation). The results are presented in Table 3 1 Activation energies, found through IRMPD experiments utilizing Paech and Williams treatment and t he BIRD experiments are in general agreement for all four peptides ( at different charge states) with one exception (the authors were unsure of the reason, and cite another identical experiment that illustrates agreement). Not counting the exception Dunba r s constant underestimated the activation energies by about ~40%. Paech and Williams effectively account for spontaneous emission as part of the radiative equilibrium of peptide molecules at the REX. However, Dunbar s model is relevant for smaller molecul es with lower density of states producing far fewer modes that spontaneously emit (i.e. stimulated emission >> spontaneous emission ). By examining key differences between the molecules dissociation kinetics, DOF, and transition states, molecular categori es are developed that aid the experimenter in determining the best course of action.


83 Molecular Categories Boltzmann distributions differ according to the size of the molecules (density of states) and the energ y of those states. Not surprisingly, the molecu lar dynamics and DOF greatly affect the nature of the distribution. In general, the smaller the molecule, the more likely that the distribution is severely truncated. However, with molecules of increasing size, the distribution begins to appear normal unti l an equilibrium state is reached. As a result, the assigned categories are somewhat fluid, as shown in Figure 3 6 for hydrocarbons, and should be approached carefully. To better understand the differences in the distributions, three molecular categories a re described in terms of their kinetics, degrees of freedom (DOF), transition states, and the models used to correct any deficiencies. 31,32 115 Small m olecule kinetics. The overall kinetic picture of unimolecular dissociation is given by, (3 5 4 ) where k 1 k 1 and k uni (= k d ) are the governing rate constants of absorption, emission, and dissociation, respectively. Each category is related directly to th e kinetic picture in Equation 3 5 4 and its effect on the Boltzmann distribution. In the small molecule category, thermal equilibrium is not maintained as k uni >>> k 1 116 119 Molecules of this size (no greater than 100 DOF) do not have the ability to reach a steady state with their surroundings, due to t he low and/or degenerate density of states. Thus, the energy builds up quickly and the lowest energy bond dissociates at a faster rate than relaxation. If the transition state is entropically favored or loose (i.e., direct bond cleavage reaction), dissocia tion is further promoted. The transition state is measured by the Arrhenius coefficient, where the entropy of activation is given by a simple rearrangement of k 1 k 1


84 (3 5 5 ) In the case mentioned A 14.5 a nd gr eater are common. 120 If these assumptions are reasonably true, the Boltzmann distribution is considered truncated and Equation 3 38 yields low activation energies, which are corrected by the m odified Tolman approach 113 states that the acti vation energy is equal to the average energy of the molecules that are reacting minus the average energy of the nonreacting reactant molecules. 121,122 Dunbar s new approach rearranged Tolman s equation in terms of the threshold energy, adding a few correct ive terms to give (3 54) where E t is the threshold energy (molecules that absorbed energy at or above the E t limit are removed), is the average energy of the population remaining (calculated from a Boltzmann d istribution) E rad is a correction factor for the dependence, and E depl is the depletion of reactant molecules within the high energy population given by E t h avg The corrective terms ( E rad + E depl ) partially cancel each other and can be represented by 300 cm 1 or just around 0.9 Kcal/mole. 123 The Arrhenius a ctivation energy is taken from the plot based on E quation 3 38 Average energy, is found by considering a n estimated value for E t and using it to truncate the calculated Boltzman n dist ribution of energies (Figure 3 7 ), and is obtained from the remaining energies. The distribution is computed from the vibrational frequencies (through computational means) and the ef fective temperature of the laser pulse through Planck s equation. Then, these values are applied to the Boltzman n equation, given by


85 (3 55) where P represents the probability that the molecule has internal energy E and W(E) is the density of states found through the direct count method of Beyer Swinehart. The average energy of the truncated Boltzma n n distribution is then determined and the value of E t is calculated using E quation 3 54 and compared to the estimated value Then E t is approximated again and is calculated until the estimated and calculated E t energies match The value for E t thus obtained is taken to be the true dissociation energy. Med ium sized molecul e kinetics The medium category finds the Boltz mann distribution as reduced, though not truncated, as the rate of dissociation is competitive with the rate of the spontaneous and stimulated emission of photons, k uni 1 124 126 In this case the density of states (~intermediate DOF) is greater than the value for an average small molecule, but it is not large enough to result in relative steady state equilibrium. Ions are removed at a fast enough rate that the Boltz mann distribution is perturbed (Figure 3 8 ). If the transition state is relatively tight with A 12.4 where simple but entropically unfavorable dissociation reactions occur (e.g., the McLafferty rearrangement ) the rates of reaction begin to disfavor d issociation. 120 This can move some of these molecules into the large molecule kinetic limit. Assuming this is not the case, m aster equation modeling is necessary to account for the distribution s defects. M aster equation modeling separates the overall mole cular kinetics into two categories, the rate of dissociation k d N i (0) and the rate of internal vibrational transitions such that (3 56)


86 where N represents the initial population fraction (Boltzmann distributi on) that corresponds to the energy level i for N i j for N J and dN is the change in the time dependent population. 124,127 The rate of dissociation is modeled by Rice, Ramsperger, Kassel, Marcus ( RRKM) theory illustrated by (3 57) where N (E E 0 ) is the sum of states h (E) is the density of states and E 0 is the RRKM threshold energy Similar to the Boltzmann calculation, the density of states can be considered through a direct count ing method with the frequency values taken from computational calculations 128 T he value of E 0 can be treated by the sudden death approach where any population above E 0 is considered to be dissociated. This is confirmed when dissociation above E 0 accelerate s rapidly thereby establishing the dissociati on point. The state transition rate constant ( k i,j ) measures the transfer of the ion population in to and out of the energy levels and is equal to the rates of absorption ( k abs ) and emission ( k emis ) of the IR radiation. The value for k abs is (3 58) In Equation 3 58, is the probability of energy occupation for the n th vibrational level in the m th mode. This probability is statistically similar to that given by E quation 3 55 The emission rate constant is (3 59) S ummation terms in E quation 3 58 and 3 59 consider all the modes of all the states with the condition that the i and j energy level difference be equal to the mode frequency. To clarify


87 the dissociation, absorption and emission r ates are added for each population within a 100 cm 1 frequency range and summed across all ranges. Using the frequency and intensities from computational calculations and substituting these values into the master equation, at a given effective temperature T dissociation of the ensemble is simulated at different times. Rate constants are then determined for the given temperature and the process is repeated for different temperatures to later form a theoretical Arrhenius plot. The m aster equation is then ad justed to better mimic the behavior of the experimental line This is done by altering the variables of the m aster equation until a fi t is found P arameters include : threshold energy E t radiation absorption ( I abs ), transition dipoles ( ) and specific tra nsition frequencies the most critical being E 0 from RRKM theory. The value for E o that satisfies the experiment is taken as the true activation energy. Accordingly the Arrhenius constant is modifi ed through transition dipole moment scaling factors and by altering the transition frequencies associated with the entropy of activation, until the intercepts are equivalent Master equation modeling and RRKM theory are used for all categories including the large molecular category discussed in the next section, which details one statistic al aspect of the IRMPD theory. Large molecule kinetics Kinetics in the large molecule category are governed by the principle k 1 >>>k uni or k 1 1 >>>k d 33,34 ,1 29 1 40 A large density of states (~high DOF) allows for the distribu tion of energy throughout many vibrational modes through intramolecular vibrational relaxation (IVR). The molecules accumulate energy more slowly and dissociate only at the high energy tail of the distribution, thus maintaining an equilibrium state with th e environment. If the transition state is very tight ( A 10 9.9 ) intricate reaction mechanism(s) create conditions that are very entropically unfavorable, further deterring dissociati on while promoting equilibrium. 120


88 Distribution of the Boltzmann energy is in a relatively steady state and the Arrhenius ener gies are reasonable in the condition otherwise known as the rapid exchange limits (REX). However, because the relationship between the temperature and the laser is modeled, the activation energy is considered relative. The type of laser involved in the ab ove processes is of some importance. Ideally infrared lasers should be precise, coherent, durable instruments, capable of high enough fluence to promote dissociation and the appropriate wavelengths to get the job done. As two types of lasers were used in the research reported in this thesis, an OPO laser for the action spectroscopy studies utilizing the messenger technique ( w ith CO 2 laser reinforcement for the CH stretches) and a CO 2 laser for the kinetics analysis, a review of their operations is presente d in Chapter 4. Figure 3 1. The potential energy diagram of the harmonic and Morse corrected anharmonic bond distances. [Reprinted with p ermission from the Free Software Foundation. Somoza, M. 2006. Morse potential. potential.png (April 2009). ]


89 Figure 3 2. Representations of the four vibrational modes of CO 2 molecule: asymmetric stretch, the doubly degenerate bending mod es, and the symmetric stretch Figure 3 3. T he coherent stepwise absorption results in the discontinuation of absorption of once resonant photons, due to the anharmonic bottleneck effect O O C C C C O O O O O O Molecular structure of Carbon Dioxide The asymmetric s tretch mode The bending mode The symmetric stretch mode


90 Figure 3 4. T he three stages of multiple photon dissociation [Reprinted with p ermission from Los Alamos Science, Los Alamos National Laboratory Lyman, J.L.; Galbraith, H.W.; Ackerhalt, J.R. 1982. Multiple Photon Excitation Los Alamos Science. (Volume 3, Page 78, Figure 11).]


91 F igure 3 5 T he plot of the rela tive intensity vs. the reciprocal of the temperature through the use of steady state approximations for four peptides; leucine enkephalin (triangles), gramicidin s (squares), bradykinin (diamonds) and melittin (circles). [Reprinted with p ermission from the American Chemical Society. Paech, K.; Jockusch, R.A.; Williams, E. R. 2002 Slow Infrared Laser Dissociation of Molecules in the Rapid Energy Exchange Limit J. Phys. Chem. A. ( Volume 106, Page 976 4, Figure 3).] Figure 3 6 The general ized effects of DOF, rate constants, and transition states on small, medium, and large molecular kinetics for hydrocarbons. [Reprinted with permission from John Wiley & Sons, Ltd. Dunbar, R.C. 2004. BIRD, Evolution, Principles,and applications Mass Spectr om. Rev. (Volume 23, Page 137 Figure 4) ]


92 F igure 3 7 The Boltzmann distribution truncated at the threshold energy (E t ) (where E depl = E t h avg ) according to the sudden death approximation for modified Tolman modeling for small molecule kinetics. [Repr inted with p ermission from the American Institutes of Physics. Dunbar, R.C. 1991 Kinetics of low intensity infrared laser photodissociation. The thermal model and application of the Tolman theorem J. Chem. Phys. (Volume 95, Page 25 42 Figure 3 ) .] Figu re 3 8 The equilibrium and nonequilibrium (depleted) Boltzmann distribution for medium molecule kinetics. [Reprinted with p ermission from American Chemical Society. Dunbar, R.C. 1994 Kinetics of Thermal Unimolecular Dissociation by Ambient Infrared Radia tion J. Phys. Chem (Volume 98, Page 8708, Figure 5). ]


93 Table 3 1 Comparison of Arrhenius activation energies from IRMPD exp eriments utilizing the Dunbar and Paech & Williams corrections, and values determined using blackbody infrared radiative dissociatio n (BIRD) experiments for four peptides: leucine enkephalin, gramicidin S, bradykinin, and melittin. The experiments are taken from the papers of Schnier et al. b 132 d ; 133 Busman et al. g; 134 Jockusch et al. c 136 h ; 135 Freitas et al. e ; 138 Paech et al. f 33 [ Reprinted with permission from the American Chemical Society. Paech, K.; Jockusch, R.A.; Williams, E. R. 2002 Slow Infrared Laser Dissociation of Molecules in the Rapid Energy Exchange Limit J. Phys. Chem. A. ( Volume 106, Page 976 6, Table 2). ]


94 C HAPTER 4 C ARBON DIOXIDE AND OPTICAL PARAMETRIC OSCILLATOR LASERS A n infrared multiple photon dissociation experiment require s a monochromatic infrared light source with a relatively high fluence in order to promote dissociation. These sources come in a var iety of types from carbon dioxide and nitrogen gas lasers to more sophisticated optical parametric oscillator and free electron lasers. Regardless of complexity light amplification by stimulated emission of radiation (LASER) consists of three basic compo nents : a lasing medium a medium and a resonan t cavity Each laser type employs different medi a to promote the emission of coherent radiation of the desired wavelengths. This in turn requires a specific means of exciting t he medium and a suitable resonant cavity to enhance the power Copious las er systems of varying complexities result. The focus of this chapter is the CO 2 and OPO lasers both of which were used in this work. Carbon Dioxide Laser In the early1960s, Patel pr oduced the first CO 2 laser at Bell Laboratories 14 1 B uilding on the earlier development of the h elium n 2 vibrational rotation transitions, Patel explored the use of the CO 2 molecule as a potential gas las ing med ium. 14 2 14 6 The first laser consisted of two tubes connected in a rectangular formation, similar to that shown in Figure 4 1 (initially without the nitrogen gas). 14 7 The top small tubing served as an inlet and excitation vessel (electrical discharge) for t he CO 2 gas and was connected on the far ends to a larger tube the resonance cavity The resonance cavity, set in a solid frame, had fixed mirror s on both ends one fully reflective and the other with a nonreflecting center of a radius ~ 5mm ( a partially r eflective mirror was also used for comparison). The excited carbon dioxide molecules entered the second chamber and emitt ed photons as the molecules relaxed to their ground state. The resonance chamber, designed to allow specific wavelengths of light to


95 un dergo constructive interference, amplified these emissions via a successive buildup of emitted photons reflected between the two mirrors. Lasing was achieved once the coherent light of a specific wavelength passed the threshold energy of the cavity. With h is initial experiments, Patel a ttained powers of 1mW in continuous wave mode, with somewhat higher pulsed powers. 14 1 Laser frequencies cover ed the P branches of two vibrational rotational transitions. Later Patel excited N 2 gas via an electrical discharge (Figure 4 1) and mixed the excited gas with the CO 2 gas to promote resonant energy transfer between the two molecules. 14 7 This initially improved where later even higher powers were achieved 1 4 8 Today, CO 2 lasers come in a v ariety of sizes. Though the basic construction is similar, major differences include improvements to the resonance chamber, closed laser systems, and tunable CO 2 systems for expanded wavelength range and power. With modern carbon dioxide lasers, all proces ses take place in a single liquid jacketed sealed resonance chamber, where electrical discharge results from a potential difference between anode and cathode posts or rings placed inside. Within the jacket, cold water is continuously circulated from a cons tant temperature bath (~20 o C) dissipating heat. The removal of the excess energy increases laser efficien cy and allows higher potentials to be used for increased energy output. Many of these lasers are also sealed, with H 2 O (~ 0.2 torr) added to the mixture to ensure the continual reprocessing of carbon monoxide to carbon dioxide (i.e., CO + H 2 O CO 2 + H 2 ) for longer operation without recharging. With tunable lasers, the totally reflecting mirror is replaced with a transparent window (i.e., zinc selenide, (ZnSe)) set at the Brewster angle. The Brewster angle allows for the complete transmission of light where the reflected (at angle B ) and refracted (at angle r ) light rays are


96 perpendicular to one another, such that B + r = 90 o Incorporating these condi gives 4 1 where n 1 n 2 are the refractive indices of the gas and the window, respectively. Equation 4 1 is rearranged to find the Brewster angle ( B ) as B 4 2 Light traveling through the Brewster window encounters a wavelength tunable diffraction grating and reflects back to the resonance chamber. By changing the grating angle, wavelengths in the R or P branches of two vibrational rotational transitions are tuned into resonance with th e optical cavity. The two vibrational rotational transitions are part of a four level system that expands available wavelengths and improves the efficiency of the CO 2 laser Four Level S ystem At the heart of a CO 2 laser is the vibrational structure of the carbon dioxide molecule. As mentioned above the linear molecule has 4 vibrational modes (3 N 5), the asymmetric and symmetric stretches and two degenerate bending modes. The lower level vibrational /rotational energy order of these modes establishes a four stage system (Figure 4 2). 14 7 ,14 9 Resonance energy transfer from N 2 to the CO 2 molecules occurs at the first energy level of the asymmetric stretch, 00 o u + ) These molecules relax ( stimulated and spontaneous emission) to two lower vibrational rotational g + ) : the first level of the symmetric stretch 10 o 0 and the second level of the bending motion s 02 o 0. T he g + levels relax (h elium gas aids th ese processes, in particular the 10 o 0 (an IR inactive mode) th at undergoes thermal relaxation ) 1 50 ,15 1 to the third


97 stage the first vibrational level of the bending motion s 01 o u ), and finally to the ground state g + ) The ability of the medium to reach the 00 o 1 excited state rapidly and to dissipate the energy from stage two quickly creates a population inversion that promotes the net production of emitted photons. This is aided by the addition of nitrogen and helium. Electrically pumped n itrogen g as provides the necessary excitation energy through collisional (near) resonant energy transfer. cross section (the lowest excited energy level 2330.7cm 1 of N 2 is very nearly resonant with 00 o 1 at 2349.16 cm 1 ) and slow deactivatio n rate (~8.8 sec @ 2torr) make N 2 ideal for the transport and transfer of energy. 1 5 2 Helium aids in the depopulation of the laser s second stage through collisional contact Also, the high thermal conductivity of He helps dissipate heat. Vibrational Rota tional Picture The laser s capabilities are further enhanced by the many rotational levels associated with each vibrational level. The rotational levels are a product of the molecules axial rotational motion (x, y, z coordinates), which spatially displace the atoms from a fixed position. As the CO 2 molecule is linear, only two assigned coordinates (by convention y and z) result in a change in the molecule s position. Simple rotational motion can be modeled by a linear rigid rotor with rotational energies of 4 3 Here, J is the rotational quantum number with values of 0, 1, 2, 3 and I is the moment of Nitrogen (g) is a homonuclear diatomic molecule that does not l ose vibrational energy through electric dipole radiation but relaxes by interacting with other molecules


98 inertia ( I=2m oxygen R 2 ). From Equation 4 3, the difference in the rotational energy levels is given by 4 4 where J J+1 is the higher rotational energy level and the rotational constant is defined by ( in Hz ) or in cm 1 ) 90 Equation 4 3 indicates that the rotational energy levels are not equal ly spaced but increase with increasing values of J Rotational levels are also more closely spaced than their vibrational counterparts ( ), with each vibrational level corresponding to a set of associated rotational levels Figure 4 3 illustrates how P ( J= 1 ) and R ( J= + 1 ) branches arise when transitions between these vibrat ional rotational levels occur. Each line within the branch corresponds to a different wavelength of light, increasing in energy from the P to the R branch accordin g to the selection rule The central Q branch, however, is forbidden ( J 0 ) for the CO 2 molecule, leaving a gap between the P and R branches. In the harmonic case, the energy ( in cm 1 ) of a particular vibrational rotational energy lev el is 4 5 To better approxi mate the energy of a vibrational rotational level, anharmoncity and bond dis tortion are taken into account. 88 Because increasing vibrational levels, the m oment of inertia increase s and the rotational constant decreases, thus decreasing the distance between each rotational level. The result is a coupling of both vibrational and rotational states through anharmonicity. The correction fo r the coupling of the rotational component is given by


99 4 6 n is the vibrational rotational coupling constant. Also, m olecular bonds are not rigid as assumed by the rigid rotor model. Distortion occurs with the rotational moti on and causes a second corrective term given as 4 7 where D e is the centrifugal distortion constant. Equations 4 3, 4 4, 4 5 and Equation 3 29 (limited to one vibrational mode) combine to yield, 4 8 The transition intensities from 00 o 1 to 10 o 0 and 02 o 0 are related to Boltzmann statistics 94 given by 4 9 Rotational levels of each vibrational level are populated according to the distribution. Laser transitions are most probable from rotational states that have maximum population at 00 o 1. This corresponds to the intensity of the laser s output at that given transition frequency Figure 4 4 gives the distribution of the corresponding P and R branches for both transitions Photon emissi ons from the first excited state of the asymmetric stretch (00 o 1) to the second excited state of the bending motion (02 o 0) and the first excited state of the symmetric stretch (10 o 0) give wavelengths of 9.174 9.354 (P)/9.443 9.773 (R) and 10.105 10.3 65 (P) and 10.441 10.885 (R) m respectively. The gaps (/) between the R and P branch are due to the

PAGE 100

100 forbidden Q branch. This phenomenon creates multiple energy transitions covering an important area of the congested fingerprint in IR spectra. Experimen tal C arbon Dioxide L asers The exper iments discussed in this work were performed on two separate laser setups: the optical parametric oscillat or (OPO) laser/CO 2 apparatus and two tunable CO 2 lasers. These were each aligned with two different FTICR mass spec trometers in the Mass Spectrometry Services lab and in the Eyler lab at the University of Florida, respectively. The OPO CO 2 beams were aligned with the ions in the cell of the FTICR mass spectrometer in order to perform a two laser experiment. Here, an of f resonant CO 2 laser was added to aid the dissociation of ions that were excited into the quasicontinuum by vibrationally resonant wavelengths from the OPO laser. This system is discussed in some detail in the next section of this chapter. The second syste m was developed after blackbody radiation photons (through heating the ICR cell) failed to dissociate phosphate groups from phosphopeptide ions. A tunable Apollo CO 2 laser in disrepair was placed into service to provide a more direct and powerful (high flu ence) source of IR photons. This arrangement included the complete set up of transfer mirrors, a laser gate, and other devices to finish the experiments. Later, an additional tunable CO 2 laser was purchased and integrated into the system. Apollo C arbon Dio xide Laser A tunable Apollo CO 2 laser, used in past experiments, was available but in need of repair. In order to facilitate these changes, the CO 2 laser was transported to the physical chemistry teaching lab with permission of Dr. Kathryn Williams and rep aired with the aid of Lawrence (Larry) Hartley Initially, the laser was checked to be sure the glass resonance cavity, hoses, etc., were in good shape. To this end, a rebuilt water chiller and connections were added, some of the old tubing replaced and se cured, and the visible high voltage connections and probes checked,

PAGE 101

101 with no notable problems associated with the cavity itself. As the existing schematics were incomplete, the electronic circuits were traced and schematics drawn. From these developed sourc es, the problems were isolated to the preamp control board, relays, and interlocks. After the replacement of a number of diodes and relays, the creation of a piggyback board that replaced worn circuits, and the reconfiguration of an interlock, the laser ap peared operational. In other words, laser plasma was evident, though the laser was not lasing due to a severely misaligned short period. The laser was moved back to the Eyler lab for integration with the FTICR mass spectrometer. Figure 4 5 illustrates the original CO 2 laser set up. With the aid of Cesar Contreras, the the connections were completed and the laser tested. Platforms were made to lift the CO 2 laser beam to the approximate center of the ICR cell window. According to plan, the focusing mirror was mounted, placed onto a sliding bracket, and secured to a long g uiding rod perpendicular to the laser beam on the east side of the table. This platform was designed to allow 360 o adjustment and movement along the rail. The focusing mirror was in line FTICR Figure 4 6. An NaCl window was added central to the output end of the CO 2 laser (west) and the focusing mirror (east). On the other sides of the salt window, a wavemeter (south) and helium neon laser mirror (north) were placed. Parallel and north of the CO 2 laser, the mutually aligned helium neon laser allowed for further tuning of the CO 2 laser and provided a visible and safe beam for the positioning of the system. Once the pieces were in place, the system was first adjusted using the helium neon laser fol lowed by optimization with the CO 2 laser. Subsequently, the power meter was placed in between the helium neon laser

PAGE 102

102 mirror and the salt window, and the readout was placed above the table. Predator the mass turned on the laser during the reaction delay sequence in the program, while the beam cell alignment was maximized by monitoring the absolute abundances of precursor and product ions in the mass spectra. A few early problems led to a number of additions to the system. Initially, feedback from the triggered CO 2 Predator program. The problem was solved by the introduction of an opto isolator box into the circuit. As the triggered beam power readings were later found to be inconsistent ( high initial output prior to settling at a given power), the trigger was later abandoned for the laser gate system (the laser continuously running) as seen in Figure 4 7. With the laser gate, a high or low voltage from Predator is sent to a control box th at is connected to a power amplifier and a rotary solenoid attached to a thin centered aluminum rod with a mirror and counter weight on each end (gate). When the voltage is high or the experiment is not running, power from the amplifier is relayed to the rotary solenoid raising the gate and closing off the laser from the cell. The power meter, now placed at the same height and in position to receive the reflected beam, measures the full power of the laser at that point. Conversely, if the voltage is low, t he gate drops during the sequence chosen (typically the reaction delay) allowing the beam to pass into the cell. By raising the voltage at the next event in the sequence, the beam is again blocked, regulating the reaction time to a single event window. Add itional features included the installation of a new antireflective coated zinc selenide Brewster and FTICR cell window to improve power transmission. A second repair was accomplished after a coolant water connection leak damaged the laser. Tracing the circ uits, it was discovered that the problem was related to the high voltage supply

PAGE 103

103 circuitry, in particular a faulty power tube. The power tube was replaced and the high voltage restored, but still the laser did not work. Upon close inspection, it was found t hat the anode inside the glass resonance chamber was damaged. Joe Caruso at the University of Florida chemistry glass shop was unable to repair the damage. However, he produced a second glass cavity, whose repairs were not as extensive, that fit the Apollo laser. After installing the second chamber, the cavity was realigned and did work but at a fraction of the original power. At that point, Dr. Eyler decided to purchase a second CO 2 laser, which was integrated into the existing system. Lasy C arbon Dioxide Laser The Lasy 20G AT tunable laser was purchased and installed as shown in Figure 4 8. The only modification to the existing set up included the addition of two other mirrors, both with the same degree of motion as previously mentioned ; the first guided t he laser beam from the Lasy CO 2 laser toward the beam of the Apollo laser, and the second mirror was set at the exact intersection of the two beams to guide the new laser along the same path. Both lasers were in turn aligned to the equipment and the cell A second gate was built for the proposed two laser experiments, utilizing the CO 2 lasers in the same manner as the double laser OPO/CO 2 experiments discussed later in this chapter and in Chapter 5. Unlike the Apollo the Lasy CO 2 laser is sealed and does n ot require a constant supply of gas. The laser also is capable of remote control, with on/off and cw/pulsed switches, power potentiometer, and wavelength control. Wavelength and power control units are wired and placed near the FTICR mputer workstation. These lasers aided in many experiments in the Eyler lab and encouraged a number of dissociation and action spectroscopy experiments, including those in described in Chapter 6.

PAGE 104

104 O ptical P arametric O scillator Laser In 1961, Franken et. al. discovered that second harmonic generation of light is produced when a relatively hi gh powered laser is passed through a material with nonlinear optical properties. 1 4 8 Optical parametric amplification and oscillation soon followed and in 1965 Giordmaine a nd Miller used this information t o build the first tunable OPO laser. 1 4 9 1 5 5 Application s of these and other advances led to stimulated Raman scattering, coherent antistokes (CARS) and inverse (IRS) and gain (SRGS) spectroscopy in addition to a host of ot her applications. 1 5 6 1 6 3 Since then, better cavity designs and the improvement of nonlinear materials have resulted in more efficient and powerful tunable lasers, providing the spectroscopist with very precise broadband energies. The OS 4000 OPO from LINOS is a continuous wave IR tunable laser whose wavelength ranges from 1.38 2.0 and 2.28 4.67 m for the signal and idler beams, respectively, at thousandths of a nm accuracy whil e producing 25 75mW per beam. 1 6 6 ,1 6 7 This section will focus on the basi c theory of the OPO laser, setup, and operation. O ptical P arametric O scillator Theory Referring to Figure 4 9, w hen a relatively high intensity pump laser beam of a set p ) is focused into any period of a periodically poled lithium niobate crys tal, signal s i ) frequencies are created, according to the relationship p s i (keeping with the law of conservation of energy). In the crystal, the signal and idler beams are amplified and the pump depleted as phase matching condition s are met. The two beams are then amplified further in the OPO resonance cavity until their power surpasses a threshold and the lasing action begins 1 6 8 In order to understand the workings of the OPO, the theory and experimental set up will now be discusse d.

PAGE 105

105 Nonlinear Optics Linear optics covers the typical behavior of light ( i.e., reflection, refraction, transmission, etc.) as it interacts with different media. As such, linear optics can be compared to the sound from speakers in an idealized stereo syste m driven at appropriate powers with no distortion. Here, the frequencies, though reflected and absorbed by the surrounding walls, etc ., are still clear and linear. Once the amplifier is turned up, however, the now overpowered speakers produce harmonic and anharmonic distortions of fundamental frequencies and the music, once clearly heard, becomes less discernable. These effects are analogous to what occurs in nonlinear optics. In nonlinear optics, rela tively powerful coherent lasers focused on a material i nduce nonlinear effects from distortion of the medium. These effects include generation of sum frequency (different multiple frequencies producing one frequency), difference frequency (a single frequency forming multiple frequencies), and harmonic frequenc y (multiples of the same frequency producing harmonic frequencies) from the fundamental (pump) frequency(s). 1 70 Both linear and nonlinear interactions are expressed in terms of the induced polarization of light given as 4 10 where o n is the nth order susceptibility, and is the electric field strength. The first order of susceptibility (1) covers linear optics. Second order susceptibility (2) corresponds t o the beginning of nonlinear behavior where second harmonic generation ( e.g. wavelength doubling from 1064 nm to 532 nm) and optical parametric oscillation (pump wavelength producing signal and idler wavelengths) occur. Second order susceptibility is a th ird degree tensor having 27 elements. 1 70 These nonlinear coefficients describe the off/near resonant interactions between the applied electromagnetic fields and the

PAGE 106

106 outer electron clouds of an atom or molecule in a medium. Here, a field stimulated change i n the electron cloud excites the entire molecule to an intermediate (virtual) state, with the annihilation of a pump photon, whereupon immediate relaxation (back to original state) creates the signal and subsequent idler photons ( p s i ). 1 7 1 ,1 7 2 W ith the lithium niobate crystal, these frequencies are adjustable as the nonlinear refractive index (important to the interactions of light with the medium) varies in a controllable manner 1 6 8 ,1 7 3 1 7 5 If the temperature is r aised or lowered (within the operation al window of 50 170 o C for the Linos OPO laser ), different signal and idler frequencies are produced. The summation of these events gives an overall statistical measurement of the induced polarization. However, unlike odd ordered susceptibilities, even ordered susceptibility tensors disappear, along with their nonlinear properties, with the inverse symmetry operations of centrosymmetric materials. Thus, production and optimization of noncentrosymmetric ( no inversion sy mmetry) crystals (e.g., LiNbO 3 K NbO 3 and Ba 2 NaNb 3 O 13 ) with high transmissivit y and optical damage thresholds is necessary to create efficient second order nonlinear effects. 1 7 6 1 7 8 Isolating the second order nonlinear terms and noting that the electric a nd magnetic fields are propagated along the z axis in a direction resonant with the cavity, Equation 4 10 simplifies to 4 11 Further expansion of the terms in Equation 4 11 reveals the interaction of the different electric fields c oupling the pump, signal, and idler beams, 4 12 4 13

PAGE 107

107 4 14 From these equations, interacting phase conditions are made apparent through the slow amplitude approximation. 1 73 1 7 9 ,1 80 The sl ow amplitude approximation treats the nonlinear wave equations above as monochromatic plane waves propagated in the z direction 1 8 1 1 8 3 E ach function is then transformed into an expression for its amplitude ( A N (z) ) change with respect to the z direction o r unit vectors ( a n ,), scalar amplitude functions ( A n ,), and wave vectors ( k n ) in the z direction, 4 15 4 16 4 17 The general equation for the amplitude scalar to the wave vector is 4 18 where n o is the nonlinear refractive index. Substituting Equations 4 15, 4 16, 4 17 into Equations 4 12, 4 13, 4 14 gives t he generated polarizablity in terms of the amplitude and wave vector functions as 4 19 4 20 4 21

PAGE 108

108 In the second approximation, the general electric field component ( ) is placed in the modified general nonlinear wave equation (from Maxwell s equations) given as, 4 22 simplified to give the equation of the slow amplitude approximation (derivatio n noted in references), 1 70 ,1 7 5 4 23 Substituting Equations 4 19, 4 20, 4 21 into Equation 4 23 leads to, 4 24 4 25 4 26 where the change in the wave vector, p k i k s = n p p n s s n i i is a measure of the coupled wave equations. Here, three way mixing takes place between the coupled waves of the applied electric fields resulting in the parametric interaction of pump, signal, and idler. This provides a mechanism for the exchange of energy among the fields. For instance, an amplitude gain for signal and idler waves (parametric amplification) and a corresponding decrease in the pump energy occurs so long as ph ase matching conditions are met (i.e., or ideally =0 ).

PAGE 109

109 Quasi Phase Matching Phase velocity mismatch (between pump, idler, and signal) occurs as velocity synchronism breaks down due to dispersion and other effects. There are two general methods for correcting these problems. 1 8 4 ,1 8 5 Trad itional phase matching involves utilizing the birefringence effect of the refractive index to reduce dispersion. The other comple mentary technique uses periodically poled regions, spaced between points of phase divergence, to bring the fields back into pha se 1 7 6 ,1 8 6 This process is kn own as quasi phase matching. As previously stated, interacting fields step out of phase as they travel through the crystal, thus affecting the gain. The maximum interaction length over which amplification of the signal and idle r fields is sustained is known as the phase coherence length. Phase coherence lengths are determined at the point where the phase of the pump and the sum of the idler and signal frequencies are 180 o out of phase Rotation of alternating coherent length seg ments (periodically poled elements in one period) in the crystal assures that the pump and the idler and signal frequencies are back in phase. These modifications are accomplished through crystal alteration or electric field poling. 1 8 7 ,1 8 8 Adaptation of th e crystal is commonly achieved by growing the crystal with periodic elements in place or by invasive alteration to an existing crystal Electric field poling applies to ferroelectric crystals (i.e., lithium niobate), where long metal electrode strips are a pplied in a periodic fashion to the top of the crystal with the bottom uniformly covered and grounded. 1 8 9 ,1 90 When a static electric field of sufficient strength is applied, the periods under the field are reversed. The periodically poled lithium niobate c rystal in the LINOS OS 4000 OPO laser operates in the latter fashion, where the depleted pump and surging signal and idler beams leave the crystal in phase to resonate in the cavity.

PAGE 110

110 Pound, Drever and Hall Technique Small frequency fluctuations in lasers are common due to variations in temperature, current, voltage, etc. Thus, a means to keep the pump laser in resonance with the cavity is critical to achieve wavelength precision and constant power. The LINOS OPO does this by considering the laser resonance cavity as a Fabry Perot interferometer and implementing the Pound, Drever, and Hall (PDH) technique 16 4 ,16 5 to lock the laser in place. With the LINOS OPO, the Nd:YAG laser is modulated to produce two 12 MHz side bands on either side of the princip a l beam. 1 6 6 Assuming the cavity and pump are generally aligned with the system these side bands enter the resonance cavity, resonate, and escape from the cavity as shown in Figure 4 10. Traveling back toward the pump laser, the beams are isolated from the incident modulated beam via a Faraday isolator (# 4 ) and directed to a photodiode (# 5 ). The signal is then compared to the modulated beam through a mixer to create an error signal. If the c ompared side bands are the same (error signal is zero) then the cavity is i n resonance. If the error signal is not zero then the control electronics send the appropriate signal to the piezo electric crystal attached to the back reflecting mirror (# 16 ) for a zero error adjustment. In the lock mode, a feedback loop keeps the pump l aser in resonance, which in turn allows the pump resonant signal and idler beams to stay in resonance. An off resonan ce intensity appears the same from either side of the resonant intensity (Gaussian like beam profile). Thus, the direction of the correctiv e change is difficult to determine. This is corrected by taking the derivative of the intensity with respect to its frequency where the on resonan ce condition is achieved when Off resonance positions are detected as either being in phase with the modulation (above resonance, positive derivative) or 180 o out of phase (below resonance, negative derivative), as shown in Figure 4 11. The

PAGE 111

111 aforementioned mixer reflects this by a positive or negative output, respectively. This in turn is de pendent on the resonance conditions inside the cavity. To better understand these conditions, a general characterization of the interaction of the laser beam components is necessary. The electric field ( E inc ) incident to the cavity is 4 27 where E o is the initial electric field, is the primary frequency, is the phase modulation of (phase dithering ), and is the modulation index Equation 4 27 is expanded through Bessel functions to give 4 28 where J o and J 1 are zero and first order Bessel functions. The final term in Equation 4 28 illustrates the inclusion of the primary frequency ( J o ) and the two side bands ( J 1 ). 1 90 The total power ( P o ) associated with the incident beam is 4 29 such that P o c +2P s where P c is the power in the carrier beam ( P c =J o 2 x P 0 ) and P s is the power in the side bands ( P s =J 1 2 x P o ). These beams are reflected in the Fabry Perot cavity (OPO resonance chamber) where the transfer function is given by 4 30 In Equation 4 30, the reflected electric field is E ref r is the amplitude reflection coefficient for the two mirrors, and fsr is the free spectral range (FSR) of the chamber. The free spectral

PAGE 112

112 range is defined by the difference between the reso nant (constructively interfering) modes in the cavity such that 4 31 where c is the speed of light and L the length of the resonance cavity. Equation 4 30, also known as the reflection coefficient is rearranged to solve for the reflected electric field E ref The incident electric field equation (Equation 4 28) is then converted to 4 32 As mentioned, the reflective field is detected by a photodetector which measures power. 1 9 1 Thus, the re flective power ( P ref. ) is found by taking the square of the absolute value of E ref (Equation 4 29 ), which when expanded is 4 33 where R e and I m are the real and imaginary parts of the equa tion, respectively. The first line of Equation 4 33 represents the unperturbed power contributions from the fundamental and sidebands. Line two gives the power addition from the interference between the waves of the carrier and sidebands. Line three is the importance is the interference or beat patterns of line 2 that provide an indirect measurement of the primary wave in terms of power. This line is isolated from the other s by the use of a mixer and low pass filter, where the sine and cosine parts are effectively separated. The real or cosine terms are used to measure low frequency modulation, where the cavity response is quicker than

PAGE 113

113 the phase modulation The mixer compares the original modulated si 4 34 o with a phase shifter. The mixer sends out a dc signa l which is passed through a low pass filter to isolate the error signal, given by 4 35 which is approximated by 4 36 as is the case here. Given the Pound, Drever, and Hall error is 4 37 with the resonance state represented by the zero point in the error signal or In the scan ning mode of the Linos OPO laser, as implemented in our laboratories an oscilloscope monitors two channels of the lock control con tinuously: the sinusoidal like error signals and the corresponding pump modes in the cavity. When switched to lock, two horizontal lines appear one at the point of the zeroed signal and the other at the top of the pump mode, indicating that the cavity is in resonance and ready for operation. O ptical P arametric O scillator Operation and Set Up The Linos OS 4000 optical parametric oscill ator laser is installed in the M ass S pectrometry S ervices laboratory at the University of Florida and aligned with the cell of a 4.7T FTICR mass

PAGE 114

114 spectrometer. The laser is placed directly behind the mass spectrometer in an adjacent room (due to lack of space) where a rectangular Installation involved the placement of two la ser tables for the OPO and related electronics, the connection of the OPO laser with the lock/temperature and YAG control modules, the calculated arrangement and careful placement of guiding mirrors for signal and idler beams, the integration of power mete r, wavemeter, iris, and oscilloscope, as well as the design and execution of nitrogen purge box. An overview of the OPO operations, setup, and tuning is given below. O ptical P arametric O scillator Operation As previously mentioned, there are three qualiti es that make up a laser: the photon laser operates under the nonlinear principles of second order optics. The medium is a periodically poled lithium niobate crystal consis ting of eighteen different poling periods, each of which correspond s A continuous wave Neodymium Yttrium Aluminum Garnet ( Nd: YAG ) laser (operating at 2 watts) acts as the pump w hose beam is guided through a series of polarize r s, mirrors, and focusing optics (Figure 4 12) to bisect one of the periodically poled periods in the crystal. For each pump photon ( E p ) lost, two photons of lower energy, signal ( E s ) and idler ( E i ) are crea ted such that E p = E s + E i These signal and idler photons continue to gain energy from the pump beam as they travel through the crystal where periodically poled regions maintain quasiphase matching. After leaving the crystal, they resonate in a chamber defined by a piezo mounted reflecting mirror in the back (#16) and a partially reflective surface on the front of the crystal ((#13) where the incident pump beam enters). The energy of the idler and signal beams increase s with the constant input of pump en ergy and multiple passes inside the cavity until the threshold energy of around 0.3W is met and laser action begins Two pairs of signal and idler beams are

PAGE 115

115 directed from the front and back of the cavity. Exiting the front of the cavity, idler l and signal go through the pump reflecting steering mirror (#10) and are separated by the signal reflecting mirror (#18) just behind it Here, idler 1 passes through and is directed to the outside as the signal is reflected by a third mirror to also exit the laser ho using (SM1). The other pair, idler 2 and signal, is taken from the cavity where a selective filter allows only idler 2 to pass which then exits from the back of the laser housing (SM2). Tuning. Before examining the laser set up, a few words about the las order. Tuning the laser occurs in three basic stages. The first two are noted in Figure 4 13. Stage 1 involves the adjustment of the crystal to one of the eighteen poling periods, based on the required wavelength range. The periodically poled crystal sits on a stage where north south adjustments are made by the turn of a knob aligning the period with the perpendicular static pump laser beam. Within that period there are strong and weak regions that are noted by strong and weak power outpu t while changing the crystal, a strong region is found (adjustment of the etalon (stage 2) is often required to acquire a stable wavelength) and the period is set. The free spectral r ange ( FSR ) of the cavity is further sampled by the FSR of the etalon in Stage 2. By adjusting the micrometer For fine differences in wavelength, the adjustmen t of the etalon (s) is sufficient to cover wavelength differences of ~thousandths of a nanometer. Co a rse adjustment or wavelength sampling for action spectroscopy is based on finding a wavelength within a broader range (i.e., one nanometer) that produces th e highest most stable power for optimal dissociation. Stage 3 is temperature tuning within the poling period. Figure 4 14 is a plot of the temperatures versus the wavelengths of each poling period for the signal (top row) and idler (bottom row). A temperat ure

PAGE 116

116 increase results in a decrease in the idler wavelength and an increase in the signal wavelength, although the total energy remains the same. Temperature tuning from 50 170 o C covers all the wavelength ranges between the poling periods with an addition al 20 % overlap, and is controlled by a separate module (in the same housing as the lock electronics) with fine and co a rse adjustments carefully monitored on the display. O ptical P arametric O scillator Set Up Two laser tables were set up in the OPO laser ro om and their heights adjusted to the corresponding height of a previously installed CO 2 laser system in the back of the FTICR As shown in Figure 4 15, the laser housing was placed on the first table with the front of the laser pointing toward the mass sp ectrometer. A separate stage was created and set on the second table to lift the temperature/lock electronics housing and oscilloscope out of the beam s path. The YAG controller was placed on a shelf under the laser table. After all the main components wer e assembled, previously configured mirrors were set into place. These stages of mirrors guided three beams of the OPO to their prospective locations. The signal beam coming out the front of the laser housing (SM1) was directed into the wavemeter placed in the far front right corner of the second table. The signal wavelength was then converted to the idler wavelength by the relationship Idler 1 was steered around the perimeter of the table to the back far right side of the laser table. Idler 2 from the back of the OPO was guided to a position side by side to idler 1 and equidistant from a path tra ced from the centerline of the ICR cell. Here, the beam s acted as two sides of an isosceles triangle on a horizontal plane raised and guided t o the center of the cell They were precisely configured based on a program that considered geometric and trigometric principles.

PAGE 117

117 Precise alignment was achieved by monitoring dissociation products from the rubidium tagged D glucosides (discussed further i n the Chapter 5) where the right mirror was adjusted until maximum dissociation was reached and then the process repeated with the left mirror. The CO 2 laser was then integrated into the system by careful adjustment of its mirror aligning the CO 2 beam ce ntral to and at an equivalent height with the OPO beams without interfering with their path s As mentioned earlier, the additional laser provided a means to incorporate a two laser experiment. This was put into use with the C H stretches of the rubidium ta gged D glucoside ions which did not readily dissociate with the OPO laser alone as will be discussed in Chapter 5. Along the path of these beams toward the ICR cell, the irises were placed to regulate the time of irradiation and measure the power. Power w as taken when the irises were closed, and the beams were reflected into a power meter. A nitrogen purge box was also designed and built in four sections to cover the OPO laser table the travel through the wall, the CO 2 laser, and the entrance into the ele ctronic region of the cell. The system was assembled with the assistance of Cesar Contreras. The CO 2 and OPO laser systems were designed and assembled to provide a means to complete the sugar and phosphopeptide experiments. As such their use spread beyon d these goals to aid many others in the Eyler and Polfer labs as tools of action spectroscopy as well as a means of fragmentation. The analysis of the action spectra for the four D glucoside isomers using the OPO/CO 2 system forms the subject of Chapter 5.

PAGE 118

118 Figure 4 1. A picture of the CO 2 reflective mirror to the right and the transmitting window to the left, micrometers for alignment, and inlets for both CO 2 and N 2 gas. [ Reprinted with permi ssion from the American Physical Society. Patel, C.K.N. 1964. Selective Excitation through Vibrational Energy Transfer and Optical Maser Action in N 2 CO 2 Physical Review Letters (Volume 13 Page 618 Figure 2 ) ] Figure 4 2. n of the four stage CO 2 chemical basis for operation. Excitation of N 2 gas and subsequent resonance transfer of energy to CO 2 molecules initiates the reaction. This energy is released from the first excited vibration le vel of the asymmetric stretch to the first and second levels of the symmetric stretch and bending motions, respectively. These vibrational rotational transitions produce photons of multiple wavelengths consistent with a set of P and R branches for each vib rational level transition. Energies from these last levels further ground state. [Reprinted with permission from the American Physical Society. Patel, C.K.N. 1964. Selec tive Excitation through Vibrational Energy Transfer and Optical Maser Action in N 2 CO 2 Physical Review Letters (Volume 13 Page 617 Figure 1 ) ]

PAGE 119

119 Figure 4 3. An iIlustration of rotational vibrational transitions where P and R branche s are created from the spacing between the two rotational vibrational levels in accordance with the selection rules. Figure 4 4. The P and R branches from 2 vibrational rotation al transitions that inspired the work for Pa 2 laser. In this Q [Reprinted with permission from the American Physical Society. Patel, C.K.N. 1964. Continuous Wave Laser Action on Vibrational Rotational Transitions of CO 2 Physical R eview (Volume 136 Page A1189 Figure 1 ) .] 7 6 5 4 3 2 1 8 6 4 2 0 V=1 V=0 P Branch Q Branch R Branch Increasing Energy J values

PAGE 120

120 Figure 4 5. The initial set up of the Apollo CO 2 laser consisting of the controller, chiller, gas m ixture, optics, power meter, wavemeter, and helium neon laser. Figure 4 6. F inal stage of the CO 2 laser beam alignment with the cell of the FTICR mass spectrometer. The iris serves as a guide and the guide rod protect s cell control electronics. ICR Cell Back Mirror Iris Guide Rod FTICR Magnet Helium Neon Laser Carbon Dioxide Laser Laser Gas Mixture Laser Mixture Gas lines Focusing Mirror Coolant Chiller Carbon Dioxide Laser Controller Coolant Lines Hi Voltage Lines NaCl Window To The FTICR (Back Mirror) Helium Neon Mir ror E W N S S Wavemeter Power Meter

PAGE 121

121 Figure 4 7. Addition of the laser gate and associated con trol electronics to the Apollo CO 2 laser to promote consistent laser power for the length of the experiment. Laser Gate Control Interface Laser Gate Power Amplifier Control Interface Signal from Predator W E N S P ower Meter

PAGE 122

122 Figure 4 8. The addition of the sealed tube tunable Lasy CO 2 laser, power amplifier, control box, and chiller with additional optics. Figure 4 9. G eneral description behavior induced by focusing a YAG pump laser (yellow) into a poling period of a periodically poled lithium niobate crystal to produce idler (blue) and signal (red) wavelengths from two sides of the cavity according to Poling P eriod Periodically Poled Period Carbon Dioxide Laser Proposed Second Laser Gate Coolant Chiller Hi Voltage Lines E W N S Coolant and Return Lines

PAGE 123

123 Figure 4 10. A picture of LINOS OPO laser with appropriate features labeled corresponding to the Pound, Drever, and Hall method for locking in the resonance cavity. (for complete assignments see Figure 4 12). Figure 4 11. D erivative of the intensity with respect to frequency of a Gaussian like beam profile, where the on resonan ce condition is achieved at the origin. [Reprinted with permission from the American Associati on of Physics Teachers. Black, E. D. 2001. An Introduction to Pound Drever Hall Laser Frequency Stabalization. American Journal of Physics (Volume 69, Page 83, Figure 6).]. Front Back

PAGE 124

124 Figure 4 12. The LINOS OPO laser with appropriate labels including the housing (1) YAG (pump) laser (2), quarter / half waveplate and half wave plate polarizers (3,6), Faraday isolator (4), phot o diode unit (5), telescope (7), pump beam steering mirror (8), iris (9), beam splitting (signal/idler from pump) and p ump steering mirror (10), three stage focusing lens (11), cover for crystal/oven (not shown (12)), o ven with PPLN crystal (13), folding mirr ors (14), etalon housings (15), p iezo mounted back or end mirror (16), first and second beam separation units (17, 18). Figure 4 13 Wavelength tuning of the OPO begins with the adjustment of the PPLN crystal or stage one. Based on the wavelength range desired, the PPLN stage is moved to a place within the corresponding period that maximizes power at a set wavelength. Stage two adjus ts the etalon to discriminate between wavelengths in its FSR Stage 1 Stage 2 1

PAGE 125

125 Figure 4 14. Temperature vs. wavelength plots for each poling period provide a means of tracking the third stage of tuning through temperature. Here a controlled change in the PPLN crystal temperature produces a different range of wavelength s in succession. The signal and idler changes are seen in the top and bottom rows, respectively.

PAGE 126

126 Figure 4 15. G eneral set up of the OPO laser including placement o f the OPO laser and the electronics, idler 1 & 2 and CO 2 alignment with the cell, signal and wavemeter placement, and the iris and power meter arrangement (only one iris is shown). Oscilloscope Temp./Lock Control YAG Control Iris Control FTICR MS Wave Meter OPO 4000 Front Nitrogen Purge Box Back CO 2 laser Power detector Wall

PAGE 127

127 CHAPTER 5 S PECTRAL AND MOLECULAR DIFFERENCES IN O METHYLATED GLYCOSIDE IS OMERS Increasingly FTICR MS is being employed to study carbohydrate composition in the gas phase, eliminating solution phase carbohydrate interactions that can com plicate the structural picture. 1 9 3 1 9 6 With mass spectrometry, a difficulty lies in different iating between the many possible biological saccharide isomers of the same m/q 80 90% of which are chiral 23 This is further complicated by a saccharide s etc The common approach of differentiation employs t andem mass spectrometry to isolate and fragment ions of known isomers (often derivatized to lock the structure), through dissociation techniques like collis i on induced dissociation (CID). 1 9 7 20 4 Here key spectral differences particular to each agmentation pattern and relative abundances provide a template for differentiation of unknown saccharides. Significant fragmentation differences are not common to all isomers however making some saccharides impossible to differentiate within the usual un certainties associated with these experiments The use of IRMPD to probe wavelength depend e nt fragmentation is one possible means to overcome the aforementioned problem. 22 Another possibility is IRMPD promoted action spectroscopy. 97 20 5 20 7 C omparison of t he a ppropriate computational work, it also gives an intimate study of the responsible gas phase structures and their intramolecular interactions. Because the i on concentration in a FTICR cell is very low a n absor ption spectrum is impossible to detect. 95 But a ction spectroscopy measures the vibrational character indirectly by wave length range (usually reported as wavenumbers ( or cm 1 )) As described in Chapter 3, t he process begins when the laser s wavelength is in resonance with a vibrational energy level, causing the transfer of multiple infrared photons. The energy associated w ith e ach absorbed

PAGE 128

128 photon spreads throughout the ion via int ra molecular vibrational re laxation (IVR) increasing the 20 8 20 9 ,2 10 where by a weakly bound metal cation saccharide complex is formed to promote the cation removal through a one or two photon absorption process. Common cations are taken from the alkali metals whose binding strengths increase with decreasing size. 21 1 21 2 Here, s imple protonation results in glycosidic bond cleavage. 21 3 Strongly bound lithium sugar complexes which were used by Polfer et al in IRMPD wavelength depend e nt fragmentation cleave across saccharide rings and at the glycosidic bond. 22 Potassiu m, rubidium, and cesium ions however, all dissociate readily without fragmenting the saccharide. Of these, r ubidium promotes clear spectra and was the cho ice for th e experiment s reported in this chapter Earlier IRMPD spectra obtained by Valle et al. expl ored the infrared rang e from 600 1700cm 1 to examine vibrational characteristics of rubidium glycoside isomers O methyl D gluco methyl D gluco methyl D galacto methyl galacto (Figure 5 1) at the F ree E lectron L aser for I nfrared eX periments (FELIX) facilit y at the FOM Institute for P lasma P hysics Rijnhuizen in the Netherlands 21 4 21 6 In these experiments, only a few questionably differentiabl e characteristics were found between the isomers ( Figure 5 2) After some consideration and computational verification, the four hydroxyl groups in the relatively isolated O H stretch region (~ 3200 3650 cm 1 ) were thought to have a greater potential for differentiation of the monosaccharide structures An o ptical parametric oscillator (OPO) laser aligned with the ions in the cell of a Fourier transform ion cyclotron resonance ( FTICR ) mass spectrometer provided the wavelength range ( 2,270 4,670 nm/4405 2141cm 1 ) and power (50 150 mW) necessary to o btain the spectra of

PAGE 129

129 the C H and O H st retches for the four complexes. 16 6 1 6 8 ,1 80 A comparison of the spectr a revealed differentiable features. Computational studies were carried out utilizing density functional theory (DFT) with the B3LYP hybrid functiona l and 6 31+G (d) and LANDL2DZ basis (for Rb) 21 7 22 1 from which zero point energies and vibrational frequencies for a number of conformers were obtained. The adjusted theoretical spectra of the lowest energy conformers were compared to the experimental spe ctra revealing candidate gas phase structures responsible for the vibrational bands observed This included the discovery that multiple conformers, differing largely in their hydrogen bond interactions, were responsible for the variation of O H stretches in the two glucoside spectra Experimental Methods Materials The four monosaccharide isomers of O m ethyl D glycoside were obtained from Dr. Brad Bendiak of the Department of Cellular and Structural Biology at the Universi ty of Colorado Health Sciences C en ter. Methanol and rubidium c hloride were purchased from Sigma Aldrich and milliQ water was acquired from the physical chemistry teaching laboratory at the University of Florida. Stock solutions were prepared by dissolving 10mg of each monosaccharide in fou r separate 10 ml solutions of 80:20 MeOH:H 2 O. Samples were taken as needed and diluted to 1x10 4 M at the same solvent ratio. Rubidium c hloride ( cationizing agent) was then added in an equimolar amount Instrumentation A tunable c ontinuous wave optical para metric oscillator (OPO) laser aligned to the cell of FTICR mass spectrometer (Bruker, Billerica, MA) with an Analytica electrospray (ESI) source ( Analytica of Branford, Inc., Branford, CT), was employed for these experiments ( Mass Spectrometry Services Labo ratory at the University of Florida ) Electrospray

PAGE 130

130 (N 2 ) flows at a rate of 35 and 155 L/hr, respectively. A 3.6k V potential difference was applied between the ESI needle and capillary. Parent ions were isolated using swept frequency ejection pulses and irradiated The products (depleted parent and rubidium ions ) were detected in a 70 500 m/q window by broadband detection. Bruker Xmass data acquisition system 7.0 was res ponsible for setting instrument parameters and data collection. The LINOS OS 4000 OPO laser was purchased from LINOS Photonics (M unich Germany) O ptical P arametric O scillator Laser The tunable c w OPO laser operates according to the nonlinear optic princip les of second order susceptibility as seen in Chapter 4 Here, e lectromagnetic radiation from the continuous wave pump laser (Nd : YAG) operating at 2W and a fixed wavelength of 1064 nm interacts with a periodically poled lithium niobate crystal in one of eighteen poling periods, to produce two different frequencies (5 1) where and are the frequenc ies of the pump, signal, and idler respectively. A gradual increase in temperature from 50 to 150 o C increase s the signal wavelengths and decreases the idler wavelengths within the capacity of the period and the temperature range. This process brings consecutive groups of wavelengths into resonance as the previous wavelengths disappear Waveleng th discrimination is further enhanced by the inclusion of an e talon in the resonance cavity. After acquisition of data for a given wavelength range, the temperature was reduced and the crystal s position changed to an adjacent poling period to introduce th e next range of wavelengths. The 18 poling periods available with the LINOS laser cover wavelength ranges of

PAGE 131

131 1.38 2.0 and 2.28 4.67 m or 7246 5000 and 4405 2141 cm 1 for signal and idler, respectively. The wavelength used for dissociation (from t he idler beam s) is calculated from the hrough a rearrangement of the conservation of energy relationship (based on Equation 5 1) between the wavelengths of the pump ( p, =1064nm), signal ( S = wavemeter), and idler ( i ) given as (5 2) F or the C H stretch region, a monochromatic (10.6 m) cw Synrad CO 2 laser ( Mukilteo, WA) beam was directed central to and on axis with the idler beams for off reso nant dissociation of OPO excited ions in the quasicontinuum. Experiment al P rocedure After preparing a solution of one of the four Rb + [ O m ethylated glycoside] complexes, the ions were introduced to the FTICR mass spectrometer via electrospray ionization, accumulated in a hexapole ion trapping region for 1s, and transferred to the cell through a series of ion optics. Once trapped in the cell, the Rb + [ O m ethylated glycoside] precursor ions were isolated and irradiated by the OPO idler beams at a set wavele ngth for 10s (for O H stretches). Mass spectra of the resulting ions were then acquired. T en scans were accumulated to improve signal to noise and the relative abundances of the precursor and product ( 85 Rb + and 87 Rb + ) ions were recorded Subsequently the OPO laser was adjusted for the next wavelength and the process was repeated. Mass spectra consisted of either the Rb + [ O m ethylated glycoside] precursor ions with no dissociation products (nonresonant wavenumbers) or the depleated precursor ions with the R b + dissociation products (resonant wavenumbers). The action spectrum was then plotted as the

PAGE 132

132 power corrected natural logarithm of the ratio versus where F corresponds to the combined relative abundance of 85 Rb + and 87 Rb + ions and P the relative abundance of the depleted precursor ions. When only the precursor ions exist the value for r educes to zero. Rubidium ions were readily dissociated by resonant absorption of infrared photons in the O H stretch region. However, the C relatively weak absorption generated little or no rubidium peaks under the same conditions (Figure 5 4A). Fluence assistance to the OPO photon resonant excited ions (in the quasi continuum ) was provided by off r esonant photons from the CO 2 laser. The off resonance condition was assured by past action spectra (Figure 5 2), which illustrated a relatively featureless absorption region at the CO 2 laser s wavelength of 10.6 m. Also, any blackbody and/or off resonant effects were avoided by changing the CO 2 laser s power and irradiation time until a product free spectrum was achieved while operating only the CO 2 laser over the total irradiation period. Following the aforement ioned protocol, the trapped and isolated ions were irradiated for ten seconds with the OPO laser, with the additional fluence of the CO 2 laser during the last three seconds. This resulted in little dissociation. A few more trials were completed until a tot al OPO irradiation time of 15 seconds, with the CO 2 laser operating concurrently for the last 7 seconds, produced sufficient fragmentation as shown in Figure 5 4B. All other conditions were the same for both O H and C H stretches. Two to three spectra wer e obtained and averaged (bin size 2cm 1 ) for each of the four Rb + [ O m ethylated glycoside] complexes.

PAGE 133

133 Computation al Methods G eneral structure s of the four isomers of O methylated D glycoside s were drawn in Hyperchem 22 2 according to the stable gas phase con formations found in the literature: the 4 C 1 conformation with the O H groups involved in a hydrogen bonding network oriented in a clockwise arrangement. 2 20 ,22 1 22 3 ,22 4 A r ubidium cation was artificially attached to the O 1 oxygen and placed above and centra l to the glycosidic ring system (at a distance of ~3 ) Ring dihedral angles and H C O H torsional angles promoting the rotation of the hydroxyl groups around the C O bonds, were defined such that the angles changed independent ly of one another. During th e algorithm while the torsional angles were randomly varied, for a total of 1000 structures. Subsequently the structures were geometrically optimized with AMBER force field (AM 1) and comparisons were made to eliminate duplicate and higher energy structures. Duplicate criteria were set to disregard conformers whose threshold energies, dihedral angles and atomic positions where within 0.05Kcal/mol, 10 o and 0.25 of another respectively The h igher energy threshold was set 15Kcal/mole above the lowest energy conformer to contain the range of possible conformers Geometric o pti mization of these remaining conformers was carried out using B3LYP/6 31+G(d) in additio n to LANDL2DZ basis 22 8 22 9 for the sugar and the rubidium ion, respectively followed by the vibrational calculations with Gaussian 03 21 7 When these structures proved inadequate to explain the existing glucoside anomeric spectra, other ring structures we re created and analyzed. Scaling factors of 0.97 and 0.96 were used for the frequencies associated with the O H and C H stretches respectively within the error of the corresponding level of theory and the results for similar saccharides. 2 20 22 1 22 4 2 30 2 3 4

PAGE 134

134 Experimental Results and Discussion T he vibrational spectra of the lowest energy conformations for these species convoluted with a Gaussian band profile of 20cm 1 were compared in the Ph.D. thesis of Cesar S. Contreras to those obtained i n the afore me ntioned experiments by Valle et al. in the 900 1400cm 1 range. Of these, the lowest energy conformer s spectrum agreed with experiment as shown in the Figure 5 5. However the experimental bands were too broad and similar to reveal any major differentiabl e features. For this reason, the O H stretches were used in the present work Glucosides O m ethyl D g lucoside The action spectr um of O methyl D g lucoside ranges from 2750 to 3750 cm 1 and includes both O H and C H stretches, which consist of four and t wo bands respectively. These bands are labeled I VI proceeding from the higher to lower wavenumbers, as seen in Figure 5 6 Even though C H stretches from 2800 to 3050 cm 1 are in general agreement with the calculated lowest energy structure, the theore tical spectrum is unable to account for all bands in the experimental O H stretch region from 3378 to 3700cm 1 In order to complete the coverage of these O H bands, the four lowest energy conformers (all within the error associated with the level of theor y) are considered and differ in their intramolecular hydrogen bond(s) and/or Rb + [ O methyl D g lucoside ] interactions. The two lowest energy conformers, 1 and 2 at 0.00 and 0.8 kcal per mole, respectively, have similar structures and vibrational spectra, as seen in Figure 5 7. Here the Rb + cation is found above the ring near the pocket created by the C 6 hydroxyl group, O 1, and the C 1 O methyl group. Conformers 1 2 and 3 exist in the 4 C 1 ring conformation. The next highest in ene rgy at 5.18 and 6.49 kcal per mole are conformers 3 and 4 respectively. The Rb + attached to conformer 3 exists outside the ring between hydroxyl groups on C 4 and

PAGE 135

135 C 6. In conformer 4 the Rb + ion is under the ring (with respect to the C 6 carbon above) between the C 2 and C 4 hydroxyl groups and O 1. This position is made favorable by the 1 C 4 half chair ring structure of 4 The assignments of the bands are as follows. Conformers 1 2 and 4 in Figure 5 7 account for band I, which is centered at 3677 with 21cm 1 full width at ha lf maximum (f whm ). This band relates to the O H stretch modes that are relatively unencumbered by inter or intramolecular bond interactions. Free O H stretches are associated with the hydroxyl group on C 2 for 1 and 2 and the hydroxyl groups on C 2 and C 4 in structure 4 Band II at 3637 cm 1 with a fwhm of 9 cm 1 is the result of either a hydrogen bond or Rb + [ O methyl D g lucoside ] interactions. A hydrogen bond between the C 4 hydroxyl hydrogen and the C 3 oxygen occurs in 1 2 and 3 Structure 3 introduces a second weaker interaction between the C 3 hydroxyl hydrogen and the C 2 oxygen. There is also rubidium interplay with the hydroxyl oxygen on C 6 for conformers 1 and 2 These motifs weaken and shift the O H stretches from band I to band II. Band III at 3560 cm 1 with a fwhm of 20cm 1 is accounted for by the 3 and 4 conformers. In s tructure 3 the rubidium cation interacts with the C 4 hydroxyl oxygen to furth er weaken the hydrogen bonded O H stretch mentioned above and this interaction corresponds to the large r peak (3597cm 1 ) in band III. The interrelation between the hydroxyl of C 6 and O methyl oxygen of C 1 creates one of two interconnected hydrogen bonds in the 4 structure which in this case moves the O H vibrational mode to the smaller peak (3540cm 1 ) in band III. Band IV is centered at 3450 cm 1 with a fwhm of 32 cm 1 With the 4 conformer s 1 C 4 structure (which tucks the C 4 hydroxyl group under the ri ng), the hydroxyl hydrogen from C 3

PAGE 136

136 is attracted to the oxygen on C 6 forming a second hydrogen bond across the ring. The network of hydrogen bonds helps shift this frequency to band IV. The C H region consists of two bands V, VI, that are made more eviden t by the two laser experiment considered earlier The C H stretches correspond to the symmetric (VI) and antisymmetric stretches (V). Lowest in energy, the symmetric modes are related combinations of degenerate C H motions from the ring, the ethyl group on C 6, and methyl group on C 1. Antisymmetric stretches are related to the ethyl group on C 6 and the methyl group on C 1 (from lowest to highest energies of vibration, respectively). These have higher predicted energies than the values in the experimental spectra. The predicted spectra of all four conformers, shown with the experimental spectrum in Figure 5 8, illustrate good overlap with all bands. O m ethyl D glucoside Th e O H stretch region of O m ethyl D glucoside correlates 9). Bands I, III, and IV have bandwidths of 21, 54, and 45cm 1 respectively The C H stretches are in similar positions as well, although the intensities are reversed and the overall wavenumber range is decreased to 2900 3050 cm 1 H stretching region of O m ethyl D glucoside ure 5 10) contains bands I and II, but the O H stretches in bands III and IV are not accounted for As band II does not exist in the experimental spectrum, two other calculated conformers are considered, 1 ( 1 C 4 ) and 2 ( O,3 S ), as seen in Figure 5 1 1 Unperturbed hydroxyl groups on C 2 and C 4 for 1 as well as on C 3 for 2 correspond to band I. Conformer 1 has only one hydrogen bond, the hydroxyl hydrogen on C 3 to the oxygen on C 6, which accounts for band III. Band IV is formed by st ructure 2 t he result of hydrogen bonding from the hydroxyl hydrogen of C 6 to the oxygen on C 4 and

PAGE 137

137 interaction of the Rb + with oxygen on C 6. Due to the lack of precise overlap between the computational conformers and the experimental bands III and I V, further computational work is necessary. Although the peak heights are reversed, the C H stretches agree in general with the assignments of Glc following the pattern of symmetric (VI) to antisymmetric (V) stretches. Galactosides O m ethyl D g alactosi de. For f our bands (II, III, V, and VI) appear at 3650, 3580, 2959, 2915 cm 1 with bandwidths of 30, 17, 17, 29 cm 1 respectively Experimental and computational spectra are illustrated in Figure 5 12 with the corresponding lowest energy structure. The general composition of the conformer is 4 C 1 with the rubidium cation above and central to the C 6 oxygen, O 1 and C 1 O methyl group The O H stretch region spans 3550 to 3680 cm 1 covering bands II and III. Relatively weak interaction of the C 2 an d C 4 hydroxyl hydrogens and the oxygen of C 3 places the C 3 and C 2 O H stretch es at the blue shifted band II (band II slightly higher in than in and includes some overlap within the region of band I) A hydrogen bond between the hydroxyl hydrogen of C 6 to the oxygen on C 4 and the C 6 oxygen connection with the rubidium cation further shifts these modes to band III. Alt hough the overall coverage is good, the theoretical peaks are a bit red shifted with slightly better vibrational mode resolution in comparison to those in the experimental spectrum. The C compu tational (2855 to 3062 cm 1 ) and experimental (2855 to 3000cm 1 ) regions indicates common spectral features with a higher energy spread for the theoretical antisymmetric stretches. O m ethyl D g alactoside. The experimental spectrum of the Gal isomer con sists of five bands centered at 3660, 3600, 3550, 2960, 2915 cm 1 with respective broad bandwidths of 30, 27, 38, 27 and 30 cm 1 Figure 5 13 includes the Gal experimental and computational spectra of

PAGE 138

138 the lowest energy conformer (including structure) rel ating to bands II III, V, and VI Similar to Gal, is blue shifted from the Glc position. The best agreement of theoretical and experimental spectra for a lpha galactose, as for is the 4 C 1 structure with the rubidium cation between O 1 and the oxygen of C 6 The lowest energy structures and computational spectra of with two notable differences: the computed Gal spectrum shows better resolution of the vibrational modes and predicts a red shift in the feature s of band III. When compared to the experimental spectrum in Figure 5 13 these distinctions become apparent. The e xperimental spectrum of Gal has two clear peaks within the blue shifted band II ( compared to only a slight splitting for ) where the vi bration of relatively free hydrogen from the C 3 hydroxyl group is of slightly higher energy than the C 2 O H stretch which weakly interact s with the C 3 oxygen. Two distinct areas of lower and higher energy also exist for the C 4 and C 6 O H stretches ( b and III), respectively. It appears that the closer rubidium interaction (due to the displaced oxygen in the axial position of the alpha sugar) with the C 6 O H stretch shifts the mode to the lower energy region of b and III while the C 4 O H stretch experi ences a corresponding slight increase in energy Overall, the computationally determined spectra of the O H stretches overlap with the experiment although a red shift is noted. The long shoulder that tails into b and IV is yet to be accounted for (experime nt pending). Alpha galac t os ide has a C H stretch pattern comparable to that of Glc. The computational and experimental spectra of the C H stretches (Figure 5 13 ) illustrate some differences in bands V and VI. However, symmetric stretches (VI) agree in position but are not of the same intensity. In the predicted spectrum, asymmetric stretches of greater energies also form distinct peaks above 2980cm 1

PAGE 139

139 Isomeric differentiation There are two general me thods of differentiating isomers from their corresponding spectra. The first is noting the wavelength( s) where the presence or absence of isomer band (s) takes place with respect to the other isomer band (s). The second is contrasting overlapping spectral features and looking for defining ratio differences. Figure 5 H stretch spectra where isomeric d ifferences are found for each band. Band I is comprised of the free O difference or a careful choice of wavenumbers (in a range > 3689cm 1 ) all ows the differentiation between the two sets of D g and absence of a peak for Due to the blue shift ed peak, if the ratio of the peaks of and/or is likely. Further comparison between for the >1, <1 conditions 5 3 and 5 4 provide s an other means of differentiation, where differences in the size of the ratios offer the potential for determining the presence of isomers indicated by the subscripts This technique is easily extended to and b etween other bands in the spectra. All isomers are accounted for in band III. The shoulder at ~3597cm 1 signifies 1

PAGE 140

140 existence A lack of dissociation in the region~3544 cm 1 and 3572cm 1 point s toward the nonexistence of all isomers with the possible exception of the respective There are also relative intensity differences between the D glycosides and each anomer. Band IV and peaks arely present a nd is completely missing. If dissociation is there, the difference in the intensities of the peaks differentiation. With the C H stretches (F igure 5 15) the intensities of band V and VI for alpha and beta anomers are inverse of each other where a simple ratio should indicate the anomeric form. The beta anomers also possess a low energy tail where alpha anomers do not. Conclusion Frequencies ( cm 1 ) of the O H and C H stretches for the ions of rubidium D glycoside isomers, Rb + [O methyl D glucoside], Rb + [O methyl D glucoside], Rb + [ O methyl D galactoside], and Rb + [O methyl D galactoside] are capable of being characterized through IRMPD i n a FTICR m ass spectrometer. Early attempts to find a means of differentiating four D g lycoside isomers in the mid IR range found little discernable difference H owever in examining the less congested O H and C H stretch regions distinctions are found. These di fferences follow patterns of H b onding and Rb + ion interactions. The highest energy band I is the result of the unperturbed motion of the O H stretch. From here, weaker hydrogen bonding along the ring and independent Rb + interaction shifts the vibr ations to band II. In band III the rubidium interaction with oxygen involved with ring hydrogen bonds further weakens the interrelated O H stretch. In addition, other combinations of hydrogen bonds and single cross ring hydrogen bonds in the 1 C 4 confo r mer s appear to also shift the frequencies to band III. The cross ring hydrogen bonding when combined with another interconnected hydrogen bond shifts the former into band IV. As expected the general trend appears to indicate that the more involved

PAGE 141

141 the rubi dium ion and the stronger the interaction(s) of the hydrogen bond(s) the weaker the O H stretch. The differentiation of all four isomers is the result. Figure 5 1. The basic structure of O methyl D gluco methyl D gluc methyl D galacto methyl galacto ) isomers ( all with a molecular weight of 194) with the C 1 O methyl and C 5 hydroxyl groups in the axial/ equatorial positions for a lpha/beta and the glucoside/ galactoside conformations respectively.

PAGE 142

142 Figu re 5 2. Action spectra of the four glycoside isomers Glc, Glc, Gal, and Gal from 600 1700cm 1 obtained by Valle et al. at FELIX [ Adapted with permission from Valle, J.J. 2005. Differentiation o f Carbohydrate S tereoisomers by I nfrared M ultiple P hoton D issociation U sing a F ree E lectron L aser and a F ourier T ransform Ion C yclotron R esonance M ass S pectrometer Ph.D. dissertation (Page 115 Figure 5 5 ) University of Florida, Gainesvil le, Florida.]

PAGE 143

143 Figure 5 3. The OPO laser general set up with the idler 1/signal and idler 2 beams coming from the front and back of the laser box (corresponding to the resonance cavity inside of the OPO 4000) respectively The signal beam travels to the wavemeter as both idler 1 & 2 are optically guided to the center of the FTICR cell. An off resonance CO 2 laser is used in th e double laser experiments. Two iris shutters (one illustrated here), regulate the beams entry and, when closed guide the beams into the power meter. Figure 5 4. Differences in the C H stretch in action spectra of Glc. A) resonant OPO laser only; B) resonant OPO laser coupled with an off resonant CO 2 laser. The CO 2 laser provides additional energy to the already excited ions by the OPO laser in the quasicontinuum. 2750 2850 2950 3050 3150 3250 3350 Wavenumbers (cm 1 ) OPO OPO & CO 2 Iris Idler 1 Signal Idler 2 Power Meter OPO & CO 2

PAGE 144

144 Figure 5 5. Theoretical (color) and experime ntal (grey) spectra of similarities between each isomer in th e 900 1400cm 1 region. [ Adapted with permission from Contreras, C.S. 2008. Carbohydrates and Amino Acids: Infrared Multiple Photon Dissociat ion Spectroscopy and Density Functional Theory Calculations. Ph.D. dissertation (Page 134 135, Figure s 6 7, 8, 9 ) University of Florida, Gainesville, Florida.] Figure 5 6. Action s pectr um of O methyl D indicat ing the general ranges and specific bands of C H (2 bands) and O H (4 bands) stretching motions. 2750 2950 3150 3350 3550 3750 Wavenumbers ( cm 1 ) OH Stretch Region ~ 3200 3650 cm 1 CH Stretch Region ~ 2850 3000 cm 1 B D I III II II IV V VI

PAGE 145

145 Figure 5 7. Comparison of the experimental (red) and theoretical spectra (with structures) of the four conformers ( 1 2 3 4 ) of O methyl D g lucoside

PAGE 146

146 Figure 5 8. Overview of the theoretical and experiment al spectra for rubidium tagged conformers 1 Glc (dk blue), 2 Glc (green) 3 Glc (aqua), and 4 Glc (dk grey) compared to the experimental spectrum ( red).All six bands are accounted for. Figure 5 9. The IRMPD spectra of noted absence of band II. 2750 2950 3150 3350 3550 3750 Wavenumbers ( cm 1 ) Exp. Conf. 1 Conf. 2 Conf. 3 Conf. 4 2750 2950 3150 3350 3550 3750 Wavenumbers (cm 1 ) Alpha Beta

PAGE 147

147 Figure 5 10. Action spectra of O H stretching region for Glc (blue) and the calculated lowest energy conformer. [Reprinted with permission from Contreras, C.S. 2008. Carbohydrates and Amino Acids: Infrared Multiple Photon Dissociation Spectroscopy and Density Functional Theory Calculations. Ph.D. dissertation (Page 141 Figure 6 19 ) Univers ity of Florida, Gainesville, Florida.]

PAGE 148

148 Figure 5 11. The IRMPD spectra and structures for conformers 1 Glc (black and left) and 2 Glc (blue green and right) and the experimental spectrum (blue) [ Adapted from Contreras, C.S. 2008. Carbohydrates and Amino Acids: Infrared Multiple Photon Dissociation Spectroscopy and Density Functional Theory Calculations. Ph .D. dissertation (Page 142 Figure 6 20, 21) University of Florida, Gainesville, Florida.] Figure 5 12. Experimental and theoretical spectra for Gal ( exp.: neon green theory: black ) and lowest energy conformer structure Wavenumbers (cm 1 )

PAGE 149

149 Figure 5 13 Experimental and theoretical spectra for Gal ( exp.: brick red theory: black ) and lowest energ y conformer structure. Figure 5 14. Spectra of the O H stretch region for all four isomers with spectral band labels. 3375 3425 3475 3525 3575 3625 3675 3725 Beta Gal Alpha Gal Alpha Glc Beta Glc Wavenumbers ( cm 1 ) I II III IV Wavenumbers (cm 1 )

PAGE 150

150 Figure 5 15. Spectra of the C H stretch region for all four isomers with spectral band labels. 2850 2870 2890 2910 2930 2950 2970 2990 3010 3030 3050 Wavenumbers (cm 1 ) Beta Gal Alpha Gal Alpha Glc Beta Glc VI V

PAGE 151

151 CHAPTER 6 THERMOCHEMICAL DYNAM ICS OF DEPHOSPHORY LATION In recent years, mass spectrometry has evolved into one of the most important tools in protein chemistry. 2 3 5 ,2 3 6 Significant to proteomics is the identification and characterization of posttranslational modifications; principally phosphorylation of the amino acids serine, threonine, and tyrosine. 2 3 7 ,2 3 8 Catalyzed by kinases, phosphorylation is an exothermic process that occurs through the transfer of a phosphate group from adenosine triphosphate (ATP) to the alcohol side chain of the amino acid. Sinc e it is r eversible, phosphorylation controls many cellular functions, like metabolism, protein synth esis, growth, and reproduction. 23 ,2 3 9 This is largely accomplished through peptide conformational changes brought about by the insertion/ interaction and rem oval /noninteraction of the phosphate group and its negatively charged oxygens (under physiological conditions) with the protein. 2 40 Understanding the kinetics and t he activation energies associated with dephosphorylation give s insight into the nature of th e phosphate p eptide relationship. 34 1 3 7 ,2 4 1 Traditionally tandem mass spectrometry is used to identifying the presence and/or sequence of amino acids bearing phosphate groups through their fragmentation products. In order to accomplish this collision indu ced dissociation (CID), electron capture dissociation (ECD), infrared multiple photon dissociation (IRMPD) etc. are employed as sole and combined fragmentation techniques within the field of mass spectromet ry 2 4 2 2 4 7 Of the approaches used Fourier trans form ion cyclotron resonance ( FTICR ) mass spectrometry is one of the most capable with MS n capability, high resolution, and mass accuracy. 44 Here, precursor ions are isolated, activated, and then dissociated. The presence of phosphopeptides is noted by the neutral phosphate loss from their precursor ions (typically HPO 3 and H 3 PO 4 which measure 8 0 a nd 98 D altons respectively ) 2 4 8 2 50 A difficulty in these techniques is the fragile posttranslational

PAGE 152

152 modification that readily dissociates requiring careful de termination of dissociation conditions. This restriction however, allows for the relative isolation of the dephosphorylation reaction and getics, and IRMPD spectroscop y. 1 3 6 ,1 3 7 ,1 3 9 ,2 4 1 ,2 5 1 In the early 1990s, Dunbar proposed that the long relatively collision free retention times of an FTICR mass spectrometer offers a means to study slow kinetic processes that develop from ions irradiated in the cell. 112 115 He theoriz ed that a continuous wave infrared laser, at low power, is capable of acting as a primary thermal source in equilibrium with the spontaneous and stimulated emission of photons from the ions in an ICR cell. Under these conditions, a Boltzmann like distribut ion of energy with an effective temperature is established. Dissociation energy bond occurs when the exceeds a threshold energy ( E t ) typically associated with the high energy tail of the distribution. By plotting the natural log of the dissociated parent ion fraction versus time dissociation rate constants (k d ) are calculated. And through the application of a modified form of the Arrhenius equation, relative activation energies are determined according to 6 1 where k d is the unimolecular rate constant, P the laser power, C a proportionality factor, and E a the relativ e activation energy as seen in C hapter 3. However, the exact relationship between the laser power and temperature is difficult to quantify The association was made clearer after McMahon and Dunbar discovered that blackbody infrared dissociation (BIRD) is possible throu gh the heating of the ICR cell. 32 117 ,2 5 2 ,2 5 4 Through master equation modeling, the discovery of the rapid exchange limit an d comparative experimental results between the BIRD and IRMPD data Paech and Williams established a means to correlate power to temperature for a series of

PAGE 153

153 var ying sized peptides, where the constant C in Equation 6 1 is given a value of (Dunbar C= qh as defined in Chapter 3) 33 120 123 13 5 ,1 3 8 ,1 3 9 Two tunable cw CO 2 lasers (the Apollo and Lasy 20G AT) were used in the course of these experiments at wavelengths of 9.294, 9.588, and 10.632 m Each wavelength was used in turn to examine the kinetic and activation energy similarities and differences for both +2 and 2 charge states of a monophosphorylated peptide ion consisting of 16 residues and a mass of 2061.8291 Da Action s pectra were also taken for both charge states. These experiments in volved th e Lasy 20G AT tunable cw CO 2 laser beam focused on the phosphopeptide ions in the cell of a 4.7T FTICR mass spectrometer. wavelength is in resonance with an ion(s) vibrational mode, resulting in energy a bsorption. The energy is then redistributed to background vibrational states through intern al vibrational relaxation (IVR) building up to eventually dissociate the weakest bond(s). 99 By measuring the fraction of the precursor ion remaining in the mass spe ctrum particular to a specific wavelength for a series of mass spectra/wavelengths an IRMPD spectrum is formed. 95 These results are used to help determine the effect if any, of wavelength choice and the ion s charge on the kinetic/energetic picture. Exp erimental Methods Materials The mon o phosphopeptide was purchased from Sigma Aldrich as part of the phosphopeptide positive control set (product number P9615) and consisted of Phe Gln pSer Glu Glu Gln Gln Gln Thr Glu Asp Glu Leu Gln Asp Lys residues taken from a n HPLC purified tryptic digest of B C a s ein. Methanol and f ormic acid were purchased from Sigma Aldrich and

PAGE 154

154 Fischer Scientific respectively and the milliQ water was obtained from the physical chemistry teaching lab at the University of Florida. Stoc k solutions were prepared by dissolving 0.1mg of the monophosphopeptide in 1000 l of milliQ water. From this supply 100 l was taken and diluted to 1000 l comprising a 50:50 MeOH:H 2 0 solution. Protonation was accomplished through the addition of 1% formi c acid and controlled deprotonation of the acidic peptide through the addition of ~1.18x 10 6 M formic acid. Instrumentation The experiments were performed on two instruments. Positively charged ion experiments were executed in the I nfinity cell of a passi vely shielded 4.7T (Magnex Scientific Ltd; Abington, U.K.) Bruker (Bruker Daltronics; Billerica, MA) FTICR mass spectrometer (BioApex 47e) united to an Analytica (Branford, CT) electrospray ionization (ESI) source and a modified heated metal capillary (120 o C) with a conical inlet. Phosphopeptide s olutions of 4.847 x10 6 M were introduced into the ESI source, via a fused sil i c a transfer line, a t a flow rate of 15 hr. and a 2.3 k V potential was applied between the ESI needle and capillary. Predator (an extension of the modular ICR data and acquisition and analysis system (MIDAS)) 25 3 and the external source control (XSRC) programs were used to control the functions of the infinity cell/data a cquisition and the ion optics, respectively. Apollo and Lasy 20G AT tunable cw CO 2 lasers were used as the source of IR photons. Negatively charged ion experiments were performed in the I nfinity cell of a n actively shielded FTICR mass spectrometer coupled to an Analytica ESI source ( Mass Spectrometry Services Laboratory at the University of Florida ) Solutions of 4.847 x10 6 M were used at a The ESI source was assisted by a nebul izer and

PAGE 155

155 desolvation gas (N 2 ) at a rate of 35 and 155 L/hr, respectively. A 2.9k V potential was applied between the ESI needle and capillary Instrument parameters and data collection were performed by the Bruker Xmass data acquisition system 7.0 A Lasy 2 0 G AT tunable cw CO 2 laser was the sole source of IR photons. Initial positive ion experiments were begun with the repaired Apollo CO 2 laser. The Lasy 20G AT CO 2 laser w as later integrated into the system, as seen in Figure 4 8 and became the laser of ch oice due to automated wavelength control and superior power stability. The laser beams were reflected from a series of gold mirrors (the second to last a focusing mirror ) to pass through a ZnSe window into the cell located directly opposite the ions entra nce. A laboratory constructed laser gate was installed to regulate the irradiation time. In the upright position, a mirror mounted on the gate redirected the beam to the power meter. When activated from allowed the beam to pass. Experimental P rocedure After preparing the appropriate solutions and injecting the sample into the source, ions were accumulated for 2s in a hexapole storage region and transferred via ion optics to the cell of the FTICR mass spe ctrometer. The phosphopeptide ion signal was found, optimized, and isolated. Positive ion isolation took place in two stages utilizing s tored wave form inverse Fourier transform (SWIFT) and chirp ejections. Negative ions were i solated with swept frequency e jection pulses. A 0.5sec delay time was introduced after each isolation step to cool the ions. Once trapped and isolated the ions were irradiated by the CO 2 laser at one of three wavelengths (9.294, 9.588, and 10.632 m ) at 1W of power for an irradiation t ime of 1s. Thirty scans were taken to improve s/n and the relative abundances of precursor and product ions. The procedure was repeated for 2s, 3s, etc., the power adjusted, and the process continued for each successive increase in power until the data for the kinetic and Arrhenius plots were complete. The

PAGE 156

156 wavelength was then changed and the process repeated until all the wavelengths were finished. Mass spectra of the precursor phosphopetide (P) and product (F) ions were analyzed and their relative abundanc es noted in an E xcel spreadsheet as seen in Table 6 1 The natural log of { [P]/([P]+[ F ])} versus irradiation time plots were created and the d issociation rate constants ( k d ) obtained from the absolute value of the slopes. Activation en e rgies ( E a ) were then determined from the modified Arrhenius plot o f the ln [ k d ] versus ln [laser Power] associated with each k d as seen in Equation 6 1. Ions for the action spectra were introduced and isolated in the cell by the methods previously mentioned. The irradiation time however was limited to 1s at 5W for positive ly charged ions and 2.5 watts for the negatively charge d ions. A mass spectrum was taken for each wavelength in the CO 2 spectrum from 9.174 9.354/9.443 9.773 and 10.105 10.365 / 10.441 10.885 m omit ting wavelengths that lacked appropriate powers (at the far ends of the P and R branches). From the mass spectrum the precursor (P) and dissociation product ions (F) were noted and plotted as the versus Experi mental Results and Discussion Phosphopeptide Mass Spectra Infrared multiple photon dissociation mass spectra for positive ly and negative ly charged phosphopeptide ions are seen in Figure 6 1 and 6 2. Irradiated at 10.6 m at 6 W for 4 sec and at 2.5 W for 3.5 se c the respective positive and negative ions mass spectra illustrate the same pattern of phosphate loss The dissociation pathway is made clearer through the knowledge that elimination is the phosphate dissociation mechanism for phosphoserine (Figure 6 3 ) where the H 3 PO 4 loss creates the serine replacement dehydroalanine and corresponding product. 25 5 2 5 6 Thus, the phosphate d issociation products are seen as originating with the precursor ion

PAGE 157

157 [P12H + ] 2 and precursor ions minus one/two water molecules ( [P12H + ] 2 H 2 O) and ([P12H + ] 2 2H 2 O). D ifferences between the spectra include evidence of phosphopeptide backbone fragmentation and dissociation products The excess internal energy in the irradiated ion cleaves the positively charged phosphopeptide i on between y 7 (m/q of 977.5 ( TENELQNK )) and b 9 (m/q of 1088.5 ( FQ p S EEQQQ )) with further dissociation at [b 9 ] H 2 O. Here the dephosphorylation and the backbone cleavage appear to be competing low energy pathways The negatively charged phosphopeptide ion s pectra do not contain evidence of competing pathways, though excess internal energy creates further fragmentation (m/q 835, 826, and 819) from dephosphorylated products (m/q 981,972, and 965, respectively) each consistent with a loss of 146 m/q (q= 2). T he relatively small mass difference, intact peptide, and the low internal energy indicate a small peptide loss, probably located at one of the termin i cleaved at the amide bond. The dipeptide consisting of p henylalanine (F) and g lutamine (Q) at m/ q of 29 2 ( y 2 ) is consi stent with the loss of the y fragment (neutral oxazolone or +1 charge) Phosphopeptide Kinetics Once the phosphopeptide IRMPD spectra were acquired, a graph comparing the fraction of the parent ion depleted over time, at different powers, was created for each wavelength. These pseudo first order kinetic plots for both the positive (a t 9.29 4, 9.588, 10.632 m ) and negative (at 9.29 4 and 10.632 m ) ions are seen in Figure 6 4 and 6 5 respectively. Readily apparent (Table 6 2 ) are the difference s between the dissociation rate constants of the negative and positive ion modes. These differences favor dephosphorylation of the negatively charged phosphopetide ions by as much as 30 to 50 times th at of their pos itively charged counterparts. Given the s imilar pattern of phosphate loss and common elimination mechanism, it would

PAGE 158

158 appear that conformational differences and the corresponding phosphate inter actions within these structures account for the kinetic disparity between the two charge states. The w ork of Yan Ting Guo et. al demonstrates one possible explanation for kinetic differences in the charge dependent ratio of trans and cis conformers, based on the orientation of the alpha carbons of proline and serine (phosphorylated and nonphosphorylated) residues in the smaller peptides, Ac ApSPK NH 2 and Ac ASPK NH 2 respectively. 25 7 Utilizing nuclear magnetic resonance (NMR) to determine equilibrium constants ( ), the authors calculated the cis ratio (%) for the nonphosphorylated and the 1 and 2 charge states of the phosphorylated peptides resulting in values of 8.4, 10.0, and 18.1, respectively. The increased cis ratio is the product of increased isomer stability due to the H bonding between the negative ly charged oxygen(s) on the phosphate ion(s) and nearby amide protons (Figure 6 6 ). Even so, the energy of the cis structure is still higher than the trans counterpart. This is communicated by K trans f or the trans ion under the same conditions Cis energy is also implied by the very nature of peptides, where the trans isomer is favored placing bulky side chain ( R) groups (attached to the alpha carbons) on o pposite sides of the C N bond. The implication is that positively charged phosphopeptide ions with neutral phosphate groups favor the trans form more than their negative complement s Therefore, positive ions do not dissociate as rapidly. The principle might help explain part of the kinetic picture in the current experiment; however, with larger systems many interactions are possible, as a pseudo first order reaction would suggest. Computational studies isolating the environment of the phosphate group could be useful in unraveling these mysteries.

PAGE 159

159 Rela tive A ctivation E nergies and I nfrared M ultiple P hoton D issociation S pectra Kinetic differences can also be a reflection of different activation energies. Examination of s 6 7 and 6 8 indicates that the corresponding ac tivation energies for the positive ions are greater than for the negative ions. This is confirmed in Table 6 3 where the relative act ivation energies are given along with their 95 % confidence limits (CI) indicating that the lower the energy required for d issociation the faster the dissociation rate. In other words, the threshold energy associated with the high energy tail of the Boltzmann distribution is lower for negative ions resulting in more ions leaving the distribution compared to the positive ions under the same conditions The large error associated with the relative activation energies might throw this idea into question. Besides the nature of the experimental conditions and the error calculation s which are believed to account for a large part of the error, the kinetic picture and mutual agreement among both positive and negative relative activation energies appear to lend credence to the above interpretation. Further comparison between the relative activation energies of positive and negative io ns and the IRMPD spectra (Figure 6 9) reveal s that differences in the IRMPD spectral intensities do not fundamentally affect the relative activation energies. The agreement follows from the idea that the laser acts as a thermal source, exciting ions throug h laser wavelength ion vibrational resonant absorption creating a Boltzmann distribution of energies. As such so long as the rapid exchange limit (REX: where the exchange of photons emitted and absorbed by the ensemble is much greater than dissociation) i s met and holds true, the number of photons absorbed by changing the power, time, or the efficiency of absorption (i.e., by choosing a vibrational mode It is believed that the large error associated with the relative activation energies of the positively charged phosphopeptide ions is in part due th e effect of the relatively small product yield/peak differences. However, both charge states suffer from the relationship between the number points that define the line of the plots and their 95% CIs. These issues can be remedied by larger parent signal an d the acquisition of more rate constants.

PAGE 160

160 with a stronger transition dipole moment) etc., only changes the number of ions crossing the energy bar rier but does not mischaracterize the activation energy. However because the exact relationship between temperature and laser power is unknown the laser power dependent Boltzmann distribution finds correlation to the temperature distribution t hrough diff erent approaches, like the Dunbar and Paech and Williams models. 31 33 113 13 9 Of these two approaches the P ae ch Williams model is considered the more accurate for the current monophosphorylated peptide, which corresponds to the same class (peptides and p roteins) of ions, is within the REX limit, and has many DOF (where spontaneous emission events are far greater than stimulated emission(s) at the resonant vibrational mode(s)). 33 R elative activation energies determined with the Dunbar approach are placed i n Table 2 2 for comparative purposes. Flora experiments on synthesized 34 and harvested phosphopeptides 13 7 [ GAG pS GAG ] 1 and [ TRDIYETD p YYRK ] 2 modeled by Dunbar s relationship gave dephosphorylation relative activation energies of 0.48 eV and 0.43eV respectively. These values are 0.17eV and 0.12eV above the highest value obtained at 9.294 m mainly 0.31 0.09, though they do fall within the error noted by both experiments. Comparisons however are made difficult due to the difference in the charge state of the first relationship, where higher char ge states t ypically correspond to greater dissociation yields and lower activation energies 13 5 and the different (nonserine) amino acid tyrosine in the second case. Differences in the size and s tructure of these molecules may also play a crucial role. The higher relative activation energy value for the positive ion taken at 9.2 9 4 m is a bit of an anomaly I t is clearly greater than the other positive ion values though it does agree within

PAGE 161

161 error Some of the discrepancy may be accounted for in the last line of the kinetics plot that once corrected through further experiments should bring the energy into line with the current trend. Conclusion Dephosphorylation of a monophosphorylated peptide re sulted in the loss of an H 3 PO 4 neutral from the precursor ion as well as subsequent water loss product ions. Illustrating competing low energy pathways (positive mode) and excess internal energy fragmentation (positive and negative modes), the phosphate di ssociation was still unencumbered by interferences. This made the analysis of the unimolecular dissociation kinetics and the relative activation energies easier. Negative ion dephosphorylation at all wavelengths provided evidence of higher rate constants t han positive ion motifs. This in turn corresponded to lower negative ion relative activation energies believed to be due to possible conformational interaction ( s ) between the negatively charged oxygen(s) of the phosphate group and its surroundings. Relativ e activation energies were also found to be independent of vibrational mode intensities (at least as measured by the dissociation action spectra ). Overall the activation energy for dephosphorylation of the 2 charge state was found to be 0. 51 0.11 eV and t he +2 charge state 0.82 0.25 eV.

PAGE 162

162 Figure 6 1 An example of an IRMPD phosphopeptide mass spectrum of the +2 charge state after CO 2 laser irradiation time of 4.0s at 10.632 m and 6W. Loss of the phosphopeptide ion occurs through elimination from the precursor ion and its water elimination products. Fragmentation of the phosphopeptide backbone is also noted by the y 7 and b 9 fragments. [P 1+2 H] +2 y 7 + [ P1+2H + 2H 2 O] +2 H 3 PO 4 [ P1+2H + ] +2 2 H 2 O [ P1+2H + ] +2 H 3 PO 4 [ P1+2H + H 2 O] +2 H 3 PO 4 b 9 + [( P1 +H + ) + Na] +2 [ P1+2H + ] +2 H 2 O [b 9 ] + H 2 O

PAGE 163

163 Figure 6 2 An example of an IRMPD phosphopeptide mass spectrum of the 2 charge s tate after CO 2 laser irradiation time of 3.5 s at 10.632 m and 2.5 W. Loss of the phosphopeptide ion occurs through elimination from the precursor ion and its water elimination products. Excess internal energy results in the loss of y 2 fragment from the d ephosphorylated peptide Figure 6 3. Diagram illustrating the elimination product for dephosphorylation of phosphoserine [ P1 2H + 2H 2 O] 2 H 3 PO 4 [ P1 2H + ] 2 H 2 O [ P1 2H + ] 2 2 H 2 O [ P1 2H + H 2 O] 2 H 3 PO 4 [ P1 2H + ] 2 H 3 PO 4 [ P1 2H + ] 2 [( P1 2H + + e ) + Na + ] 2 [ P1 2H + 2H 2 O H 3 PO 4 ] 2 y 2 [ P1 2H + H 2 O H 3 PO 4 ] 2 y 2 [ P1 2H + H 3 PO 4 ] 2 y 2

PAGE 164

164 A B lnP f@ 3 15 W = ( 9 4 )x 10 3 t 0 019 0 011 lnP f@ 3 65 W = ( 0 013 0 005 )t 0 029 0 010 lnP f@ 4 15 W = ( 0 026 0 009 )t 0 03 0 02 lnPf @ 4 70 W = 0 05 0 03 t 0 005 0 067 0 18 0 16 0 14 0 12 0 1 0.08 0 06 0.04 0.02 0 0 0 5 1 1 5 2 2 5 3 3 5 4 4 5 lnP f time (sec) 3 15 Watts 3.65Watts 4.15Watts 4 70 Watts ln P f@ 1 93 Watts = ( 2 1 0 8 )x 10 3 t ( 1 67 0 26 )x 10 2 ln P f@ 3 64 Watts = ( 7 5 )x 10 3 t ( 2 8 1 1 )x 10 2 ln P @ 4 00 Watts = ( 0 021 0 005 )t 0 025 0 015 ln P f@ 4 60 Watts = ( 0 035 0 019 )t 0 02 0 04 0 2 0 15 0 1 0.05 0 0 0.5 1 1.5 2 2 5 3 3 5 4 4 5 5 lnP f time (sec) 1 93 Watts 3.64 Watts 4 00 Watts 4.6 Watts

PAGE 165

165 C Figure 6 4 Pseudo first order kinetic plots (w/ 95% confidence limits) measurin g the dephosphorylation of positively charged (+2) phosphopeptide ions over time at the give n powers for A) 9.29 4 m (w/ selected error bars depicting general range of error in these experiments) B) 9.588 m and C) 10.632 m wavelengths. The absolute val ue of each slope gives the corresponding unimolecular dissociation rate constant for the given CO 2 laser power. lnP f@4.00W = (0.019 0.003)t + 0.007 0.009 lnP f@ 4 50 W = ( 0 035 0 021 )t + 0 01 0 05 lnP f@5.00W = (0.049 0.011)t + 0.004 0.027 lnP f@ 6 00 W = ( 0 07 0 03 )t 0 09 0 10 0 5 0 45 0 4 0.35 0 3 0 25 0 2 0.15 0 1 0 05 0 0 05 0 0 5 1 1 5 2 2 5 3 3.5 4 4 5 5 lnP f time (sec) 4 00 Watts 4 50 Watts 5 00 Watts 6.00 Watts

PAGE 166

166 A B Figure 6 5 Pseudo first order kinetic plots (w/ 95% confidence limits) measuring the dephosphorylation of neg atively charged ( 2) phosphopeptide ions over time at the given powers for A) 9.29 4 m and B) 10.632 m wavelengths. The absolute value of each slope gives the corresponding unimolecular dissociation rate constant for the given CO 2 laser power. lnP f @0.78Watts = (0.08 0.07)t + 0.13 0.26 lnP f @ 1 03 Watts = ( 0 16 0 04 )t + 0 14 0 11 lnP f@ 1 50 Watts = ( 0 29 0 07 )t + ( 4 14 )x 10 4 lnP f@ 1 74 Watts = ( 0 51 0 15 )t + 0 10 0 21 1 2 1 0 8 0.6 0 4 0.2 0 0 2 0 0 5 1 1.5 2 2 5 3 3 5 4 4 5 5 lnP f time (sec) 0.78 Watts 1 03 Watts 1 5 Watts 1 74 Watts lnP f@ 1 50 Watts = ( 0 058 0 021 )t + 0 15 0 09 lnP f@2.00Watts = (0.10 0.04)t + 0.05 0.09 lnP f@ 2 50 Watts = ( 0 20 0 06 )t + 0 13 0 17 lnP f@ 3 50 Watts = ( 0 37 0 05 )t 0 013 0 011 1 8 1 6 1.4 1 2 1 0 8 0 6 0 4 0 2 0 0 2 0 1 2 3 4 5 6 7 8 lnP f time (sec) 1 50 Watts 2 00 Watts 2 50 watts 3.50 Watts

PAGE 167

167 Figure 6 6: Cis trans conformers from phosphopeptide study illustrating a possible reason fo r the differences in kinetics and relative Arrhenius activation energies observed in the results reported in this dissertation [Reprinted with permission from Springer. Guo, Y.; Li, Y.; Zhu, Z.; Zhao, Y. 2005. Effect of the Phosphate Group with Different Negative Charges on the Conformation of Phosphorylated Ser/Thr Pro Motif. Int. J. Pept. Res. Ther. (Volume 11, Page 160, Figure 1).] A lnk obs = (4.2 1.7)ln P Watts 9.6 2.3 6 5 4 3 2 1 0 1 1.1 1 2 1 3 1 4 1.5 1 6 1 7 1.8 ln k obs (s 1 ) ln Power (Watts)

PAGE 168

168 B C Figure 6 7 Arrhenius relative activation energy plots (w/ 95% confidence limits) for unimolecular dissociation of phosphate neutrals from positively charged (+2) phophopetide ions at A) 9.2 9 4 m, B) 9.588 m, and C) 10.632 m wavelengths ln k obs = ( 3 1 2 0 )lnP Watts 8 3 2 6 9 8 7 6 5 4 3 2 1 0 0 0 5 1 1 5 2 2 5 ln k obs (s 1 ) ln Power (watts) lnk obs = ( 3 1 1 7 )lnP Watts 8 2 2 7 8 7 6 5 4 3 2 1 0 0 5 0.7 0 9 1 1 1.3 1 5 1 7 1 9 ln k obs (s 1 ) ln Power (Watts) :

PAGE 169

169 A B Figure 6 8 Arrhenius relative activation energy plots (w/ 95% confidence limits) for unimolecular disso ciation of phosphate neutrals from negatively charged ( 2) phophopetide ions at A) 9.2 94 m and B ) 10.632 m wavelengths. lnk obs = ( 2 2 0 6 )lnP Watts 1 99 0 23 4 3.5 3 2 5 2 1 5 1 0 5 0 0 8 0 6 0.4 0.2 0 0 2 0 4 0 6 0.8 ln k obs (s 1 ) ln Power (Watts) lnk obs = ( 2 2 0 7 )lnP Watts 3 8 0 6 4 3 5 3 2 5 2 1.5 1 0 5 0 0 0 2 0.4 0 6 0 8 1 1 2 1 4 ln k obs (s 1 ) ln Power (Watts)

PAGE 170

170 A B Figure 6 9 The IRMPD (Action) s pectra (with selected error bars depicting general range of error) for both positively (A) and negatively (B) charged phosphopeptides covering the majority of the tunable CO 2 wavelength range from 9.193 to 9.354 (Blue), 9.458 to 9.714 (Red), 10.1 25 to 10.365 (Green) and 10.4 76 to 10. 787 (Purple) m

PAGE 171

171 Table 6 1. Example of the conversion of the product and fragment ions from a mass spectrum to a point on a kinetic plot of ln {[P]/([P]+[F])} versus time This point represents the dephosphorylation of a positively charged (+2) monophosphopeptide ion irradiated for 4.0s at 6W and 10.632 m Table 6 2 Unimolecular rate constants for dephosphorylation of both negatively and positively charged phosphopetide ions at the given wavelengths and corresponding powers. Isotopes m/ q 1031.5 1022.5 1013.5 982.5 973.5 964.5 1088.5 977.5 614.3 858.4 1032.0 1023.0 1014.0 983.0 974.0 965.0 1089.5 978.5 615.3 859.4 1032.5 2023.5 1014.5 983.5 974.5 965.5 1090.5 979.5 616.3 860.4 100.79 0 4.661 1.353 5.75 9 1.883 0.512 6.408 4.539 8.087 1.416 65.94 8 5.379 1.3 10 7.34 7 2.762 1.086 4.39 4 2.959 3.686 1.021 24.63 4 3.412 0.483 4.43 2 2.090 1.342 1.2211 0.680 1.207 0.784 Relative Abundance s 191.375 16.599 27.212 12.023 8.179 12.9 81 3.221 207.974 63.616 {[ P ]/([ P ]+[ F ])} ln{[ P ]/([ P ]+[ F ])} vs. 10.632 m 0. 766 0.26 7 Unimolecular dissociation rate constants, k d (s 1 ) / Power (W) Positive Ion s 9.294 m 0.009/2.15 0.013/3.15 0.026/4.15 0.048/4.70 9.588 m 0.002/1.92 0.007/3.64 0.021/4.00 0.035/4.60 10.632 m 0.019/4.00 0.035/4.50 0.049/5.00 0.070/6.00 Negative Ion s 9.294 m 0.077/0.78 0.160/1.03 0.287/1.50 0.513/1.74 10.632 m 0.058/1.50 0.102/2.00 0.20/2.5 0.372/3.50

PAGE 172

172 Table 6 3 Relative Arrhenius activation energies with 95% confidence limits utilizing both Model Positive Ion Rel. E a (eV) 9.294 m 0. 59 0.24 Average 1.0 0.4 Average 9.588 m 0.4 2 0.28 0.46 0.14 0. 7 1 0.5 0.82 0.25 10.632 m 0.39 0.21 0.7 4 0.4 Negative Ion Rel. E a (eV) 9.294 m 0. 31 0.09 Average 0.5 1 0.15 Average 10.632 m 0. 28 0.08 0.29 5 0.06 0. 52 0.15 0.51 5 0.11

PAGE 173

173 CHAPTER 7 CONCLUSION S AND FUTURE WORK Infrared multiple photon dissociation of ions in the cell of a Fourier transform ion cyclotron resonance mass spectrometer has led to the cre ation of action spectra for four rubidium D glycoside isomer complexes and subsequent structural elucidation based on frequency agreement with theoretical structures as well as th e determination of thermochemica l properties for positively and negatively c harged monophosphorylated peptide ions. These experiments also introduced three laser systems into the Eyler and Polfer labs: two carbon dioxide gas lasers and an optical parametric oscillator lase r that are still being utilized for a myriad of experiments The OPO laser takes advantage of the nonlinear properties of a periodically poled lithium niobate crystal that exhibits second order susceptibility characteristics. Key among these properties is the ability to convert the pump laser beam photon (Nd:YAG l aser) into two separate photons (signal and idler) of differing wavelengths t he sum of whose energy equals that of the original photon. The laser, covering a range of 1 .38 2.0 and 2.28 4.67 m for the respective signal and idler beams was acquired in order to explore the vibrationally uncongested O H, N H, and C H stretch regions with IRMPD spectroscopy. The hope was to find a means to differentiate saccharide isomers, D glycoside s in particular, through differences in the ir O H stretch modes. Initial experiments began with the setup of the OPO, which included the design and alignment of external transfer optics, U niblitz shutters, power meter, gas purge box etc. Once established, with the aid of Cesar Contreras, action spectra of four rubidium tagged D glycoside isomers were obtained and revealed differentiable bands. Upon comparison of the action spectra to vibrationally consistent computational structures, the assigned vibrational motions appeared to

PAGE 174

174 correspond to an extensive hydrogen bonding netw ork made more complex through rubidium inte ractions and ring structure fluct uations. This culminated in the determination that up to four conformers were responsible for the action spectra of the Rb + [ O methyl D glucoside ] complex. Accepting such a wide range of conformers for one isomer and the general difficulty in finding adequate matches between the IRMPD spectra and the DFT determined frequencies has led to the implementation of different strate gies. One inherent weakness in the past was the conformational search, wher e the applied semi empirical AM 1 force field did not work well with the rubidium tagged glucoside conformer s To correct for this shortcoming, a template of each geometrically opti mized isomer was created from which a rubidium cation was added to one of many predetermined zones. Using the Gausview 3.0 program, this process was repeated for different quadrants until the coverage was complete. The resulting geometries from the many ne wly created conformers were then saved as Cartesian coordinates where the Gaussian job files were created and executed. Another potential problem was finding an effective core potential that could properly model the unique interaction of the rubidium ion w ith isomeric states. Here the LANDLDZ basis was replaced with SDD ( Stuttgart/Dresden) ECP/ basis as the result of recent success with similar sugars. These calculations are currently in progress and hopefully will yield results that are more amicabl e. As a long term projec t, the expanded study of the D hexose ions is in order with a nod to the study of di and tri saccharide isomers of these components. Development of a library of sugar action spectra and schemes for differentiation can then be sta rted and added to with each successive experiment, perhaps resulting in the development of diagnostic soft ware. T he additional examination of structure through vibrational spectra comparison could expand into

PAGE 175

175 other arenas to help unlock sugar s complex rol e with proteins lipids and other biologically relevant molecules. After failed attempts at acquiring unimolecular dissociation rate constant information from black body infrared radiation dissociation (BIRD) procedures, a tunable Apollo CO 2 laser was rep aired and pressed into service. The repair and the laser set up (table, optics, windows wave meter etc.) were accomplished with the help of Lawerence Hartley and Cesar Contreras respectively. Later a Lasy tunable CO 2 laser was purchased and incorporated into the design. Both infrared lasers serve as the source of IR photons for a number of experiments. Of particular interest was the determination of the unimolecular kinetics and relative activa tion energies of positively (+2 ) and negatively ( 2) charged monophosphorylated peptide ions. Comparison of the slopes of the kinetic plots revealed that the positive ions dissociated at a much slo wer rate than the negative ions, a pattern that held up for all positive/negative ions recorded at different wavelengths Arrheni u s activation energies appeared to confirm that the higher dissociation yields of the negative ions were directly related to their lower activation energies. Kinetics determined at other wavelengths confirmed this result. The 95 % confidence limit s suggest that further expansion of the kinetic s for each wavelength would r esult in more points for the ir Arrhenius plot line and serve as a m eans to make the data clearer The interesting question is why this disparity exists in the first place and how can it best be modeled? Along these lines further research exploring the interactions of phosphopeptide ions with th eir surroundings is warranted. Taking a cue from studies of larger proteins, regions most likely to interact with phosphate group (domains) can be extracted and the relationship studied to later incorporate each local difference into the larger picture. Here additional exploration of the

PAGE 176

176 effects of the possible cis isomer should also be examined and kinetic modeling implemented to further bre ak down the constituents of the pseudo first order rate constants.

PAGE 177

177 L IST OF REFERENCES (1) Kunz, H Angew. Chem. Int. Ed. 2002 47 4439. (2) Ramberg, P. Chemical Structure, Spatial Arrangement: The Early History of Stereochemistry, 1874 1914 ; Ashgate Publishing : Vermont, 2003 (3) Jonsson, A. P.; Cell. Mol. Life Sci. 2001 58, 868 (4) Zala, J. Mass. Spectrom. Rev. 2004 23 ,161. (5) Benedict, S. R. J Biol. Chem 1925 64 207 (6) Benedict, S. R. J Biol. Chem 1926 68 759. (7) Benedict, S. R. J. Biol. Chem 1928 76 457. (8) Benedict, S. R. J. Biol. Chem 1931 92 141. (9) Tollens, C. Z. Physiol. Chem. 1908 56 1908. (10) Tollens, C. Z. Physiol. Chem. 1909 61 95. (11) Fischer, E Ber. Disch. Chem. Ges 1891 24 1836. (12) Fischer, E Ber. Disch. Chem. Ges 1891 24 2683. (13) Cahn, R.; I ngold, C.; Prelog, V. Angew. Chem. Int. Ed. Eng. 1966 5 385. (14) Davis, B. G. ; Fairbanks, A. J. Carbohydrate chemistry ; Oxford University: New York, 2002 (15) Lee, Y. C.; Lee, R. T. J. Chin. Chem. Soc 1999 46 283. (16) Rao, V. S. R. Conformation of Carbohydrate s ; Harwood Academic Publishers: Amsterdam, 1998. (17) Barrows, S. E.; Dulles, F. J.; Cramer, C. J.; French, A. D.; Truhlar, D. G. Carbohydrate Res 1995 276 219. (18) McNaught, A. D. Pure & Appl. Chem 1996 68 1919. (19) Shin, S. K.; Beauchamp, J. L. J. Am. Chem. Soc 1990 112 2057. (20) Stephenson, J.L.; Booth, M.M.; Shalosky, J.A.; Eyler, J.R.; Yost, R.A. J. Am. Soc. Mass Spectrom. 1994 5 886.

PAGE 178

178 (21) Kirkwood, D. A.; Stace, A.J. J. Am. Chem. Soc 1999 120 12316. (22) Polfer, N. C.; Valle, J. J.; Moore, D. T.; Oomens, J .; Eyler, J. R.; Bendiak, B. Anal. Chem 2006 78 670. (23) Nelson, D. L.; Cox, M.M. Lehninger Prin. of Biochem. Worth Publishers: New York, 2000 (24) Alberts, B.; Johnson, A.; Lewis, J.; Raff, M.; Roberts, K.; Walter, P. Molecular Biology of the Cell:Reference E dition ;Taylor & Francis Inc.:New York, 2007 (25) Kinter, M. Protein Sequencing and Identification Using Tandem Mass Spectrometry ; John Wiley & Sons:New York, 2000 (26) Creighton, T. E. Proteins: Structures and Molecular Geometry ; W. H. Freeman: New York, 1993. (27) Pauling, L.; Corey, R. B.; Branson, H. R. Proc. N. A. S. 1951 37 205. (28) Nesloney, C. L.; Kelley J.W.; Biorg. & Med. Chem 1996 4 739. (29) Neurath, H. J. Phys. Chem 1940, 44, 296. (30) Jahnke, W. Widmer, H.; Cell. Mol. Life. Sci 2004 61 580. (31) Dunbar, R. C. J. Phys. Chem 1994 98 8705. (32) Dunbar, R. C.; McMahon, T. B. Science 1998 279 194. (33) Paech, K.; Jockusch, R.A.; Williams, E.R.; J. Phys. Chem. A. 2002, 106 9761. (34) Flora, J.W.; Muddiman, D.C. J. Am. Soc. Mass Spectrom. 2004, 15 121. (35) Marshall, A.G. A cc. Chem. Res 1985, 18 316. (36) Comisarow M. B.; Marshall, A.G. J. Mass Spectrom. 1996 3 1 381. (37) Comisarow, M. B.; Marshall, A. G. Can. J. Chem. 1974 52 1997. (38) Comisarow, M. B.; Marshall, A. G. J. Chem. Phys. 1975 62 293. (39) Comisarow, M. B.; Marshall, A G. Chem. Phys. Lett 1974 25 282. (40) Lawrence, E. O.; Livingston, M. S. Pys. Rev 1932 40 19. (41) Sommer, H,; Thomas, H. A.; Hipple, J. A. Phys. Rev 1951 82, 697.

PAGE 179

179 (42) Wobschall, D. Rev. Sci. Instrum 1965 36 466. (43) Cooley, J. W.; Tukey, J. W. Am. Math. Soc 1965 19 297. (44) Marshall, A. G.; Hendrickson, C. L; Jackson, G. S. Mass Spectrom. Rev 1998 17 1 (45) Marshall, A. G.; Hendrickson, C. L. Int. J. Mass Spectrom 2002 215 59 (46) Jackson, G. S.; Hendrickson, C. L.; Reinhold, B. B.; Marshall, A. G. Int. J. Mass Spectrom. Ion Processes 1997 165/166 327. (47) Guan, S.; Marshall, A. G. Int. J. Mass Spectrom Ion Processes 1996 157/158 5. (48) Schweikhard, L.; Marshall, A. G. J. Am. Soc. Mass Spectrom 1993 4 433. (49) Jackson, G. S.; White, F. M.; Guan, S.; Marsha ll, A. G. J. Am. Soc. Mass Spectrom 1999 10 759. (50) Comisarow, M. B.; Marshall, A. G. J. Mass Spectrom 1996 31 581 (51) Caravatti, P.; Allemann, M. Org. Mass Spectrom. 1991, 26 514. (52) Jackson, G.; Canterbury, J. D.; Guan, S.; Marshall, A. G. J. Amer. Soc. Mass Spectrom. 1997 8 283 (53) Comisarow, M. B. J. Chem. Phys 1978 69 4097. (54) Limbach, P. A.; Grosshans, P. B.; Marshall, A. G. Anal. Chem 1993 65 135 (55) Field, F.H.; Franklin, J.L. Electron Impact Phenomena and the Properties of Gaseous Ions. Academic Press: New York, 1957. (56) Mrk, T.D. Int. J. Mass Spectrom. Ion Processes. 1982, 45 125. (57) Mrk, T.D. Electron Impact Ionization in Gaseous ion Chemistry and Mass Spectrometry Futrell, J.H., ed. John Wiley and Sons: New York, 1986. (58) Munson, M.S.B.; Field, F.H. J. Am. Chem. Soc. 1966, 88 2621. (59) Munson, M.S.B. Int. J. Mass Spectrom. 2000, 200 243. (60) Harrison, A.G. Chemical Ionization Mass Spectrometry, 2nd ed CRC Press: Boca Raton, 1992

PAGE 180

180 (61) Dunbar, R.C. Ion Photodissociation in Gas Phase Ion Chemistry 1st e d Bowers, M.T., ed. Academic Press: New York, 1979 Vol. 2. (62) Robb, D.B.; Covey, T.R.; Bruins, A.P. Anal. Chem. 2000, 72 3653. (63) Raffaeli, A.; Saba, A. Mass Spectrom. Rev. 2003, 22 318. (64) Juhasz, P.; Costello, C.E. Rapid Commun. Mass Spectrom. 1993, 7 343 (65) Karas, M.; Ehring, H.; Nordhoff, E.; Stahl, B.; Strupat, K.; Hillenkamp, F.; Grehl, M.; Krebs, B, Org. Mass Spectrom 1993 28 1476 (66) Pitt, J. J.; Gorman, J. J. Rapid Commun.Mass Spectrom. 1996 10 1786 (67) Tanaka, K.; Waki, H.; Ido, Y.; Akit a, S.; Yosh ida, Y.; Yoshida, T. Rap Commun Mass Spectrom 1988 2 151. (68) Karas, M.; Bachmann, D.; Hillenkamp, F. (1985). Anal. Chem. 1985 57 2935. (69) Karas M, ; Hillenkamp, F. Anal. Chem. 1988 60, 2299. (70) Mamyrin, B. A. Int. J. Mass Spectrom. Ion Process 1994 131 1. (71) Goss, J.H. Mass S pectrometry, 1st ed Springer: New York, 2004 (72) Glckmann, M.; Pfenninger, A.; Krger, R.; Thierolf, M.; Karas, M.; Horneffer, V.; Hillenkamp, F.; Strupat, K. Int. J. Mass Spectrom. 2001, 210/211 121. (73) Karas, M.; Glckmann, M.; Schfer, J. J. M ass Spectrom. 2000, 35 1. (74) Westmacott, G.; Ens, W.; Hillenkamp, F.; Dreisewerd, K.; Schrenberg, M. Int. J. Mass Spectrom. 2002, 221 67. (75) Harvey, D. J. Int. J. Mass Spectrom 2003 226 1 (76) Cole, R. B. J. Mass Spectrom 2000 35 763. (77) Kebarle, P.; Tang, L. Anal. Chem 1993 65 972A (78) Van Berkel, G. J.; Zhou, F. Anal. Chem 1995, 67 2916 (79) Kebarle, P.; Ho, Y. In Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation and Applications Cole, R.B., ed. John Wiley & Sons: New York, 1997 (80) Dole, M.; Mack, L.L.; Hines, R.L.; Mobley, R.C.; Ferguson, L.D.; Alice, M.B. J. Chem. Phys 1968, 49 2240.

PAGE 181

181 (81) Mack, L.L.; Kralik, P.; Rheude, A.; Dole, M.J. Chem. Phys 1970 52 4977. (82) Iribarne, J.V.; Thomson, B.A. J. Chem. Phys 1976 64 2287. (83) Thomson, B.A.; Iribarne, J.V. J. Chem. Phys 1979 71 4451. (84) Siuzdak, G. Mass Spectrometry for Biotechnology ; Academic Press: New York, 1996 (85) Lyman, J. L.; Galbraith, H. W.; Ackerhalt, J. R. Los Alamos Science 1982 3 66 (86) Gatz, C. R. Introduction to Quantum Che mistry Charles E. Merrill Publications: Ohio 1971 (87) Pauling, L.; Wilson, B. E. Introduction to Quantum Mechanics Dover Publications, Toronto, 1968 (88) Atkins, P. W.; Friedman, R. S. Molecular Quantum Mechanics Oxford University Press: New York, 2001 (89) Levine, I. N. Quantum Chemistry Prentice Hall: New Jersey, 2000. (90) Atkins, P.W. Physical Chemistry W.H. Freeman and Company: New York, 1998 (91) Herzberg, G. Molecular Spectra and Molecular Structure Spectra of Diatomic Molecules D. Van Nostrand Company Inc :New York, 1953 (92) Herzberg, G. Molecular Spectra and Molecular Structure Infrared and Raman Spectroscopy D. Van Nostrand Company Inc.:New York, 1951 (93) Messiah, A. Quantum Mechanics Vol. II John Wiley & Sons Inc.:New York, 1978 (94) Wilson E. B.; Decius, J.C.; Cross, C. C. Molecular Vibrations Dover Publications:Toronto, 1955 (95) Oomens, J.; Sartakov, B. G.; Meijer, G.; Helden G. V. Int. J Mass Spectrom 2006 254 1 (96) Dunbar, R. C. Int. J. of Mass. Spectrom 2000, 200 571. (97) Duncan, M. A. Int. J. of Mass. Spectrom. 2000, 200 545. (98) Bloembergen N.; Yabonovitch E. Physics Today 1978 May, 23. (99) Bagratashvili,V. N.; Letokhov, V. S.; Makarov, A. A.;Ryabov, E. A. Multiple Photon Infrared Laser Photophysics and Photochemistry Gordon Breach, New York, 1985

PAGE 182

182 (100) Reis ler, H.; Wittig, C. Ann.Rev.Phys.Chem. 1986 37 307. (101) Merchant, V. E.; Isenor, N. R. IEEE J. of Quan. Elect. 1976 12 603. (102) Grant, E. R.; Schulz, P. A.; Subdo, A. S.; Shen,Y. R.; Lee, Y. T. Phy. Rev, Le t t. 1978 40 115. (103) Gray, S. K.; Stine, J. R. Laser Chem. 1985, 5 209. (104) Mukamel, S.; Jortner, J. J. of Chem. Phys. 1974 60 4760. (105) Felker, P. M.; Zewail, A. H. Phys. Rev. Le t t 1984 52 301. (106) Dunbar, R. C. J. Am. Chem. Soc. 1971 93 4354. (107) Shin, S. K.; Beauchmp, J. L. J. Am. Chem. Soc. 1990, 112, 2057 (108) Shin, S. K.; Beauchmp, J. L. J. Am. Chem. Soc. 1990, 112, 2066. (109) Peiris, D. M.; Cheeseman, M. A.; Ramanathan, R.; Eyler, J. R. J. Phys. Chem. 1993 97 7839. (110) Stephenson, M. M.; Booth, J. A.; Shalosky, J. R.; Eyler, J. R. J. Am. Soc. Mass Spectrom. 1994 5 886. (111) Watson, C. H.; Zimmerman, J. A.; Bruce, J. E.; Eyler, J. R. J. Phys. Chem. 1991 95 6081. (112) Uechi, G. T.; Dunbar, R. C. J.Chem.Phys. 1990 93 1627. (113) Dunbar, R.C. J. Chem. Phys 1991 95 15. (114) Uechi, G. T.; Dunbar, R. C. J.Chem.Phys. 1993 98 7 887 (115) Dunbar, R.C. Mass Spec.Rev. 2004, 23 127 (116) Lin, C. Y.; Dunbar, R. C J. Phys. Chem 1996 100 19655. (117) Lin, C. Y.; Dunbar, R. C. ; Haynes, C. L.; Armentrout, P. B.; Tormer, D. S.; McMahon, T. B J. Phys. Chem 1996 100 19659. (118) Rodriguez Cruz, S. E. ; Jockusch, R. A.; Williams, E. R J. Am Chem. Soc. 1999a 121 8886. (119) Rodriguez Cruz, S. E.; Jockusch, R. A.; Williams, E. R J. Am Chem. Soc. 1999b 121 8898.

PAGE 183

183 (120) Price, W.D.; Williams, E.R. J. Phys. Chem. A. 1997 101 8844 (121) Tolman, R. C J. Am. Chem. So c 1920 42 2506. (122) Tolman, R. C. Statistical Mechanics and Applications to Physics and Chelistry Chemical Catalog:New York, 1927 (123) Dunbar, R. C. J. Phys. Chem 1994 98 2707. (124) Stevens, S. M.; Dunbar, R. C.; Price, W. D.; Sena, M; Watson, C. H.; Nichols L. S.; Riveros, J. M.; Richardson, D. E.; Eyler, J. R. J. Phys. Chem. A 2002 106 9686. (125) Price, W. D.; Schnier, P. D.; Willimas, E. R. J. Phys. Chem. B 1997 101 664. (126) Jockusch, R. A.; Williams, E. R. J. Phys. Chem 1998 102 4543. (127) Wendell, F. Unimo lecular Reactions: A Concise Introduction Cambridge Press:New York, 2003 (128) Robinson, P. J.; Holbrook, K. A. Unimolecular Reactions Wiley Interscience:New York, 1972. (129) Price, W.D.; Williams, E.R. J. Phys. Chem. A 1997 101 8844. (130) Price, W. D.; Schnier, P. D.; Jockusch, R. A.; Strittmatter, E. R.; Willimas, E. R. Anal. Chem. 1996 118, 10640. (131) Price, W. D.; Schnier, P. D.; Willimas, E. R. Anal. Chem 1996 68 859. (132) Schneir, P. D.; Price, W. D.; Jockusch,R. A.; Williams, E. R. J. Am. Chem. Soc ., 1996 118 7178. (133) Schnier, P. D.; Price, W. D.; Strittmatter, E. F.; Williams, E. R. J. Am. Soc. Mass Spectrom 1997 8 771. (134) Busman, M.; Rockwood, A. L.; Smith, R. D. J. Phys. Chem 1992 96 2397 2400. (135) Jockusch, R. A.; Schnier, P. D.; Price, W. D.; Strittmatter E. F.; Demirev, P. A., Williams, E. R. Anal. Chem 1997 69 1119 (136) Jockusch, R. A.; Paech, K.; Williams, E. R. J. Phys.Chem. A 2000 104 3188. (137) Flora, J. W.; Muddiman, D. C. J. Am. Chem. Soc. Com 2002 124 6646.

PAGE 184

184 (138) Freitas, M. A.; Hendrickson, C. L.; M arshall, A. G. Rap. Com. Mass Spectrom 1999 13 1639. (139) Freitas, M. A.; Hendrickson, C. L.; Marshall, A. G. J. Am. Chem. Soc. 2000 122 7768. (140) Hannis, J. C.; Muddiman, D. C. J. Mass Spectrom 2002 219 139. (141) Patel, C. K. N. Phys. Rev 1964 136, A1187. (142) Javan, A.; Bennett, W. R.; Heriott, D. R. Phys. Rev. Le t t 1961 6 106. (143) Bennett, W. R.; Faust, R. A.; McFarlane, R. A.; Patel, C. K. N. Phy. Rev. Le t t 1962 8 470 (144) Patel, C. K. N.; Bennett, W. R.; Faust, W. L.; McFarlariane, R. A. Phys. Rev. Le t t 196 2 9 102. (145) Adel, A.; Barker, E. F. Phys. Rev 1933 44 185. (146) Adel, A Astrophys. J. 1941 94 375. (147) Patel, C. K. N. Phys. Rev. Le t t 1964 13 618. (148) Franken, P. A.; Hill, A. E ; Peters, C. W.; Weinreich, G, Phys. Rev. Le t t 1961 7 118. (149) Kingston, R. H. Proc. IRE 1962 50 472. (150) Moeller, G.; Rigden, J. D. Appl. Phys. Let. 1965 7 274. (151) Witteman, W. J. IEEJ. Quan. Elect 1969 QE 2 375. (152) Kroll, N.M. Phys. Rev. 1962 127 1507. (153) Akhmanov, S. A.; Khokhlov, R. V. J. Expt. Theor. Phys. 1963 16 252. (154) W ang, C. C.; Ravette, Appl. Phys. Rev 1965 6 169. (155) Giordmaine, J. A.; Miller, R. C. Phys Rev. Le t t 1965 14 169. (156) Woodbury, E. J.; Ng, W. K. Proc. IRE 1962 50 2367. (157) Eckhardt, G.; Hellwarth, R. W.; McClung, F. J.; Schwarz, S. E. ; Weiner, D.; Woodbur y, E. J. Phys. Rev. Le t t 1962 9 455. (158) Hellwarth, R. W. Phy. Rev. 1963 130 1850.

PAGE 185

185 (159) Maker, P. D.; Tekhune, R. W. Phys. Rev 1965 137 A801. (160) Begley, R. F.; Harvey, A. B.; Byer, R. L. Appl. Phys. Le t t 1974 25 387. (161) Tolles, W. M.; Nibler, J. W.; McDon ald, J. R.; Harvey, A. B. Appl. Spec. 1977 31 253. (162) Jones, W. J.; Stoicheff, B. P. Phys. Rev. Le t t 1964 13 657. (163) Gorner, H.; Maier, W.; Kaiser, W. J Ram. Spec 1974 2 363. (164) Pound R. V. Rev. Sci. Instrum 1946 17 490 (165) Drever, R. W. P.;Hall, J. L. ;Kowalski, F. V.;Hough, J.;Ford, G. M.;Munley, A. J.;Ward, H Appl. Phy. B: Las. a nd Opt 1983 31 97. (166) Mueller, F. OS 4000 Operating Manual LINOS Photonics:Munch, 2007. (167) Mueller, F. OPOtalk_customer LINOS Photonics: Munich 2007 (168) Harris, S. E. Proc. o f the IEEE 1969 57 2096. (169) Eason, R. W.; Miller, A. Signal Processing London:Chapman & Hall, 1993 (170) Boyd, R. W. Nonlinear Optics London:Elsevier, 2008 (171) Mollow, B. R.; Glauber, R. J. Phys. Rev. 1967 160 1076. (172) Mollow, B. R.; Glauber, R. J. Phys. Rev. 1967 160 1097. (173) He, G. S.; Song, H. L. Physics of Nonlinear Optics Singapore, World Scientific Pub. Co., 1999. (174) Hoden, M V.; Warner, J. Phys. Le t t. 1966 22 243. (175) Bergman, J. C.; Ashkin, A.; Ballman, A. A.; Dziedzic, J. M.; Levinstein, H. J.; Smith, R. G. Appl. Phys. Le t t 1968 12 92. (176) Myers, L. E.; Erkhardt, R. C.; Fejer M. M. ; Byer, R. L.; Bosenberg, W. R.; Pierce, J. W. J. Opt. Soc. Am ., 1995 12 2102. (177) Biaggio, I.; Kerkoc, P.; Wu, L. S.; Gunter, P.; Zysset, B. J, Opt. Soc. Am 1992 9 507. (178) Smith, R. G.; Geusic, J. E.; Levinstein, H. J.; Rubin, J. J.; Singh,S. ; Van Uitert, L. G. Phys. Rev. 1966 146 187. (179) Baumgartner, R. A.; Byer, R. L IEEE J. of Quan. Elec. 1979 QE 15 432.

PAGE 186

186 (180) Schiller, S.; Schneider, K.; Mlynek, J. J. Opt. Soc. Am 1999 1 6 1512. (181) Farzad, M. H. Pramna, J. of Phys 2007 69 395. (182) Sanborn, J. Z.; Hellings, C.; Donnelly, T. D. J. Opt. Soc. Am. B 2003 20 152. (183) Lui, H.; Kung, A. H. J. Opt. Soc. Am 2008 16 9714. (184) Pochi, Y.; Yariv, A.; Hong, C. J. Opt. Soc. Am. 1977 67 423 (185) Yariv, A; Pochi, Y. J. Opt. Soc. Am. 1977 67 438. (186) Bloembergen, N.; Sievers, A. J. Appl. Phys. Le t t 1970 17 483. (187) Houe, M.; Townsend, P. D. J. Phys.D: Appl. Phys 1995 28 1747. (188) Fiore, A.; Berger, V.; Rosencher, E.; Bravetti, P.; Nagle, J. Let ters to Nat ure 1998 391 463. (189) Fejer, M. M.; Magel, G. A.; Jundt, D. H.; Byer, R. L. IEEE J. Quant. Elec. 1992 25 2631. (190) Laurell, F. Opt. Mat 1999 11 235. (191) Black, E. D. Am. J. Phys 2001 69 79. (192) Black E. Notes on Pound Drever Hall Technique Pasa dena, 1998 (193) Schalley, C. A. Mass Spectrom. Rev. 2001 20 253. (194) Harvey, D. J. Mass Spectrom. Rev. 1999 18 349. (195) Zala, J. Mass Spectrom. Rev. 2005 23 161. (196) Park, Y.; Lebrilla, C. B. Mass Spectrom. Rev. 2005 24 232. (197) Stobiecki, M. Phytochemistry 2000 54 237. (198) Zhou, Z.; Ogden, S.; Leary, J. A. J. Org. Chem. 1990 55 5444. (199) Hofmeister, E. H.; Zhou, Z.; Leary, J. A. J am. Chem Soc 1991 113 5964. (200) Cancilla, M.; Penn, S.; Carroll, J.; Lebrilla, C. J. Am. Chem. Soc 1996 118 6736. (201) Cancilla, M.; Wong A.; Voss, L.; Lebrilla, C. Anal. Chem 1999 71 3206.

PAGE 187

187 (202) Garozzo, D.; Impallomeni, G.; Montaudo, G.; Spina, E. Rapid Commun. Mass Spectrom 1992 6, 550 (203) Asam, M. R.; Glish, G. L. J. Am. Soc. Mass Spectrom. 1997 8 987. (204) Kim, M. S.; McLafferty, F. W. J Am. Chem. Soc. 1978 100 327. (205) MacAleese, L.; Maitre, P. Mass Spectrom. Rev. 2007 26 583. (206) Eyler, J. R. Mass Spectrom. Rev. 2008 28 448 (207) Polfer, N. C.; Oomans, J. Mass Spectrom. Rev. 2008 28 468. (208) Bagratashvili, V. N.; Letokov, V. S.; Makarov, A. A.; Ryabov, E. A. Multiple Photon Infrared Laser Photophysics and Photochemistry Harwood: Churchester, 1985 (209) Okumura, M.; Yeh, L. I.; Lee, Y. T. J. Chem. Phys. 1985 83 3705. (210) Okumura, M.; Yeh, L. I.; Myers, J. D.; Lee, Y. T. J. Chem. Phys. 1986 85 2 326. (211) Cancilla, M. T.; Penn, S. G.; Carroll, J. A.; Lebrilla, C. B. J. Am. Chem. Soc. 1996 118 6736. (212) Bothner, B.; Carmichael, L.; Staniszewski, K.; Sonderegger, M.; Siuzdak, G. Spectroscopy 2002 16 71 (213) Ngoka, L. C.; Gal, J.F.; Lebrilla, C. B. Anal. Chem. 1994 66 692. (214) Oepts, D.; van der Meer, A. F. G.; van Amersfoort, P. W. Infrared Phys. Tech. 1995 36 297. (215) Valle, J. J.; Eyler, J. R.; Oomens, J.; Moore, D. T.; van der Meer, A. F. G.; von Helden, G.; Meijer, G.; Hendrickson, C. L.; Marshall, A. G .; Blakney, G. T. Rev. Sci. Instr. 2005 76 023103/1. (216) Valle, J. J. Differentiation of Carbohydrate Stereoisomers by Infrared Multiple Photon Dissociation using a Free Electron Laser and a Fourier Transform Ion Cyclotron Resonance Mass Spectrometer Ph.D. Dissertation, University of Florida, Gainesville, Florida 2005 (217) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J., J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J.

PAGE 188

188 E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Kom aromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; D.02 ed.; Gaussian, Inc.: Wallingford CT, 2004 (218) Cl ark, T.; Chandrasekhar, J.; Spitznagel, G.; Schleyer, P. J. Comp. Chem. 1983 4 294. (219) Rassolov, V.; Pople, J.; Ratner, M.; Windus, T. J. Chem. Phys. 1998 109 1223. (220) Momany, F. A.; Appell, M.; Willett, J. L.; Bosma, W. B. Carbohydr. Res. 2005 340 1638 (221) Momany, F. A.; Appell, M.; Willett, J. L.; Schnupf, U. B.; Bosma, W. B. Carbohydr. Res. 2006 341 525. (222) HyperChem(TM) Professional; 7.51 ed.; Hypercube, Inc. (223) Zuccarello, F.; Buemi, G. Carbohydr. Res. 1995 273 129 (224) Schnupf, U.; Willett, J. L.; Bosma W. B.; Momany, F. A. Carbohydr. Res. 2007 342, 196 (225) Contreras, C. S. Carbohydrates and Amino Acids: Infrared Multiple Photon Dissociation Spectroscopy and Density Functional Theory Calculations Ph.D. Dissertation, University of Florida, Gainesville, F lorida 2008 (226) Hemmingsen, L.; Madsen, D. E.; Esbensen, A. L.; Olsen, L.; Engelsen, S. B. Carbohydr. Res., Molecular Modeling of Carbohydrates 2004 339 937. (227) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M. ; Spellmeyer D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995 117 5179. (228) Russo, T.; Martin, R.; Hay, P. J. Phys. Chem. 1995 99 17085. (229) Wadt, W.; Hay, P. J. Chem. Phys 1985 82 270. (230) Russell, D.; Johnson, I. NIST Computational Chemistry Comp arison and Benchmark Database; NIST Standard Reference Database Number 101, 2006

PAGE 189

189 (231) French, A.; Brady, J. Computer M odeling of Carbohydrate M olecules American Chemical Society: Washington, DC, 1990 (232) Momany, F. A.; Appell, M.; Strati, G.; Willett, J. L. Ca rbohydr. Res. 2004 339 553 (233) Appell, M.; Strati, G.; Willett, J. L.; Momany, F. A. Carbohydr. Res. 2004 339 537. (234) Appell, M.; Willett, J. L.; Momany, F. A. Carbohydr. Res. 2005 340 459. (235) Aebersold, R.; Mann, M Nature 2003 422 188. (236) Aebrsold, R.; Goodlett. D. R. Chem. Rev 2001 101 269. (237) Sun, H.; Tonks, N. K. Trends Biol. Sci. 1994, 19 480 (238) Shi, S. D.H.; Hemling, M. E.; Carr, S. A.; Horn, D. M.; Lindh, I.; McLafferty, F.W. Anal. Chem. 2001 73 19 22. (239) Hunter T. Cell 1995 80 225. (240) Selfton, B.N.; Hunter, T. Protein Phosphorylation Academic Press:San Diego, 1998. (241) Flora, J. W.; Muddiman, D. C. J. Am. Mass. Spectrom 2004 15 121. (242) Little, D. P.; Speir, J. P.; Senko, M. W.; Oconnor, P. B.; McLafferty, F. W. Anal. Chem. 1994 66 2809 2815. (243) F lora, J. W.; Muddiman, D. C. Anal. Chem. 2001 73 13305. (244) Carr, S. A.; Huddleston, M. J.; Annan, R. S. Anal. Biochem. 1996 239 180. (245) Cooper, H. J.; Hakansson, K.; Marshall, A. G. Mass Spectrom. Rev 2005 24 201. (246) Green M. K.; Lebrilla, C. B. Mass Spec trom Rev. 1997 16 53. (247) Creese, A. J.; Cooper, H. J. J. Am. Soc. Mass. Spectrom 2008 19 1263. (248) DeGnore, J. P.; Qin, J. J. Am. Soc. Mass Spectrom. 1998, 9, 1175. (249) Carr, S. A.; Huddleston, M. J.; Annan, R .S. Anal. Biochem. 1996 239 180. (250) Huddleston, M J.; Annan, R. S.; Bean, M. F.; Carr, S. A. J. Am. Soc. Mass Spectrom 1993 4 710. (251) Affolter, M.; Watts, J. D.; Krebs, D. L. Anal. Biochem. 1994, 223, 74.

PAGE 190

190 (252) Tonner, D. S.; Thlmann, D.; McMahon, T. B. Chem. Phys. Lett. 1995 233 324. (253) Senko, M. W.; Canterbury, J. D.; Guan, S. H.; Marshall, A. G. Rapid Commun. Mass Spectrom. 1996 10 1839. (254) McMahon, T. B.; Tholmann, D.; Tonner, D. S.; Salahub, D. R.; Wei, D. J. Am. Chem. Soc. 1995 117 12819. (255) Degnore, J. P.; Qin, J. Am. Soc. Ma Sppect rom 1998 98 1075. (256) Ger l t, J. A.; Gassman, P. G. J. Am Chem, Soc 1992 114 5928. (257) Guo, Y. T.; Li, Y. M.; Zhu, Z. T.; Zhao, Y. F. Int. J Pept. Res. Ther. 2005 2 159.

PAGE 191

191 BIOGRAPHICAL SKETCH Wright (Lee) Pearson III was born in Coral Gables, Flor ida to Wright L. Pearson, Jr. and Betty Pearson The oldest of two boys, he grew up primarily in Miami, Florida with his family After graduating from Miami Lakes High School, he attended Miami Christian College and was th roughout his college career. Lee graduated from Miami Christian College with dual degree s in psychology and theology. Following his schooling, he gained considerable work experience in a variety of industries In June of 1995 Lee resolved to co ntinue his education by pursuing a Doctorate of Philosophy in p hysical c hemistry. He briefly attended Miami Dade College, where he earned an award for Chemistry Student of the Year in 1999. Following his acceptance to the University of Florida, he continue d to earn other awards the Hypercube Scholar award in 2000 for post baccalaureate research in computational chemistry and Chemistry Teaching awards in 2002 and 2004. Upon completion of his Ph.D. program in the summer of 2009, Lee pursue d postdoctoral resea rch in spectroscopy