Citation
Exploring Desorption/Ionization on Porous Silicon Mass Spectrometry and Its Applications

Material Information

Title:
Exploring Desorption/Ionization on Porous Silicon Mass Spectrometry and Its Applications
Creator:
LI, QIAN ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Etching ( jstor )
Ionization ( jstor )
Ions ( jstor )
Isomers ( jstor )
Mass spectra ( jstor )
Mass spectroscopy ( jstor )
Molecules ( jstor )
Porphyrins ( jstor )
Signals ( jstor )
Silicon ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Qian Li. 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:
2/28/2006
Resource Identifier:
436098799 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

EXPLORING DESORPTION/IONIZATI ON ON POROUS SILICON MASS SPECTROMETRY AND ITS APPLICATIONS By QIAN LI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Qian Li

PAGE 3

This document is dedicated to my parents, for their love and support.

PAGE 4

ACKNOWLEDGMENTS I would like to extend my thanks and appreciation to those who have contributed to my achievement, for their support, encouragement and guidance. First, I would like to express my deepest gratitude to my parents, for letting me travel to the United States when they needed a daughter’s company the most. They have been supportive of every decision I have made, and have been proud of me for the smallest success I have accomplished. My interest in chemistry was inspired by my high school chemistry teacher, who was a serious woman with impeccable logic, and believed that “the best way of learning is through your own hand.” It remains the guideline for my 12-year study in chemistry. I am also indebted to my previous school, Nanjing University. The Chemistry Department there provided one of the best programs and was equipped with excellent lecturers. We received rigorous trainings in math, physics, as well as different divisions of chemistry. “You might not remember all you are taught, but you will at least know that is knowledge exists, and know where to look for it,” one of the lecturers said. In Spring 2000, I started my study in the Chemistry Department, University of Florida. I would like to thank Dr. Winefordner for having accepted me into his group (JDW group), for being understanding and supportive all the time. I will miss the “summer memos” which I usually can not wait to share with friends. Dr. Nico Omenetto is a precious presence to my graduate student life. I am grateful for the strength and courage brought out by his words. I also want to thank Dr. Ben Smith, who has this iv

PAGE 5

ability to explain the most complicated things in the simplest way, and to incorporate the passion for science into the daily life. Most of my research was done in the mass spectrometry service lab, where I received advice from Dr. Powell. He provided the guidance and expertise for this dissertation, and showed me the wonderful world of mass spectrometry. I thank him for being patient with the questions I had, teaching me little by little each day, the mass spectrometry and much more. Dr. Harrison’s scientific writing class, as well as the questions he asked on our research group meetings, have benefited me greatly. There is nothing more encouraging than the fact that your research constantly draws people’s interest. Also, I would not forget my fellow graduate students with whom I have shared my journey. The “orphans” from the mass spectrometry service lab, and the past and previous members of JDW/Harrison/Omenetto group. The discussions over language, religion, culture and life, in addition to science, made my stay at UF more colorful , enjoyable and rewarding. I have had pleasant collaborations with Dr. Alonso Ricardo from the Benner group and Dr. Hubert Gill from the Scott group. The fruitful scientific conversations contributed to the success of our projects. Finally, I would like to thank my husband, Zehui Cao, for being my closest companion for the past six years. It is his support and tolerance that made the completion of this dissertation possible. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES .............................................................................................................ix LIST OF FIGURES .............................................................................................................x ABSTRACT .....................................................................................................................xiii CHAPTER 1 INTRODUCTION............................................................................................................1 Historical Overview and Background..........................................................................1 Early Years of Mass Spectrometry........................................................................1 Ionization Techniques...........................................................................................2 Time-of-Flight Mass Analyzer.....................................................................................5 Background and Principle.....................................................................................5 Delayed-Extraction................................................................................................8 Reflectron Kinetic Energy Analyzers....................................................................9 Laser Desorption/Ionization Techniques....................................................................10 Laser Desorption/Ionization................................................................................10 Matrix-Assisted Laser Desorption/Ionization.....................................................11 Porous Silicon and Its General Applications..............................................................14 Porous Silicon Fabrication..................................................................................14 Properties of Porous Silicon................................................................................17 General Applications of Porous Silicon..............................................................19 Desorption/Ionization on Porous Silicon Mass Spectrometry (DIOS-MS)................21 Principles.............................................................................................................21 Characteristics.....................................................................................................22 Mechanism..........................................................................................................23 Applications and Development of DIOS-MS......................................................25 2 FUNDAMENTALS AND APPLICABILITIES OF DIOS-MS.....................................30 Introduction.................................................................................................................30 Production of PSi........................................................................................................30 Anodic Etching....................................................................................................31 Metal-Assisted Etching.......................................................................................32 vi

PAGE 7

Comparison of Two Methods..............................................................................33 Surface Evaluation of PSi...........................................................................................35 FT-IR...................................................................................................................35 Auger Electron Spectroscopy..............................................................................36 SEM.....................................................................................................................37 Instrumentation and Sample Preparation in DIOS-MS Analysis...............................41 Instrumentation....................................................................................................41 Laser....................................................................................................................41 Mass Analyzer.....................................................................................................42 Target Plate..........................................................................................................42 Vacuum System...................................................................................................43 Sample Preparation..............................................................................................43 SEM Evaluation of PSi with Sample...................................................................44 Applicability of DIOS-MS.........................................................................................46 Individual Analytes.............................................................................................46 Mixtures...............................................................................................................48 Polymer Analysis.................................................................................................51 Reflectron Mode..................................................................................................52 Negative Mode....................................................................................................53 Limitations of DIOS-MS............................................................................................57 Background and Contaminants............................................................................57 Reduced Response to Analytes with Certain Structures.....................................59 Reduced Response to Peptides............................................................................60 Aging and Re-etching.................................................................................................63 Concept of Aging................................................................................................63 DIOS-MS on Aged and Re-etched PSi...............................................................64 SEM of Aged and Re-etched Surfaces................................................................66 Contact Angle Measurements..............................................................................67 Effect of Acid Content on DIOS-MS Signal..............................................................69 Conclusions.................................................................................................................71 3 DIOS-MS STUDIES ON PENTOSE-BORATE COMPLEXES...................................74 Introduction.................................................................................................................74 Experimental Methods................................................................................................75 Chemicals and Reagents......................................................................................75 Detection of [Pentose+K] + Adduct Ions..............................................................76 Detection of Pentose-Borate Bomplexes.............................................................76 Competition Experiments between 1,4-Anhydroerythritol and Each Pentose Isomers.............................................................................................................76 Competition Experiments between 13 C 5 -Ribose and Each Pentose Isomers.......76 Results and Discussion...............................................................................................77 Detection of [Pentose+K] + Adduct Ions..............................................................77 Detection of Pentose-Borate Complexes.............................................................80 Competition Experiments 1,4-Anhydroerythritol and Pentose Isomers..............84 Competition Experiments 13 C 5 -Ribose and Pentose Isomers..............................87 Conclusions.................................................................................................................90 vii

PAGE 8

4 DIOS-MS STUDY OF PORPHYRIN/PORPHODIMETHENES AND ITS IMPLICATIONS........................................................................................................92 Introduction.................................................................................................................92 Porphyrins and Porphodimethenes......................................................................92 Mass Spectrometry Analysis of Porphyrins and Their Derivatives....................93 DIOS-MS of Porphyrins and Porphodimethenes................................................96 Experimental Methods................................................................................................96 Sample Preparation and Mass Calibration..........................................................96 Results and Discussion...............................................................................................97 Free Porphodimethenes Ligands and Their Pd Complexes.................................97 Porphyrins with Different Chelating Metals.....................................................102 Effect of Laser Power........................................................................................106 Porphyrins with Similar Masses........................................................................108 Conclusions...............................................................................................................109 5 CONCLUDING REMARKS........................................................................................111 LIST OF REFERENCES.................................................................................................116 BIOGRAPHICAL SKETCH...........................................................................................124 viii

PAGE 9

LIST OF TABLES Table page 2-1. Molecular weights and structures of three tested molecules......................................46 2-2. Components of four mixtures.....................................................................................48 3-1. Volumes of the benzoic acid and KOH solution (both at 10 mM) used for constructing the mixtures, and the resulting concentration of related species.........82 4-1. Structures of four porphodimethene free ligands.......................................................97 4-2. Representative structure of four porphodimethene Pd complexes.............................99 4-3. DIOS-MS results of four free ligands and four Pd complexes.................................100 4-4. Structures and DIOS-MS results of metalloporphyrins with different ligands and chelating metal.......................................................................................................102 ix

PAGE 10

LIST OF FIGURES Figure page 1-1. Scheme of a linear TOF mass analyzer operating in the positive mode.......................6 1-2. Delayed-extraction in a linear TOF instrument at the positive detection mode...........9 1-3. Scheme of a TOF mass spectrometer equipped with reflectrons...............................10 1-4. Schematic of a MALDI process.................................................................................12 1-5. Structure of porous silicon..........................................................................................18 1-6. Room temperature photoluminescence spectra of crystalline silicon and PSi...........19 1-7. SEM images of microreactors....................................................................................21 1-8. Schematic of a DIOS-MS process..............................................................................22 2-1. Instrument set-up for anodic electrochemical etching................................................31 2-2. Images of crystalline silicon and porous silicon.........................................................33 2-3. SEM images of PSi prepared by metal-assisted etching............................................33 2-4. FT-IR spectrum of PSi................................................................................................36 2-5. Auger electron spectrum of PSi..................................................................................37 2-6. SEM images of two types of PSi morphology...........................................................38 2-7. Three regions of PSi spot fabricated with metal-assisted etching..............................40 2-8. SEM images of PSi with incompletely etched areas..................................................41 2-9. Modification of a MALDI plate into a DIOS plate....................................................43 2-10. Photograph of a PSi chip..........................................................................................44 2-11. SEM images of PSi surfaces with or without samples.............................................45 2-12. DIOS mass spectra of three individual analytes.......................................................47 x

PAGE 11

2-13. DIOS mass spectra of four mixtures.........................................................................49 2-14. DIOS mass of a polymer sample..............................................................................51 2-15. DIOS mass spectra of HP mix in the reflectron mode.............................................52 2-16. Negative DIOS mass spectra of three samples.........................................................54 2-17. DIOS mass spectra of two compounds in both the positive mode and negative mode.........................................................................................................................56 2-18. DIOS mass spectrum of angiotensin II in the presence of keratins..........................58 2-19. MALDI and DIOS mass spectra of caffeine and cortisol.........................................59 2-20. MALDI and DIOS mass spectra of a peptide mixture.............................................61 2-21. DIOS mass spectra of bradykinin and angiotensin II of different amount...............62 2-22. DIOS mass spectra of angiotensin II, bradykinin, and a peptide mixture................65 2-23. SEM images of aged and re-etched PSi surfaces.....................................................67 2-24. Water contact angles on four different surfaces.......................................................69 2-25. Effect of TFA concentration on DIOS-MS signal....................................................71 3-1. A chart of DIOS-MS signal of [pentose + K] + ...........................................................77 3-2. DIOS mass spectra of ribose with different concentration of K + ...............................78 3-3. DIOS mass spectra of five pentose isomers in the presence of K + .............................79 3-4. Formation of the pentose-borate complex..................................................................80 3-5. DIOS mass spectra of ribose/Na 2 B 4 O 7 mixtures........................................................81 3-6. Negative DIOS-MS of benzoic acid/KOH mixtures .................................................83 3-7. Competition reaction between pentose isomers and 1, 4-anhydroerythritol in the presence of borate ion..............................................................................................84 3-8. DIOS mass spectra of mixtures of pentose isomers and 1, 4-anhydroerythritol in the presence of borate ion..............................................................................................85 3-9. Results of four sets of competition experiments for four aldopentose isomers..........86 3-10. Competition reaction between pentose isomers and 13 C 5 -labeled ribose in the presence of borate ion..............................................................................................87 xi

PAGE 12

3-11. DIOS mass spectra of the mixtures of pentose isomers and 13 C 5 -ribose in the presence of borate ion..............................................................................................88 3-12. Results of competition experiments for four sets of experiments for 13 C 5 -ribose with pentose isomers........................................................................................................89 4-1. Basic structure of a) porphyrins and b) porphodimethenes........................................92 4-2. DIOS mass spectra of a) 4-1 , b) 4-2, c) 4-3, and d) 4-4............................................98 4-3. Experimental and theoretical (inserted, [M+H] + ) isotopic distribution of 4-1...........99 4-4. Two sets of DIOS mass spectra of a) 4-5 and b) 4-7................................................101 4-5. DIOS mass spectra of a) 4-9 and b) 4-10.................................................................104 4-6. DIOS mass spectra of a) 4-9, b) 4-11 and c) their theoretical isotopic distribution of corresponding [M] + ................................................................................................106 4-7. DIOS mass spectra of a) 4-13, b) 4-15 and c) the theoretical isotopic distribution of corresponding [M] + ................................................................................................106 4-8. Total ion intensity of [M] + and [M+H] + , and ion intensity ratio of [M] + /[M+H] + at different laser attenuation of compounds a) 4-11 and b) 4-13...............................107 4-9. The formation and DIOS mass spectrum of 4-20.....................................................109 xii

PAGE 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 EXPLORING DESORPTION/IONIZATION ON POROUS SILICON MASS SPECTROMETRYAND ITS APPLICATIONS By Qian Li August, 2005 Chair: James D. Winefordner Major Department: Chemistry Desorption/ionization on porous silicon mass spectrometry (DIOS-MS) is a surface-enhanced laser desorption/ionization technique featured with low chemical background, where no additional matrix compounds are needed. A wide range of analytes have been investigated using DIOS-MS, and the technique has found its applications in areas such as enzymatic reaction monitoring, forensics, and protein characterization. In this work, two porous silicon (PSi) manufacturing procedures, anodic etching and metal-assisted etching, are carried out. The latter is chosen to be the routine method due to improve reproducibility and efficiency, as well as less contact with hazardous hydrofluoric acid. The fundamental performance of DIOS-MS is evaluated on such processed PSi. The advantages of DIOS-MS, including low chemical background and easy sample preparation, are demonstrated with a variety of analytes. The limitations of the technique are also discussed. Background ions are detected with lower sample xiii

PAGE 14

amount and/or aged PSi surface. When compared to MALDI, DIOS-MS has reduced response to peptides and compounds with certain structures. The first application of DIOS-MS is to determine the binding preference of the borate ion to four pentose isomers. An important question in the evolution of ribonucleic acids (RNAs) is why ribose is favored over other pentose isomers. Boron is proven to play important roles in pentose synthesis and stabilization, and it may also account for the fact that ribose is the favored building block in RNA synthesis. Two competition schemes are designed, and binding preference is determined to be ribose>lyxose>arabinose> xylose. This work illustrates the potential of DIOS-MS in the analyses of non-volatile, small molecules in delicate chemical equilibria. The possible dramatic change of pH associated with the introduction of a MALDI matrix, which may disturb the equilibria of interest, is avoided. DIOS-MS is also applied for the analyses of porphyrins, porphodimethenes and their metal complexes. Molecular species are detected with DIOS-MS while other available MS techniques failed to provide consistent molecular weight information. For certain compounds, both [M] +• and [M+H] + are observed, and the internal mass calibration for accurate m/z measurements is essential. The effects of laser power and oxidation potentials of the analytes on the formation of [M] +• are discussed. xiv

PAGE 15

CHAPTER 1 INTRODUCTION Desorption/ionization on porous silicon mass spectrometry (DIOS-MS) is one of the newly appeared topics in the area of mass spectrometry. To better understand the root and the emergence of DIOS-MS, the related history of mass spectrometry is reviewed in brief. Historical Overview and Background Early Years of Mass Spectrometry Mass spectrometry is a technique to obtain the molecular weight and related information on compounds through measuring the mass-to-charge value of the corresponding ions. The technique has a rich tradition of discovery and development from the early effort of physicists to the recent endeavors of scientists of a wide range of areas. It can be traced back to 1803, when John Dalton introduced the atomic theory to explain the properties of matter. 1 About a century later, J. J. Thompson obtained ions from a gas discharge and created first mass spectrum. It was no more than a series of parabolic lines, then termed “positive rays”, recorded on a photographic plate; yet, it symbolized the start of a great field of science. 2 The first application of mass spectrometry is often recognized as the discovery of neon isotopes by F. W. Aston, a student of J. J. Thompson and a Nobel Prize laureate in physics. He gave an early definition of the mass spectrometer as: “an apparatus in which the focused beam of rays is brought up to a fixed slit, and there detected and measured electrically.” In addition, he also commented that “mass spectrograph is best restricted to 1

PAGE 16

2 those forms of apparatus capable of producing a focused mass spectrum of lines on a photographic plate.” 3 With the advance of modern mass spectrometry, those statements might not hold the entire truth, but they certainly served as a guideline for many scientists early on. As a result of the research, exact masses of the elements and relative abundances of the isotopes were precisely measured, leading to the establishment of a standard scale for atomic weights by International Union for Pure and Applied Chemistry (IUPAC) in late 1950s. 4 If the first few decades of mass spectrometry saw the major contributions from physical chemists, the scenario has changed since. The flourish of the petroleum industry offered the technique a new stage of application after World War II. In the 1960s, coupling gas chromatographs to mass spectrometry further extended the scope of the technique. 5 In the past 20 years, mass spectrometry has experienced the fastest development with a large scope of exciting innovations. Currently, a variety of mass spectrometers, ionization techniques, as well as options for coupling to other analytical devices are available to researchers of many disciplines. Ionization Techniques A mass spectrometer is often divided into three components: an ion source, an analyzer and a detector. All three have witnessed their own course of development, often associated with the advances in physics, electronic engineering, and computer science. At present, there are five major types of mass analyzers: magnetic sector, quadrupole, ion trap, time-of-flight, and Fourier-transform ion cyclotron resonance mass analyzer. Although operating under different principles, mass analyzers all share a common aspect that an electric or magnetic field is required to manipulate ions to achieve ion

PAGE 17

3 separation and detection. Being the main device for the current study, the time-of-flight mass analyzer will be discussed in more detail later. As for detectors, commonly used are electron multipliers and array detectors. They experienced relative little change over the years; however, their performance is crucial in accomplishing the desired response time, sensitivity, and accuracy. Another important component of a mass spectrometer is the ion source, which reflects the specific ionization technique used in a mass spectrometry analysis. As DIOS is considered a novel ionization technique, a brief evolution of this subject in mass spectrometry is described here. The ion source converts neutral molecules to ions. It has been of great interest to mass spectrometrists to design and improve the ways of making ions, necessitated by distinct properties of different analytes. The first ion source, as mentioned earlier, was a low pressure gas discharge devised by J. J. Thompson. Later, F. W. Aston, an experienced glass blower, improved the design of discharge tubes from cylindrical-shaped into spherical shaped, and a milliampere current was obtained at 40 kV for an “indefinite period”. 6 Using this ion source, coupled with a double-focusing mass analyzer, a resolving power of 2,000 was achieved. As a result, minute differences were eventually noticed between the actual masses of the elements and the whole number multiples of the mass of hydrogen. Also pioneered in the ionization techniques was the electron ionization (EI) ion source, first introduced by A. J. Dempster in 1918 7 and then improved by W. Bleakney in 1929. 8 For EI, gaseous phase neutral molecules collide with a guided electron beam at ~70 eV to form positively charged radical molecular ions, [M] +• .

PAGE 18

4 Closely related to EI is the chemical ionization (CI) ion source. In CI, an electron source is still required, and an additional reagent gas is introduced to the ionization chamber. The role of the reagent gas is to first become reagent ions through an EI process, and then to ionize the analyte molecule by protonation. CI is a “softer” ionization technique compared to EI, and often produces intact protonated molecules which contain the most important information of mass spectrometry analysis: molecular weight. Currently, EI and CI both remain as valuable methods in the analysis of organic molecules. An additional ionization technique involves an atom beam, resulting in “secondary” analyte ions. Fast atom bombardment (FAB) was introduced in the 1970s; 9 it employs a beam of atoms with a kinetic energy of several thousand electronvolts to bombard the sample dissolved in a liquid matrix. The process causes the desorption/ionization of the analyte. When a primary ion beam is used instead of the atom beam as the energy source, the technique is referred as liquid secondary ion mass spectrometry (LSIMS). Along with other desorption/ionization techniques, such as field desorption (FD), plasma ( 252 Cf) desorption (PD), etc., FAB was developed in an effort to analyze nonvolatile and/or thermally labile compounds incompatible with EI or CI. However, it was the introduction of matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI) in the late 1980s that had the greatest impact on the analysis of these difficult samples. These two ionization techniques enabled researchers to investigate the area of biomolecules, e.g., proteins, peptides and oligonucleotides, which the developments in biochemistry demanded.

PAGE 19

5 As an ion source for mass spectrometry, ESI was first introduced by John Fenn and co-workers in 1985, and the featured analysis of non-volatile compounds. 10 In this technique, the analyte is mixed with a solvent and sprayed from a narrow tip across an electric field. After the solvent evaporates from charged fine droplets generated in the spray, bare ions are eventually formed, sampled and analyzed. Aiming at providing a “soft” ionization process as well, MALDI represents a different approach and will be described later in the chapter. In short, both ESI and MALDI are useful tools in the post-genomic era. They have expanded the horizon of mass spectrometry and provided ample research opportunities, especially for protein analysis in life science, as shown in numerous publications. In summary, mass spectrometry is an ever growing area that embraces and contributes to many disciplines. A quick glance at the proceedings of the 52nd Annual conference of American Society for Mass Spectrometry (ASMS) would give one an idea of how many areas mass spectrometry affects. 11 In addition to its traditional beneficiaries, such as organic chemistry, petroleum chemistry, geochemistry, and environmental chemistry, research activities are now carried out in biochemistry, polymer chemistry, pharmaceutical industry, food industry, and forensics. Time-of-Flight Mass Analyzer Background and Principle Because of its similarity to MALDI in terms of instrument configuration, DIOS is usually coupled with a time-of-flight (TOF) mass analyzer commonly used for MALDI. TOF is conceptually the simplest mass analyzer; it separates ions based on their arrival times drifting through a field-free region with the initial velocities acquired in the same electric field.

PAGE 20

6 Fig. 1.1 shows the scheme of a linear TOF instrument operating in the positive mode. Ions, generated upon laser radiation, are repelled and accelerated toward a field-free fight tube between the source and the detector. Experiencing the same electric field, ions with different mass-to-charge ratios acquire the same kinetic energies, and correspondingly, different initial velocities before they enter the flight tube. Ions with smaller m/z value have higher initial velocity, and reach the detector first. DetectorSourceField free region d Laser+ 20 kV Figure 1-1. Scheme of a linear TOF mass analyzer operating in the positive mode. In a simplified fashion, the following Eq. 1-1 describes the factors that determine the time of flight of an ion in a TOF instrument, where m is mass of the ion (kg), z is the number of electron charges, d is the drift distance (m), V is the accelerating potential of the electrical field (V) and e is the electron charge (c). Vedzmt22 (1-1) This straightforward concept contributed to the early arrival of the TOF mass analyzer. It was first introduced in 1946 by Stephens, 12 and refined by Wiley and McLaren in 1955 to give birth to the first commercial instrument. 13 However, the

PAGE 21

7 application of the TOF analyzer was limited in its early years (1960s to 1980s), due to the unfulfilled requirement for pulsed ion source and fast electronics. Not until the 1990s did TOF enjoy the renaissance of becoming a mainstream mass analyzer. A few important advances in related areas were responsible for the return, a converging of events often referred as “perfect timing”. 14 The first major development is the introduction of matrix-assisted laser desorption/ionization (MALDI) 15;16 with a pulsed ion source. Discrete packages of ions of all the mass-to-charge values are generated in nanoseconds, sent to the flight tube, and then separated and detected; no gating or pulsed extraction is necessary. Moreover, MALDI is capable of ionizing biomolecules of high molecular weights, which may exceed the mass range of other popular mass analyzers, such as the quadruple analyzer and the ion trap. Ions with higher m/z values place little constraint on the TOF instrument, for bigger molecules simply take longer time to move through the flight tube. It requires only microseconds to extend the mass range considering the dependence of . zmt/2 Another development was in the electronics and computer technology that facilitated the recording and processing of mass spectra in the microsecond time frame, e.g., nanosecond digitizers and focal plane detectors. Also, affordable pulsed lasers and improved ion optics have increased the popularity of the TOF instrument. In addition, there were two major improvements in the TOF analyzer that played important roles in its revitalization: the delayed-extraction and reflectron kinetic energy analyzers. Both of these implements, featured in the instrument used for this research, drastically improved the resolution of the mass analyzer, thus providing enhanced analytical performance.

PAGE 22

8 Delayed-Extraction Delayed-extraction, also known as time-lag focusing, was first incorporated to the TOF instrument by Wiley and McLaren in the 1950s. 13 The importance of the technique came to light when it was coupled to a MALDI ion source by several research groups. 13;17-20 When a MALDI-TOFMS instrument is operated in the continuous extraction or dc extraction mode, ions produced by a laser pulse are immediately extracted by an electric field, created by continuously applying a dc voltage to the source plates. The drawback of this operation mode is that ions maintain the initial kinetic energy distribution resulting from the laser radiation. Consequently, this distribution leads to a spread of arriving time and causes poor resolution. Fig. 1.2 shows the principle of delayed extraction. In contrast to continuous extraction, ions are allowed to expand into a field-free region between the firstt and the second plate 1 and plate 2; the initial kinetic distribution is converted into a spatial spread. After a few hundred nanoseconds, an electric field of several kilovolts is created between the two plates by manipulating the voltage in either plate. Ions are then extracted to the downstream electric field and further accelerated. For ions of the same m/z value, the less energetic ions move slower during field-free expansion and stay behind. However, they acquire more kinetic energy during extraction and eventually reach the detector at the same time as the more energetic ions.

PAGE 23

9 201020 0 in kV Detector 201017 0 in kV 100-400 ns laterUpon laser radiationSource Figure 1-2. Delayed-extraction in a linear TOF instrument at the positive detection mode. The gray circles represent ions with higher initial kinetic energy. Reflectron Kinetic Energy Analyzers The Russian scientist Mamyrin first described the concept of the reflectron for a TOF mass analyzer in his 1966 dissertation and later in his articles. 21;22 In the 1980s, TOF instruments equipped with reflectrons were available from several manufacturers, e.g., Leybold-Heraeus., Bruker Daltonics, Finnigan MAT. 23 A reflectron, also referred as an electrostatic reflector, consists of a series of grids or ring electrodes. Located at the end of the linear flight tube, it generates a retarding electric field and acts as a mirror by deflecting ions to another flight tube. Like the delayed extraction, a reflectron corrects the kinetic energy dispersion, shown in Fig.1-3. The idea, according to Mamyrin, was to “neutralize” the energy spread outside the ion source. Ions of the same mass-to-charge value reach the linear detector at slightly different times owing to their initial kinetic energy distribution. While in the decelerating field created in the reflectron, this distribution is compensated for since it takes longer

PAGE 24

10 time for the more energetic ions to go in and come out of the “mirror”. Consequently, isobaric ions will reach the reflectron detector as a much better focused ion package. Linear detector Linear detector Reflectron detectorSourceSourceReflectron t0t1t2t0t1t3t4 Figure 1-3. Scheme of a TOF mass spectrometer equipped with reflectrons. Laser Desorption/Ionization Techniques Laser Desorption/Ionization The ultimate goal of ionization methods in mass spectrometry is to generate gas phase ions. In EI or CI, the ionization process consists of two relatively distinct steps: vaporization and ionization. First, volatile samples in a condensed phase release gas phase molecules due to rapid heating. Then, electron-molecule or ion-molecule interactions create the ions for mass analysis. In addition to the two-step ionization passway, there is another type of ionization technique – desorption/ionization technique – in which vaporization and ionization occur almost at the same time. This category includes fast atom bombardment (FAB), liquid secondary ionization mass spectrometry (LSIMS), field desorption (FD), plasma desorption (PD), and later on, matrix-assisted laser desorption ionization (MALDI).

PAGE 25

11 Common to the desorption/ionization techniques is the rapid addition of energy on samples of condensed phase and the subsequent generation of ions. The energy source can be an atom or ion beam, a high-temperature plasma, or laser radiation. Desorption/ionization techniques appeared later than EI or CI, and their developments were driven by the need to analyze non-volatile molecules with high molecular weights. These techniques have broadened the applications of mass spectrometry, especially in biochemistry, medicinal chemistry and life science. At the same time, the advance in the ion source has stimulated efforts to improve mass analyzers. Among the advances in desorption/ionization techniques, laser desorption/ionization (LDI) has attracted attention dating back to the 1970s, when Vastola et al. reported the laser mass spectra of sodium hexylsulfonate salts with no observable fragmentation. 24 Generally, laser pulses of ~10 6 -10 10 W/cm 2 are focused on a sample surface of about ~10 -3 -10 -5 cm 2 , and a plume of ions and neutral molecules is created from the laser ablation for mass analysis. 25 Since its first introduction, LDI remained as an effective method in generating gas phase ions. However, it was the discovery of MALDI in the late 1980’s, after many important developments in the LDI technique, has drastically changed the picture of mass spectrometry. Matrix-Assisted Laser Desorption/Ionization There are two research groups responsible for the introduction of MALDI. German scientist Franz Hillenkamp and Japanese scientist Koichi Tanaka are both recognized and have won awards for their contribution to bringing the technique to light. Since its formal debut in 1988, numerous papers have been published on various aspects of the technique,

PAGE 26

12 including applications, mechanisms, sample preparation methods, and designs of coupling to different mass analyzers and separation techniques. Below is a brief description of the principle of the MALDI technique, the roles of the matrix, and how the concept of the matrix would lead to other desorption ionization techniques, such as DIOS. A common MALDI process is described in Fig. 1-4. The analyte is mixed with a matrix compound in liquid phase and the mixture is spotted on the supporting surface (usually stainless steel) and dried. As the result of co-crystallization, a “solid solution” is formed in which the analyte molecules are embedded in excess matrix molecules and are well isolated. Upon interaction with the UV laser radiation, matrix molecules in the condensed phase are excited, and the rapid heating of the crystals is induced after having accumulated a large amount of energy. Localized sublimation of the matrix crystals results in an expanding matrix plume, carrying intact analyte molecules/ions into the gaseous phase. Laser, 337 nmMatrix moleculeAnalyte molecule Figure 1-4. Schematic of a MALDI process. A final mixture of the analyte and the matrix compound is achieved on a MALDI target, usually a stainless steel plate, after combining the two components via various routes. Laser radiation at 337 nm is used to radiate the mixture and create ions in the resulting plume, including both matrix ions and desired analyte ions.

PAGE 27

13 Preformed analyte ions, meaning charged analyte molecules formed in the condensed phase before laser radiation, can be a source of ionization. In addition, ionization may be due to gas phase ion interactions as well as other proposed chemical or physical pathways. 26 Although the MALDI mechanism is still under discussion, one important observation is that little fragmentation accompanis the process. Since there is almost no direct impact of laser energy on analyte molecules, and they are cooled down during the plume expansion into the vacuum, analyte molecules experience little fragmentation. Thus, MALDI is recognized as a “soft” ionization method, and this attribute has led to its successful analysis of non-volatile, heat-labile, high molecular weight biomolecules. The matrix compounds play significant roles in the MALDI process. They are usually simple organic acids with high molar absorption coefficients ( > 10,000 L/mol cm) at the incident laser wavelength. The functions of a MALDI matrix includes: 1) absorb laser energy, prevent the direct laser impact on the analyte, and desorb the analyte through plume expansion; 2) isolate the analyte molecules and prevent the production of analyte clusters; 3) provide a proton-rich environment and promote ion formation. Commonly used matrices include 2,5-dihydroxybenzoic acid (DHB), -cyano-4-hydroxycinnamic acid (CHCA), sinapinic acid, 3-hydroxypicolinic acid (HPA), dithranol, etc. MALDI-MS can be applied to answer research questions involving the mass determination of peptides, proteins, oligosaccharides, oligonucleotides and polymers. Coupled with the other methodologies and protein database searches, protein identification and quantification can help one understand the phenomena in areas that are

PAGE 28

14 otherwise unapproachable. In addition, the introduction of MALDI, and the concept of the matrix has inspired other ionization techniques. One such technique is desorption/ionization on porous silicon (DIOS). If a matrix compound creates a soft ionization/desorption environment, are there other materials having the similar properties that fit for this goal? In 1999, Wei and coworkers proposed the idea of using a porous silicon (PSi) surface as a platform for mass spectrometry analysis. They were able to observe peptides and drug molecules down to low femtomole levels. 27 The attractive advantage of DIOS is the much-reduced background in the lower mass region, where in MALDI, the presence of matrix peaks usually hinders the identification of the desired signal. At the same time, DIOS retains the some of the advantages of MALDI. First, it is also a “soft” ionization technique, thus suitable for the analysis of non-volatiles with little fragmentation. Second, DIOS generates singly-charged ions, promising simplicity in spectrum interpretation. DIOS can be readily implemented in any MALDI compatible instruments, and is commonly coupled to a TOF analyzer. In a DIOS analysis, a PSi chip can be attached to a slightly modified MALDI plate using double-sided tape; the plate is inserted into the ion source the same way as in a MALDI experiment. Before more detailed discussion on DIOS-MS and its applications, the physical and chemical properties of PSi is briefly described. Porous Silicon and Its General Applications Porous Silicon Fabrication First introduced in 1956, anodic electrochemical etching has been the most commonly used procedure in porous silicon fabrication. 28 Other routes such as metal-assisted etching, 29 stain etching, 30 agglomeration of fine Si powder 31 have also been

PAGE 29

15 reported. Silicon substrates are high-quality single crystals with controlled types and amounts of dopant (boron, arsenate, phosphorus, etc.). Although there are different etching apparatus designs in anodic etching, the experiment generally involves a Teflon cell, a hydrofluoric acid (HF)-containing etching solution and two electrodes, the anode being the silicon substrate and the cathode a platinum wire. 32 The etching parameters include the applied voltage or current, the etching duration, the etching solution composition (especially HF concentration), and with or without illumination. These conditions can be adjusted to achieve desired porous silicon morphologies and may vary for silicon substrates of different doping type (nor p-) and resistivity (). As an example, the PSi used in DIOS-MS can be fabricated through anodic etching using n-type, phosphorus-doped, low resistivity (0.01-0.02 cm) crystalline silicon, with a current maintained at ~50 mA/cm 2 for ~1-2 min under tungsten filament illumination (~50 mW/cm 2 ). 33 Metal-assisted etching is not a widely used method in PSi fabrication; however, it is the main approach used in this research (a comparison with anodic etching will be discussed in Chapter 2). The technique does not require electrical current or illumination; instead, it begins with coating the crystalline silicon with a thin layer (nanometer thickness) of Au or other precious metals. Then the coated chip is immersed in an etching solution, usually composed of HF, H 2 O 2 and an organic solvent, such as ethanol. Despite the different etching route, the chemical change on the surface was suggested to follow the HF-Si chemistry for both methods. In the anodic etching, incident photons are believed to initiate the chemical reaction (Eq. 1-2). Electrons facilitate

PAGE 30

16 hydrogen gas evolution (Eq. 1-3). SiF 4 is generated in the following step (Eq. 1-4) which further reacts with HF to yield the final product H 2 SiF 6 (Eq. 1-5, 1-6). I. Reactions for anodic etching: Cathode Reaction: Si + 2HF + 2h SiF 2 + 2H + (1-2) 2H + + 2e H 2 (g) (1-3) Anode Reaction: SiF 2 + 2HF SiF 4 + H 2 (g) (1-4) Si F 4 + 2HF H 2 SiF 6 (1-5) Overall Reaction Si + 6HF H 2 SiF 6 + 2H 2 (g) (1-6) II. Reactions for metal-assisted etching Cathode Reaction: H 2 O 2 + 2H + 2H 2 O + 2h + (1-7) 2H + + 2e H 2 (g) (1-8) Anode Reaction: Si + 4HF + 4h+ SiF 4 + 4H + (1-9) Si F 4 + 2HF H 2 SiF 6 (1-10) Overall reaction: Si + H 2 O 2 + 6HF 2H 2 O + H 2 SiF 6 + H 2 (g) (1-11) In metal-assisted chemical etching, 29 the deposited Au metal coating appears as nanometer-sized (~10 nm) islands which may act as local cathodes while H 2 O 2 provides

PAGE 31

17 positive “holes” (h + ) for the redox reactions. The silicon chip itself acts as the “anode”. The localized electrochemical process is shown in Eq. 1-7 – 1-11. The mechanisms for pore development during PSi formation are not well understood. The size and shape of the pores are affected by above-mentioned experimental factors. A few theories have been proposed on the pore formation process, which is roughly divided into two stages: pore initiation and pore propagation. The pore initiation stage may involve one or a combination of a few of the following factors: dopant-induced dissolution, 34 interface instability, 35 hydrogen induced defects, 36 diffusion effects, 37 and vacancy supersaturation. 38 In the pore propagation stage, image force effects, 39 hole diffusion, 40 charge transfer, 41 quantum confinement, 42 and surface tension 43 are possible factors in the formation of the columnar structure of the porous silicon. Properties of Porous Silicon The porous construction of PSi is schematically shown in Fig. 1-5a. 44 The size of the pores is nanometers while the depth of the porous layer is microns. The structure, sometimes considered as a system of interconnected quantum wells, is also referred to as “quantum sponge” structure, 45 for the dimension of the pores fall within a quantum confinement range. The phenomenon of “quantum dots” is also attributed to the fact that the size of certain materials is reduced to the “magical” range and become photo-luminescence. 42 A variety of topography measurements, such as atomic force microscopy (AFM), scanning electron microscopy (SEM), and scanning tunneling microscopy (STM) have been used to reveal the different types of pores, 32 shown in Fig. 1-5b.

PAGE 32

18 1 2 3 4 5 a b Figure 1-5. Structure of porous silicon. a) Schematic structure of porous silicon layer on crystalline bulk silicon (adopted from Reference 42); b) pore types (adopted from Reference 30): (1,2) blind, or dead-end, (3) interconnected or branched, (4) totally isolated or “closed”, (5) “through” pores. According to the International Union for Pure and Applied Chemistry (IUPAC), pore sizes are classified by the pore width as micropore (2 nm), mesopore (2-50 nm) and macropore (50 nm). The majority of electrochemically and metal-assisted etched PSi is mesoporous. The “porosity” is defined as the fraction of the apparent volume of the sample attributed to pores, ranging from ~4% up to ~95%. Another concept used to describe the physical appearance of PSi is the surface area, which varies from ~1 m 2 /g in macroporous surfaces to ~800 m 2 /g in wholly microporous surfaces. 32 The chemical composition of PSi surface is a controversial topic. Many have demonstrated that the freshly-etched PSi surface contains Si-H x groups, typically using IR absorption 46;47 and nuclear magnetic resonance (NMR) technique. 48 Fluorine was also observed on PSi, 49 although in what form the element was incorporated is still under debate. Another fact that interests researchers is the rapid oxidation of the material, which could happen within the first few minutes after PSi fabrication. 50;51 As a result, a structure resembling siloxene [(Si 6 H 6 O 3 ) n ] was proposed. 52 Moreover, the presence of carbon, its

PAGE 33

19 content reaching 10% (C/Si) in aged samples, was observed. Ambient air was identified to be the source of contamination. 50 General Applications of Porous Silicon Recent research activities have been prompted by the discovery of strong room-temperature photoluminescence (PL) of PSi in the visible by Canham in 1990. 42 The luminescence is more efficient than the infrared luminescence of crystalline bulk silicon as shown in Fig. 1-6. 53 A significant fraction of the more than 2,000 published papers on this phenomenon seek clarification of the source of the emission. More than thirty different theories have been proposed; they can be grouped into four categories: 1) quantum confinement; 2) nanocrystal surface states; 3) specific defects or molecules; 4) structurally disordered phases. 32 Figure 1-6. Room temperature photoluminescence spectra of crystalline silicon and PSi. (taken from reference 44) Despite the obscure origin of photoluminescence in PSi, fresh opportunities have emerged for applications, especially in physics and material science. The first to be noticed is the optoelectonic area, which is a combination of optics and electronics. 54 Bulk silicon is not suitable for major components in optoelectronic devices such as a light

PAGE 34

20 emitting devices or optical waveguides, due to its poor light emission properties and low quantum efficiency. However, silicon technology is well-established in electronic device manufacture, so it is of great importance to be able to utilize the experience and knowledge of silicon technology in optoelectonics. The discovery of intense visible PL of PSi bridges the gap, allowing the manufacture of Si-based optoelectronic parts 55;56 and integrated circuits. 57 PSi is also used in sensors, owing to its high surface area. Molecules adsorbed to PSi surface may change the capacitance, resistance or the PL properties of the material. Accordingly, several transducer schemes have been proposed. 58-60 In addition, applications of PSi as passive optical components, like interference filters and diffraction gratings, can be designed through fabricating the material. 61;62 Besides the new involvement with mass spectrometry, the potential for PSi to serve as supporting material in biomedical fields are currently under investigation. From the material science point of view, problems in biocompatibility, material stability and surface-biofluid interactions hinder the applications. 32 However, chemically functionalized surfaces showed less degradation. PSi surfaces modified with 1-dodecyne through the Si-H bond were tested in an environment mimicking human blood plasma and displayed less oxidative destruction. 63 In addition to its ability to simply surviving the blood plasma, PSi can be used to support living cultures of mammalian tissues. 64 Comparing to other silicon-based substrates, PSi showed a higher viability for the cells of interest. The result suggested the possible interface between electronic circuitry and human neural circuitry.

PAGE 35

21 Another interesting application of PSi is in miniaturized Total Analytical Systems (-TAS). These systems have “microreactors” created by electrochemical etching of prefabricated silicon channels (Fig.1-7), followed by the immobilization of catalyst molecules on the PSi surface. 65 By converting crystalline silicon into PSi, the surface area is greatly increased, leading to up to 350-fold gain in enzymatic activity. The analysis can be integrated into a MALDI-TOF instrument with the whole process automated. a b Figure 1-7. SEM images of microreactors. a) a 32-channel structure of lithographically fabricated silicon; b) cross section of the structure after the anodic etch (taken from Reference 62) With the recognition of PSi’s ability to assist desorption/ionization in mass spectrometry, applications of the material in biochemistry are broadened and represent a good deal of research interest. Desorption/Ionization on Porous Silicon Mass Spectrometry (DIOS-MS) Principles The basis of the DIOS model is simplistic and straightforward. The process parallels MALDI in that it is also a solid surface ionization technique. Instead of a stainless still target, a porous silicon surface is used in DIOS as the substrate for sample deposition without additional matrix compounds (Fig.1-8). 27

PAGE 36

22 SourceTOF mass analyzer Laser 337 nm Analyte ions Figure 1-8. Schematic of a DIOS-MS process. A 337 nm N 2 laser is used to irradiate the dried sample residue. Ions generated are extracted and accelerated to a TOF mass analyzer. More commonly, the DIOS substrate is patterned to have an array of PSi spots, allowing the analyses of multiple samples. The porous structure plays a similar role in DIOS as the matrix does in MALDI. The matrix isolates analyte molecules and dissipates the laser energy. Likewise, the PSi has the functions of accommodating analyte molecules in its nano-sized pores, and absorbing laser energy at the same time (absorption coefficient at 337 nm is on the order of 10 5 cm -1 42 ). Characteristics DIOS-MS has demonstrated performances comparable to MALDI-MS, with intact molecules being detected at femtomole or attomole levels with little or no fragmentation. 27;33 More importantly, the absence of the matrix allows analysis of low molecular weight samples without signal interference caused by the matrix. The absence of the matrix also allows for simple sample preparation. In MALDI sample preparation, the solvent, the matrix to analyte ratio, as well as the type and amount of acid modifier added are variables that affect the performance of the technique. It can be time-consuming to determine suitable sample preparation protocols for

PAGE 37

23 particular types of compounds. On the contrary, sample preparation in DIOS is less complicated and the only step is to dissolve the sample in a proper solvent. In addition, not needing an externally introduced matrix avoids the possible destruction of the analyte by the acidic matrix. Acid-sensitive samples are not compatible with MALDI, since the matrix compounds are organic acids having a pH around 2-3; however, it poses no problem for DIOS. This also means the technique is suitable for analytical problems where the inherent equilibrium of interest is sensitive to pH. In terms of instrumentation, DIOS can be adapted to most existing mass spectrometers with a MALDI source. The only modification is on the MALDI plate, usually made from stainless steel, where a recession having the size of a DIOS (PSi) chip is machined in the center of the sampling surface to accommodate the chip. Compared to MALDI, the extra step in DIOS analysis is to fabricate the PSi chip. The raw material, silicon wafers, are inexpensive and the fabrication procedure is relatively easy with an option of batch processing. DIOS chips are not recommended for re-use. Besides the advantages discussed here, DIOS does have the drawback of limited mass range. Optimal performance of DIOS is typically obtained for molecules less than 3000 Da, although the detection of an18 kDa molecule was reported. 66 Moreover, despite the fact that there are no matrix peaks, signal suppression from background ions of unidentified sources can be problematic. 67 Mechanism While researchers are gaining experimental knowledge of the technique, the mechanism behind the phenomenon is also a focus of interest. After comparing the effects of UV and IR lasers, different surface morphologies (plain, rough and porous),

PAGE 38

24 different substrate materials (silicon or carbon), Sergey Alimpiev and coworkers concluded that three aspects need to be considered to understand the DIOS process. 68 The first was the role of the laser radiation. The local temperature on PSi surface caused by UV laser pulses was calculated to be ~ 800 K; desorption was believed to be the result of the rapid heating. However, the production of protonated molecules, [M+H] + , was not simply explained by the photo-ionization of certain reagent molecules followed by chemical ionization of the analyte, since IR lasers also generated similar [M+H] + where the laser energy was not enough to initiate the photo-ionization. As a result, desorption of pre-formed ions seemed a more plausible route. The second was the importance of surface morphology. From the very beginning, it has been noticed that signals were observed from porous silicon surfaces but not cystalline silicon surface. Although the photoluminescent property of PSi was well-known, independent research groups have shown that it was not related to DIOS performance. Still, the UV absorbing nature of PSi was responsible for the desorption processes, suggested by Alimpiev and coworkers. 68 It was also suggested that in addition to the fact that the high surface area in porous or rough silicon surface retained the analyte, the surface morphology might have two more functions. 1) The lower heat capacity and lower heat conductivity of PSi, compared to plain silicon, further increased the surface tempetature. 69;70 2) Experimental support was obtained by Alimpiev et al. that the surface roughness played a role in physically separating preformed analyte ions from their counter ions, as charge separation on the surface was crucial for the desorption of preformed analyte ions. 68 The authors also indicated that additional evidence was needed to confirm the proposed mechanism.

PAGE 39

25 The last observation was that the analytes needed to have an aqueous ionization constant (pK a ) value larger than 4 for [M+H] + ions to be observed in DIOS. In other words, basic species with larger pK a values would have a better change of being ionized. After correlating the signal responses of a variety of compounds to their aqueous phase basicities and gas phase basicities, it was concluded that the former property was more important to a DIOS process. This also indicates that solvent is present on PSi surface to stabilize the preformed ions via solvating. As a result, DIOS is not truly “matrix-free” due to the ubiquitous existence of residual solvent. Overall, DIOS is an exciting addition to the toolbox of ionization techniques. Due to its unique features, DIOS-MS is used to analyze small organic molecules, 27 peptides, 27 polymers, 71;72 carbohydrates 73 in areas such as protein identification, 66;74 enzyme activity monitoring, 66;75 quantitative analysis, 76 and direct tissue analysis. 67 Furthermore, the technique is undergoing advancements including PSi substrate modifications, incorporation of IR lasers, 77-80 atmospheric pressure analysis, 81;82 as well as options of coupling to separation techniques 74 and MS/MS 83 analysis. Representative applications and recent developments of DIOS-MS are briefly reviewed next. Applications and Development of DIOS-MS DIOS-MS was first applied by Suizdak’s group to characterize small molecules, using the technique’s inherent advantage of reduced chemical background in the low mass region. 27 Analytes with ionizable functionalities were routinely detected with little or no fragmentation up to 3000 Da. The mass spectra were relatively clean and free of interfering ions. Samples at femtomole or attomole levels were observed with good signal-to-noise ratios. Noticeably here, the requirement of certain “functionalities” is

PAGE 40

26 consistent with the proposed mechanism discussed above where a pK a greater than 4 is necessary. Suizdak’s group proceeded to employ the technique in protein characterization, which included both functional and structural analyses. 66 In their work, functional assays were performed on an esterase, a glycosidase, a lipase, and exo-, endo-proteases by using enzyme-specific substrates. Enzymatic reactions were sustained on the DIOS surfaces in incubation chambers, and the formation of enzyme reaction products was monitored. As an example, choline, the enzymatic product of acetylcholinesterase and acetylcholine, was quantified over time using an internal standard; the result revealed the pseudo-first-order kinetics of the reaction. The researchers continued to test several enzyme inhibitors by including them in the initial mixtures. The effectiveness of the inhibitors were acquired according to the change in the ion signal of choline, the enzymatic reaction product. The research demonstrated that DIOS-MS was suitable for screening enzymatic inhibitors and was superior to radioactive isotope labeling or fluorescent labeling which either involved hazardous reagents or reagents with modified structures. Since peptides also fall into the effective mass range of DIOS, protein identification has also been carried out using standard peptide fingerprinting methods. One advantage acclaimed was that DIOS was more tolerant to biological contaminants commonly present in peptides digests, thus less purification was required. 66 Sweedler et al. has also explored the possibility of direct assay of tissues and cells with DIOS-MS. 67 The atrial gland sliced to expose fresh tissue was pressed against the PSi for ~2 s. Solvent was applied to help transport analyte into the pores. Similarly, a buffer solution of dissociated neuron cells was spotted on PSi surface. Mass spectra

PAGE 41

27 obtained with DIOS-MS in the range of 1000-4000 Da showed peptide and hormone peaks for the two samples, as well as other unknown signals. The authors suggested DIOS offered a promising alternative to MALDI in tissue mass spectral imaging. While DIOS-MS was demonstrated to be a useful tool in mass spectrometry analysis, efforts from various research groups have contributed to the developments in instrumentation and methodology of the technique. First, quantitative DIOS-MS analysis was attempted. 76 The key point was to use electrospray deposition, which allowed homogenous sample deposition on the PSi surface. Electrospray was accomplished through a fused-silica column of 100 m i.d. with a pulled tip of ~5 m, and a potential difference between the capillary and the DIOS chip. Significant improvements in quantitation as well as much higher sample-to-sample reproducibility were achieved compared to traditional dried-droplet sample deposition method. In addition to quantitative analysis, the electrospray deposition also facilitated the off-line coupling of DIOS-MS to LC separation. 74 In this work, electrospray deposition generated a spatially preserved linear track of the separated sample on a DIOS chip. Improved sequence coverage was achieved with DIOS approach for two model proteins, compared to nanoLC/ESI-MS. 74 This was mainly due to, suggested by the authors, the lessened problem in signal suppression in DIOS-MS. The nature of off-line coupling imposed no time limit on peptide analysis, otherwise data dependent scanning used in on-line techniques might miss peptides of lower intensities. Although it was an off-line approach, the average time for MS analysis the DIOS chip was satisfactory: ~ 2 min/lane

PAGE 42

28 for ~2 lane/sample. With simultaneous LC separations and automated deposition, a 10-fold increase in throughput was achieved over conventional LC/MS. With the attempts in modifying the front-end of DIOS analysis, Suizdak et al. have also investigated MS/MS analysis using a TOF/TOF mass analyzer. 83 The ability to provide high mass accuracy and fragmentation information was an additional advantage for the technique in small molecule characterization and protein identification. The feasibility of atmospheric pressure (AP) DIOS-MS has been investigated by MassTech Inc. (Columbia, MD), which has developed commercial AP-MALDI sources. Minimal modification was required to convert an AP-MALDI source to an AP-DIOS source, and an ion trap mass analyzer was used to provide MS n capability. Although not comparable to vacuum DIOS, the sensitivity (sub-picomole range for standard peptide mixtures) was still reasonably good. In addition, improved signals in lower mass range allow for significantly better sequence coverage for tested tryptic digests. In another set of experiments by the same research team, quantitative analysis of drug molecules with AP-DIOS-MS was carried out. Reasonable reproducibility, and good linearity in a dynamic range of 3 orders of magnitude were achieved. Murray et al. have evaluated the IR laser as the energy source and obtained interesting results. 77-80 While previous work by other groups had suggested that having IR absorbing solvent would improve DIOS signal on a PSi surface, they reached the conclusion that enhancements in the efficiency of desorption/ionization could be obtained on both porous and crystalline silicon. Since the mechanism for UV-DIOS is still not clear, and IR-DIOS may follow different principles, understanding the theoretical meanings and practical significance of IR-DIOS remains an intriguing task.

PAGE 43

29 Among recent research involving DIOS-MS, one breakthrough was the chemical modification of PSi surface. 84 The surface functional group, Si-OH, was generated on a purposely-oxidized PSi and subjected to silylation reactions. The importance of such procedure was reflected on the much improved sensitivity gained for certain molecules, such as the detection of 480 molecules (~800 yoctomole) of des-arg-bradykinin. Moreover, it was claimed that the stability of the DIOS chip was improved, and the surface was more resistant to oxidation and acid/base hydrolysis. The authors also proposed that the surface could be modified with a variety of functional groups for different purposes. 84 For example, on-spot sample clean-up was demonstrated on a C 18 derivatized PSi surface. In this case, a small amount of washing solution was applied to the dried sample spot and quickly withdrawn to remove hydrophilic salts and contaminants while hydrophobic analytes were retained on the PSi surface. It worked in a similar way as a Zip-Tip (Millipore Corporation, Billerica, MA), and was termed as “z-touch” for being a manipulation on the z-axis perpendicular to a surface. While C 18 or pentafluorophenol derivatized surfaces displayed a signal boost for hydrophobic analytes, surfaces modified with an amine group were amenable to hydrophilic analytes. There continues to be research concerning DIOS-MS as well as utility of PSi-related material in mass spectrometry. As a result, the understanding of the mechanism and applications of the technique to various problems are forthcoming. This dissertation will start with a fresh assessment of DIOS-MS, including PSi preparation, surface evaluation, as well as a discussion on the applicability and limitations of the technique, then followed by several unique applications of DIOS-MS.

PAGE 44

CHAPTER 2 FUNDAMENTALS AND APPLICABILITIES OF DIOS-MS Introduction This chapter describes the author’s effort in developing DIOS-MS in an independent laboratory setting, including the porous silicon surface fabrication and the adaptation to a MALDI-TOF instrument. Two approaches of PSi fabrication for mass spectrometry analysis were compared, and the metal-assisted etching was selected over the anodic etching. Next, PSi surface prepared by metal-assisted etching was evaluated using a variety of techniques, including FT-IR, Auger electron spectroscopy and SEM. Various analytes and instrumental operating modes were examined to explore the applicability of DIOS-MS using PSi prepared by metal-assisted etching. Also, the limitations of DIOS-MS were discussed through examples. A noticeable feature of PSi was the aging of the surface, which reportedly caused deteriorated DIOS-MS signals. 85 The phenomenon was investigated and a method referred as “re-etching” was applied in an attempt to counteract the effect. Furthermore, strategies to improve the DIOS-MS signal were examined. Production of PSi As mentioned in Chapter 1, there are different approaches of PSi fabrication. 28-31 In DIOS-MS analysis, a popular method is anodic electrochemical etching. 27;33 In addition, Sweedler et. al. suggested that PSi generated from H 2 O 2 MEtal-assisted (HOME) etching 30

PAGE 45

31 was also suitable for DIOS analysis. 86 In this work, both schemes were examined, and the optimal etching method and conditions were determined. Anodic Etching First, anodic etching was carried out on an assembly (Fig. 2-1a) composed of an optical system and an etching cell. The optical system included a white light source (150 w halogen lamp), a patterning mask and three glass lenses. The mask was a black plastic slide with an array of small transparent spots and numbers. The transparent areas allowed light to pass and focus on the crystalline silicon surface. Only the illuminated surface developed PSi. As a result, chips with an array PSi spots (e.g., 4x6) and numbers were produced, which became a registration system for sample deposition. Power SupplyDigital Current MeterResistor Etching CellMicroscope Stage Light SourceLens OneMaskLens TwoLens ThreePlatinum Wire (cathode)Teflon Cell TopSilicon ChipGold Plate (anode)Teflon Cell Bottom O-ringab Figure 2-1. Instrument set-up for anodic electrochemical etching. a) Assembly of the optical system and the etching cell; b) electrical circuit for PSi fabrication.

PAGE 46

32 Silicon chips (2 cm 2 cm) were cut from a low resistivity (0.005-0.02 cm), n-type, (111) oriented crystalline silicon wafer (Silicon Sense, Inc. Nashua, NH). A cylindrical etching cell forms (Fig. 2-1a) when a silicon chip was sandwiched between two Teflon cell parts, with a gold sheet beneath the silicon chip as the anode, and a platinum wire immersed in the etching solution (HF/ethanol, v:v =1:1) as the cathode. An o-ring was used to seal the cell. Fig. 2-1b shows the electrical circuit for the anodic etching. A resistor (10 k) was included to better control the etching current, which was monitored with a digital multimeter. To achieve a sharp image, the etching cell was placed on a microscope stage for fine adjustment of the vertical position. PSi was etched under white light illumination at a current density of ~4 mA/cm 2 for ~1 min. Metal-Assisted Etching The PSi surface was also manufactured with the HOME (H 2 O 2 MEtal) etching method, or simply referred as the metal-assisted etching. Optimized etching conditions were as following: a silicon chip was cut from a low resistivity (0.005-0.02 cm), n-doped, (111) oriented crystalline silicon wafer (Silicon Sense, Nashua, NH). Using a patterning aluminum mask, the chip was sputter-coated with a thin layer of Au in a Hummer TM 6.2 sputter coater (Ladd Research, Williston, VT). The Au-coated silicon chip was immersed in an etching solution composed of 49% HF: 30% H 2 O 2 : ethanol (1: 1: 1 by volume) for 20 s. PSi was developed in the areas coated with Au, and an arrayed chip of 16 PSi spots (approximately 800 m in diameter) was produced.

PAGE 47

33 Comparison of Two Methods Both anodic etching (Fig. 2-2) and metal-assisted etching (Fig. 2-3) were able to generate PSi structure, as indicated in the SEM images obtained on a Hitachi S-4000 FE-SEM instrument (Hitachi High Technologies America, Schaumburg, Illinois). Figure 2-2. Images of crystalline silicon and porous silicon. a) SEM image of crystalline silicon; b) SEM image of PSi prepared by anodic etching; and c) optical image of a sample spot prepared by anodic etching using the patterning mask. Figure 2-3. SEM images of PSi prepared by metal-assisted etching (cross section view) at magnifications of: a) 100 k; b) and c) 50.0 k; d) 30 k b a c a b d c

PAGE 48

34 For most experiments carried out in this dissertation, arrayed PSi chips were needed for multiple sample analyses. They were chips with an array of PSi spots surrounded by un-etched, crystalline silicon. It was achieved, as previously described, through either using a mask during Au-coating step in the metal-assisted etching, or including a mask in the optical system during the anodic etching. Although PSi made from both methods did not exhibit noticeable difference in DIOS-MS performance, metal-assisted etching became the routine method to produce arrayed PSi chips due to a few advantages. First, metal-assisted etching is more efficient. The sputter coating process is the most time-consuming step, which requires 20 min to purge the sputter chamber with Argon. However, the actual coating time is less than 1 min. Also, batch processing is possible, when a set of six crystalline silicon chips are Au-coated simultaneously, and etched one by one in the same etching solution. On average, the fabrication of one PSi chip using metal-assisted etching is ~10 min. On the other hand, anodic etching involves constant assembly and disassembly of the etching cell for each PSi chip production. The alignment of the optical components is tricky, while precise focusing of the patterning mask is crucial for reproducible light projection for PSi production. A light source is necessary in the anodic etching; however, the quality of the patterned PSi chip sometimes suffers from the reflection and diffraction of the light in the cell filled with the aqueous solution. Moreover, leaking of the etching solution from the etching cell may occur, thus prevent a smooth operation. In short, the anodic etching encounters more problems, and generally demands much longer time in PSi production.

PAGE 49

35 Second, the metal-assisted etching is safer; it renders less contact with the hazardous hydrofluoric acid (HF). In this method, the actual etching is carried out in a Teflon beaker, instead of a Teflon cell which required assembly and disassembly while still wet with HF. As mentioned above, leaking of etching solution can happen, and HF could damage the nearby equipments, or even harm the operator. In conclusion, the metal-assisted etching is a more suitable PSi fabrication method in the current laboratory setting, and is chosen to be the routine approach. Surface Evaluation of PSi After a preliminary examination of PSi using SEM, other techniques were also used to investigate the property of PSi surface fabricated with the metal-assisted etching. FT-IR First, PSi surface was subjected to FT-IR measurement to understand its chemical composition through the characteristic absorption pattern (Fig. 2-4). The analysis was carried out within one day of PSi fabrication. Carbon dioxide in the air, as well as residual water and organic solvent (mainly ethanol from the etching solution) on the surface account for certain major absorption peaks, e. g., 3800-3500 cm -1 ( O-H of H 2 O, O-H of ethanol); 2900 cm -1 ( C-C of ethanol); 2370 cm -1 ( C=O of CO 2 ); 1500 cm -1 ,1600 cm -1 ( O-H of H 2 O); 1400,1200,1080 cm -1 ( O-H of ethanol). Despite the interferences mentioned above, other absorptions are indicative to the PSi surface chemical composition. The band around 3000-3500 cm -1 is assigned to SiO-H stretch, and 1100-1000 cm -1 to Si-OH stretch. Although the fresh surface was suggested to be silicon hydride, 33 the Si-H stretch at 2080-2200 cm -1 was not observed, which is likely due to the rapid conversion from Si-H to silicon oxides (including Si-OH) in the

PAGE 50

36 air. 87 In addition, the presence of H 2 O 2 in the etching solution may also contribute to the observed Si-OH stretch. % Transmittance Wavenumber (cm 1 ) Figure 2-4. FT-IR spectrum of PSi Auger Electron Spectroscopy If IR measurement helped understand the chemical bands formed on PSi surface, Auger electron spectroscopy shed light on the elemental composition of the material. The technique involves bombarding the target surface with a finely focused electron beam. The energy transfer leads to the excitation of a core electron into a higher energy level. At this point, one of the possible relaxation modes is to emit an Auger electron, which is characteristic of the parent element. An energy spectrum of the detected electrons displays peaks indicative to the elements present. The measurements were conducted on AES Perkin-Elmer PHI 660 Scanning Auger Multiprobe (Physical Electronics division, Perkin-Elmer Co. Eden Prairie, MN) and one AES spectrum is shown in Fig. 2-5.

PAGE 51

37 Kinetic Energy (eV)dN(E)Min: -3490Max: 2025 50250 450 650 850 1050 1250 1450 1650 1850 2050 Si1Si2C1O1K1 Figure 2-5. Auger electron spectrum of PSi. Silicon was observed as expected. The presence of oxygen peak was again attributed to the rapid oxidation of the surface. Another source of oxygen may be the adsorption of carbohydrates from the atmosphere, and the observation of the carbon peak is in agreement with the postulation. One can argue that the residual solvent, namely ethanol, could be responsible for the carbon and oxygen peaks. This is unlikely since the PSi samples were stored in the vacuum chamber of 10 -9 torr for two days. In this measurement, no fluorine peak was detected, which corresponded to the proposed etching reaction described in Chapter 1.The final fluorine-containing product, H 2 SiF 6 , remains dissolved in the etching solution. The Auger electron spectrum shown here also contains a weak potassium peak; the ubiquitous ion is a common contamination and appears in many DIOS-MS spectra. SEM SEM has proved to be a powerful tool in PSi surface evaluation. It was first used to confirm the successful fabrication of PSi with different etching methods. In the course of this work, the technique was repeatedly carried out to ensure the quality of PSi produced.

PAGE 52

38 Overall, PSi with homogenous morphology was produced reproducibly. However, heterogeneity of PSi, reflected in three aspects discussed below, was observed during the screening. a b c d Figure 2-6. SEM images of two types of PSi morphology. All four images were acquired at 50 k magnification and from different PSi samples. Image a) and c) are top views, while b) and d) were cross-section views. Image a) and b) represent one type of morphology, and c) and d) another. First, PSi acquired different overall morphologies, even though the etching conditions remained unchanged. Shown in Fig. 2-6 are two types of PSi surfaces most often obtained following the same etching protocol. Fig. 2-6a and 2-6b represent the type of surface composed of tightly packed silicon columns. Another type of morphology, shown in Fig. 2-6c and 2-6d, is featured by interconnected silicon walls. Visual

PAGE 53

39 examination of the images indicated that the second type had higher void space content, consequently, a higher porosity. The diversity of the morphology observed here did not have an apparent impact on the DIOS-MS signal. We were able to continue using the same PSi fabrication procedure even though different morphology may rise. The development of different morphologies can be explain by the fact that more than one mechanism of pore initiation and pore propagation may exist, as discussed in Chapter 1. The randomness of PSi morphology is therefore attributed to different pore-formation pathways. In addition to morphology variations on different PSi spots/samples, close examination revealed that the same PSi spot exhibited heterogeneity. Shown in Fig. 2-7 is a PSi spot fabricated using the standard metal-assisting etching. Three regions are identified on the spots: the inner region, the middle ring, and the outer ring. The inner region possesses a morphology of larger pores, comparing to the other two regions. Overall, the porosity follows the order: inner ring region > outer ring region > middle ring region. This configuration is highly reproducible and the brighter line of the middle ring is sometimes visible with the human eye. The effect is not fully explained. It may associate with the uneven Au deposition pattern obtained during sputter coating prior to etching. It was noticed that in DIOS-MS analysis, better signals were usually obtained at the edge of the spot. Is it because that the morphology of less porosity at the edge promotes DIOS signal? It is unlikely, however, because PSi spot of overall less porosity also demonstrates stronger signal at the edge. Another possibility is the discriminated sample distribution where more analyte exists in the edge than in the center.

PAGE 54

40 b a c d Figure 2-7. Three regions of PSi spot fabricated with metal-assisted etching. Image a) shows the PSi spot; image b), c) and d) are collected in the inner ring, middle ring, and outer ring region, respectively. Another observation of heterogeneous morphology was the occasionally incomplete PSi formation. The image of the 3k magnification (Fig. 2-8a) reveals a “patched” PSi surface. At first, the patches were suspected to be dust particles or other solid contaminations. However, a close view at the 50k magnification (Fig. 2-8b) shows these patches were islands of less-processed porous silicon, areas which did not undergo the etching reaction to the same extend as the surrounding area. Again, these PSi surfaces were obtained following the same etching protocol. Such occasional deviation on PSi morphology seemed to be inevitable and unpredictable, based on the SEM images collected periodically. Possible contributing factors include wear of the Au source (an aluminum ring target with inch of Au layer upon purchasing) in the sputter coating machine, and fluctuation of the glow discharge

PAGE 55

41 current during Au sputter coating. Both could lead to the variation of the thickness of thmetal layer, which eventually affect the formation of PSi. e f PSi with incompletely etched areas. a) At 3 k magnification; b) at 50 k magnification. In sulightly different morphologies among different PSi spots s an aration in DIOS-MS Analysis In thed. InstrIOS-MS analysis was performed on a Bruker Daltonics (Billerica, MA) Reflex II MAL Figure 2-8. SEM image a b s o mmary, PSi may acquire s or within the same spot, even if the same manufacture procedure is carefully executed. Despite that, no dramatic impact on the DIOS-MS signal is caused by thevariation of the morphology. At the same time, the SEM technique is demonstrated aimportant tool to gain insight of PSi surfaces. Instrumentation and Sample Prep is section, important aspects of routine DIOS-MS analysis are describ umentation Laser D DI-TOF mass spectrometer retrofitted with delayed extraction. In the ion source region, energy was provided by a nitrogen laser operated at 337 nm with a 3 ns pulse duration. The adjustable repetition rate was set at 3 Hz. The laser (~200 J/pulse) wasguided and focused by an iris, and a series of mirrors and lenses. In addition, a neutral

PAGE 56

42 density filter was installed in the optical path to adjust the laser energy. The final laser energy irradiating the DIOS sample was estimated to be a few J/pulse at a laser spot siof ~200 m diameter. To achieve impro ze ved resolution, the ion source was equipped with a delayed extraclong y ith the ion source was a time-of-fight mass analyzer operated in either the liner the mass spectrometer for the DIOS analysis, a modification on the MAL steel tion module of three operational modes, short (100 ns), medium (200 ns) and (400 ns). In DIOS-MS analysis, a short delay time was employed, since the ions of interest had relatively low mass-to-charge values. A DIOS-MS spectrum was usuallobtained as a sum of 50 laser shots. Mass Analyzer Coupled w ear mode or the reflectron mode. Mostly analyses were performed in the linear mode for better sensitivities; however, the reflectron mode was called upon where bettresolution was required. On the other hand, DIOS-MS was more commonly operated at the positive ion detection mode; the negative ion mode was also investigated and has demonstrated unique characteristics. Target Plate To adapt DI sample plate was made, as shown in Fig. 2-9. The top of a regular stainlessMALDI plate was milled in the center to accommodate a PSi chip, which was attached to the target with a piece of double-sided tape. The target was loaded into the ion source following the same loading process as a MALDI target .

PAGE 57

43 Regular stainless steel MALDI plate DIOS plate with notched surface Figure 2-9. Modification of a MALDI plate into a DIOS plate Target insertion was realized by a series commands conducted on a manual control module. The main steps were: 1) activate the electronics and magnet that holds the stainless steel target; 2) pump the probe region to reach a pressure of ~ 10 -2 torr; 3) open the lock valve between the probe region and the ion source region, and insert the target; 4) deactivate the magnet and leave the target in the ion source region; 5) retreat the probe and close the lock valve. Vacuum System The vacuum system was divided by two pneumatic valves into three regions: probe region, ion source region and analyzer region. High vacuum regions, i.e., the ion source region and the analyzer region, were each sustained by turbo pumps backed by individual rough pumps. DIOS-MS analysis was performed at a pressure of ~10 -6 torr. Sample Preparation Analytes were prepared at ~ 0.1 mM if not otherwise stated. Aliquots of 0.2 or 0.5 L sample solutions were deposited on a 16-well array chip (Fig. 2-10). For the sample solutions that spread, usually in organic solvents, smaller volume (0.2 L) was used to avoid cross contamination caused by too much spread. On the other hand, sample solutions with high water content needed sufficient volume (0.5 L) to facilitate pipette spotting. For most analytes, a certain percentage of organic solvent was used to reduce the surface tension of the droplet, so that the analyte was carried into the pores. A typical

PAGE 58

44 solvent system was composed of 50% acetonitrile : 50 % deionized water : 0.1% trifluoro acetic acid (TFA). Still, the property of the sample and the goal of the analysis determined the solvent for DIOS sample preparation. For example, pentose analyses, described in following chapters, were carried out in pure aqueous solutions, and organometallic molecules were analyzed in pure organic solvents (chloroform or dichloromethane). Figure 2-10. Photograph of a PSi chip. It is worth noticing that the sample preparation is much easier in DIOS than in MALDI. Not to mention all the delicate MALDI sample preparation techniques developed through the years, even the most straightforward method, the dried-droplet method, involves selecting the right matrix, choosing a solvent system compatible for both the matrix and analyte, optimizing the matrix to analyte ratio, etc. On the contrary, the sample preparation step in DIOS-MS only requires dissolving samples in proper solvents, which is less time-consuming and less laborious. SEM Evaluation of PSi with Sample The appearance of the PSi surface deposited with sample was investigated. For samples at typical testing concentrations ( < 0.1 mM), it was expected that a thin layer of

PAGE 59

45 sample would form on a PSi surface and no apparent solid residue would be observed. It was shown to be valid by the SEM technique (Fig. 2-11). Shown in Figure 2-11a, b, c are, respectively, a PSi surface without sample, a PSi surface deposited with 200 pmol caffeine (1.0 mM, 0.2 M), and a PSi surface first deposited with 200 pmol caffeine and then cleaned by sonication. In this experiment, a concentration of 1.0 mM, 10-fold higher than the routine concentration, was intentionally used in an effort to observe possible residues on the surface. Even so, as shown in the images, no apparent solid residue was seen on the second surface. In fact, the appearances of the three PSi surface were almost identical. Figure 2-11. SEM images of PSi surfaces with or without samples. a) PSi without sample; 2) PSi deposited with 200 pmol (0.2 L, 1.0 mM) caffeine; 3) PSi deposited with 200 pmol and then cleaned by sonication.

PAGE 60

46 Since the surface area of these PSi surfaces is unknown, it is difficult to calculate the surface concentration. However, the SEM results indicate that a very thin layer of analyte molecules formed on the PSi surface without developing patches or chunks of solid residue. The observation is significant because it illustrates the importance of the high surface area of PSi. The laser radiation is therefore able to penetrate through the thin sample layer and impinge upon the PSi, which has a high absorption coefficient and low heat conductivity. 68 An effective local temperature increase is maintained and facilitates the ionization/desorption of the analyte of a presumably low surface concentration. Applicability of DIOS-MS DIOS-MS has been used for a variety of analytes, as described in Chapter 1. However, previous research activities were mostly carried out using PSi surfaces prepared by anodic etching. In this section, PSi processed using metal-assisted etching is examined in DIOS-MS analyses of different types of samples, as well as different mass spectrometer operating modes. Individual Analytes Individual analytes were first tested. DIOS Mass spectra of a simple organic molecule, an enzyme substrate and a peptide are shown in Fig. 2-12. Their chemical information is listed in Table 2-1. Table 2-1. Molecular weights and structures of three tested molecules. Compound Name and empirical formula Structure m/z of [M+H] + (average, Da) a 2,9-dimethyl-1,10-phenanthroline C 14 H 12 N 2 NNCH3CH3 209.3

PAGE 61

47 Table 2-1. Continued. b N-benzoyl-L-phenylalanyl-L-valyl-L-arginine-4-nitroanilide (hydrochloride) C 33 H 40 N 8 O 6 (HCl) 645.7 c bradykinin fragment 1-7 C 35 H 52 N 10 O 9 Arg-Pro-Pro-Gly-Phe-Ser-Pro 757.9 Figure 2-12. DIOS mass spectra of three individual analytes. a) A small organic molecule; b) an enzyme substrate; and c) a peptide. Clean mass spectra were obtained for the three listed molecules with mass accuracy of ~ 0.1% for [M+H] + . The experiments indicate the matrix-free advantage of DIOS-MS. Without interfering matrix peaks, indicative analyte peaks are identified unambiguously.

PAGE 62

48 Moreover, the PSi of the metal-assisted etching demonstrates similar capability in DIOS-MS analysis as the more commonly used surfaces processed through the anodic etching. Mixtures After the successful studies of individual compounds, four mixtures were applied to DIOS-MS to investigate its potential for identification of individual species in mixture (Fig. 2-13). The components of the mixtures are listed in Table 2-2. The first three mixtures are composed of compounds of the same type including drug molecules, peptides, fluorine containing standards, respectively. The fourth sample is a mixture of compounds from different categories. For each mixture, the number of components varies. Table 2-2. Components of four mixtures. Mixture Component Empirical Formula of Component [M+H] + avg Amount per Spot (pmol) Pseudocphedrine HCl C 10 H 15 NO HCl 166.2 2.0 Procainamide HCl C 13 H 21 N 3 HCl 236.3 1.5 Amitriptyline HCl C 20 H 23 N HCl 278.4 2.0 Verapamil HCl C 27 H 38 N 2 O 4 HCl 455.6 0.5 a Reserpine C 33 H 40 N 2 O 9 609.7 1.0 Angiotensin I fragment 11-14 C 21 H 35 N 7 O 6 482.6 30 angiontensin I fragment 1-7 C 41 H 62 N 12 O 11 900.0 30 ACTH fragment 4-10 C 44 H 59 N 13 O 10 S 963.1 30 Angiontensin I C 62 H 89 N 17 O 14 1297.5 30 ACTH fragment 1-16 C 89 H 133 N 25 O 22 S 1938.2 15 b ACTH fragment 1-39 C 210 H 315 N 57 O 57 S 4583.2 6.0 HP-0621 C 12 H 18 O 6 N 3 P 3 F 12 622.2 7.5 HP-0921 C 12 H 18 O 6 N 3 P 3 F 24 922.2 10 HP-1251 C 30 H 18 O 6 N 3 P 3 F 48 1522.3 7.5 HP-2121 C 42 H 18 O 6 N 3 P 3 F 72 2122.4 11 c HP-2721 C 54 H 18 O 6 N 3 P 3 F 96 2722.5 13.5 Arginine C 6 H 14 N 4 O 2 175.2 50 Reserpine C 33 H 40 N 2 O 9 609.7 30 d Angiotensin II C 50 H 72 N 13 O 12 1297.5 10 Mixture a included five drug molecules with molecular weights lower than 700 Da. All five singly-charged protonated molecule peaks ([M+H] + ) were observed, and the background signals were minimal (Fig. 2-13a). Since most of the drug molecules are

PAGE 63

49 organic compounds with molecular weights in the lower mass region, and the majority of them have basic function sites for proton attachment, DIOS-MS is ideal for such molecules. In this example, signal suppression among components was not a limiting factor, and simultaneous measurement of multiple samples with DIOS-MS was possible. Although DIOS-MS was capable of providing molecular weight information for all the components in mixture a, two peptides in mixture b were not observed (Fig. 2-13 b). They were angiotensin I fragment 11-14 and adrenocorticotrophic hormone (ACTH) fragment 4-10. Mixture b was a custom-made peptide mixture to mimic protein digests. It was not entirely surprising that a few compounds were “missing”, which is also common in both ESI-MS and MALDI-MS for protein digest analysis. Figure 2-13. DIOS mass spectra of four mixtures. a) Mixture of five drug molecules in methanol : water = 50%: 50%; b) mixture of six peptides in acetonitrile : water : TFA = 50% : 50% : 0.1%; c) HP standard mixture in acetonitrile : water = 95% : 5%; d) mixture of an amino acid, and a drug molecule and a peptide in acetonitrile : water : TFA = 50% : 50% : 0.1%.

PAGE 64

50 If this example exhibited the limited strength of DIOS-MS in proteomics, the technique still possesses an inherent advantage due to its “matrix-free” attribute in analyzing peptide samples: the ease in coupling to separation techniques. The fact that no matrix addition step is required may place DIOS in a more competitive position compared to MALDI in certain situations. It is also encouraging to see that ACTH, a molecule of 39 amino acids, generated a signal of considerable intensity. Successful DIOS-MS analyses of compounds larger than 3000 Da have been rarely reported; yet, it is important that DIOS-MS is not limited to the suggested mass range (< 3000 Da). 66 The third mixture (mixture c) tested was a commercial tuning mix for ESI-MS from Agilent (Agilent Technologies, Inc., Palo Alto, CA), often referred as the “HP mix”, for “HP” is part of the nomenclature of the components. The compounds in the HP mix were fluorinated hydrocarbon derivatives. Five components of relatively higher concentrations are list in Table 2. In this dissertation, the HP mix was often used as a mass calibration standard for analytes of higher molecular weight (> 1000 Da). On the other hand, to achieve mass calibration within 1200 Da, mixture d was constructed. Notice the first three mixtures consisted compounds of the same type, mixture d was purposely made from three compounds distinct in their structure: a single amino acid, a drug molecule and a peptide of eight amino acid residue. The mass spectrum (Fig. 2-13d) presented all three protonated molecules, indicating simultaneous analysis of samples of different type was possible.

PAGE 65

51 Polymer Analysis To test the applicability of DIOS-MS for polymer analysis, experiment was performed on a polymer sample (Fig. 2-14). It was a commercial NaRbCsI/polyethylene glycol (PEG) solution composed of 0.25 mg/mL PEG1000, 0.2 mg/mL NaI, 0.2 mg/mL RbI, and 0.05 mg/mL CsI. The addition of the cations was needed to facilitate the ionization of PEG, and their individual concentrations were not further optimized. Figure 2-14. DIOS mass spectra of a polymer sample. a) DIOS mass spectrum of NaRbCsI/PEG; b) expanded mass spectrum between 460-1300 m/z. In the DIOS mass spectrum, most peaks below 400 Da are cations and cation iodine adducts (Fig. 2-14a). For signals above 400 Da, there are two dominant series of polymer peaks (Fig. 2-14b), both separated by 44 Da, the mass of the repetition unit of PEG (-CH 2 CHOH-). The first series (482, 526, 470, etc.) is identified as sodium adducts of PEG

PAGE 66

52 (150+23+44n), and the second series (500, 544, 588, etc) corresponds to rubidium adducts (150+85.5+44n). DIOS-MS was previously applied to forensics by analyzing polymer samples collected at crime scenes. 72 The successful analysis of the polymer sample in our lab demonstrates the application is not limited by the PSi manufacture method or the specific mass spectrometer used. Since polymer MALDI-MS often encounters difficulties in matrix/analyte co-crystallization, polymer DIOS-MS may become significant in overcoming such problems. Reflectron Mode Resolution is important in any analytical measurements, and mass spectrometry is no exception. For the instrument used for DIOS-MS analysis, resolution was improved by operating in the reflectron mode. Figure 2-15. DIOS mass spectra of HP mix in the reflectron mode. Inserted were zoom-in views of individual peaks with unit mass resolution.

PAGE 67

53 Described in Chapter 1, a reflectron is a series of ion mirrors that reduce the kinetic energy distribution of isobaric ions. Shown in Fig. 2-15, isotopic distribution of analyte peaks are observed for the HP mix. In this mode, sensitivity is often sacrificed due to framentation. However, it is sometimes important to obtain improved mass resolution, when unit mass resolution is required for certain applications. Negative Mode The possibility of conducting DIOS-MS in the negative ion mode was also investigated. Shown in Fig. 2-16 are negative DIOS mass spectra of three samples: succinic acid, NaRbCsI/PEG, and HP mix. The analyses of these samples or similar analytes have not been reported. The first sample is a simple organic acid in a basic solution. The DIOS-MS signal is believed to be the result of desorption of the deprotonated negative ion: HOOCCH 2 CH 2 COO . While most literature focused on the analysis of drug molecules, peptides and proteins in DIOS-MS research, little has been done on simple organic molecules. The significance of the negative mode experiment exists in the potential applications in detecting simple organic acids of biological importance. For instance, the geometric distributions of such species, e.g., succinic acid in biofilms 88 or picolinic acid in bacterial samples, 89 can be monitored using DIOS imaging mass spectrometry. 90 Compared to MALDI-MS in which the addition of matrix may disturb the original distribution of the analyte, DIOS-MS gains information on the intact geometry without the matrix intrusion.

PAGE 68

54 Figure 2-16. Negative DIOS mass spectra of three samples. a) Succinic acid (5 mM in 5 mM KOH); b) NaRbCsI/PEG (m/z of 127, 277, 339, 427, 489 were attributed to I , [NaI+I] , [RbI+I] , [2NaI+I] , [RbI+CsI+I] ; and c) the HP mix. The second spectrum (Fig. 2-16b) was obtained from the NaRbCsI/PEG mixture. Interestingly, not a single polymer related peak was observed. Instead, signals of negatively charged salt adducts appeared in the spectrum with very good S/N. This result, along with detection of cation adduct species of NaRbCsI/PEG in the positive mode, will allow the qualitative or quantitative analysis of inorganic compounds, providing an additional approach to inorganic mass spectrometry. The HP mix is a commercially available mass calibration standard for ESI-MS in both the positive and negative modes. It has been evaluated in the positive DIOS-MS mode, and the analysis in the negative mode was also conducted (Fig 2-16c). A cleaner

PAGE 69

55 mass spectrum was obtained with improved S/N in the negative mode than in the positive mode. Most observed peaks matched listed masses (m/z 1306, 1806, 2306, 2516), whose structures are not known. These three examples have demonstrated the viability of negative DIOS-MS. Moreover, a general observation was that ion peaks in negative DIOS mass spectra had better S/N. It has been suggested that at times DIOS-MS exhibited better performance in the negative mode, 86 although no explicit data were reported. In this research, although no extensive investigation was carried out in the negative mode of DIOS-MS, the following two examples demonstrated the negative detection mode could generate more indicative ion signals than the positive detection mode (Fig. 2-17). Both compounds are caffeine structure analogues, and the chemical structure of theophylline is shown in Fig. 2-17b. In the positive mode, the DIOS mass spectrum for theophylline showed peaks of m/z 181, 203 and 219, corresponding to [M+H] + , [M+Na] +, and [M+K] + . The presence of cation adducts complicated the spectrum, and reduced the signal intensity of the desired [M+H] + peak. However, the compound displayed a single peak ([M] ) of 180 in the negative DIOS mode, which is preferred over a multiple peaks for easier spectrum interpretation. As for the Br-containing caffeine analogue, all analyte related peaks were suppressed by the background signals in the positive mode, and the acquisition of molecular weight information was impossible. Yet, when the analysis was conducted in the negative mode, a doublet of the molecular ion at m/z 258, 260 representing the Br isotopic pattern was observed.

PAGE 70

56 Figure 2-17. DIOS mass spectra of two compounds in both the positive mode and negative mode: a) and b) were theophylline (C 7 H 8 N 4 O 2 , m.w. 180, structure shown); c) and d) were a Br-containing caffeine structure analogue (C 7 H 7 N 4 O 2 Br, m.w. 259.1, structure unknown) The improved signal in negative DIOS-MS may result from the fact that desorption/ionization of the background species is discriminated in the negative mode. 86 The origin of the background peaks (shown in Fig. 2-17c) is still under debate; the proposed sources include hydrocarbon adsorbed from air, contaminants leached from storage containers, and Si ion clusters ablated from the porous surface. 86 Effort was taken in this research to reveal the identity of the background peaks. However, the limited resolution of the current mass spectrometer hindered the exact mass measurements of the ions of interest, thus their identities remain unknown.

PAGE 71

57 In negative DIOS-MS, molecular ions took the form of [M] . The hypothesis proposed here is that upon laser ablation, a highly energetic plume is formed due to a rapid local temperature rise similar to a MALDI process. 26 Electrons, along with neutrals and ions, are present in the plume. As part of the gas-phase interactions, electron capture by M resulted the formation of [M] ions which are extracted into the mass analyzer and detected. Although negative DIOS-MS has not been a main research focus, it has the potential to provide complimentary information to positive DIOS-MS, and has proven to be superior in certain cases in providing molecular weight information. Moreover, it may help understand the micro-environment of the desorption/ionization, as well as the mechanism of the technique. Limitations of DIOS-MS In the previous section, it was demonstrated that DIOS-MS is a versatile platform for mass spectrometry analyses with different types of analytes and different operating modes. DIOS, because of its “matrix-free” feature, is considered advantageous compared to MALDI, a closely related ionization technique. However, the inherent limitations of DIOS might account for the fact that MALDI is a more widely used ionization technique in general. Background and Contaminants Although there is no matrix initially introduced to samples, the presence of background ions can still be a problem (Fig. 2-17c). The background signals interfere with low-mass ions, preventing an unambiguous identification of an analyte. One way to overcome this problem is to increase the analyte concentration to achieve the optimized spectrum quality.

PAGE 72

58 As mentioned earlier, the possible background may be caused by hydrocarbon adsorbed from air, contaminants leached from storage containers, and Si ion clusters ablated from the porous surface. 86 Extended storage of PSi usually resulted in increased background signals. Often referred as “aging”, this phenomenon will be elaborated in following sections. In addition to above-mentioned background species, DIOS is prone to contamination introduced by the operator. Extreme cautions should be taken while working with DIOS chips to prevent any contact with any potential sources of contamination. Gloves are required for handling the chips; containers and tools (e.g., Teflon -coated tweezers) are stored in the dedicated hood. Even so, undesired signals from contaminates may still appear in a DIOS mass spectrum. An often encountered incidence is the observation of ions around 500 Da (Fig. 2-18), which are identified as keratins originated from human hair or skin. DIOS-MS has a very high response to these contaminants. In fact, a way to test the viability of a commercial DIOS chip is to look for these peaks after contacting the PSi surface with the finger. 91 Figure 2-18. DIOS mass spectrum of angiotensin II in the presence of keratins. Peaks around 500 Da are assigned as keratin contaminants.

PAGE 73

59 Reduced Response to Analytes with Certain Structures Under our current laboratory setting, the whole protein detection with DIOS-MS was not successful, and the largest molecule observed was ~ 4500 Da. The limited mass range hindered the analysis of larger molecules. Moreover, reduced response to analytes of certain structures further reduced the number of suitable samples for DIOS-MS. Figure 2-19. MALDI and DIOS mass spectra of caffeine and cortisol. a) Caffeine, m.w. 194; and b) cortisol, m.w 362 in DHB. Inserted are DIOS mass spectra. The chemical structures of the two molecules are also shown. First, it is was noticed that several analytes detectable with MALDI-MS were less successful with DIOS-MS analysis. Despite the matrix ions present in the lower mass

PAGE 74

60 region, protonated molecules for caffeine and cortisol were both identified in MALDI-MS (Fig. 2-19 a and b). However, for the DIOS-MS analyses of caffeine and cortisol, either the [M+H] + ion was completed suppressed by background ions (Fig. 2-19a, inserted), or the cation adducts were detected along with [M+H] + (Fig. 2-19b, inserted). The low aqueous basicity of caffeine (pK b ~ 14) may account for the DIOS-MS results. As discussed in Chapter 1, aqueous phase basicity was suggested to play an important role, and the desorption of preformed ions was believed to be the major passway of gas phase ion formation. 68 Accordingly, compounds with high pK b value generate the protonated molecules at much lower yields in the acid/base equilibrium in water. As a result, fewer desired analyte ions (protonated molecules) will be promoted to gas phase during the thermal process of desorption. Interestingly, caffeine was reportedly observed with DIOS-MS. 27 However, we believe that the peak attributed to caffeine was in fact a fragment of reserpine, one of the two compounds included with caffeine as the sample solution. Instead of the theoretical m/z of the protonated caffeine (195), the controversial peak had a m/z of 196, which corresponded to the value of the most intense fragment in reserpine’s EI mass spectrum in National Institute of Standards and Technology (NIST) database. Also, in the same reported mass spectrum, another peak noted by the authors as anonymous matched the second major fragment of reserpine. This further confirmed that fragmentation of reserpine did occur, and the assignment of caffeine was inaccurate. Reduced Response to Peptides Another disadvantage of DIOS was the reduced sensitivity for peptides compared to MALDI. Fig. 2-20 shows a DIOS mass spectrum and a MALDI mass spectrum of the same peptide mixture, its six peptide component described previously in Table 2-2. The

PAGE 75

61 sample had been stored in the refrigerator for several months, and began to degrade due to decomposition of the peptides, incorporation of salts and adsorption of the peptides onto the container wall. Even so, using MALDI, we were able to observe five peptides with considerable S/N. On the contrary, only three peptides were detected using DIOS-MS with decreased spectrum quality. Figure 2-20. MALDI and DIOS mass spectra of a peptide mixture. a) DIOS mass spectrum; and b) MALDI mass spectrum obtained with the same peptide mixture using the same mass spectrometer. In light of this, individual peptides with a range of concentrations were tested to investigate the detection limit of DIOS-MS for representative compounds. Two analytes, bradykinin and angiotensin II, at various amount (200 fmol, 2 pmol, 20 pmol) were detected with DIOS-MS. Except for the adjustment on laser power, other

PAGE 76

62 detection parameters were kept the same. As the amount decreased (Fig. 2-21), the S/N of the analyte signal decreased as expected; at the same time, the background noise increased due to the higher laser power required for smaller sample deposition. In this research, the minimum analyte amount observed was ~5 fmol for angiotensin II. The overwhelming background posed difficulties in detecting small amount of analytes with molecular weight less than 700 Da. Figure 2-21. DIOS mass spectra of bradykinin and angiotensin II of different amount. a) Bradykinin (average m.w. 757.9), and b) angiotensin II (average m.w. 1047.2) at the amount of 20 pmol, 2 pmol, and 200 fmol. In DIOS-MS, similar to any laser desorption technique, quantitative analysis is difficult. Construction of a standard calibration curve is problematic due to the limited dynamic range and poor reproducibility. Shot-to-shot fluctuations, non-linearity of the detector response, as well as variations in gas-phase ion interactions, are the main causes.

PAGE 77

63 For successful quantitative DIOS analysis, incorporating certain form of internal standard is very important. We have carried out a semi-quantitative analysis using a pseudo-internal standard, and the experiment is elaborated in Chapter 3. Aging and Re-etching After investigation of the applicabilities and the limitations of the DIOS-MS, important aspects of DIOS analysis, aging and re-etching of the PSi chip,are addressed. Concept of Aging The aging of PSi, initially recognized as a process of “slow oxidation”, was first noted in 1965 by Beckmann. 92 Later, the adsorption of contaminants from the atmosphere was identified as another consequence. 93 Currently, aging of PSi is used to describe the slow conversion of the freshly etched surface to a contaminated native oxide layer over time. The phenomenon has received detailed study, and the particular attention is given to how photoluminescence of the material evolved as a result. Similarly, the impact of aging on DIOS-MS analysis has been noticed, 85 and decreased performance was speculated to be caused by absorbed contaminants, and the changed physical properties of PSi surface. Recently, a process referred to as “double-etching”, 72 in which the fresh PSi is oxidized with ozone and etched again in a HF solution, is adopted to minimize the deterioration of the surface. However, no detailed results were published to describe the effect of aging and double etching on DIOS. Here, the aging phenomenon was investigated for PSi prepared by metal-assisting etching. A treatment similar to “double etching” was attempted to counteract the aging effect. DIOS mass spectra, as well as SEM images, were obtained on the different surfaces of interest. In addition, contact angles were collected for freshly etched, aged, and re-etched surfaces.

PAGE 78

64 DIOS-MS on Aged and Re-etched PSi In this experiment, the PSi chips were stored in sealed plastic containers after fabrication for 1 to 14 days. The aged chips were immersed in the same solution used in the first etching (49% HF : 30% H 2 O 2 : ethonal = 1:1:1) for 20 s in an attempt to restore the same chemical structure of the first etching. The duration of re-etching was an empirical decision based on the SEM results discussed next. Longer durations were avoided because of the excessive removal of the porous layer. DIOS-MS analysis for three sets of experiments with two peptides and a peptide mixture are shown in Fig. 2-22. The top spectra in each set were collected from fresh PSi surfaces, the middle spectra from aged surfaces (aging durations denoted in the figure) and, the bottom spectra from re-etched surfaces.

PAGE 79

65 Figure 2-22. DIOS mass spectra of angiotensin II, bradykinin, and a peptide mixture. a)-c) Angiotensin II; d)-f) bradykinin; and g)-i) a peptide mixture. Spectra of a), d) and g) were collected from freshly etched PSi, b) and g) from PSi aged for 24 hours, h) from PSi aged for 14 days, and c), f) and I) from re-etched PSi surfaces.

PAGE 80

66 The DIOS-MS results suggested additional background peaks appeared after one day of storage. The background peaks were not entirely reproducible, while longer storage time led to increased peak intensities, as well as additional background peaks. Daily variation of the air-borne contaminants might have contributed to the different background profiles. Also indicated here, re-etching was able to remove the background peaks to a certain extent or even completely. However, it was achieved at the loss of the overall S/N and the appearance of anonymous peaks in the molecular ion region. Briefly, the overall spectrum quality was degraded as a result of re-etching. The PSi chips used for this work were therefore either freshly etched or aged in air without further re-etching. SEM of Aged and Re-etched Surfaces To gain insight into the effect of re-etching, SEM images of the surfaces re-etched for different durations were acquired. Shown in Fig. 2-23 are an aged (a) and three re-etched PSi surfaces (b-d). The aged surface does not display dramatic morphological change compared to the fresh surface. In fact, the surface oxidation or the adsorption of environmental contaminants was not expected to cause any surface morphology change detectable by SEM. On the other hand, the aged surface after re-etching treatment showed apparent differences. The porous feature gradually disappeared when the top layers “dissolved” during the HF treatment. The morphology change of the re-etched surfaces is used to explain the reduced background and deteriorated DIOS-MS performance. During re-etching, the porous structure loses the exposed top portion, along with the oxidized layers and trapped contaminants. Even though the background ions are reduced, the surface turns into a

PAGE 81

67 “flatter” morphology, and this “less-porous” surface is less effective in assisting desorption/ionization in DIOS-MS. a b d c Figure 2-23. SEM images of aged and re-etched PSi surfaces. a) Aged PSi for 24 h; b), c), and d) PSi surfaces re-etched for 1 min, 4 min and 2 hr, respectively. Contact Angle Measurements Wettability, one important feature that affects the interactions between solutions and surfaces, was investigated on surfaces of crystalline silicon, freshly etched, aged and re-etched PSi. The concept of “wettability” differs from hydrophobicity in a way that the former is determined not only by the chemical characteristics, but also the morphology. As early as in 1930’s94 and 1940’s,95 results have been published relating the surface roughness to

PAGE 82

68 water wettability, which became the theoretical foundation for the development of the water resisting materials. In DIOS-MS researc h, PSi sometimes was recognized as hydrophobic based on the obserat ropriate solvet angle, defined as the anM-PLUS series contact angle meter (Tantec Inc. S vation that aqueous droplets beaded up on the surface, and the phenomenon was attributed to the formation of silicon hydrite (Si-H) terminal. 86 However, we believe ththe apparent “increased hydrophobicity” of PSi compared to crystalline silicon is a combined effect of the surface chemical composition, and roughened surface. Understanding the wettability of the surface will help determine the app nt or solvent system. Solutions of high organic solvent concentration, thus, low surface tension, tend to spread more on a surface. On the contrary, aqueous solutions usually form round droplets on the same surface. In DIOS-MS, proper organic/water ratios for the solvent are important for the analysis, since the ultimate goal of the solvent is to carry the analyte to the whole PSi surface area and into the pores not too much (spreading), or too little (beading up and drying into thick residue). One way to learn about the wettability is to measure the contac gle between a substrate surface and the tangent line at the point of contact of the liquid droplet liquid/solid/gas interface. When water is used as a testing liquid, larger contact angle indicates lower wettability. The experiments were done on a CA chaumburg, IL), and on four different surfaces: crystalline silicon, freshly etched PSi, aged PSi, and re-etched PSi, using deionized water as test liquid (Fig. 2-24). The averages and standard deviations were calculated from 10 measurements.

PAGE 83

69 contact angles of various surfaces389275120020406080100120140c-Sifreshly etchedPSiaged PSi in airre-etched PSicontact angle (degree) Figure 2-24. Water contact angles on four different surfaces. The results above show that all PSi surfaces have lower wettability than crystalline silicon. The PSi surface becomes more wettable when aged, and the wettability decreases after re-etching. This is explained by the fact that the hydrophilicity of the PSi surface increased due to the formation of the oxidized layer and the adsorbed contaminants. Re-etching removes such effect. However, there is no satisfying explanation for the fact that the re-etched surface exhibited even lower wettability than the freshly etched PSi. The knowledge gained in the surface wettability helped choose the proper solvent in the experimental design. A certain amount of organic solvent is often preferred in DIOS-MS to allow the distribution of the analyte into the pores. When only pure aqueous solutions are examined, the usage of aged PSi chips is less restricted and sometimes preferred, due to its improved wettability toward water. Effect of Acid Content on DIOS-MS Signal For peptide samples, DIOS-MS sometimes suffers from low sensitivity and salt tolerance, which may limit its application in proteomics. To improve DIOS-MS signals for peptide samples, various efforts were attempted, including incorporation of

PAGE 84

70 ammonium citrate as the salt sequestering compound, coating the PSi surface with gold nano-particles as additional energy absorbing media, and examination of conductive tape and non-conductive tape. Mixed results were achieved and no significant enhancements in spectrum quality were gained for all above mentioned approaches. Despite that, the content of acid modifier was shown to have an impact on DIOS-MS signals of peptide samples. An angiotensin solution stored up to 18 months was analyzed in the presence of different amounts of acid modifier: trifluoroacetic acid (TFA). As a result of the extended storage, cation adducts formed due to the inclusion of salts. The aged solution was intentionally used to mimic biological samples with ubiquitous salt content. When there was no TFA added (Fig. 2-25a), only cation adduct ions of angiotensin I were observed, while [M+H] + was not detected as all. After introducing increasing amounts of acid (protons) to the final solution, the re-established equilibrium led to an increased concentration of protonated molecules as preformed ions, and might have also favored the formation of such ions in gas phase ion interactions. Consequently, the protonated molecule was observed as a major signal when 0.1% TFA was used (Fig. 2-25b). However, more acid modifier did not necessarily mean an over enhanced signal. As shown in Fig. 2-25c, d, the protonated molecule signal actually decreased at higher TFA concentration (0.5% and 1%); at the same time, contaminant peaks (keratins) appeared. Similar results were obtained on parallel experiments of a peptide mixture. The optimum concentration of TFA acid modifier was about ~0.1-0.2%, with a corresponding pH of ~2.7-3.0.

PAGE 85

71 Figure 2-25. Effect of TFA concentration on DIOS-MS signal. Mass spectra a, b, c, and d correspond to TFA concentration of 0, 0.1%, 0.5%, and 1%. Optimal TFA concentration is ~0.1% Conclusions In this chapter, the two PSi production methods, anodic etching and metal-assisted etching, were described in detail. Despite the fact that anodic etching has been the major approach for PSi production in DIOS-MS research, metal-assisted etching, being safer and more efficient, became the routine method. SEM images illustrated PSi morphology featured by ~30-60 nm diameter pores. The PSi surface was evaluated with FT-IR, Auger electron spectroscopy (AES) and SEM. FT-IR revealed the formation of the Si-OH bond due to the presence of hydrogen peroxide in the etching solution or rapid post-etching oxidation. The AES results

PAGE 86

72 confirmed the presence of the oxygen, which could originate from the oxidation and/or the adsorption of air-borne hydrocarbon contaminants. To learn more about the physical characteristics of PSi, SEM was employed again and heterogeneous morphologies on different PSi spots and within the same spots were observed. These results were explained by the randomness of the pore formation and the uneven Au deposition during metal-assisted etching. No obvious impact on the DIOS-MS signal was resulted from the morphology variations. After the production and evaluation of the PSi substrate, instrumentation and other topics related to DIOS-MS analysis were discussed. The laser, mass analyzer, sample target, and vacuum system were the relevant components of the instrument system used in current DIOS-MS studies. One important advantage of this technique was simple sample preparation, and the routine procedure for sample preparation was described. The appearance of PSi with deposited samples was evaluated with SEM. A thin sample layer was believed to be crucial for DIOS-MS. Experiments were then designed to explore the applicability of the technique. Individual analytes, mixtures (drug molecules, peptides, fluorine containing standards, and compounds of different types), and a polymer sample were evaluated. In addition, the reflectron mode and the negative mode of DIOS-MS were carried out to survey the versatility of the DIOS-MS. At the same time, limitations of the technique were also discussed. The features of DIOS-MS are summarized as: 1) DIOS-MS is a suitable tool for a variety of compounds at picomole level, with ease of sample preparation and low chemical background. 2) The ability to obtain mass spectra in the reflectron and the

PAGE 87

73 negative ion modes extends the range of DIOS-MS applications. Particularly, the negative mode is promising with low chemical background which made the “matrix-free” feature more significant. 3) The technique still suffers background (especially at higher laser power and for aged surfaces), which may be caused by desorption of adsorbed air-borne contaminants, or Si clusters ablated upon laser radiation. 4) Compared to MALDI-MS, the technique has reduced response to peptides and molecules with low aqueous basicity. Despite the drawbacks of the DIOS-MS, unique applications are designed to take advantage of the “matrix-free” sample preparation as demonstrated for carbohydrate and organometalics analyses in following chapters. This chapter also addressed an important aspect of DIOS-MS, aging and re-etching of the PSi chip. Re-etching was able to remove the background peaks to a certain extent or even completely, but at a price of the decreased S/N. As a result, the PSi chips used for this work were either freshly etched or aged in air without further re-etching. Moreover, results of contact angle measurement served as a guideline for the choice of solvent systems. Finally, to improve DIOS-MS signals for peptide samples, efforts were attempted and the content of acid modifier was shown to have an effect on DIOS signal. A certain amount of TFA enhanced the spectrum quality (~0.1%); however, desired signals deteriorated when too much acid modifier were added (~0.5-1%).

PAGE 88

CHAPTER 3 DIOS-MS STUDIES ON PENTOSE-BORATE COMPLEXES Introduction DIOS-MS is a surface-enhanced laser desorption/ionization technique featured with simple sample preparation where no additional matrix compounds are needed. DIOS can be readily employed in mass spectrometers with a MALDI ion source, and the system is capable of high throughput analysis. The technique has found applications in many areas, and a wide range of analytes has been investigated by DIOS-MS, including peptides, drug molecules, protein digests, carbohydrates 33;66;73;83;85 Researchers have also explored the technique under atmospheric pressures, 81;82 using an infrared laser, 77;79 as well as in areas of forensics, 71;72 polymer, 71;96 quantitative analysis, 76 and automated enzymatic activity assays. 75 Most recently, surface modifications on porous silicon introduced new research opportunities, 97;98 underlined by the detection of ~480 molecules. 84 In this chapter, DIOS-MS was used as an ionization technique for simple carbohydrate molecules in both the positive and negative modes. Negatively charged pentose-borate complexes of low molecular weights were successfully observed in a parallel, high throughput fashion. The results were used to determine, for the first time by mass spectrometry, the relative binding affinities among pentose-borate complexes. An important question in the evolution of ribonucleic acids (RNAs) is why ribose is favored over other pentose isomers. It has been shown that borate ion, present in tourmalines and evaporites in deserts, plays an important role in the synthesis of pentoses (5-carbon monosaccharides) from simple organic precursors: formaldehyde and 74

PAGE 89

75 glycolaldehyde in the laboratory. 99 The borate ion also stabilizes pentoses through the formation of boron-diol complexes, which prevents them from undergoing “browning” degradation. 100 Verchere and Hlaibi presented the first comprehensive analysis that considered the effect of boron in the conformational equilibrium of carbohydrates, and later these authors included the effect in the calculation of the association constants by using the combination of potentiometric titration and 11 B and 13 C NMR spectroscopy. 101;102 While boron has been recognized for its importance in pentose synthesis and stabilization, explicit structural interactions need to be investigated. Could boron also account for the fact that ribose, among the four isomers (ribose, xylose, arabinose, lyxose), is the favored building block for RNA? Does boron have a higher binding affinity to ribose than to the other isomeric pentoses, so that ribose is stabilized and enriched to be available for RNA synthesis? These questions were approached with DIOS-MS. This application illustrates the “matrix-free” advantage of DIOS-MS, i.e., no matrix compounds are added. The absence of the matrix not only eases the background problems present in MALDI-MS, but also allows the analysis free of the chemical environments introduced by the matrix, such as low pH. The unique features suggest DIOS-MS as a promising technique in applications where MALDI-MS is not compatible with the experimental conditions, or the analyte ions are in the low mass range. Experimental Methods Chemicals and Reagents All reagents were purchased from Sigma-Aldrich Co. (St. Louis, MO) if not otherwise stated. 13 C 5 -ribose was purchased from Omicron Biochemicals Inc. (South

PAGE 90

76 Bend, IN). All solutions were aqueous, and deionized water was used. Crystalline silicon wafers were purchased from Silicon Sense Inc. (Nashua, NH) Detection of [Pentose+K] + Adduct Ions The protocol of PSi chip production and the parameters of mass spectrometry analysis were described previously. A potassium acetate (KAc) master solution (100 mM) was diluted into a series of solutions of 0.5, 1.0, 2.0, 10, 50 mM. Pentose solutions (10 mM) were mixed with the KAc solutions at equal volume respectively to achieve molar ratios of pentose: K + = 20:1, 10:1, 5:1, 1:1, 1:5, 1:10. The five pentose isomers tested were D-ribose, L-arabinose, L-xylose, D-xylose, and L-lyxose. Detection of Pentose-Borate Bomplexes Each pentose isomer (10 mM, 20 L) was mixed separately with a sodium borate solution (Na 2 B 4 O 7 , 10 mM, 2.5 L). The signal at m/z 307 was monitored in the negative mode. Further dilutions were made to determine the optimum analyte concentration. Competition Experiments between 1,4-Anhydroerythritol and Each Pentose Isomers 1,4-Anhydroerythritol (AHE, 0.1 M, 100 L) and the corresponding pentose isomer (0.1 M, 100 L) were mixed in an Eppendorf tube containing deionized water (700 L) and vortexed thoroughly. A sodium borate solution (Na 2 B 4 O 7 , 0.025 M, 100 L) was added, and the resulting mixture was vortexed and equilibrated for 2 h at room temperature. An aliquot of 0.5 L was spotted on the DIOS plate, and allowed to dry before DIOS-MS analysis. Ions of m/z 251, 261, 307 were monitored in the negative linear mode. Competition Experiments between 13 C 5 -Ribose and Each Pentose Isomers 13 C 5 -Ribose (10 L, 0.1 M) and each unlabeled pentose isomers (0.1 M, 10 L) were mixed separately in Eppendorf tubes containing deionized water (70 L) and

PAGE 91

77 vortexed thoroughly. A sodium borate solution (Na 2 B 4 O 7 , 0.025 M, 70 L) was added, and the resulting mixture was vortexed and equilibrated for 2 h at room temperature. An aliquot of 0.5 L was spotted on the DIOS plate, and allowed to dry before DIOS-MS analysis. Ions of m/z 307, 312, 317 were monitored in the negative reflectron mode. Results and Discussion Detection of [Pentose+K] + Adduct Ions Although DIOS-MS was capable of detecting a variety of compounds, direct analysis of monosaccharides was difficult and sometimes cationization was essential to form detectable adducts. 73 In the effort to detect pentoses, no significant ion signal associated with any pentose isomers was detected when measured individually. However, in the presence of potassium ion, pentose isomers (m.w. 150) formed adduct ions [pentose + K] + of m/z 189, which were detectable with DIOS-MS. K+-pentose adduct signal at different K+ : pentose ratios02000400060008000100001200010:15:11:11:51:101:20[K+] : [pentose]S/N of K+-pentose adduct L-arabinose D-ribose D-xylose L-xylose L-lyxose Figure 3-1. A chart of DIOS-MS signal of [pentose + K] + . Signal of m/z 189 at different K + : pentose molar ratios for five tested pentose isomers were plotted.

PAGE 92

78 Interestingly, it was observed that the pentose to K + molar ratio affected the ion peak profile and the overall spectrum quality. Fig. 3-1 showed the [pentose + K] + signal of m/z 189 at different K + : pentose ratios. When K + ion was present in excess, little adduct signal was detected. Moreover, an equal concentration of the cation and the pentose did not give the best signal, as one might expect. As an example, the DIOS mass spectra of ribose with different concentration of K + are shown here (Fig. 3-2). The aqueous ribose (0.1 mM) and the aqueous potassium acetate solution (0.1 mM) were mixed at different volume ratios as described in the experimental section. Improved spectra were obtained at lower K + ion concentrations. Figure 3-2. DIOS mass spectra of ribose with different concentration of K + . The aqueous ribose solution (0.1 mM) and the aqueous potassium acetate (0.1 mM) were mixed at volume ratios of 20:1, 10:1, 5:1, 1:1, and 1:5.

PAGE 93

79 In this experiment, when there was more K + (pentose to K + molar ratios less than 1), m/z 189 of the adduct ion was observed at low intensities, and with other anonymous peaks. It appeared that although potassium ion assisted the detection of neutral pentose molecules by forming charged adducts, higher K + concentration in the solution suppressed adduct ion signals. The result was consistent with electrospray ionization mass spectrometry, where addition of a small amount of salt was sometimes necessary to achieve the desired ionization. 103 However, too much salt also quenched the signal. Figure 3-3. DIOS mass spectra of five pentose isomers in the presence of K + . Five pentoses (m.w. 150), D-ribose, L-arabinose, L-xylose, D-xylose, L-lyxose, were detected as adduct ions [pentose + K]+ of m/z 189 in DIOS-MS. Clean mass spectra were obtained where K + and [Pentose + K] + peaks are the only prominent signals.

PAGE 94

80 At a pentose to K + ratio of 10:1, all five pentose isomers tested in the experiment (D-ribose, L-arabinose, L-xylose, D-xylose, L-lyxose) were detected with DIOS-MS as potassium adducts with good S/N in clean mass spectra, where the only other prominent signal was from K + at m/z 39 (Fig. 3-3). Detection of Pentose-Borate Complexes After pentose isomers were successfully detected in the positive mode through the formation of cation adducts, the analytes of interest, dehydration products of borate ion and pentose, were measured. The mixture of pentose/Na 2 B 4 O 7 produced an m/z 307 pentose-borate complex ion (Fig. 3-4). The formation of the spirane structure, confirmed by NMR techniques, was a result of the coordination between the cis-diol groups and boron. Figure 3-4. Formation of the pentose-borate complex. In the presence of borate ion, pentoses adopted the cyclic five-member ring structure, and formed the spirane complex through the interaction between the boron atom and two diol moieties. The pentose complexes were successfully detected in the negative DIOS-MS and the mass spectrum of the ribose-borate is shown in Fig. 3-5, where the ions of m/z 307 were the only dominant peaks. The volumes of two reactant solutions (pentose/Na 2 B 4 O 7 ) were calculated to give boron : pentose a molar ratio of 1 : 2. At first, the analysis was carried out at a lower pentose/Na 2 B 4 O 7 concentration (1 mM borate ion, 2 mM pentose,

PAGE 95

81 Fig. 3-5a). However, improved S/N was achieved with higher initial reactant concentration (10 mM borate ion, 20 mM pentose, Fig. 3-5b), which was used in future experiments. Compared to peptide analysis reported and practiced in our laboratory, the sensitivity of the pentose-borate complex detection was considerably reduced. This might be caused by the aqueous nature of the test solutions. In this application, aqueous solutions were required in order to retain the native conditions used in the synthesis. However, such solutions didn’t spread well due to its high surface tension, thus hindered the analyte to distribute further into the pores. Figure 3-5. DIOS mass spectra of ribose/Na 2 B 4 O 7 mixtures. The spectra were obtained in the negative linear mode at different initial pentose/Na 2 B 4 O 7 concentrations: a) 1 mM borate ion, 2 mM pentose, and b) 10 mM borate ion, 20 mM pentose. In both spectra, the m/z 307 ion was the predominant signal while S/N was improved at a higher pentose/Na 2 B 4 O 7 concentration.

PAGE 96

82 Demonstrated in these experiments, the advantage of using DIOS-MS over other MS techniques is the minimal sample preparation which retained the native pH value (~ 8.5) essential for the stability of the complex. Same for the following competition experiments, it is critical not to have any external chemicals which may potentially affect the equilibrium of the system. To determine a mass calibration system is important for all mass spectrometry studies, and it is also true in this application. Here, a few requirements were to be met: 1) signals in the negative operation mode; 2) signals in the low mass range; 3) simple spectra to avoid misinterpretation; and 4) signals with high intensities to survive the reflectron mode. In this case, a benzoic acid (C 6 H 5 COOH)/potassium hydroxide (KOH) mixture was used for the negative mode mass calibration. A benzoic acid solution was first tested with DIOS-MS, and the result was not satisfying. To solve the problem, a basic species was added to increase the concentration of the C 6 H 5 COO ion. As result, benzoic acid was mixed with KOH at different volumes to determine the optimized concentration ratios in the final mixtures (Table 3-1). The final concentration of the two ions of interest, C 6 H 5 COO and K + , are also listed in the table. Table 3-1. Volumes of the benzoic acid and KOH solution (both at 10 mM) used for constructing the mixtures, and the resulting concentration of related species. Volume used Resulting mixture Mixture Benzoic acid (mL) KOH (mL) Final conc. ratio (BA:KOH) [C 6 H 5 COO ] (mM) [K + ] (mM) a 5.0 1.0 5:1 2.2 1.7 b 2.5 1.0 2.5:1 3.0 2.9 c 1.0 1.0 1:1 5.0 5.0 d 1.0 2.0 0.5:1 3.3 6.7 f 1.0 4.0 0.25:1 2.0 8.0

PAGE 97

83 Representative DIOS-MS spectra are shown in Fig. 3-6 for the five listed mixtures. Ions of C 6 H 5 COO and [2C 6 H 5 COO +K + ] were arbitrarily selected for mass calibration. The C 6 H 5 COO ion (m/z 121) was observed in all the experiments. When concentration of the K + ion was relatively low as in mixture a and b, [2C 6 H 5 COO +Na + ] ion was observed, although no sodium salt was intentionally added. As the concentration of the K + ion increased, the [2C 6 H 5 COO +K + ] ion signal first increased and then decreased, which matched the change of C 6 H 5 COO concentration (Table 3-1). In short, the optimum signals of C 6 H 5 COO and [2C 6 H 5 COO +K + ] were obtained when the volume ratio was 1:1, which was used as the final calibration mixture. Figure 3-6. Negative DIOS-MS of benzoic acid/KOH mixtures. The five mass spectra were obtained from benzoic acid/KOH mixtures at different initial volume ratios. Spectra a-e corresponded to mixture a-e in Table 3-1. The initial concentration of 10 mM was used for the two solutions.

PAGE 98

84 Competition Experiments 1,4-Anhydroerythritol and Pentose Isomers 1,4-Anhydroerythritol (AHE), which possesses a cis-diol moiety as does pentose, was used as a reference compound to compete with individual pentose isomers. Shown in Fig. 3-7 is the suggested reaction scheme. When mixed at 1:1 ratio, both AHE and the pentose were capable of forming the spirane complexes composed of two identical competing molecules and one boron atom. At the same time, a third type of the complex acquired a structure of both two competing molecules and one boron atom. The yields of the three complexes reflect the binding affinities of AHE and pentose with boron. Accordingly, four pentose isomers of interest (ribose, xylose, lyxose, arabinose) were made to compete against AHE. The results of the four sets of experiments were used to study the relative binding affinity to boron among different pentose isomers. Figure 3-7. Competition reaction between pentose isomers and 1, 4-Anhydroerythritol in the presence of borate ion. Three negatively charged complexes, suggested structures shown here, have m/z of 215, 261, and 307. Here, ion intensities of the pentose-borate complex (m/z 307) and AHE-borate complex (m/z 215) were used for comparison. Although the relative responses of the various complexes in this experiment were not known, an assumption was made that four pentose-borate complexes had similar ionization efficiencies. It was worth noting that the AHE-borate complex (m/z 215) was not exactly an internal standard, since its

PAGE 99

85 concentration in the four equilibra were different. However, the ionization efficiency of the AHE-borate complex was not a concern. The possible ionization efficiency differences between ion m/z 307 and m/z 215 was cancelled out since the final evaluation was among four pentose-borate complexes after they were first compared to the AHE-borate complex. 0%10%20%30%40%50%60%m/z 215m/z 261m/z 307factions of peak intenstities OOHOHOHOH D-riboseOOHOHOHOH D-xylose 0%10%20%30%40%50%60%70%m/z 215m/z 261m/z 307factions of p eak intenstities ba Figure 3-8. DIOS mass spectra of mixtures of pentose isomers and 1, 4-anhydroerythritol in the presence of borate ion. a) Xylose; and b) ribose. The structure of xylose and ribose are also shown here. The inserted bar charts presented three individual ion intensity percentages in the summed intensities of three ions (m/z 215, 261, 307). Shown in Figure 3-8 are DIOS-MS mass spectra of mixtures of xylose and ribose with AHE in the presence of borate ion. As demonstrated in the mass spectra and the corresponding diagrams, the ion intensity fraction of m/z 307 in the ribose mixture is

PAGE 100

86 much higher than that of m/z 215, while for the xylose mixture, the m/z 307 ion is less intense. This indicated that ribose bound better to boron than xylose. In the bar charts, the ion intensity fraction of each complex was calculated as a percentage of the total ion intensity of all three complexes formed in the equilibrium. Intensity fractions were plotted as the averages from six spectra acquired on the same DIOS spot, each spectrum consisting of 50 laser shots. The error bars represented the relative standard deviations (RSDs). pentose competition experiment with AHE0.0%10.0%20.0%30.0%40.0%50.0%60.0%70.0%m/z 215 m/z 261m/z 307percentage of ion intensity xylose arabinose lyxose ribose Figure 3-9. Results of four sets of competition experiments for four aldopentose isomers of interest (ribose, xylose, lyxose, arabinose). Ribose/AHE/borate mixture showed highest ion intensity percentage at m/z 307 than the other three aldopentoses. The results of four sets of competition experiments are summarized in Figure 3-9. The ribose/AHE/borate mixture showed the highest ion intensity fraction at m/z 307 than the other mixtures. It was concluded that ribose has a higher affinity to boron than the other pentose isomers. The order of affinities was determined as ribose (1.0, 9%) > lyxose (0.73, %) > arabinose (0.44, %) > xylose (0.30, %); the first numbers

PAGE 101

87 in the parentheses were ion intensity fractions of m/z 307 normalized against that of ribose, and the second numbers were the RSDs. Competition Experiments 13 C 5 -Ribose and Pentose Isomers In addition to using a structural analogue as a competition reagent to investigate the relative binding affinities of pentose isomers toward boron, a 13 C-labeled ribose was used in a direct competition experiment against other pentose isomers. Shown in Figure 3-10 is the suggested reaction scheme. Figure 3-10. Competition reaction between pentose isomers and 13 C 5 -labeled ribose in the presence of borate ion. When 13 C 5 -labeled ribose was used, the resulting complexes had m/z of 307, 312 and 317 instead with 13C depicted as black dots. The isotopically labeled pentose allowed direct comparison between the pentose isomers. Naturally, 13 C 5 -ribose was chosen to be the reagent. Similar to the previous experiment, the competition reagent was mixed with each of the four pentose isomers (ribose, xylose, lyxose, arabinose) at a 1:1 ratio in the presence of the borate ion. Ions of m/z 307, 312, 317 were monitored in the negative reflectron mode instead (Figure 3-11). The ion intensity ratios of m/z 307 to m/z 317 were plotted in Figure 3-12. The average values and standard deviations of ion intensities were calculated from six DIOS-MS mass spectra obtained on the same spot, each being a sum of 50 laser shots.

PAGE 102

88 As shown in Figure 3-11, ions at m/z 307, 312 and 317 were observed in all four mixtures. Ion signals at m/z 311, 316 (one mass unit lower than the expected m/z values) were believed to associate with non-fully 13 C 5 -labeled species. In addition, the ion intensity ratio of m/z 307 to m/z 317 was about one in the 13 C 5 -ribose/ribose/borate mixture, which was expected, since the isotopically labeled species should behave very similarly to its counterpart in terms of chemical reactivity and ionization efficiency. Due to the inherent impurity problem of the isotopic labeling, the initial ratio of regular ribose to fully-labeled 13 C-ribose was in fact greater than one to begin with. As a result, more ribose-borate complex formed at equilibrium so that the ion intensity ratio of m/z 307 to m/z 317 was slightly greater than 1. Figure 3-11. DIOS mass spectra of the mixtures of pentose isomers and 13 C 5 -ribose in the presence of borate ion. The mixtures were 13 C 5 -ribose with ribose, xylose, lyxose, arabinose, respectively.

PAGE 103

89 More significantly, the results showed that in the other three mixtures, the 13 C 5 -ribose-borate complex (m/z 317) displayed higher ion intensity than the borate complexes of competing pentose isomers (xylose-borate, arabinose-borate and lyxose-borate, m/z 307). With the assumption that the involved complexes had similar ionization efficiencies, it was reasonable to say that the 13 C 5 -ribose-borate complex had a higher yield in the corresponding equilibrium, meaning that 13 C 5 -ribose had a better binding affinity to boron. Fig. 3-12 showed the preferential order of binding to boron: ribose (1.00, %) > lyxose (0.60, %) > arabinose (0.34, %) > xylose (0.19, %); the first numbers in the parentheses were ion intensity fractions of m/z 307 normalized against that of ribose, and the second numbers were the corresponding RSDs. pentose competition experiment with C-13 ribose0.000.200.400.600.801.001.201.40ion intensity ratio of m/z 307 : m/z 317 ribose lyxose arabinose xylose Figure 3-12. Results of competition experiments for four sets of experiments for 13 C 5 -ribose with pentose isomers. Ion intensity ratios of m/z 307 to m/z 317 in four mixtures and the preferential order of binding to boron: ribose>lyxose>arabinose>xylose. The results confirmed the observation in the AHE competition experiments. Notice that the error bars were not overlapping in Fig. 3-12, unlike in Fig. 3-9, so the result was of higher confidence.

PAGE 104

90 As previously mentioned, stability constants of borate complexes have been determined by using a combination of potentiometric titration and 11 B NMR spectroscopy, and the reported stability constant trend for aldopentoses was: ribose > xylose > lyxose > arabinose. 101;102 Although the published order didn’t agree with our result, the ribose nonetheless was determined to have the highest stability constant. The degradation rates of pentose isomers under alkaline/colemanite conditions were determined using 1 H NMR technique by estimating the loss of selected signals. The results also demonstrated that all the aldopentoses were stabilized by boron to a different extent, following the trend: ribose > lyxose > arabinose > xylose. 104 Conclusions We applied DIOS-MS to monosaccharide analysis. First, the pentose isomers were detected as potassium adducts in the positive mode and borate complexes in the negative mode. It was demonstrated that only a small amount of cation species was required for successful detection of the cation adduct. A pentose structural analogue was used to compete with individual pentoses in forming borate complexes to determine the binding preference among the four pentose isomers. The DIOS-MS detection was carried out in the negative mode, and a proper mass calibration system was developed. Finally, 13 C 5 -labeled ribose was included in another competition experiment which further confirmed the first competition results. Ribose exhibited higher affinity to boron than other aldopentoses, and the binding preference was determined to be ribose>lyxose>arabinose >xylose. The result indicated that the favored binding between ribose and boron could be an important factor in RNA evolution.

PAGE 105

91 This work also illustrated the potential of DIOS-MS in the analysis of non-volatile, heat-labile, small molecules in delicate chemical equilibria. Without external matrix compounds, background signals were not a limiting factor and the possible dramatic change in pH, which could disturb the equilibria of interest, was avoided.

PAGE 106

CHAPTER 4 DIOS-MS STUDY OF PORPHYRIN/PORPHODIMETHENES AND ITS IMPLICATIONS Introduction Porphyrins and Porphodimethenes Tetrapyrrolic macrocycles are present in nature as part of important enzymes or proteins for diverse functions including light-harvesting, electron transfer, oxygen transfer. 105 Ubiquitous in living systems, porphyrins are among the most abundant tetrapyrrols. They are highly conjugated ligands, and their basic structure is illustrated in Fig. 4-1a. NNHNHNRRRR NNHNNH ab Figure 4-1. Basic structure of a) porphyrins and b) porphodimethenes. Also investigated in this chapter are porphodimethenes (Fig. 4-1b). They differ from porphyrins in having two saturated carbons at mesopositions in the structure. Porphodimethenes are known as intermediates formed during the natural redox-reactions of porphyrins, and they can be used as precursors in synthesizing otherwise inaccessible novel porphyrins. 106 The first synthetic scheme for air-stable porphodimethene production was reported in 1974. 107 Later, other research groups have suggested alternative routes to obtain such 92

PAGE 107

93 species. Among them, Michael Scott and coworkers pioneered the synthesis of spiro-tricyclic porphydimethenes, 108 and continued in designing novel porphyrin species. A variety of analytical techniques, such as X-ray crystallography and 1 H NMR, have been applied to confirm the successful synthesis. Here, DIOS-MS analysis was used as the mass spectrometry approach to characterize those molecules. In this collaboration work, we also investigated the metal complexes of porphyrin and porphodimethenes. The inclusion of metals into porphyrin/porphodimethene macrocycles are commonly used in syntheses, and the resulting complexes usually possess different properties and reactivities compared to their free-base precurcors. It is thus significant to inspect the DIOS-MS behavior of the metal complexes in addition to the free-base ligands. Mass Spectrometry Analysis of Porphyrins and Their Derivatives Most ionization techniques have been applied to the characterization of porphyrins and their metal complexes with mass spectrometry. EI 109 and CI 110 were first used in porphyrin analysis, and only mild fragmentation was observed due to the stable macrocyclic nucleus. In addition to the confirmation of molecular weights, EI-MS helped discover novel fragmentation pathways for porphyrin derivatives, which contributed to understanding the properties of the species. 111 Even though EI and CI were adequate for mass spectrometry analysis in early porphyrin researches, the successful synthesis of more complex porphyrin structures demanded softer ionization techniques to avoid the fragmentation of the peripheral functional groups. At the same time, new ionization techniques were introduced to the mass spectrometry field, and many of them have been applied to porphyrin analysis including fast atom bombardment (FAB), 112;113 liquid secondary ionization mass

PAGE 108

94 spectrometry (LSIMS), 114 field desorption (FD), 113 laser desorption ionization (LDI), 115-117 matrix-assisted laser desorption/ionization (MALDI), 112;113;115;117-120 electrospray ionization (ESI) 121 and two-step laser desorption/photo-ionization. 122;123 Among all ionization techniques mentioned above, LDI and MALDI received most of the attention. Some researchers believed that there was no fundamental need of adding matrix for porphyrin analysis, since this type of compound has sufficient absorption at the laser wavelength normally used in mass spectrometry analysis. 124 However, examples were shown that with the addition of matrix, fragmentation was reduced. In some cases, such as the analysis of large porphyrin assemblies, MALDI-MS was the only approach to obtain molecular ion information. 115;117;118 Common MALDI matrices, e.g., 2,5-dihydroxybenzoic acid (DHB) 118 and 4-hydroxy--cyano-cinnamic acid 117 were proven to be effective. Innovative matrices, e. g., 1,4-benzoquinone 124 and 2,7-dimethoxynaphthalene, 119 were also used to facilitate MALDI analysis. Despite the advantage of being a soft ionization method, MALDI has its limitations in the sample preparation step, where choice of matrix can be time-consuming, and the addition of matrix may cause undesired demetallation. 118;124 Moreover, poor solubility of porphyrins in the solvents used for the matrix is another problem. Even with a special solvent system (toluene : ethanol = 1:1), layered deposition was still required, which resulted in increased fragmentation. 118 On the other hand, LDI is a more straightforward approach for the analysis of porphyrins; however, mass calibration poses difficulties to the analysis. Due to the lack of suitable calibration compounds, “porphyrin standards” are sometimes used for mass calibration, and they are assumed to form [M] + . 120 However, it is possible for the

PAGE 109

95 “porphyrin standards” to generate both the cation radical ([M] + ) and the protonated molecule ([M+H] + ), as demonstrated in the literature as well as in this work. For example, in the work of Jones et al. and Srinivasan et al., both types of porphyrin molecular species, porphyrin cation radicals ([M] + ) and protonated porphyrins ([M+H] + ) were observed using LDI, even when the same ionization technique was employed. 116;124 Therefore, a full understanding of the involved porphyrin ionization chemistry is needed before using porphyrin standards as mass calibration compounds. In fact, MALDI also reportedly gave rise to both forms of ions for porphyrins. 115;118 In a similar instance, Lidgard, et al. pointed out that this technique posed pitfalls for interpretation of mass spectra of calix[3]indoles, another type of large molecules with conjugated benzene ring systems. While most analytes generated [M+H] + or [M+Na] + , a particular calyx[3]indole molecule gave the cation radical. 125 This cautionary note is significant in porphyrin analysis too, given the well-documented fact that porphyrins are capable of forming both [M] + and [M+H] + ions. Careful mass calibration using structurally non-related compounds is very important to ensure the correct assignment of observed molecular species. In conclusion, porphyrins consist of an important category of compounds, and mass spectrometry has been widely used for their characterization. Among all the ionization techniques investigated, LDI and MALDI are often employed. However, LDI usually generates more fragmentation, and sometimes calibration can be a problem. On the other hand, MALDI is a softer method; however, it might suffer from the incompatibility with matrix compounds and insolubility in solvents.

PAGE 110

96 DIOS-MS of Porphyrins and Porphodimethenes In this report, DIOS-MS was introduced as an additional soft ionization technique in porphyrin analysis. For porphyrin and porphodimethenes, DIOS-MS is superior to LDI being a softer ionization method. Moreover, internal mass calibration with independent compounds (such as the HP mix) is carried out to ensure correct m/z values. When compared to MALDI-MS, DIOS-MS represents a simpler approach with the “matrix-free” sample preparation. There is no need to spend time determining an optimal matrix compound and the sample deposition protocol. As a result, throughput is improved. In addition to the advantage of DIOS as a suitable ionization method for porphyrins and porphyrin derivatives of interest, the reflectron detection mode on the TOF analyzer allows for the acquisition of characteristic isotopic distributions of the inquired molecules. The information is used to confirm the empirical formula of analytes. Briefly, DIOS-MS provides simplified mass spectra with accurate molecular weight information as well as isotopic distributions for porphyrin analysis. After successful analysis of a variety of porphyrin samples, the effect of oxidation potential of these compounds on the formation of [M] + and [M+H] + ions will be discussed. Experimental Methods Sample Preparation and Mass Calibration The protocol of PSi chip production and the parameters of mass spectrometry analysis were described previously. Porphodimethenes, porphyrin and their metal complexes analyzed in this research were obtained from Dr. Michael Scott’s research group. Minimum sample preparation was required for DIOS-MS analysis. Samples (a

PAGE 111

97 few milligrams) were dissolved in 200 L chloroform, and an aliquot of 0.5 L test solution was spotted on individual PSi spots. For most of the DIOS-MS measurements here, internal mass calibration standards were included to ensure the accurate assignment of the observed ions. For internal mass calibration, 0.2 L HP mix (Agilent Technologies, Inc., Palo Alto, CA) was deposited on a PSi spot and dried prior to sample deposition. Results and Discussion Free Porphodimethenes Ligands and Their Pd Complexes First, four free-base porphodimethene (4-1, 4-2, 4-3, and 4-4 in table 4-1) were examined with DIOS-MS. The HP mix (components detailed in Chapter 2) was used for external mass calibration. Compound 4-1 and 4-2 differed in the substitution alkyl group, and compound 4-3 and 4-4 were -cis, -trans isomers. The first pair (4-1, 4-2) and the second pair (4-3, 4-4) had different ring structures attached to the carbons at the mesopositions. Table 4-1. Structures of four porphodimethene free ligands. Compounds Basic structure Substitution group R 4-1 CH3CH3CH3 4-2 NNHNHNOORR t-But-Bu

PAGE 112

98 NHNNNHOOR Table 4-1. Continued. 4-4 R CH3CH3CH 3 Figure 4-2. DIOS-MS of a) 4-1 , b) 4-2, c) 4-3, and d) 4-4. The monoisotopic masses of [M+H] + of the four compounds are 905.4, 1045.5, 953.4 and 953.4, respectively. Clean DIOS mass spectra of the four tested free-base porphodimethenes were obtained with almost no fragmentation (Fig. 4-2), despite the fact that these polycyclic vicinal diketone compounds possess usually fragile tertiary carbons at the meso-position. Another observation was that the externally mass-calibrated isotopic distributions of

PAGE 113

99 tested compounds matched the theoretical values of radical cations [M+H] + with a mass accuracy of 0.5 Da (Figure 4-3). Figure 4-3. Experimental and theoretical isotopic distribution of 4-1. Inserted mass spectrum in blue is the theoretical isotopic distribution of [M+H] + . To investigate the effect of inclusion of metal into porphodimethenes, palladium complexes of the these four free-base porphodimethenes (4-5, 4-6, 4-7, and 4-8 in Table 4-2) were subjected to DIOS-MS analysis. The HP mix was used for internal mass calibration. DIOS-MS results of four free ligands and four complexes are summarized in Table 4-3. Table 4-2. Representative structure of four porphodimethene Pd complexes. Compound Basic structure Ligand (L) 4-5 4-1 4-6 4-2 4-7 4-3 4-8 NNNNRRPd Pd-L 4-4

PAGE 114

100 Table 4-3. DIOS-MS results of four free ligands and four Pd complexes. Compound Molecular species observed Fragmentation 4-1 [M+H] + No 4-2 [M+H] + No 4-3 [M+H] + No 4-4 [M+H] + No 4-5 [M+H] + Yes 4-6 [M+H] + Yes 4-7 [M+H] + Yes 4-8 [M+H] + Yes As opposed to the observation with free-base porphodimethenes, fragmentation, including characteristic losses of 28, 56, occurred for the Pd complexes. Shown in Fig. 4-4 were the representative DIOS-MS mass spectra of Pd complexes (4-5 and 4-7). The fragmentation was attributed to macrocyclic distortion caused by coordination of the metal and the nitrogen. It was postulated that the loss of 28 and 56 corresponded to the loss of one and two -COgroups, which resulted in the collapse of the two cyclic systems at the mesoposition to the center tetra-pyrrole structure. Interestingly, the corresponding free-base ligands did not exhibit the characteristic loss, and only the Pd complexes underwent the fragmentation upon laser radiation. The observation might indicate a novel synthetic route of making porphyrin derivatives using Pd complexes. Based on this hypothesis, an experiment was set up to reproduce the conversion simply by projecting a UV radiation on the Pd complexes. The expected product was not obtained, which might be due to the difficulty in completely recreating the stringent conditions achieved in a DIOS process in a synthetic lab setting. However, a light-initiated rearrangement was proven to be responsible for the conversion from metallo-porphodimethenes into metallo-porphyrins, which have also been subjected to DIOS-MS and the results were discussed in the next section. 106

PAGE 115

101 Figure 4-4. DIOS mass spectra of a) 4-5 and b) 4-7. In each set , the top mass spectrum shows the observation of molecular species with fragmentation in the presence of the internal mass calibration compounds HP622 and HP922 from the HP mix (components detailed in Chapter 2). The bottom mass spectrum in each set is the expanded molecular species region. Inserted (in blue) are theoretical isotopic distribution of the corresponding [M+H] + . It is to be stressed that the observed molecular species were determined as the protonated molecular [M+H] + , using either the external and internal mass calibration.

PAGE 116

102 Better accuracy was achieved with internal mass calibration, and it was applied in most of the following experiments. Porphyrins with Different Chelating Metals In addition to porphodimethenes free ligands and metal complexes, porphyrin derivatives with different chelating metals (copper, palladium and zinc) were also investigated. The non-planar, sheet-like metallo-porphyrins, bearing large, fused exocyclic ring systems, were obtained through light-initiated oxidative rearrangements of corresponding metallo-porphodimethenes, a – 2e and +2H process. Initially, attempts of treating free-base porphodimethenes with various oxidants failed to give free-base porphyrins. This was because porphodimethenes coordinated with late transition metals possess lower positive first oxidation potentials than their free-base counterparts. In fact, cyclic voltammograms of some metallated derivatives of a free-base orphodimethenes revealed that these species exhibited lower positive first oxidation potentials by at least 0.3 V. 106 The structures and DIOS-MS results of selected metalloporphyrins were summarized in Table. 4-4. Table 4-4. Structures and DIOS-MS results of metalloporphyrins with different ligands and chelating metal. For most of compounds, both , [M] + and [M+H] + were observed. Compound Metal Molecular species Ligand 4-9 Cu [M] + , [M+H] + 4-10 Pt [M] + , [M+H] + NNHNHNORRO Ligand I –trans

PAGE 117

103 Table 4-4. Continued. 4-11 Cu [M] + [M+H] + 4-12 Pt [M+H] + NNHNHNOORR Ligand II -cis 4-13 Zn [M] + [M+H] + 4-14 Pt [M+H] + Ot-But-BuNNHNHNRRt-But-BuO Ligand III -trans 4-15 Zn [M] + [M+H] + 4-16 Pt [M] + t-But-BuOt-But-BuNNHNHNRRO Ligand IV -cis 4-17 Zn [M] + [M+H] + Ot-But-BuNNHNHNRRt-But-BuO Ligand V -trans 4-18 Zn [M] + [M+H] + t-But-BuOt-But-BuNNHNHNRRO Ligand IV -cis DIOS-MS was able to provide useful information, including molecular weights and isotopic distribution, to confirm the successful synthesis of the compounds. For example,

PAGE 118

104 Fig. 4-5 showed the DIOS mass spectra of 4-9, 4-10 (Pd and Cu complexes of ligand I), and the theoretical isotopic distribution of the two compounds. Despite the fragmentation, molecular species were observed. Internal mass calibration ensured the correct m/z values. After comparing with the theoretical isotopic distributions, it was revealed that these molecular species were actually composed of [M] + and [M+H] + . Most of metalloporphyrins analyzed here exhibited the formation of [M] + , in addition to [M+H] + . It was believed that this process of losing one electron was facilitated by the low positive first oxidation potentials of the metalloporphyrins. Figure 4-5. DIOS mass spectra of a) 4-9 and b) 4-10. Inserted (in blue) are expanded regions of molecular species. The theoretical isotopic distributions of two compounds are also shown (c).

PAGE 119

105 The effect of oxidation potentials of porphyrins on mass spectrometry analysis has been investigated by Vandell et al. using ESI. 121 It was found that, for binary mixtures of metalloporphyrins whose oxidation potentials differed by less than 0.1 V, the resulting ion abundance of each species was directly proportional to concentration. For binary mixtures whose oxidation potentials differed by more than 0.1 V, relative abundances of the radical cations of each metalloporphyrin were determined by the oxidation potential as well as the concentration, with the analyte of lowest oxidation potential being ionized preferentially. In DIOS-MS analysis, as listed in Table 4-3, all porphodimethenes and their metal complexes only formed [M+H] + . On the other hand, porphyrin derivatives have fused exocyclic-ring systems (Tabel 4-4), in contrast to porphodimethenes which have saturated carbons at -meso positions (Table 4-1). The highly conjugated ring systems lead to the lower oxidation potentials of porphyrin derivatives. As a result, upon the laser radiation, porphyrin derivatives give rise to [M] + ions while porphodimethenes do not . However, this theory was not followed for porphyrins with small differences in oxidation potentials, such as -trans, -cis porphyrin isomers. In general, -trans isomers possess lower first oxidation potentials compared to their -cis counterparts, by a small value (~0.1 V). 106 Yet, the preliminary data of DIOS-MS showed a higher portion of [M+H] + for –trans isomers for most isomer pairs. Examples are given in Fig. 4-6 and 4-7. In Fig. 4-6, the -trans and -cis isomers of a metalloporphyrin, 4-9 and 4-11 displayed distinct isotopic distribution. When compared to the theoretical distribution in Fig. 4-6c, DIOS mass spectrum of 4-11 (Fig. 4-6a) shows a higher portion of [M] + ions than that of 4-9 ( Fig. 4-6b). A similar example is shown in Fig. 4-7. In this case, the difference in

PAGE 120

106 experimental distributions of the two isomers was less evident; however, the portion of [M] + could be extracted with calculation for comparison. The result showed, again, the -cis isomer created more [M] + ions. Figure 4-6. DIOS mass spectra of a) 4-9, b) 4-11 and c) the theoretical isotopic distribution of corresponding [M] + . Figure 4-7. DIOS mass spectra of a) 4-13, b) 4-15 and c) the theoretical isotopic distribution of corresponding [M] + . Effect of Laser Power It is known that the laser power contributes to the signal intensity change in DIOS-MS, as in MALDI, LDI or any laser desorption ionization technique. Will laser power affect the distribution of [M] + and [M+H] + in case of the porphyrin? The preliminary results showed that the laser power did not exhibit any obvious effect on [M] + formation. Demonstrated here are DIOS signals of compounds 4-13 and 4-15 collected at various laser powers (Fig. 4-8).

PAGE 121

107 a) 4-15 -cis0100020003000400050006000700046485052545658606264laser attenuationtotal ion intensity [M], [M+H0.00.51.01.52.0ion intensity ratio [M]/[M+H ] b) 4-13 -trans 01000200030004000500060007000464850525456586062laser attenuationtotal ion intersity [M], [M+H ] 0.00.51.01.52.0ion intensity ratio n [M]/[M+H] Figure 4-8. Total ion intensity of [M] + and [M+H] + , and ion intensity ratio of [M] + /[M+H] + at different laser attenuation of compounds a) 4-11 and b) 4-13. Laser attenuation of 50 was not available on the instrument. The laser power was varied by changing the attenuation settings of the neutral density filter in the optic system. The intensity of [M] + and [M+H] + were calculated by deconvoluting the experimental isotopic distribution using Eq. 4-1 and 4-2. In these equations, I [M]+ and I [M+H]+ are the ion intensities used for plotting, I E1 and I E2 are intensities of the first and second peak of the experimental isotopic distribution, and

PAGE 122

108 I T2 /I T1 is the intensity ratio of the first and second peak in the theoretical isotopic distribution. I [M]+ = I E1 Eq. 4-1 I [M+H]+ = I E2 – I E1 (I T2 /I T1 ) Eq. 4-2 For both compounds investigated, total ion intensities increased as the laser power was increased. However, despite the considerable fluctuation, [M] + /[M+H] + values did not exhibit an increasing or decreasing response to the laser power change. Also shown in the two charts, compound 4-15 with the -cis conformation demonstrated the relative higher [M] + /[M+H] + , as previously illustrated in Fig. 4-7. These preliminary data show that laser power did not have obvious effect on [M] + /[M+H] + of metalloporphyrins in this analysis. Since the theory of relating the oxidation potentials to the formation of [M] + can not fully explain the behavior of porphyrins/porphodimethenes and their metal complexes in DIOS-MS, it is likely that there are additional mechanisms governing the formation of [M] + ions. Porphyrins with Similar Masses Listed in Table 4-4, metal complexes of ligand III and IV were obtained from their precursors of ligand V and VI through an oxidation reaction (removal of four hydrogen atoms and formation of two additional bonds). Different by a mass of 4, their molecular species were easily differentiated with DIOS-MS. However, in one experiment, the analyte (1133.5 Da) could not be attributed to either the precursor (1135.5 Da) or the oxidation product (1131.5 Da). With repeated trials and careful internal mass calibration, it was determined that the analyte was a partial oxidation product with only two hydrogen atoms removed and one additional bond formed. (Fig. 4-9) The structure was later confirmed by X-ray crystallography.

PAGE 123

109 CuOt-But-BuNNNNRRt-But-BuO CuOt-But-BuNNNNRRt-But-BuO 4-20 m.w. 1133.5 (monoisotopic)4-19 m.w. 1135.5 (monoisotopic) Figure 4-9. The formation and DIOS mass spectrum of 4-20. a) A partial oxidation reaction resulted a product 2 Da lower than the starting material and 2 Da higher than the expected product; b) DIOS mass spectrum of 4-20 (left) and the theoretical isotopic distribution (right). Again, being the soft ionization method, DIOS-MS was capable of generating characteristic molecular species with high accuracy, which helped identify unanticipated products in organic synthesis. Conclusions DIOS-MS was feasible for the analysis of porphyrins, porphodimethenes and their metal complexes. For all the samples tested (some were included in Table 4-3 and 4-4), molecular species were detected with minimum sample preparation. In addition, the technique was able to provide isotopic distribution information to confirm the successful synthesis of these compounds. For certain compounds, fragmentation was observed.

PAGE 124

110 It was observed that porphyrin derivatives gave rise to [M] + ions while porphodimethenes did not. The phenomenon was explained by the fact that porphyrins have lower oxidation potential (~0.3 V). however, other mechanisms may contribute to the formation of [M] + , considering the controversial result with -cis, -trans isomers. Moreover, laser power was shown not to change the [M] + /[M+H] + ratio according to preliminary results. Finally, an important feature of this research was the utility of an independent internal mass calibration for accurate m/z measurements, which contributed to the correct assignment of [M] + or [M+H] + ions. With the confidence level achieved, novel synthetic products with similar mass to known compounds were identified.

PAGE 125

CHAPTER 5 CONCLUDING REMARKS Mass spectrometry is a useful tool in life science, drug discovery, organic synthesis, etc. Developments in this area have equipped scientists with an expanding collection of instruments and methodologies. DIOS-MS is a recent addition to the mass spectrometry field. It is a surface-enhanced laser desorption/ionization technique featured by low chemical background and no additional matrix compounds. This dissertation evaluated the performance of DIOS-MS, including its advantages and limitation. The two PSi production methods, anodic etching and metal-assisted etching, were compared. Despite the fact that anodic etching has been the major approach for PSi production in DIOS-MS research, metal-assisted etching, being safer and more efficient, became the routine method. SEM images illustrated that PSi morphology featured by ~30-60 nm diameter pores. The PSi surface was evaluated with FT-IR, Auger electron spectroscopy (AES) and SEM. FT-IR revealed the formation of the Si-OH bond due to the presence of hydrogen peroxide in the etching solution or rapid post-etching oxidation. The AES results confirmed the presence of the oxygen, which could originate from the oxidation and/or the adsorption of air-borne hydrocarbon contaminants. To learn more about the physical characteristics of PSi, SEM was employed again and heterogeneous morphologies on different PSi spots and within the same spots were observed. These results were explained by the randomness of the pore formation and the 111

PAGE 126

112 uneven Au deposition during metal-assisted etching. No obvious impact on the DIOS-MS signal was resulted from the morphology variations. After the production and evaluation of the PSi substrate, instrumentation and other topics related to DIOS-MS analysis were discussed. The laser, mass analyzer, sample target, and vacuum system were the relevant components of the instrument system used in current DIOS-MS studies. One important advantage of this technique was simple sample preparation, and the routine procedure for sample preparation was described. The appearance of PSi with deposited samples was evaluated with SEM. A thin sample layer was believed to be crucial for DIOS-MS. Experiments were then designed to explore the applicability of the technique using PSi prepared with metal-assisting etching. Individual analytes, mixtures (drug molecules, peptides, fluorine containing standards, and compounds of different types), and a polymer sample were evaluated. In addition, the reflectron mode and the negative mode of DIOS-MS were carried out to survey the versatility of the DIOS-MS. At the same time, limitations of the technique were also discussed. The characteristics of DIOS-MS developed in our lab include: 1) DIOS-MS is a suitable tool for a variety of compounds at picomole level, with ease of sample preparation and low chemical background. 2) The ability to obtain mass spectra in the reflectron and the negative ion modes extends the range of DIOS-MS applications. Particularly, the negative mode is promising with low chemical background which made the “matrix-free” feature more significant. 3) The technique still suffers background (especially at higher laser power and for aged surfaces), which may be caused by desorption of adsorbed air-borne contaminants, or Si clusters ablated upon laser radiation.

PAGE 127

113 4) Compared to MALDI-MS, the technique has reduced response to peptides and molecules with low aqueous basicity. This chapter also addressed an important aspect of DIOS-MS, aging and re-etching of the PSi chip. Re-etching was able to remove the background peaks to a certain extent or even completely, but at a price of the decreased S/N. As a result, the PSi chips used for this work were either freshly etched or aged in air without further re-etching. Moreover, results of contact angle measurement served as a guideline for the choice of solvent systems. To improve DIOS-MS signals for peptide samples, efforts were attempted and the content of acid modifier was shown to have an effect on DIOS signal. A certain amount of TFA enhanced the spectrum quality (~0.1%); however, desired signals deteriorated when too much acid modifier were added (~0.5-1%). Despite the drawbacks of the DIOS-MS, two applications were investigated to demonstrate the utility of the technique. First, we applied DIOS-MS to monosaccharide analysis. The pentose isomers were detected as potassium adducts in the positive mode and borate complexes in the negative mode. It was shown that only a small amount of cation species was required for successful detection of the cation adduct. A pentose structural analogue was used to compete with individual pentoses in forming borate complexes to determine the binding preference among the four pentose isomers. The DIOS-MS detection was carried out in the negative mode, and a proper mass calibration system was developed.

PAGE 128

114 A different approach using 13 C 5 -labeled ribose in another competition experiment further confirmed the first competition results. Ribose exhibited higher affinity to boron than other pentose isomers, and the binding preference was determined to be ribose>lyxose>arabinose >xylose. The result indicated that the favored binding between ribose and boron could be an important factor in RNA evolution. This work also illustrated the potential of DIOS-MS in the analysis of nonvolatile, heat-labile, small molecules in delicate chemical equilibria. Without external matrix compounds, background signals were not a limiting factor and the possible dramatic change in pH, which could disturb the equilibria of interest, was avoided. In the second application, DIOS-MS was determined to be feasible for the analysis of porphyrins, porphodimethenes and their metal complexes. For all the samples tested (some were included in Table 4-3 and 4-4), molecular species were detected with minimum sample preparation. In addition, the technique was able to provide isotopic distribution information to confirm the successful synthesis of these compounds. For certain compounds, fragmentation was observed. It was observed that porphyrin derivatives gave rise to [M] + ions while porphodimethenes did not. This was the first observation of the radical cation in DIOS-MS. The phenomenon was explained by the fact that porphyrins have lower oxidation potential (~0.3 V). however, other mechanisms may contribute to the formation of [M] + , considering the controversial result with -cis, -trans isomers. Moreover, laser power was shown not to change the [M] + /[M+H] + ratio according to preliminary results. An important feature of this research was the utility of an independent internal mass calibration for accurate m/z measurements, which contributed to the correct

PAGE 129

115 assignment of [M] + or [M+H] + ions. With the confidence level achieved, novel synthetic products with similar mass compared to known compounds were identified. The investigation on DIOS-MS has shown that, despite its inherent limitations such as the presence of background ions, the technique is a method of choice for suitable applications. In mass spectrometry, it is common that a particular ionization technique is found most effective for certain types of molecules. Similarly, a particular mass spectrometry approach can be superior than others depending on the experiment conditions and goals of specific applications. Researchers need to be familiar with the options available and select the right tool for the problems at hand. This work on DIOS-MS has expanded the utility the technique, and has shown the technique as a promising mass spectrometry method.

PAGE 130

LIST OF REFERENCES 1. Grayson, M. A. Measuring Mass: From Positive Rays to Proteins, Chemical Heritage Press: Philadelphia, 2002, Chapter 1. 2. Kiser, R. W. Introduction to Mass Spectrometry and Its Applications, Prentice-Hall: Englewood Cliffs, 1965. 3. Aston, F. W. Mass Spectra and Isotopes, Longmans, Green & Co.: New York, 1942, Chapter V. 4. De Laeter, J. R. International Journal of Mass Spectrometry 1998, 178, 1-7. 5. Middleditch B. S. Practical Mass Spectrometry, Plenum Press: New York, 1979. 6. Aston, F. W. Mass Spectra and Isotopes, Longmans, Green & Co: New York, 1942, Chapter VI. 7. Dempster, A. J. Physical Review 1918, 11, 316-24. 8. Bleakney, W. Physical Review 1929, 34, 157-60. 9. Watson, J. T. Introduction to Mass Spectrometry, Lippincott-Raven: Philadelphia, 1997. 10. Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B. Anal. Chem. 1985, 57, 675-79. 11. Proceedings of 52 nd Annual Conference of American Society for Mass Spectrometry, American Society for Mass spectrometry, 2004. 12. Stephens, W. E. Physical Review 1946, 69, 691. 13. Wiley, W. C.; Mclaren, I. H. Review of Scientific Instruments 1955, 26, 1150-57. 14. Guilhaus, M.; Mlynski, V.; Selby, D. Rapid Communications in Mass Spectrometry 1997, 11, 951-62. 15. Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. International Journal of Mass Spectrometry and Ion Processes 1987, 78, 53-68. 16. Tanaka, K.; Waki, H.; Ido, Y.; Akita, S.; Yoshida, Y.; Yohida, T. Rapid Communications in Mass Spectrometry 1988, 2(8), 151-53. 116

PAGE 131

117 17. Brown, R. S.; Lennon, J. J. Analytical Chemistry 1995, 67, 1998-2003. 18. Whittal, R. M.; Li, L. Analytical Chemistry 1995, 67, 1950-54. 19. Vestal, M. L.; Juhasz, P.; Martin, S. A. Rapid Communications in Mass Spectrometry 1995, 9, 1044-50. 20. Colby, S. M.; King, T. B.; Reilly, J. P. Rapid Communications in Mass Spectrometry 1994, 8, 865-68. 21. Mamyrin, B. A.; Karataev, V. I.; Shmikk, D. V.; Zagulin, V. A. Zhurnal Eksperimentalnoi I Teoreticheskoi Fiziki 1973, 64, 82-89. 22. Mamyrin, B. A.; Shmikk, D. V. Zhurnal Eksperimentalnoi I Teoreticheskoi Fiziki 1979, 76, 1500-05. 23. Mamyrin, B. A. International Journal of Mass Spectrometry 2001, 206, 251-66. 24. Vastola, F. J.; Mumma, R. O.; Pirone, A. J. Organic Mass Spectrometry 1970, 3, 101-04. 25. Edmond de Hoffmann; Vincent Strooband Mass Spectrometry, Principles and Applications, Second ed. Wiley: New York, 2001. 26. Zenobi, R.; Knochenmuss, R. Mass Spectrometry Reviews 1998, 17, 337-66. 27. Wei, J.; Buriak, J. M.; Siuzdak, G. Nature 1999, 399, 243-46. 28. Uhlir, A. Bell System Technical Journal 1956, 35, 333-47. 29. Li, X.; Bohn, P. W. Applied Physics Letters 2000, 77, 2572-74. 30. Archer, R. J. Journal of Physics and Chemistry of Solids 1960, 14, 104-10. 31. Beale, M. I. J.; Chew, N. G.; Uren, M. J.; Cullis, A. G.; Benjamin, J. D. Applied Physics Letters 1985, 46, 86-88. 32. Canham L. T. Properties of Porous Silicon, Short Run Press Ltd.: London, 1997. 33. Shen, Z. X.; Thomas, J. J.; Averbuj, C.; Broo, K. M.; Engelhard, M.; Crowell, J. E.; Finn, M. G.; Siuzdak, G. Analytical Chemistry 2001, 73, 612-19. 34. Dubin, V. M. Surface Science 1992, 274, 82-92. 35. Kang, Y.; Jorne, J. Journal of the Electrochemical Society 1993, 140, 2258-65. 36. Allongue, P.; Devilleneuve, C. H.; Pinsard, L.; Bernard, M. C. Applied Physics Letters 1995, 67, 941-43.

PAGE 132

118 37. John, G. C.; Singh, V. A. Physical Review B 1995, 52, 11125-31. 38. Corbett, J. W.; Shereshevskii, D. I.; Verner, I. V. Physica Status Solidi A-Applied Research 1995, 147, 81-89. 39. Beale, M. I. J.; Benjamin, J. D.; Uren, M. J.; Chew, N. G.; Cullis, A. G. Journal of Crystal Growth 1986, 75, 408-14. 40. Smith, R. L.; Chuang, S. F.; Collins, S. D. Journal of Electronic Materials 1988, 17, 533-41. 41. Gaspard, F.; Bsiesy, A.; Ligeon, M.; Muller, F.; Herino, R. Journal of the Electrochemical Society 1989, 136, 3043-46. 42. Canham, L. T. Applied Physics Letters 1990, 57, 1046-48. 43. Kompan, M. E.; Kuzminov, E. G.; Kulik, V. B.; Novak, I. I.; Beklemyshev, V. I. Jetp Letters 1996, 64, 748-53. 44. Feng Z. C.; Tsu R. Porous Silicon, World Scientific: Singapore, 1994. 45. Lehmann, V.; Gosele, U. Advanced Materials 1992, 4, 114-16. 46. Rao, A. V.; Ozanam, F.; Chazalviel, J. N. Journal of the Electrochemical Society 1991, 138, 153-59. 47. Brandt, M. S.; Fuchs, H. D.; Stutzmann, M.; Weber, J.; Cardona, M. Solid State Communications 1992, 81, 307-12. 48. Brandt, M. S.; Ready, S. E.; Boyce, J. B. Applied Physics Letters 1997, 70, 188-90. 49. Earwaker, L. G.; Farr, J. P. G.; Grzeszczyk, P. E.; Sturland, I.; Keen, J. M. Nuclear Instruments & Methods in Physics Research Section B-Beam Interactions with Materials and Atoms 1985, 9, 317-20. 50. Grosman, A.; Ortega, C.; Siejka, J.; Chamarro, M. Journal of Applied Physics 1993, 74, 1992-96. 51. Loni, A.; Simons, A. J.; Canham, L. T.; Phillips, H. J.; Earwaker, L. G. Journal of Applied Physics 1994, 76, 2825-32. 52. Fuchs, H. D.; Stutzmann, M.; Brandt, M. S.; Rosenbauer, M.; Weber, J.; Breitschwerdt, A.; Deak, P.; Cardona, M. Physical Review B 1993, 48, 8172-83. 53. Stutzmann, M.; Brandt, M. S.; Rosenbauer, M.; Weber, J.; Fuchs, H. D. Physical Review B 1993, 47, 4806-09.

PAGE 133

119 54. Kalkhoran, N. M.; Namavar, F.; Maruska, H. P. Applied Physics Letters 1993, 63, 2661-63. 55. Richter, A.; Steiner, P.; Kozlowski, F.; Lang, W. IEEE Electron Device Letters 1991, 12, 691-92. 56. Loni, A.; Canham, L. T.; Berger, M. G.; Arens-Fischer, R.; Munder, H.; Luth, H.; Arrand, H. F.; Benson, T. M. Thin Solid Films 1996, 276, 143-46. 57. Hirschman, K. D.; Tsybeskov, L.; Duttagupta, S. P.; Fauchet, P. M. Nature 1996, 384, 338-41. 58. Stievenard, D.; Deresmes, D. Applied Physics Letters 1995, 67, 1570-72. 59. Anderson, R. C.; Muller, R. S.; Tobias, C. W. Sensors and Actuators A-Physical 1990, 23, 835-39. 60. Lauerhaas, J. M.; Sailor, M. J. Science 1993, 261, 1567-68. 61. Berger, M. G.; Thonissen, M.; ArensFischer, R.; Munder, H.; Luth, H.; Arntzen, M.; Theiss, W. Thin Solid Films 1995, 255, 313-16. 62. Lerondel, G.; Romestain, R.; Vial, J. C.; Thonissen, M. Applied Physics Letters 1997, 71, 196-98. 63. Buriak, J. M.; Stewart, M. P.; Canham, L. T.; Reeves, C. L.; Newey, J. P.; Houlton, M. R.; Cox, T. I. Abstracts of Papers of the American Chemical Society 1999, 218, U414-U415. 64. Bayliss, S. C.; Heald, R.; Fletcher, D. I.; Buckberry, L. D. Advanced Materials 1999, 11, 318-21. 65. Drott, J.; Lindstrom, K.; Rosengren, L.; Laurell, T. Journal of Micromechanics and Microengineering 1997, 7, 14-23. 66. Thomas, J. J.; Shen, Z. X.; Crowell, J. E.; Finn, M. G.; Siuzdak, G. Proceedings of the National Academy of Sciences of the United States of America 2001, 98, 4932-37. 67. Kruse, R. A.; Rubakhin, S. S.; Romanova, E. V.; Bohn, P. W.; Sweedler, J. V. Journal of Mass Spectrometry 2001, 36, 1317-22. 68. Alimpiev, S.; Nikiforov, S.; Karavanskii, V.; Minton, T.; Sunner, J. Journal of Chemical Physics 2001, 115, 1891-901. 69. Benedetto, G.; Boarino, L.; Brunetto, N.; Rossi, A.; Spagnolo, R.; Amato, G. Philosophical Magazine B-Physics of Condensed Matter Statistical Mechanics Electronic Optical and Magnetic Properties 1997, 76, 383-93.

PAGE 134

120 70. Benedetto, G.; Boarino, L.; Spagnolo, R. Applied Physics A-Materials Science & Processing 1997, 64, 155-59. 71. Shen, Z. X.; Thomas, J. J.; Siuzdak, G.; Blackledge, R. D. Journal of Forensic Sciences 2004, 49, 1028-35. 72. Thomas, J. J.; Shen, Z. X.; Blackledge, R.; Siuzdak, G. Analytica Chimica Acta 2001, 442, 183-90. 73. Compton, B. J.; Siuzdak, G. Spectroscopy-An International Journal 2003, 17, 699-713. 74. Prenni, J. E.; Shen, Z. X.; Trauger, S.; Chen, W.; Siuzdak, G. Spectroscopy-An International Journal 2003, 17, 693-98. 75. Shen, Z. X.; Go, E. P.; Gamez, A.; Apon, J. V.; Fokin, V.; Greig, M.; Ventura, M.; Crowell, J. E.; Blixt, O.; Paulson, J. C.; Stevens, R. C.; Finn, M. G.; Siuzdak, G. Chembiochem 2004, 5, 921-27. 76. Go, E. P.; Shen, Z. X.; Harris, K.; Siuzdak, G. Analytical Chemistry 2003, 75, 5475-79. 77. Dutta, S. M.; Rousell, D. J.; Murray, K. K. Abstracts of Papers of the American Chemical Society 2003, 225, U233. 78. Laboy, J. L.; Kim, J. K.; Murray, K. K. Abstracts of Papers of the American Chemical Society 2003, 225, U121. 79. Rousell, D. J.; Dutta, S. M.; Murray, K. K. Abstracts of Papers of the American Chemical Society 2003, 225, U121. 80. Rousell, D. J.; Dutta, S. M.; Murray, K. K. Abstracts of Papers of the American Chemical Society 2003, 225, U824. 81. Huikko, K.; Ostman, P.; Sauber, C.; Mandel, F.; Grigoras, K.; Franssila, S.; Kotiaho, T.; Kostiainen, R. Rapid Communications in Mass Spectrometry 2003, 17, 1339-43. 82. Laiko, V. V.; Taranenko, N. I.; Berkout, V. D.; Musselman, B. D.; Doroshenko, V. M. Rapid Communications in Mass Spectrometry 2002, 16, 1737-42. 83. Go, E. P.; Prenni, J. E.; Wei, J.; Jones, A.; Hall, S. C.; Witkowska, H. E.; Shen, Z. X.; Siuzdak, G. Analytical Chemistry 2003, 75, 2504-06. 84. Trauger, S. A.; Go, E. P.; Shen, Z. X.; Apon, J. V.; Compton, B. J.; Bouvier, E. S. P.; Finn, M. G.; Siuzdak, G. Analytical Chemistry 2004, 76, 4484-89.

PAGE 135

121 85. Lewis, W. G.; Shen, Z. X.; Finn, M. G.; Siuzdak, G. International Journal of Mass Spectrometry 2003, 226, 107-16. 86. Kruse, R. A.; Li, X. L.; Bohn, P. W.; Sweedler, J. V. Analytical Chemistry 2001, 73, 3639-45. 87. Canham, L. T.; Houlton, M. R.; Leong, W. Y.; Pickering, C.; Keen, J. M. Journal of Applied Physics 1991, 70, 422-31. 88. Vanderhoeven, J. S.; Franken, H. C. M. Caries Research 1982, 16, 375-83. 89. Snyder, A. P.; Thornton, S. N.; Dworzanski, J. P.; Meuzelaar, H. L. C. Field Analytical Chemistry and Technology 1996, 1, 49-59. 90. Todd, P. J.; Schaaff, T. G.; Chaurand, P.; Caprioli, R. M. Journal of Mass Spectrometry 2001, 36, 355-69. 91. DIOS-Target Plates Additional User Notes, Waters Co., 2004. 92. Beckmann, K. H. Angewandte Chemie-International Ed. 1965, 4, 968-&. 93. Canham, L. T.; Houlton, M. R.; Leong, W. Y.; Pickering, C.; Keen, J. M. Journal of Applied Physics 1991, 70, 422-31. 94. Wenzel, R. N. Journal of Industrial and Engineering Chemistry 1936, 28, 988. 95. Cassie, A. B. D.; Baxter, S Transactions of the Faraday Society 1944, 40, 546-51. 96. Arakawa, R.; Shimomae, Y.; Morikawa, H.; Ohara, K.; Okuno, S. Journal of Mass Spectrometry 2004, 39, 961-65. 97. Meng, J. C.; Averbuj, C.; Lewis, W. G.; Siuzdak, G.; Finn, M. G. Angewandte Chemie-International Edition 2004, 43, 1255-60. 98. Meng, J. C.; Siuzdak, G.; Finn, M. G. Chemical Communications 2004, 2108-09. 99. Breslow, R. Tetrahedron Letters 1959, 22-26. 100. Ricardo, A.; Carrigan, M. A.; Olcott, A. N.; Benner, S. A. Science 2004, 303, 196. 101. Chapelle, S.; Verchere, J. F. Tetrahedron 1988, 44, 4469-82. 102. Verchere, J. F.; Hlaibi, M. Polyhedron 1987, 6, 1415-20. 103. Roussis, S. G.; Proulx, R. Analytical Chemistry 2002, 74, 1408-14. 104. Ricardo, A. Bioorganic Molecules in the Cosmos and the Origins of Darwinian Molecular Systems, Doctoral Dissertation, 2005.

PAGE 136

122 105. Kadish, K.; Smith, K. M.; Guilard, R. The Porphyrin Handbook, Academic Press: Burlington, 1999. 106. Gill, H. S.; Harmjanz, M.; Scott, M. J. Abstracts of Papers of the American Chemical Society 2004, 227, U1557. 107. Buchler, J. W.; Puppe, L. Annalen der Chemie-Justus Liebig 1974, 1046-62. 108. Harmjanz, M.; Gill, H. S.; Scott, M. J. Journal of the American Chemical Society 2000, 122, 10476-77. 109. Jackson, A. H.; Kenner, G. W.; Smith, K. M.; Aplin, R. T.; BUDZIKIE.H; e; ssi, C. Tetrahedron 1965, 21, 2913-&. 110. Gallegos, E. J.; Sundararaman, P. Mass Spectrometry Reviews 1985, 4, 55-85. 111. Laycock, J. D.; Ferguson, J. A.; Yost, R. A.; Quirke, J. M. E.; Rohrer, A.; Ocampo, R.; Callot, H. Journal of Mass Spectrometry 1997, 32, 978-83. 112. Schafer, M.; Budzikiewicz, H. Journal of Mass Spectrometry 2001, 36, 1062-68. 113. Frauenkron, M.; Berkessel, A.; Gross, J. H. European Mass Spectrometry 1997, 3, 427-38. 114. Carlson, J. B.; Vouros, P. Journal of Mass Spectrometry 1996, 31, 1403-08. 115. Morkved, E.; Kjosen, H.; Vedde, J.; Uggerud, E. European Journal of Mass Spectrometry 2002, 8, 35-39. 116. Jones, R. M.; Lamb, J. H.; Lim, C. K. Rapid Communications in Mass Spectrometry 1995, 9, 921-23. 117. Fenyo, D.; Chait, B. T.; Johnson, T. E.; Lindsey, J. S Journal of Porphyrins and Phthalocyanines 1997, 1, 93-99. 118. Green, M. K.; Medforth, C. J.; Muzzi, C. M.; Nurco, D. J.; Shea, K. M.; Smith, K. M.; Lebrilla, C. B.; Shelnutt, J. A. European Mass Spectrometry 1997, 3, 439-51. 119. Aiello, I.; Di Donna, L.; Ghedini, M.; La Deda, M.; Napoli, A.; Sindona, G. Analytical Chemistry 2004, 76, 5985-89. 120. Stulz, E.; Mak, C. C.; Sanders, J. K. M. Journal of the Chemical Society-Dalton Transactions 2001, 604-13. 121. Vandell, V. E.; Limbach, P. A. Journal of Mass Spectrometry 1998, 33, 212-20. 122. Dale, M. J.; Costello, K. F.; Jones, A. C.; LangridgeSmith, P. R. R. Journal of Mass Spectrometry 1996, 31, 590-601.

PAGE 137

123 123. Antolovich, M.; Nguyen, T. H.; Paddonrow, M. N.; Willett, G. D. Organic Mass Spectrometry 1992, 27, 1034-41. 124. Srinivasan, N.; Haney, C. A.; Lindsey, J. S.; Zhang, W. Z.; Chait, B. T. Journal of Porphyrins and Phthalocyanines 1999, 3, 283-91. 125. Lidgard, R. O.; McConnell, D. B.; Black, D. S.; Kumar, N.; Duncan, M. W. Journal of Mass Spectrometry 1996, 31, 1443-45.

PAGE 138

BIOGRAPHICAL SKETCH Qian Li was born in Jiang Su province, China, on February 18th, 1974 to Fashun Li and Guiying Zhang. She was in her second grade of the elementary school when the family moved to Ya’an, a beautiful small town in Sichuan Province. At age 17, she was admitted to Chemistry Department, Nanjing University, China. Seven years later, she received the bachelor’s degree and the master’s degree in organic chemistry. After a brief break from school, she joined the Ph.D. program at the Department of Chemistry, University of Florida, in Spring 2001. 124