Boca Raton London New York
CRC Press is an imprint of the Taylor & Francis Group, an informa business
CRC Press Tayl...
39 downloads
627 Views
16MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Boca Raton London New York
CRC Press is an imprint of the Taylor & Francis Group, an informa business
CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2010 by Taylor and Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number: 978-1-4200-8373-6 (Hardback) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright. com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Practical aspects of ion trap mass spectrometry / edited by Raymond E. March, John F.J. Todd. p. cm. -- (Modern mass spectrometry) Includes bibliographical references and index. ISBN 0-8493-4452-2 (vol. 1) 1. Mass spectrometry. I. March, Raymond E. II. Todd, John F.J. III. Series. QD96.M3P715 1995 539.7’.028’7--dc20 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com
95-14146
To ion trappers, young and old, everywhere.
Contents Preface.......................................................................................................................xi Volume IV Contents................................................................................................xxi Editors.....................................................................................................................xxv Contributors...........................................................................................................xxix
part I Ion Reactions Chapter 1 Ion/Ion Reactions in Electrodynamic Ion Traps....................................3 Jian Liu and Scott A. McLuckey Chapter 2 Gas-Phase Hydrogen/Deuterium Exchange in QuadrupoleIon Traps.............................................................................................. 35 Joseph E. Chipuk and Jennifer S. Brodbelt Chapter 3 Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry in a Quadrupole Linear Ion Trap....................................... 59 Graeme C. McAlister and Joshua J. Coon
Part II Ion Conformation and Structure Chapter 4 Chemical Derivatization and Multistage Tandem Mass Spectrometry for Protein Structural Characterization........................ 83 Jennifer M. Froelich, Yali Lu, and Gavin E. Reid Chapter 5 Fourier Transform Ion Cyclotron Resonance Mass Spectrometry in the Analysis of Peptides and Proteins........................................... 121 Helen J. Cooper Chapter 6 MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies: Wine Astringency Approached by ESI-IT-MS............. 153 Benoît Plet and Jean-Marie Schmitter
vii
viii
Contents
Chapter 7 Structure and Dynamics of Trapped Ions.......................................... 169 Joel H. Parks Chapter 8 Applications of Traveling Wave Ion Mobility-Mass Spectrometry.....................................................................................205 Konstantinos Thalassinos and James H. Scrivens
part III Ion Spectroscopy Chapter 9 The Spectroscopy of Ions Stored in Trapping Mass Spectrometers.................................................................................... 239 Matthew W. Forbes, Francis O. Talbot, and Rebecca A. Jockusch Chapter 10 Sympathetically-Cooled Single Ion Mass Spectrometry................. 291 Peter Frøhlich Staanum, Klaus Højbjerre, and Michael Drewsen Chapter 11 Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!....................................................................... 327 Fernande Vedel
part IV Practical Applications Chapter 12 Boundary-Activated Dissociations (BAD) in a Digital Ion Trap (DIT)....................................................................................... 367 Francesco L. Brancia, Luca Raveane, Alberto Berton, and Pietro Traldi Chapter 13 The Study of Ion/Molecule Reactions at Ambient Pressure with Ion Mobility Spectrometry and Ion Mobility/ Mass Spectrometry.......................................................................... 387 Gary A. Eiceman and John A. Stone Chapter 14 The Role of Trapped Ion Mass Spectrometry for Imaging.............. 417 Timothy J. Garrett and Richard A. Yost
ix
Contents
Chapter 15 Technology Progress and Application in GC/MS and GC/MS/MS................................................................. 439 Mingda Wang and John E. George III Chapter 16 Remote Monitoring of Volatile Organic Compounds in Water by Membrane Inlet Mass Spectrometry................................ 491 Romina Pozzi, Paola Bocchini, Francesca Pinelli, and Guido C. Galletti Author Index.........................................................................................................509 Subject Index......................................................................................................... 513
Preface This monograph is Volume V of a miniseries devoted to (i) practical aspects of applications of mass spectrometry for the study of gaseous ions confined in ion traps, and (ii) treatments of the theory of ion confinement in each ion-trapping device. Volumes I–III were published in 1995 under the title Practical Aspects of Ion Trap Mass Spectrometry. Volume III, Chemical, Environmental and Biomedical Applications, is a companion to Volumes I and II, subtitled Fundamentals of Ion Trap Mass Spectrometry and Ion Trap Instrumentation, respectively. Volumes I–III are concerned principally with the history, theory, and applications of the quadrupole ion trap and, to a lesser degree, of the quadrupole mass filter. Volume V, published in 2009 under the title Practical Aspects of Trapped Ion Mass Spectrometry, and subtitled Applications, is a compa nion to Volume IV, subtitled Theory and Instrumentation. The contents of Volume IV are given following the conclusion of this preface. The history of the quadrupole ion trap was presented in tabular form in Chapter 2 of Volume I as “The Ages of the Ion Trap” and, upon revisiting this table, one is struck by the spectacular progress that has been made in the ion-trapping field since 1995. In the Preface to Volume II, we noted two exceptional landmarks in this history: first, the invention of the quadrupole ion trap (and quadrupole mass filter) by Wolfgang Paul and Hans Steinwedel, which was recognized by the award of the 1989 Nobel Prize in Physics, in part, to Wolfgang Paul and Hans Dehmelt; and, second, the discovery announced in 1983 of the mass-selective instability scan by George C. Stafford, Jr. On these two landmarks rested the entire field of ion trap mass spectrometry. One of the table entries for 1990 was “Electrospray Ionization (Van Berkel, Glish, and McLuckey),” and Chapter 3 of Volume II was devoted to “Electrospray and the Quadrupole Ion Trap.” A further contribution, entitled “Electrospray/Ion Trap Mass Spectrometry – Applications,” by Hung-Yu Lin and Robert D. Voyksner, appeared as Chapter 14 in Volume III. The advent of electrospray ionization and its ready compatibility with ion-trapping devices has brought about a revolution in the accessibility of covalent compounds for examination by mass spectrometry in general and by quadrupole ion trap mass spectrometry in particular. For their development of soft desorption ionization methods for mass spectrometric analyses of biological macromolecules, John Fenn and Koicho Tanaka received the Nobel Prize in Chemistry for 2002. We add our congratulations and thanks to these Nobelists and to those from the mass spectrometry community. The enormous impact that electrospray ionization has made in biochemistry in general, and in the study of proteins in particular, is remarkable. Virtually every mass spectrometry laboratory is now equipped with electrospray ionization; compounds for which derivatization was previously essential for examination by electron impact can now be examined facilely in solution by direct infusion to an electrospray ionization source. As testament to this situation, more than half of the chapters presented in Volumes IV and V are concerned with the use of electrospray ionization. The practice of trapping gaseous ions and the applications thereof have expanded considerably during the past decade or so, in part due to the use of electrospray xi
xii
Preface
ionization but also as witnessed by the substantial growth in popularity of quadrupole ion traps and of Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometers, instruments that hitherto were regarded as being rivals rather than complementary technologies. In addition, we have seen the nascence of new methods for trapping ions, such as the Orbitrap™, the digital ion trap (DIT), the rectilinear ion trap (RIT), and the toroidal ion trap. Furthermore, during this period, there have been significant advances in the development and application of the quadrupole ion trap and of the quadrupole mass filter, both standalone and in concatenation with other mass spectrometric instruments, for example, with Fourier transform ion cyclotron resonance and with time-of-flight (TOF) mass spectrometers. New and/or modified existing methods for ion processing have been developed and applied; these methods include electron capture dissociation (ECD), electron transfer dissociation (ETD), charge inversion, proton transfer reaction (PTR), electron transfer (ET), and ion attachment (IA). Other recent advances involving the coupling of ion mobility spectrometry (IMS) with mass spectrometry have brought about the introduction of high-field asymmetric waveform ion mobility spectrometry (FAIMS) and traveling wave ion mobility mass spectrometry (TWIM-MS). Indeed, so many advances have occurred in the ion-trapping field that we needed to consider a somewhat broader definition of ion trapping compared with what has been employed hitherto; after several iterations, we arrived at the definition proposed in Section 1.1 of Volume IV, “an ion is ‘trapped’ when its residence time within a defined spatial region exceeds that had the motion of the ion not been impeded in some way.” Clearly, this definition includes those various forms of ion mobility spectrometry mentioned above. Armed with this definition of ‘trapped ions,’ it seemed appropriate to the editors that a further volume in this mini-series could be undertaken, not limited to quadrupole devices but encompassing advances in all aspects of trapped ion mass spectrometry. When a commercial product has achieved a degree of market acceptance, which we believed was the case for the three volumes of Practical Aspects of Ion Trap Mass Spectrometry, one is reluctant to lose the connectivity within the miniseries upon embracing an expansion of the field in question. Fortunately, a minor word change to Practical Aspects of Trapped Ion Mass Spectrometry saved the day. With this small but significant change in title, the expanded field could be considered and included within the ‘practical aspects of ion trapping’ rubric. The collective response to our subsequent approaches to potential authors in the expanded ion-trapping field was near overwhelming, so much so that in fact two monographs, Volumes IV and V, have resulted from this endeavor. Volume IV is entitled Theory and Instrumentation and is composed of six parts: Fundamentals, New Ion Trapping Techniques, Fourier Transform Mass Spectrometry, Quadrupole Rod Sets, 3D-Quadrupole Ion Trap Mass Spectrometry, and Photochemistry of Trapped Ions. Volume V is entitled Applications and features four parts: Ion Reactions, Ion Conformation and Structure, Ion Spectroscopy, and Practical Applications. Part 1. Ion Reactions is composed of three chapters in which ion reactions, that is, ion/neutral reactions or ion/ion reactions, are examined. Several ion-trapping devices have the capability for examining reactions of ions with neutral species and other ionic species where the extent of the reaction is monitored by the mass
Preface
xiii
spectrometric function of the instrument. The quadrupole, or electrodynamic, ion trap is inherently compatible with the study of ion/ion reactions due to its unique ability to store simultaneously ions of both polarities in overlapping regions of space. In Chapter 1 is presented a review of the instrumental requirements for the study of ion/ion reactions. Particular emphasis is given to the use of an electrodynamic ion trap for the study of multiply-protonated peptide molecules with anions. The trapped ions assume characteristic sets of m/z-dependent frequencies of motion in the oscillating quadrupole field of the ion trap, which allows ready manipulation of ions for ion isolation and activation, both of which are common elements in a tandem mass spectrometric experiment. The ‘tandem-in-time’ nature of the ion trap MSn experiment provides well-defined conditions for ion/ion reactions and permits determination of ion genealogy. A bath gas, such as helium at ca 1 mTorr, intended originally to cool the ions to the center of the trap so as to enhance both sensitivity and mass resolution upon mass analysis, improves ion/ion reaction efficiencies by maximizing the spatial overlap and minimizing the translational energies of the two ion clouds. Chapter 2 is focused on a particular type of ion/neutral reaction, namely that in which hydrogen atoms in the ion exchange with deuterium in the neutral reaction partner. The quadrupole ion trap mass spectrometer is well suited for investigations of such hydrogen/denterium (H/D) exchange reactions because the kinetics of reactions can be monitored accurately by varying the ion storage time. As illustrated in this chapter, applications of H/D exchange in quadrupole ion traps range from those involving small organic molecules, especially involving comparisons of isomers, to larger biological molecules for which conformational effects play a significant role. Chapter 3 considers the prospect of utilizing multiple ion-manipulation methodologies, which are available with ion trap mass spectrometers, to achieve whole protein sequence analysis; such analysis is described as top-down proteomics. The basis of this approach is the implementation of multi-functional tools for systematic ion manipulation and processing, where ion/ion reactions such as electron transfer, proton transfer, and ion attachment, represent one family of such tools. These technologies are inter-meshed with conventional ion trap processing methodologies of ion isolation and collision-induced dissociation. Concatenation of MSn scan functions from these individual components can constitute a versatile approach that promises to accelerate markedly the field of large molecule mass spectrometry. The practical application of these processing methodologies in a linear ion trap requires modification of the ion trap electronics to allow for the superimposition of a radiofrequency voltage on the end lenses, which allows for charge-sign independent trapping. Part 2. Ion Conformation and Structure presents discussions of structural characterization of proteins and peptides using quadrupole ion trap mass spectrometry, Fourier transform ion cyclotron resonance mass spectrometry, and the novel method known as traveling wave ion mobility mass spectrometry. In addition to the observation of collective fluctuations of the molecular substructures within biomolecules, the organization of atoms in small ion clusters is investigated using electron diffraction. In Chapter 4 is discussed the ‘bottom-up’ or ‘shotgun’ tandem mass spectrometric approach to protein identification and characterization, which is the complementary method to top-down proteomics that is discussed in Chapter 3. In order to overcome the limitations of bottom-up proteomics, chemical derivatization strategies are
xiv
Preface
explored that direct the fragmentation of protonated peptides to the formation of sequence product ions. Such strategies can be employed to direct fragmentation to the formation of non-sequence product ions. Chapter 5 provides an overview of Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry and its applications in the structural characterization of peptides and proteins. The principles of FT-ICR, that is, ion motion, ion excitation/ detection, and instrumental considerations, are discussed and an explanation of the features of FT-ICR that make it so suitable for peptide/protein analysis is presented. New methods for the fragmentation of peptide and protein ions in FT-ICR mass spectrometry, such as sustained off-resonance irradiation collision-induced dissociation (SORI-CID), infrared multiphoton dissociation (IRMPD), blackbody infrared radiative dissociation (BIRD), surface-induced dissociation (SID), and electron capture dissociation (ECD), are described in detail. Innovative hybrid FT-ICR instruments, which have recently become available, are reviewed. In conclusion, the chapter discusses the applications of FT-ICR in ‘bottom-up’ and ‘top-down’ proteomics. Chapter 6 is devoted to the tandem mass spectrometric investigation of supramolecular assemblies of peptides with non-covalently-bonded polyphenols. The quest of the specific investigation recounted in this chapter was to gain insight at a molecular level into the interaction of polyphonies with proline-rich peptides, and to develop a future analytical methodology for the evaluation of astringency, specifically the astringency of wine. Two relevant points of interest are (i) polyproline peptides are subjects of intensive study, as is shown in the following chapter, Chapter 7, and (ii) the chemistry of proteins with non-covalently-bonded ligands is under examination because of the possibility of facile transport of ubiquitous compounds of doubtful environmental value into organs such as the liver. Analysis of these supramolecular assemblies of proline-rich peptides with a wide range of flavonoids (polyphenols), by means of energy-resolved mass spectrometry (ERMS), led to the creation of a relative affinity scale of the proline-rich peptides for the flavonoids examined. Chapter 7 describes quadrupole ion trap studies of the organization of atoms in small ion clusters and the observation of collective fluctuations of the molecular substructures within biomolecules. The introduction of new ion sources, in particular metal-cluster aggregation sources and electrospray ionization, have provided unique opportunities to produce ion beams composed of metal atom clusters and biomolecules, respectively. Metal clusters are formed with a single charge but in a broad array of masses corresponding to the number of atoms, whereas biomolecular ions are generated for a single species in an ensemble of charge states. These studies take advantage of the advances in ion trap technology for flexible and reliable ion-cloud manipulation of higher mass ions required for the electron diffraction of, for example, Ag55+ and Au21−, and for fluorescence measurements of dye-derivatized polyproline peptides. The results presented here enunciate clearly the ways in which these methods have contributed to our understanding of how the atoms are organized in small metal clusters and of the temperature dependence of local fluctuations of biomolecular conformations. In a drift cell, ions migrate through a counter-flowing buffer gas in the presence of a low electric field. The use of the drift cell in this manner is often referred to as ion mobility spectrometry (IMS), which is now a well-established analytical technique that is employed throughout the world for the detection of explosives, drugs, and
Preface
xv
chemical warfare agents. Ion mobility measures the time it takes for ions to traverse a drift tube. Ion separation occurs as a result of interactions between these ions and the buffer gas; the extent of separation depends not only on the mass and charge, as may be anticipated, but on the shape (or conformation) of the ion, which is unique to ion mobility spectrometry. The study of ion/molecule reactions in IMS is discussed in Chapter 13. Two alternative approaches have been introduced recently; these are high-field asymmetric waveform ion mobility spectrometry (FAIMS) and traveling wave ion mobility spectrometry (TWIMS). In Volume IV, Chapter 5 was devoted to a discussion of FAIMS. In Chapter 8 is presented an account of traveling wave ion mobility spectrometry. Unlike drift cell ion mobility experiments, where a constant low electric field is applied to the mobility cell, traveling wave ion mobility spectrometry uses a traveling wave comprising a series of transient direct current voltages to propel ions through a stacked-ring ion guide (SRIG) to which radiofrequency voltages have been applied to consecutive electrodes. The SRIG consists of a series of ring electrodes that are arranged orthogonally to the ion transmission axis, and opposite phases of radiofrequency voltage are applied to adjacent rings. When a transient direct current potential, superimposed upon this radiofrequency potential, is applied to one pair of adjacent ring electrodes, ions are propelled through the SRIG. The transient direct current potential moves along ring electrode pairs across the length of the SRIG at regular time intervals, generating a sequence of traveling waves (T-Waves). This particular configuration of SRIG is referred to as a traveling wave ion guide (TWIG). A concatenation of three TWIGs has been incorporated within a Q-TOF geometry to create the Synapt™ HDMS system, a commercial instrument incorporating ion mobility separation. Most applications using the Synapt have focused on studying the conformation of proteins and protein complexes. Among the applications discussed here is a study of the prion protein, a fibril-forming protein involved in prion diseases. Prions are a class of fatal, infectious, neurodegenerative diseases that affect both humans and animals. Part 3. Ion Spectroscopy. In Chapter 9, we return to the theme of ion photodissociation, which was included also in Volume IV, Part 6, in an exploration of trapped-ion photodissociation, electron photodetachment, and fluorescence. Trapped-ion fluorescence may offer an alternative approach for the elucidation of ion conformation. Whereas these spectroscopic experiments require high ion densities, much attention is directed to the spectroscopic study of single ions confined in an ion trap. Chapters 10 and 11 are illustrative of such studies, with the former devoted to the study of a single molecular ion in a linear ion trap and the latter to a single atomic ion in P aul-type ion traps. While both types of studies require extensive cooling of the subject ion, once such cooling has been achieved, the ions can remain confined for many hours. Chapter 9 contains a discussion of practical aspects of experimental design for the pursuit of photodissociation, electron photodetachment, and fluorescence of trapped, mass-selected organic ions. A review is given of the wide range of possible spectroscopic experiments that can be combined fruitfully with the ion storage and massselective capabilities provided by ion-trapping devices for the scrutiny of molecular ions. Details of the modification of a quadrupole ion trap together with the results from extensive modeling of the apparatus are presented. Photodissociation is the
xvi
Preface
fragmentation of an ion due normally to the absorption of light in a narrow wavelength range. In this chapter, the excitation source is stepped or scanned through a range of wavelengths, while monitoring ion intensities as a function of excitation wavelength, in order to construct an optical spectrum. When the extent of photodissociation is monitored as a function of excitation wavelength, the process is termed ‘action’ or ‘consequence’ spectroscopy. The application of action spectroscopy, which has been used to generate vibrational (infrared) and vibronic (ultraviolet-visible) spectra of mass spectrometric precursor and product ions, is discussed. Fluorescence spectroscopy, in which radiative emission from activated ions can be monitored using a photon detector, is shown here to be highly sensitive to a chromophore’s local environment, making it an excellent probe of ion conformation. In Chapter 10, the novel technique of sympathetically-cooled single ion mass spectrometry (SCSI-MS) is described; this technique relies on the measurement of the resonant excitation frequency of one of the two oscillatory modes of a trapped and crystallized linear two-ion system consisting of one laser-cooled atomic ion of known mass and the a priori unknown atomic or molecular ion, whose mass is to be determined. The mass of the unknown ion can be deduced from this measured frequency. The crystallization of the two-ion system results from the sympathetic cooling of the unknown ion through the Coulomb interaction with the laser-cooled ion; the two-ion system is aligned along the axis of the linear ion trap. Resonant excitation can be promoted by applying a sinusoidally-varying electric field along this axis. The resonance frequencies are determined by monitoring fluorescence from the laser-cooled ion while scanning the period of the applied driving force. When the period is equal to the period of one of the two oscillatory modes of the two-ion system, that is, the centerof-mass mode where the ions move in phase, or the breathing mode where the ions move with opposite phase, the motion of the ions becomes highly excited. Examples of molecular ions examined here are CaO+, MgD+, and MgH+. Chapter 11 gives a review of atomic clocks of the future using single ions confined in relatively small ion traps; small or miniature ion traps have been discussed in Volume IV, Chapter 2. The purpose of Chapter 11 is to expound upon the specific topic of atomic clocks utilizing ion traps and the new challenges engaged presently for the measurement of time with extremely high precision. A major part of physics is dedicated permanently to the enhancement of measurement and more precise definitions of the fundamental units. Among them, the time unit, the second, is one of the most crucial units necessary for the advancement of knowledge. The time unit was the first for which the definition put aside any material systems in that Greenwich Mean Time (GMT) was defined, in 1884, on the assumption that one second is equal to 1/86,400 of the mean solar day. As most of the fundamental constants can be related either to a time or to a frequency measurement, the quest for the detection of the smallest possible time variation in these constants, that is, the attainment of a time variation measurement of these constants at the 10 −17–10−18 level, is being continued. Highly accurate clocks are not merely a convenience; they are a necessity for such fundamental problems as the local position invariance, baseline interferometry, observation of the so-called ‘gravitational red-shift’, and for the ground-positioning system (GSP)-Galileo systems that require a panoply of atomic clocks located in satellites as well on the Earth’s surface. All of the current
Preface
xvii
research into types of ion clocks and their development are covered explicitly in this chapter. Part 4. Practical Applications presents five practical examples of trapped-ion technology that reflect the wide diversity of applications of trapped-ion devices. Yet there is a common thread that links these applications, and it is the existence of such a thread that justifies the publication of the Volume IV and Volume V monographs. This common thread links the efforts, foresight, and business acumen of manufacturers with the knowledge and experimental skills of researchers to bring forth instruments at an affordable price that will enhance and protect the well-being of mankind. Such a claim is not an overstatement, as is shown by the final three chapters, Chapters 14–16. In the Preface to the first edition of Quadrupole Storage Mass Spectrometry,* a monograph that may be familiar to some of the more curious graduate students, we wrote “There is now abundant evidence of the application to the health services of mass spectrometric techniques with concomitant high sensitivity and resolution for toxicological studies; studies of metabolism and incipient disease; environmental problems; the quality of food, well water, and materials; forensic sciences; and so forth. Thus, the advent of the ion trap detector permits a much greater use of mass spectrometric techniques not only in the technically developed countries but also in those countries which are technically less advanced.” Chapter 12 affords an example of industry–university cooperation with a description of the commercially-available digital ion trap, which is the subject of Chapter 4 in Volume IV, employed for the fragmentation of mass-selected ions by boundary-activated dissociation, a technique that was discovered in an academic research laboratory. Chapter 13 gives an account of the utilization of ion mobility spectrometry and ion mobility/mass spectrometry for the study of basic ion/molecule reactions at ambient pressure, which is the pressure regime used commonly for the detection of explosives, drugs, and chemical warfare agents. Chapter 14 is concerned with a novel application, that of imaging mass spectrometry wherein thin tissue sections are analyzed directly and permit the creation of chemically-selective images of intrinsic chemical distributions. This technique allows characterization of known compounds from a variety of tissues, and the identification of unknown chemical signatures for a variety of studies such as disease progress or pharmaceutical studies. Chapter 15 permits a review of the progress made in the instrumentation for gas chromatography/ion trap mass spectrometry since the introduction by Finnigan MAT of the first commercial gas chromatograph/Ion Trap Detector™. Ion traps have found an important application as in situ chemical analyzers for a broad range of fields such as homeland security, industry, and environmental monitoring applications. Such devices can be used in marine science, where there is a high demand for monitoring natural compounds and the ever-increasing quantities of compounds of anthropogenic origin that enter rivers, lakes, and oceans. Chapter 16 presents a detailed account of the prolonged remote monitoring of volatile organic compounds in field waters by membrane inlet mass spectrometry (MIMS) using a quadrupole ion trap.
∗ March R.E., Hughes R.J., Todd J.F.J., Quadrupole Storage Mass Spectrometry, 1989. New York, Wiley
Interscience.
xviii
Preface
Chapter 12 describes the novel operation of a digital ion trap (DIT) for the determination of the boundaries of the stability diagram and for the utilization of boundary-activated dissociation. Ion motion in a Paul or quadrupole ion trap driven by a rectangular wave quadrupolar field was described in the early 1970s, but it was not until 2000 that the mass-selective resonance method with the ion secular frequency under digital operation conditions was described. The circuits of the digital ion trap switch very rapidly between discrete direct current high voltage levels in order to generate the trapping waveform voltage applied to the ring electrode. An alternative ion activation method, boundary-activated dissociation, proposed in 1991, entails moving the working point (that is, the point (az, qz) on a quadrupole ion trap stability diagram defined by the magnitudes of the trapping parameters az and qz) of a mass-selected ion species close to one of the boundaries of the stability diagram. This method can be realized with the combined effect of suitable direct current and radiofrequency potentials applied to the ion trap electrodes. Under these conditions, dissociation of the mass-selected ion species can be induced. In order to evaluate the performance of the digital ion trap for boundary-activated dissociation experiments, the real shape of the stability diagram needed to be determined. As an ion species undergoes fragmentation when its working point is moved close to a stability boundary, this behavior was used to map the boundaries of the stability diagram for a digital ion trap. In the digital ion trap, variation of the duty cycle of the rectangular waveform readily allows the introduction of the direct current component for boundary-activated dissociation experiments. Regrettably, direct current power supplies are no longer made available in commercial ion trap instruments. Chapter 13 gives an introduction to the principles of ion mobility spectrometry, together with an overview of the type of information obtainable from ion mobility studies at atmospheric pressure and the variety of experimental methods employed in such studies. It is shown that thermodynamic data, which are obtainable from these studies and are suitable for tabulation, include standard enthalpies, entropies, and free energies; such data, when obtained at a specified temperature, can be regarded as universally applicable when all participants are at thermal equilibrium. Thermal equilibrium is established readily in an ion mobility spectrometer at ambient pressure because each ion experiences more than 1010 collisions per second with neutral atoms or molecules of the supporting gas atmosphere. In addition, the residence time of an ion in an ion mobility spectrometer operating at atmospheric pressure is ca 5–50 ms, which allows the study of the interactions of ions with molecules at very low concentrations. Illustrated here is the further advantage of thermochemical determinations obtained by ion mobility spectrometry in that the available temperature range, from sub-ambient to more than 500 K, is far greater than that available with many other experimental methods. In the lower electrostatic field conditions of ion mobility spectrometry, thermal conditions always prevail for ions. Hence, this technique permits ready determination of ion/molecule reaction rate constants, including those for clustering reactions, for both positive and negative ions, in the temperature range from below ambient to at least 600 K. Electron association and detachment reactions are studied more easily in an ambient pressure ion mobility spectrometer than the more conventional swarm-beam method.
Preface
xix
In Chapter 14 is described the principles and instrumental approaches of imaging mass spectrometry; it also provides real-world examples of the capabilities of quadrupole ion traps (QITs) and linear ion traps (LITs) for modern imaging mass spectrometry. Imaging mass spectrometry permits direct analysis of thin tissue sections from which chemically-selective images of intrinsic chemical distributions are created. Imaging mass spectrometry using matrix-assisted laser desorption/ionization (MALDI) provides the ability to characterize and to localize compounds within tissue sections, identifying potential but unknown markers of diseases such as cancer, or to determine where an administered drug (and its metabolites) has localized in a tissue section. An advantage that mass spectrometry provides to the field of imaging is the identification of non-targeted compounds. In a typical fluorescence experiment, a fluorophore must be administered that binds to a given compound for ready identification of that specific compound. With imaging mass spectrometry, the targeted compound can be localized and other compounds that may localize with the targeted compound can be identified, thus providing a more complete understanding of the chemical signature of the specific state under investigation. Examples are discussed of the remarkable utility of sequential tandem mass spectrometry to effect MSn for the identification and structural characterizations of compounds from intact tissues. In addition, MSn is invaluable for the identification of isobaric ions: for example, m/z 828 in one sample was shown to consist of four isobaric ions and each was identified using MS3. Chapter 15 is devoted to a review of the development of the quadrupole ion trap as a detector for compounds eluting from a gas chromatograph, together with an account of the progress made by Varian Inc. in ion trap technology for gas chromatography/ mass spectrometry and gas chromatography/tandem mass spectrometry. Described here is the new type of non-linear ion trap, that is, the field is made non-linear by the superimposition of a dipole and higher-order multipoles upon the quadrupole field by a switchable electric circuit. A detailed discussion of ion traps with electrically-induced non-linear fields is given in Volume IV, Chapter 14. In the ion trap, both dipole and quadrupole supplemental fields are applied to the two end-cap electrodes with their frequencies tuned to βz = 2/3; the resonance of each ion species in turn with both the dipole and quadrupole supplemental fields results in improved mass resolution, higher scan speed, and extended charge capacity. Electron impact ionization and chemical ionization, within the ion trap and external to it, are discussed. The chapter is replete with many examples of applications and contains 47 figures. Chapter 16 describes the monitoring of some 16 volatile organic compounds that are included in the European Union Directive 98/83 and classified as being potentially deleterious to human health when present in drinking water. The technique employed was that of membrane inlet mass spectrometry (MIMS) combined with a quadrupole ion trap. While this technique is well known for its simplicity and sensitivity, no previous account has been published of the implementation of this technique to work unattended for months. Four instruments were deployed in unmanned sites, where they monitored volatile organic compounds (VOCs) in natural waters and wastewater during a period exceeding 1 year for each instrument. The instruments were equipped with software that facilitated the automatic operation of each analysis, the identification and quantitation of VOCs from the raw mass spectra, and the transmission of the results to a remote control room via an Internet connection. In the remote control
xx
Preface
room, a personal computer displayed the results as bar graphs and was programed to activate alarms when set concentration thresholds were exceeded. The chapter discusses laboratory performance and field performance: the former in terms of sensitivity, reproducibility, linearity tests, and comparison with purge-and-trap combined with gas chromatography/mass spectrometry; and the latter in terms of data output, most frequent maintenance operations and technical failures, and overall stability of the four remotely-controlled instruments. The longest period of unattended remote monitoring was of 526 days, of an industrial wastewater treatment plant. We wish to thank the many people who have assisted us in one way or another with myriad tasks that must be carried out in order to arrive at the publication of a monograph from a collection of manuscripts in a variety of formats and styles. First of all, to our contributors, without whom this monograph would not have appeared in print. We give thanks for their individual inspiration; we thank them for the fruits of their labors, and for their patient toleration of the idiosyncrasies of our editing, often involving repeated iterations between the two of us and the authors themselves. The 16 chapters that constitute Volume V have originated from 36 authors and co-authors; a total of 91 authors and co-authors contributed to Volumes IV and V. For many of these co-authors this project has been a novel experience, thus we thank our lead authors for responding to our urging that they collaborate with young scientists in their laboratories. From where else will the monographs of tomorrow originate? At CRC Press, we thank Fiona Macdonald, Pat Roberson, Rachael Panthier, Lindsey Hofmeister, Hilary Rowe, and Jennifer Derima; at Datapage, we thank Ramkumar Soundararajan, the Project Manager. Finally, we express our sincere appreciation for the tolerance of our respective spouses, Kathleen March and Mavis Todd, and for their patience, support, and sacrifices while this project, known informally as ‘PRATIMS’, took over our lives. Raymond E. March John F.J. Todd
Volume IV Contents Practical Aspects of Trapped Ion Mass Spectrometry Volume IV: Theory and Instrumentation Edited by Raymond E. March and John F.J. Todd Table of Contents part I Fundamentals Chapter 1.
An Appreciation and Historical Survey of Mass Spectrometry Raymond E. March and John F.J. Todd
Chapter 2.
Ion Traps for Miniature, Multiplexed and Soft Landing Technologies Scott A. Smith, Chris C. Mulligan, Qingyu Song, Robert J. Noll, R. Graham Cooks, and Zheng Ouyang
Part II New Ion Trapping Techniques Chapter 3.
Theory and Practice of the Orbitrap™ Mass Analyzer Alexander Makarov
Chapter 4.
Rectangular Waveform Driven Digital Ion Trap (DIT) Mass Spectrometer: Theory and Applications Francesco Brancia and Li Ding
Chapter 5.
High-Field Asymmetric Waveform Ion Mobility Spectrometry Randall W. Purves
Chapter 6.
Ion Traps with Circular Geometries Daniel E. Austin and Stephen A. Lammert xxi
xxii
Volume IV Contents
part III Fourier Transform Mass Spectrometry Chapter 7 Ion Accumulation Approaches for Increasing Sensitivity and Dynamic Range in the Analysis of Complex Samples Mikhail E. Belov, Yehia M. Ibrahim, and Richard D. Smith Chapter 8 Radio Frequency-Only-Mode Event and Trap Compensation in Penning Fourier Transform Mass Spectrometry Adam M. Brustkern, Don L. Rempel, and Michael L. Gross Chapter 9 A Fourier Transform Operating Mode Applied to a ThreeDimensional Quadrupole Ion Trap Y. Zerega, J. Andre, M. Carette, A. Janulyte, and C. Reynard
part IV Quadrupole Rod Sets Chapter 10 Trapping and Processing Ions in Radio Frequency Ion Guides Bruce A. Thomson, Igor V. Chernushevich, and Alexandre V. Loboda Chapter 11 Linear Ion Trap Mass Spectrometry with Mass-Selective Axial Ejection James W. Hager Chapter 12 Axially-Resonant Excitation Linear Ion Trap (AREX LIT) Yuichiro Hashimoto
part V 3D-Quadrupole Ion Trap Mass Spectrometry Chapter 13 An Examination of the Physics of the High-Capacity Trap (HCT) Desmond A. Kaplan, Ralf Hartmer, Andreas Brekenfeld, Jochen Franzen, and Michael Schubert Chapter 14 Electrically-Induced Nonlinear Ion Traps Gregory J. Wells and August A. Specht
Volume IV Contents
xxiii
Chapter 15 Fragmentation Techniques for Protein Ions Using Various Types of Ion Trap J. Franzen and K. P. Wanczek Chapter 16 Unraveling the Structural Details of the Glycoproteome by Ion Trap Mass Spectrometry Vernon Reinhold, David J. Ashline, and Hailong Zhang Chapter 17 Collisional Cooling in the Quadrupole Ion Trap Mass Spectrometer (QITMS) Philip M. Remes and Gary L. Glish Chapter 18 ‘Pressure Tailoring’ for Improved Ion Trap Performance Dodge L. Baluya and Richard A. Yost Chapter 19 A Quadrupole Ion Trap/Time-of-Flight Mass Spectrometer Combined with a Vacuum Matrix-Assisted Laser Desorption Ionization Source Dimitris Papanastasiou, Omar Belgacem, Helen Montgomery, Mikhail Sudakov, and Emmanuel Raptakis
part VI Photochemistry of Trapped Ions Chapter 20 Photodissociation in Ion Traps Jennifer S. Brodbelt Chapter 21 Photochemical Studies of Metal Dication Complexes in an Ion Trap Guohua Wu, Hamish Stewart, and Anthony J. Stace
Editors Raymond E. March, PhD, DSc, D(hc), FCIC, is presently Professor Emeritus of Chemistry at Trent University in Peterborough, ON, Canada. He obtained a BSc (Hons) in Chemistry from Leeds University in 1957; a PhD from the University of Toronto in 1961 (supervised by Professor John C. Polanyi, Nobelist 1986); a DSc from Leeds University in 2000; and an honorary doctorate (D(hc)) from l’Université de Provence in 2008. From 1954 to 1957, he was a Cadet Pilot in the Leeds University Air Squadron Royal Air Force Volunteer Reserve (RAFVR) and, from 1958 to 1963, a Flight Lieutenant in the Royal Canadian Air Force (Auxiliary) (RCAF). From 1960 to 1961, he held a Canadian Industries Limited Research Fellowship. From 1962 to 1963, he was a Post-Doctoral Fellow with Professor H.I. Schiff at McGill University, and a Research Associate from 1963 to 1965, during which time he lectured at McGill University and Loyola College. In 1965, he joined the faculty of Trent University where he has conducted independent research for some 44 years in gas-phase kinetics, optical spectroscopy, gaseous ion kinetics, analytical chemistry, nuclear magnetic resonance spectroscopy, and mass spectrometry. Kathleen and Ray have been married for 51 years; they have three daughters, Jacqueline, Roberta, and Sally with spouses Paul, Stuart, and Lauren, respectively, and nine grandchildren, Shawn, Jessica, Thomas, Daniel, Rebecca, Sara, James, Madeline, and Carson, in order of appearance. Dr. March has published and/or co-authored over 170 scientific papers and some 75 conference presentations in the above areas of research with emphasis on mass spectrometry, both with sector instruments and quadrupole ion traps. Dr. March is a co-author with Dr. Richard J. Hughes and Dr. John F.J. Todd of Quadrupole Storage Mass Spectrometry, published in 1989. A second edition of Quadrupole Storage Mass Spectrometry, co-authored by Dr. March and Dr. John F.J. Todd was published in 2005. Dr. March and Dr. John F.J. Todd co-edited three volumes entitled Practical Aspects of Ion Trap Mass Spectrometry, published in 1995. Volume IV in the series Practical Aspects of Trapped Ion Mass Spectrometry is in press. Dr. March is a co-author with Oscar V. Bustillos and André Sassine of A Espectrometria de Massas Quadupolar, published in Portuguese in 2005. Professor March is a Fellow of the Chemical Institute of Canada and a member of the American, British, and Canadian Societies for Mass Spectrometry. In 2009, he received the Gerhard Herzberg Award of the Canadian Society for Analytical Sciences and Spectroscopy (CSASS). In
xxv
xxvi
Editors
1995, he received the Distinguished Faculty Research Award from Trent University, and the Canadian Mass Spectrometry Society presented him with the Recognition Award and, in 1997, with the Distinguished Contribution Award. Dr. March is a member of the Editorial Advisory Boards for Rapid Communications in Mass Spectrometry, the Journal of Mass Spectrometry, and the International Journal of Mass Spectrometry. In 1975, Dr. March was an Exchange Fellow (NRC-CNRS) at Orsay, France, with Professor Jean Durup; in 1983, an Exchange Fellow (NRCRoyal Society of London) in Swansea, Wales, with Professor J.H. Beynon; in 1989 and 1992, a Visiting Professor, Université de Provence, Marseille, France, with Prof Fernande Vedel; in 1993 and 1995, a CNRS Visiting Professor, Université Pierre et Marie Curie, Paris, France, with Prof Jean-Claude Tabet; and in 1999, a Visiting Professor, Université de Provence, Marseille, France, with Yves Zerega. In 1987, Dr. March was a Distinguished Lecturer at the Universities of Berne, Neuchatel, and Lausanne, in Switzerland. Dr. March’s research in the field of mass spectrometry and gas-phase ion chemistry involved the development and application of mass spectrometric instruments, particularly quadrupole ion trap mass spectrometers and hybrid mass spectrometers, for both fundamental studies and the formulation of analytical protocols for the determination of compounds of environmental interest. His current research interests are focused within Trent University’s Water Quality Centre (www. trentu.ca/wqc/). As a founding member of the Water Quality Centre his principal research interest lies in the mass spectrometric and nuclear magnetic resonance spectroscopic investigation of natural compounds that, having been formed by plants, may enter waterways and/or the water table. His current research involves the study of flavonoids and flavonoid glycosides; such compounds are often found in those products that have become known as neutraceuticals. Electrospray ionization combined with tandem mass spectrometry permits the investigation of ion fragmentation at high mass resolution and the derivation of possible ion fragmentation mechanisms using ion structures; these studies are supported by theoretical calculations carried out in collaboration with Professor E.G. Lewars. An important aspect of this research is the development of appropriate analytical protocols for flavonoid glycosides in water and in plant extracts. Nuclear magnetic resonance (NMR) studies of flavonoids and metabolites, carried out in collaboration with Professor D.A. Ellis and Dr. D.C. Burns, have permitted a rationalization of chemical shifts with product ion mass spectra and the development of a predictive model for 13C chemical shifts in flavonoids. At present, Dr. March is carrying out an investigation of volatile compounds formed by Ash trees in response to an attack by the Emerald Ash Borer. These researches are supported by the Natural Sciences and Engineering Research Council of Canada (Discovery Grants Program), the Canada Foundation for Innovation, the Ontario Research and Development Challenge Fund, Ontario Ministry of Natural Resources, and Trent University. Dr. March has enjoyed longterm collaborations with Professor John Todd, with colleagues at l’Université de Provence and l’Université Pierre et Marie Curie (France), and with colleagues in Padova (Italy).
Editors
xxvii
John F.J. Todd, BSc, PhD, CChem, FRSC, CEng, FInstMC, is currently Emeritus Professor of Mass Spectroscopy at the University of Kent, Canterbury, U.K. He obtained his Class I Honours BSc degree in Chemistry in 1959 from the University of Leeds, from whence he also gained his PhD degree and was awarded the J.B. Cohen Prize in 1963; he was a member of the radiation chemistry group led by the late Professor F.S. (later Lord) Dainton, FRS. From 1963 to 1965, he was a Fulbright Research Scholar and Post-Doctoral Research Fellow in Chemistry with the late Professor Richard Wolfgang at Yale University. In 1965, he was one of the first faculty members appointed to the then new University of Kent at Canterbury, U.K. John and Mavis Todd have been married for 46 years and have three sons: John (Andrew), Eric, and Richard, two daughters-in-law Dorota and Marie, and six grandchildren, Alice, Max, Maja, Luke, Daniel, and Lara. Professor Todd’s research interests, spanning some 44 years, have encompassed positive and negative ion mass spectral fragmentation studies, gas discharge chemistry, ion mobility spectroscopy, analytical chemistry, and ion trap mass spectrometry. His work on 3D quadrupole (Paul) ion traps commenced in 1968, when he first developed the “Quistor/Quadrupole” instrument for the characterization of the behavior of ions confined in radiofrequency electric fields and as a vehicle for the study of gasphase ion chemistry. As a consultant to Finnigan MAT during the 1980s and 1990s, he was a member of the original team that developed the first commercial ion trap mass spectrometer. In another consultancy role, Professor Todd is involved currently with one of the most extended single mass spectrometric investigations ever undertaken: the use of an ion trap mass spectrometer for the isotope ratio measurement of cometary material as part of the “Rosetta” project (launched 2004, scheduled arrival at its target comet in 2014). Professor Todd has published and/or co-authored some 116 scientific papers and over 118 conference contributions, concentrating mainly on various aspects of mass spectrometry. With Professor Dennis Price, he co-edited four volumes of Dynamic Mass Spectrometry and he edited Advances in Mass Spectrometry 1985 (which contained the proceedings of the 10th International Mass Spectrometry Conference, Swansea, at which he was also a plenary lecturer). In addition, he was an editor of the International Journal of Mass Spectrometry and Ion Processes from 1985 to 1998, has served on the Editorial Boards of Organic Mass Spectrometry/Journal of Mass Spectrometry and Rapid Communications in Mass Spectrometry, and is currently a member of the Board for the European Journal of Mass Spectrometry. With Dr. Raymond E. March, Dr. Todd co-edited three volumes entitled Practical Aspects of Ion Trap Mass Spectrometry, published by CRC Press in 1995. In addition, Dr. Todd was a co-author with Dr. Raymond E. March and Dr. Richard J. Hughes of Quadrupole Storage Mass Spectrometry, published by Wiley in 1989; a second edition of Quadrupole Storage Mass Spectrometry, co-authored by Dr. Todd and Dr. Raymond E. March was published in 2005. Volume IV in the series Practical Aspects of Trapped Ion Mass Spectrometry is in press.
xxviii
Editors
Professor Todd is a Chartered Chemist and a Chartered Engineer, and has served terms as Chairman and as Treasurer of the British Mass Spectrometry Society. In 1988, he was a Canadian Industries Limited Distinguished Visiting Lecturer at Trent University, Peterborough, Ontario. In 1997, Dr. Todd was awarded the Thomson Gold Medal by the International Mass Spectrometry Society for “outstanding contributions to mass spectrometry,” and in 2006 he was awarded the Aston Medal by the British Mass Spectrometry Society, of which he is also a Life Member. In 2008, he was accorded Honorary Life Membership of the Royal Society of Chemistry. Outside the immediate confines of his academic work, Professor Todd was appointed as Master of Rutherford College, University of Kent (1975–1985), and as the first Chairman of the newly created Canterbury and Thanet Health Authority (UK National Health Service) between 1982 and 1986. During the period 1995–2006 he was the founding Chairman of the newly established Board of Governors of St Edmund’s School Canterbury, and until August 2007 he was a Governor of Canterbury Christ Church University; he was admitted as an Honorary Fellow of Canterbury Christ Church University in 2008. From 1979 to 1989, Professor Todd was Chairman of the Mass Spectrometry Sub-Committee, Commission I.5 of the International Union of Pure and Applied Chemistry (IUPAC), and between 1995 and 2007 he was Chairman of the Management Advisory Panel for the EPSRC National Mass Spectrometry Service Centre, based at the University of Wales Swansea. He has enjoyed long-term collaborations with co-editor Professor Raymond March, with colleagues at Finnigan MAT in the United Kingdom and the United States, and with groups in Nice (France) and Padova and Torino (Italy).
Contributors Alberto Berton CNR-ISTM Corso Stati Uniti 4 Padova, Italy
Michael Drewsen Department of Physics and Astronomy University of Aarhus Aarhus, Denmark
Paola Bocchini Department of Chemistry ‘G. Ciamician’ University of Bologna Bologna, Italy
Gary A. Eiceman Department of Chemistry and Biochemistry New Mexico State University Las Cruces, New Mexico
Francesco L. Brancia Shimadzu Research Laboratory (Europe) Manchester, United Kingdom
Matthew W. Forbes Department of Chemistry University of Toronto Toronto, Ontario, Canada
Jennifer S. Brodbelt Department of Chemistry and Biochemistry University of Texas at Austin Austin, Texas Joseph E. Chipuk Department of Chemistry and Biochemistry University of Texas at Austin Austin, Texas Joshua J. Coon Department of Chemistry and Biomolecular Chemistry University of Wisconsin Madison, Wisconsin Helen J. Cooper School of Biosciences University of Birmingham Edgbaston, Birmingham, United Kingdom
Jennifer M. Froelich Department of Chemistry Michigan State University East Lansing, Michigan Guido C. Galletti Department of Chemistry ‘G. Ciamician’ University of Bologna Bologna, Italy Timothy J. Garrett Department of Medicine University of Florida Gainesville, Florida John E. George III Varian Inc., Scientific Instruments Walnut Creek, California Klaus Højbjerre Department of Physics and Astronomy University of Aarhus Aarhus, Denmark xxix
xxx
Rebecca A. Jockusch Department of Chemistry University of Toronto Toronto, Ontario, Canada Jian Liu Department of Chemistry Purdue University West Lafayette, Indiana Yali Lu Department of Chemistry Michigan State University East Lansing, Michigan Graeme C. McAlister Department of Chemistry and Biomolecular Chemistry University of Wisconsin Madison, Wisconsin Scott A. McLuckey Department of Chemistry Purdue University West Lafayette, Indiana Joel H. Parks The Rowland Institute at Harvard Cambridge, Massachusetts Francesca Pinelli Department of Chemistry ‘G. Ciamician’ University of Bologna Bologna, Italy Benoît Plet European Institute of Biology and Chemistry University of Bordeaux Pessac, France Romina Pozzi Department of Chemistry ‘G. Ciamician’ University of Bologna Bologna, Italy
Contributors
Luca Raveane CNR-ISTM Corso Stati Uniti 4 Padova, Italy Gavin E. Reid Department of Chemistry, Biochemistry and Molecular Biology Michigan State University East Lansing, Michigan Jean-Marie Schmitter European Institute of Biology and Chemistry University of Bordeaux Pessac, France James H. Scrivens Department of Biological Sciences University of Warwick Coventry, United Kingdom Peter Frøhlich Staanum Department of Physics and Astronomy University of Aarhus Aarhus, Denmark John A. Stone Department of Chemistry Queens University Kingston, Ontario, Canada Francis O. Talbot Department of Chemistry University of Toronto Toronto, Ontario, Canada Konstantinos Thalassinos Department of Biological Sciences University of Warwick Coventry, United Kingdom
xxxi
Contributors
Pietro Traldi CNR-ISTM Corso Stati Uniti 4 Padova, Italy Fernande Vedel Physique des interactions ioniques et moléculaires (PIIM) Université de Provence Marseille, France
Mingda Wang Varian Inc., Scientific Instruments Walnut Creek, California Richard A. Yost Department of Chemistry University of Florida Gainesville, Florida
Part I Ion Reactions
Reactions in 1 Ion/Ion Electrodynamic Ion Traps Jian Liu and Scott A. McLuckey Contents 1.1 Introduction.........................................................................................................3 1.2 Tools for the Study of Ion/Ion Reactions............................................................4 1.2.1 Ion/Ion Reactions in Three-Dimensional (3D) Quadrupole Ion Traps..............................................................................5 1.2.2 Ion/Ion Reactions in Linear Ion Traps (LITs).........................................9 1.2.3 Ion/Ion Reactions in Hybrid Instruments.............................................13 1.3 Methodologies/Applications..............................................................................15 1.3.1 Charge State Manipulation: Proton Transfer........................................15 1.3.1.1 Macromolecule Mixture Analysis..........................................15 1.3.1.2 Precursor Ion Charge State Manipulation..............................16 1.3.1.3 Simplification of Product Ion Mass Spectra..........................17 1.3.2 Charge Inversion...................................................................................19 1.3.3 Metal–Ion Transfer................................................................................19 1.3.4 Electron Transfer Dissociation (ETD)..................................................21 1.4 Conclusions....................................................................................................... 24 References...................................................................................................................25
1.1 INTRODUCTION Interactions between gas-phase ions of opposite polarities occur commonly in various environments such as the atmosphere, plasmas, flames [1–4], etc. It has been more than a century since the first study of the interaction between oppositely-charged ions, which can be dated back to the work by Thomson and Rutherford [5]. However, the study of ion/ion reactions can be challenging particularly when the reactions take place between singly-charged cations and anions, as was the case in the majority of the early studies, because the products are neutral species and, therefore, difficult to analyze and detect. With the advent of electrospray ionization (ESI) [6–9] and its propensity for producing multiply-charged ions from high mass molecules, attention has been directed to the reactions between oppositely-charged ions involving multiply-charged ions, which produce charged products readily amenable to study by mass spectrometry. Consequently, a rapidly-growing range of reaction phenomena are being observed in the ion/ion reactions of multiply-charged ions, which is permitting new insights to be drawn regarding ion/ion reaction thermodynamics and dynamics. While ion/ion 3
4
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
reactions involving large multiply-charged ions are not important in plasmas or flames, for example, they are enabling an expanding array of new analytical applications, particularly in bioanalysis. Although ion/ion reactions can be implemented readily under various conditions, including near atmospheric pressure [10–15], ion/ion reactions performed in electrodynamic ion traps afford the opportunity for selective ion manipulation due to the mass/charge-dependent frequencies of motions executed by stored ions. Therefore, ion traps allow for the study of ion/ion reactions within the context of tandem mass spectrometry (MSn) experiments. As a result, many ion/ion reaction studies have been undertaken in electrodynamic ion trap-based instruments. Several reviews [16–18] of ion/ion reactions involving multiply-charged ions have been published, that focus on aspects of instrumentation, applications, reaction phenomena, and fundamentals including thermodynamics and kinetics. Rather than providing a comprehensive discussion of all aspects related to ion/ion chemistry, this chapter aims primarily to provide a brief description of the instrumentations, methodologies, and applications of ion/ion reactions in ion traps, especially within the context of biomolecule analysis, with particular emphasis on developments since the publication of recent reviews.
1.2 TOOLS FOR THE STUDY OF ION/ION REACTIONS Fundamental requirements for any tool intended for ion/ion reaction studies are the ability to generate ions of opposite polarities within a single experiment, and to furnish an interaction environment delivering good spatial and temporal overlap for the oppositely-charged ion populations. The environment for the ion/ion interaction can be created either outside or inside a mass spectrometer. The first ion/ion reactions involving multiply-charged ions, for example, were demonstrated at near atmospheric pressure (ca 2 Torr) using a Y-tube flow reactor [10,11], which admit, into separate inlet arms of the reactor, ions of opposite polarities produced by two ion sources, for example, ESI, and discharge sources. The ion/ion reactions took place once the two ion streams merged in the outlet arm of the reactor, which was coupled to the interface of the mass spectrometer, before sampling into the instrument. Implementation of ion/ion reactions external to the mass spectrometer separates physically the ionization process and ion/ion reactions from the mass analysis step. Advantages derived from this separation include the simplicity with which such ion/ion reactors can be adapted to any mass spectrometer coupled with ESI, independent optimization of mass analysis, and virtually no limits are imposed by the characteristics of the mass analyzer on the kinds of ions that can be used as reactants. However, reaction conditions can be difficult to define in reactors operating at near atmospheric pressure due to the existence of a complicated reaction environment, where a mixture of ions, solvent vapors, and atmospheric gases are present in the reaction region. This situation can lead to ambiguities in the determination of mechanisms that give rise to products in some cases. Moreover, implementing ion/ ion reactions outside a mass spectrometer does not allow for a true tandem mass spectrometric experiment to be performed involving an ion/ion reaction between mass analysis stages. Many of the drawbacks associated with the implementation of
Ion/Ion Reactions in Electrodynamic Ion Traps
5
ion/ion reactions outside a mass spectrometer can be overcome by using an electrodynamic ion trap as a reaction vessel, at the cost of somewhat greater experimental complexity.
1.2.1 Ion/Ion Reactions in Three-Dimensional (3D) Quadrupole Ion Traps The majority of the early studies of ion/ion reactions under low pressure (ca 1 mTorr) conditions were performed with three-dimensional (3D) ion traps (that is, conventional Paul traps), which store ions in three dimensions by a radio-frequency (RF) voltage applied to a central ring electrode sandwiched between two end-cap electrodes. The 3D ion trap is inherently compatible with the study of ion/ion reactions due to its unique ability to store simultaneously ions of both polarities in overlapping regions of space [19,20]. The trapped ion assumes a characteristic set of m/z-dependent frequencies of motions in the oscillating quadrupole field of the ion trap, which allows ready manipulation of ions of specific massto-charge ratios for ion isolation and activation, both of which are common elements in a tandem mass spectrometric experiment. The ‘tandem-in-time’ nature of the ion trap MSn experiment [21,22] provides well-defined conditions for ion/ion reactions and is particularly useful in the determination of ion genealogy. Furthermore, the use of a bath gas, such as helium at ca 1 mTorr in the ion trap, intended originally to cool the ions translationally into the center of the trap to improve the mass resolution for the mass analysis [23], also improves ion/ion reaction efficiencies by maximizing the spatial overlap and minimizing the translational energies of the two ion clouds [24]. While the ion trap is particularly well-suited to serve as a reaction vessel for ion/ion reactions, it places constraints on the range of reactions that can be studied. For example, all of the reactant and product ions must fall within the limited range of m/z-values that can be stored simultaneously in an ion trap. The normal operation of the ion trap places a lower limit to the mass-to-charge value for ion storage, also known as low mass cut-off (LMCO) [25] of the ion trap, which is defined sharply by the operating RF (that is, RF frequency and amplitude) and ion trap dimensions. Any ion having an m/z-value less than the LMCO assumes an unstable trajectory and will be ejected from the ion trap. In the absence of a DC field, all ions having m/z-values greater than the LMCO lie within the region of ion stability of the ion trap. However, ions of different mass-to-charge values experience different trapping potentials in the ion trap, as approximated by the so-called pseudo-potential trapping well ( Dz ) [25], which is defined also by the amplitude and frequency of the RF operating voltage, and the ion trap dimensions. The magnitude of the pseudo-potential trapping well is approximately inversely proportional to the mass-tocharge ratio at qz < 0.4. When the kinetic energy of an ion is close to or exceeds the magnitude of ( Dz ), it cannot be stored efficiently. Therefore, the shallow pseudo-potential trapping well associated with ions of high mass-to-charge ratios sets a practical upper limit to the range of ions that can be stored in the ion trap. Note, however, that a trapping mechanism for ions of high mass-to-charge ratio, in addition to that provided by the oscillating quadrupolar field, can be created from the electrostatic field created by ions of lower m/z-value stored in the ion trap. This mechanism has been referred to as ‘trapping by proxy’ [26] and can be important when the magnitude of Dz for the ions of high m/z-value is too low for efficient ion storage. Ions of low m/z-value are stored in
6
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
deeper pseudo-potential wells than the ions of high m/z-value, and, when the density of ions of low m/z-value is sufficiently high, oppositely-charged ions of high m/z-value can be prevented from escaping the ion trap. As a result, the upper limit of the m/z-value for mutual ion storage can be increased when a high-density of ions of low m/z-value of opposite polarity is available. For a related reason, the simultaneous presence of ions of each polarity can affect mass analysis, particularly for the ions of lower charge densities. The electric field of the higher density ion population can affect significantly the motions of the lower density ion population such that they do not oscillate at the frequencies expected based solely on the presence of an oscillating quadrupolar electric field [27]. This complication makes it necessary to eject the ion population of higher density prior to mass analysis of the ion population of lower density when the ion trap serves as both a reaction vessel and a mass analyzer. Despite these constraints, the advantages associated with physical separation of the ionization process from the ion/ion reaction region and ready manipulation of ions of specific m/z-values by use of an ion trap instrument make it a powerful tool in the study of ion/ion reactions. The 3D ion trap can be used directly without any modification as a reactor for ion/ ion reactions, but accommodations must be made to facilitate the admission and transmission of ions of each polarity into the ion trap. Ion admission into the ion trap is accomplished usually in one of these two ways: (i) introducing ions of one polarity through a hole in an end-cap electrode and admitting ions of opposite polarity into the trap through the aperture in the ring electrode; or (ii) directing ions of each polarity sequentially through a hole in one of the end-cap electrodes with the guidance of a ‘turning’ quadrupole. By use of either ion introduction or transmission approach, a variety of ion sources have been adapted to the 3D ion trap to generate ions of different types (such as, positive vs negative; open-shell vs closed-shell) for various applications. For example, shown schematically in Figure 1.1 is one of the most commonly-used designs [28] for the study of ion/ion reactions between multiply-charged cations and singly-charged anions. This instrument incorporates an ESI source to produce multiply-charged cations, which were introduced into the ion trap through the end-cap electrode adjacent to the source, and an atmospheric sampling glow discharge ionization (ASGDI) source [29] to generate singly-charged negative ions, which entered the ion trap through an aperture in the ring electrode. A similar configuration was adopted in a modified Hitachi M-8000 ion trap mass spectrometer [30] (San Jose, CA, USA) with improved figures-of-merit for mass analysis. A typical experimental sequence with this instrumental setup involved cation injection, ion isolation, anion injection, mutual storage of both polarities, and mass analysis through resonance ejection [31]. Particular emphasis was directed to proton transfer reactions for charge-state manipulation of multiply-protonated protein/peptide ions, both precursor and product ions, using proton transfer reagents such as anions derived from perfluoro-1,3-dimethylcyclohexane (PDCH). Recently, a setup similar to that shown in Figure 1.1 has been particularly useful in the study of another important reaction phenomenon, electron transfer dissociation (ETD) [32,33], by reacting multiply-protonated polypeptides with radical anions produced by the ASGDI source from species such as sulfur dioxide and nitrobenzene [33,34]. An apparatus also similar to that shown in Figure 1.1 has been developed by Glish and co-workers [35] to study the ion/ion reactions of iron and iron-containing ions with oppositely-charged peptide and protein ions, in which the multiply-charged
7
Ion/Ion Reactions in Electrodynamic Ion Traps Ion trap analyzer Guard ring Electrospray needle
Electron multiplier
Protein sample infusion
Conversion dynode Gate lens
Application of high voltage DC pulser (gate lens)
Positive ions Negative ions
Inlet for air/PDCH vapor
Needle valve
PDCH/air vapor
FIGURE 1.1 Schematic of a quadrupole ion trap mass spectrometer for ion/ion reactions between multiply-charged cations generated by ESI (ion injection through the ion entry endcap electrode) and singly-charged anions produced by ASGDI (ion injection through the ring electrode). (Reproduced from Stephenson J.L.; McLuckey, S.A., Int. J. Mass Spectrom. Ion Processes. 1997, 162, 89–106. With permission from Elsevier.)
cations were produced via ESI and introduced into the ion trap through the hole in an end-cap electrode while the singly-charged anions were formed using laser desorption from a stainless steel surface and admitted into the ion trap via the aperture in the ring electrode. In addition to reagent ions, electrons from a gated filament have been introduced also into the 3D ion trap [36] through a hole in the ring electrode for an in situ formation of singly-charged positive ions [37], as demonstrated in the first study of ion/ion reactions using a 3D ion trap. Despite its utility in the study of multiplycharged anions reacting with singly-charged cations, this approach is not well-suited to the study of ion/ion reactions of multiply-charged cations due to the inefficient nature of in situ anion formation with electrons from a heated filament [38]. Significant progress has been made in the exploration of gas-phase ion/ion chemistry with the development of the apparatus described above; however, ion injection via the ring electrode [36] in these approaches places constraints on the range of species available for study due to the existence of strong electric fields near the ring electrode, which lead to both low efficiency in ion transmission (approximately two orders of magnitude lower than axial injection through the end-cap electrode) and a high propensity for ion fragmentation [30]. The constraints imposed by the use of the ring electrode for ion introduction can be removed by admitting ions of both polarities through an end-cap electrode with the use of a DC ‘turning’ quadrupole [39], which directs the ions along a common ion path into the ion trap. As one example of
8
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
such an implementation, ions of opposite polarity were produced from two different ion sources that were both perpendicular to the axial dimension of the ion trap and 180º from each other [39]. The employment of a ‘turning’ quadrupole allows for flexible combinations of up to three ion sources, including combinations like ESI/ESI/ atmospheric pressure chemical ionization (APCI), ESI/ASGDI, ESI/corona discharge ionization, and ESI/ESI, the latter of which allowed for the first study of ion/ion reactions between multiply-charged positive and negative ions [39,40]. As the scope of ion/ion reaction studies expands, it is sometimes desirable that product ions from an ion/ion reaction be ‘processed’ further by various other reagent ions produced from different ion sources (that is, sequential ion/ion reactions involving different reagent ions). A simple example of such an application is charge reduction of a protein complex formed by gas-phase ion/ion reactions, which requires two ESI sources for multiply-charged positive and negative protein ion formation and another ion source (for example, ASGDI source) to generate the charge-reducing reagent [41]. Such experiments can be accomplished readily in a 3D ion trap by taking advantage of its MSn capability, provided enough ion sources can be coupled to the instrument and ions of different types can be delivered sequentially to the ion trap in a timely manner. Such an apparatus has been developed [41], as shown in Figure 1.2, with up to four independent ion sources, one of which is an ASGDI source capable of delivering ions through the ring electrode and the other three can be any combination of ESI, ASGDI, and corona discharge ionization sources, from which ions are introduced sequentially into the ion trap via the end-cap electrode guided by a DC turning quadrupole. As the analytical utility of ion/ion reactions has become increasingly apparent, 3D ion traps intended for ion/ion reaction applications, such as ETD, have become commercially available, such as the Bruker HCTultra™/Agilent 6340 ETD ion trap.
Glow discharge source
Turning quad
–ESI source
ESI/discharge source
Ion trap
Tube lens + ESI source
FIGURE 1.2 Schematic diagram of multiple-source quadrupole ion trap mass spectrometer. (Reproduced from Badman, E.R.; Chrisman, P.A.; McLuckey, S.A., Anal. Chem. 2002, 74, 6237–6243. With permission from American Chemical Society.)
Ion/Ion Reactions in Electrodynamic Ion Traps
9
This ion trap instrument has a chemical ionization source located orthogonally to an octopole ion guide, which admits fluoranthene anions formed in the chemical ionization source and transmits them to the entrance of the end-cap electrode [42]. The positive polypeptide ions produced from ESI are admitted axially to the ion trap through the octopole ion guide and the entrance of the end-cap electrode.
1.2.2 Ion/Ion Reactions in Linear Ion Traps (LITs) Linear or two-dimensional (2D) quadrupole ion traps have been adapted widely for ion storage in systems comprised of Fourier transform ion cyclotron resonance (FT-ICR) spectrometers [43–45], time-of-flight (TOF) [46–48], and standard 3D ion trap mass spectrometers [49,50] due to the improved trapping efficiency of the linear ion trap (LIT) and its increased ion capacity relative to a 3D ion trap. With the development of the LIT as a stand-alone mass spectrometer [51,52], interest has been directed to the implementation of ion/ion reactions in such instruments. However, a LIT that uses DC trapping voltages at the ends of the multi-pole array, which is the usual way of operating a LIT, is not suitable for simultaneous trapping of ions of both polarities, which is required in an ion/ion reaction implemented in the mutual storage mode. Methods have been developed to provide mutual ion storage of both polarities in the axial direction of the LIT by creating RF barriers at the two ends of the LIT quadrupole array. A straightforward approach is to apply an auxiliary RF voltage on the two containment lenses of the LIT, which produces the required RF barriers for ions of both polarities in the axial direction. This method has been demonstrated [32] by the Hunt group on a modified LIT mass spectrometer (Finnigan LTQ™ mass spectrometer, Thermo Electron, San Jose, CA, USA), as shown in Figure 1.3, in the first study of ion/ion ETD. An alternative approach to create an RF barrier to effect mutual ion storage along the axial direction is to unbalance the RF potential applied to the LIT quadrupole array by subtracting a fraction of RF amplitude applied to one set of opposing rods and adding the same amount of RF to other pair of rods [53,54]. The degree of RF unbalance is determined by the fraction of the RF amplitude subtracted from one opposing set of rods and applied to the other. When the quadrupole array and the containment lenses share the same DC offset level, unbalancing the RF on the rods creates an oscillating RF in the axial direction in the fringing region of the rod set, which is equivalent to applying RF to the containment lenses. This approach has been applied successfully to the third quadrupole (Q3) LIT of a modified hybrid triple quadrupole/LIT (Q-Trap 2000, Applied Biosystems/MDS Sciex, Concord, ON, Canada) [54]. However, flexible control (turning ‘On’ and ‘Off’) of the RF barrier created by the unbalance of the RF is not straightforward to achieve in an MSn experiment. As a result, ion injection into the LIT and the performance of mass analysis by massselective axial ejection (MSAE) [55] may be affected adversely when the existing RF barrier at the end of the rod set cannot be applied or removed in a timely manner. To address this issue, subsequent work on the same apparatus involved superposing to the containment lenses of both collision cell (Q2) and Q3 LIT an auxiliary RF, as indicated in Figure 1.4, which can be controlled readily by the scan function during an experiment. Various ion/ion reactions have been demonstrated successfully with this setup in both the Q2 and Q3 linearion traps (LITs) with a much higher (roughly one order
10
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
Front
Center
Back
ESI
CI
+
–
Front lens
Back lens
(b)
0V Precursor ions moved to front section +
–10 V
(c)
+5 V
Anion injection
+
–
(d)
+5 V –
+ (e)
Ion/ion reaction
–
0V
–
+
0V
– (f )
–
End reaction and scan out +
–
0V
FIGURE 1.3 Schematic of a LIT instrument and experimental steps involved in ion/ion reactions (a) Injection of cations from ESI. (b) Transfer of cations to the front section of LIT. (c) Injection of reagent anions to the center section of LIT from a chemical ionization source. (d) Isolation of positively charged precursor ions and reacting reagent anions by a broadband signal. (e) Mutual storage of both polarities in the center section of LIT. (f) Removal of anions leaving the positively charge product ions for subsequent mass analysis. (ions of opposite polarities are injected along the axial direction from either end of the LIT). (Reproduced from Syka, J.E.P.; Coon, J.J.; Schroeder, M.J.; Shabanowitz, J.; Hunt, D.F., Proc. Natl. Acad. Sci. USA. 2004, 101, 9528–9533. With permission from National Academy of Sciences.)
of magnitude) reaction rate observed in the former, presumably due to the higher bath gas pressure in Q2 (ca 5 mTorr in Q2 and ca 5 × 10−5 Torr in Q3) leading to better ion cloud overlap and reduced relative velocities. Different schemes have been used for the introduction of oppositely-charged ions into the LIT. In the modified LTQ instrument, shown schematically in Figure 1.3,
11
Ion/Ion Reactions in Electrodynamic Ion Traps
Fused silica capillary
Curtain plate Orifice
Skimmer ~
Capillary holder
Q0
Nano-ESI tip
Q1
~
Q2
IQ2
Q3
IQ3
Triggered +/– HV
Triggered –/+ HV
FIGURE 1.4 Schematic of a triple quadrupole/LIT mass spectrometer modified by superposing an auxiliary RF on IQ2 and IQ3 containment lenses. (Note the pulsed dual ESI source is also illustrated.) (Reproduced from Xia, Y.; Liang, X.R.; McLuckey, S.A., J. Am. Soc. Mass Spectrom. 2005, 16, 1750–1756. With permission from Elsevier.)
positive ions formed from ESI are introduced from one end of the LIT and the negative ions produced from the chemical ionization source are admitted from the other end of the LIT [32]. Mass analysis via radial ion ejection [51] through slots in one pair of opposing rods facilitates this arrangement because ion detection occurs in the x or y-plane providing ready access to both ends of the LIT for ion sources. This arrangement for ion admission of both polarities along the z-axis of the LIT from the front and rear ends of the instrument leads to high ion injection and trapping efficiencies. Distinct from the LTQ LIT, LITs such as those marketed by ABI and MDS Sciex employ mass-selective axial ejection for mass analysis, which requires that a detector be placed near one end of the quadrupole array [52]. As a result, only one end of the LIT can be used for ion admission along the z-axis of the instrument. Because of this constraint, efforts have been made to deliver ions of both polarities from the same end of the instrument to gain the advantages associated with ion injection along the z-axis of the instrument. However, it is noteworthy that proton transfer ion/ion reactions have been implemented [56] in the same instrument by introducing multiply-charged positive ions from the front end of the instrument and admitting singly-charged negative ions from the side of the Q3 quadrupole coupled with an ASGDI source, but with lower ion injection efficiency and higher ion fragmentation due to the RF field along the radial direction of the Q3 LIT. The first demonstration of ion injection of both polarities via a common atmosphere/vacuum interface of an LIT employed sonic spray ionization (SSI) [57–59] to produce from the same solution both positive ions and negative ions, which were injected sequentially into the LIT [60]. Despite its relative simplicity and effectiveness for implementing a variety
12
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
of ion/ion reactions, SSI usually gives lower ionization efficiency relative to other spray methods; in addition, deleterious matrix effects on the analyte ions associated with the addition of the species intended to provide the reagent ions may become severe. The limitations associated with SSI can be largely overcome by the use of a pulsed dual polarity ionization source that allows for independent optimization of the ionization conditions for each ion polarity. The pulsed dual ESI (see Figure 1.4) and ESI/APCI sources were developed subsequently for such purposes [61,62]. High reproducibility of ion signals of both polarities and minimum interference between the two ion sources were noted when the two alternately-pulsed ion sources were placed in front of the same atmospheric/vacuum interface. Almost all types of ion/ ion reactions demonstrated previously can be accomplished using the dual ESI and ESI/APCI sources, which include reactions involving single proton transfer, single electron transfer, multiple proton transfer, etc. [61,62] It can be inferred readily from the implementation of the pulsed dual source that more than two emitters can be adapted to an LIT by sharing the common atmospheric/vacuum interface and ion path, provided that each emitter can be controlled independently by the instrument software. For example, a pulsed triple ionization source has been demonstrated successfully [63] for a sequential ion/ion reaction where two different reagents were desired. Compared to SSI, pulsed double and triple ionization sources give higher ionization efficiency and enable a wider choice for ion/ion reaction combinations, which is critical for the exploration of a broad range of ion/ion chemistries. In addition to mutual storage ion/ion reactions, transmission mode ion/ion reactions have been demonstrated also in the LIT as an alternative option, in which reactions occur when ions of one polarity pass through the quadrupole array with relatively low kinetic energy while ions of the other polarity are stored in the LIT using DC potentials applied to the containment lenses [56,64–67]. The effectiveness of the transmission mode reaction is attributed largely to the high ion injection/ transmission efficiency along the axial direction of the LIT and the fast kinetics associated with ion/ion reactions. The extent to which a reaction can proceed is determined by a number of factors such as the number density of the pre-stored ions, the kinetic energy, and the transmission time of the transmitted ion, etc. A variety of ion/ion reaction phenomena have been demonstrated in transmission mode, which include proton transfer, charge inversion, and electron transfer on both a triple quadrupole/LIT and a hybrid LIT/TOF instrument [56,64–67]. The results show that the extent of ion/ion reaction is similar when either polarity is trapped while passing the other, and the extent can be comparable also to that obtained with a mutual storage ion/ion reaction. Because of the omission of a mutual storage period, transmission mode ion/ion reactions can enjoy the advantage of improved instrument duty cycle [67]. However, the biggest advantage associated with the transmission mode approach is that minimal instrument modification is required for the implementation of ion/ion reactions in most commercial instruments. With the use of an LIT as a reaction vessel, transmission mode ion/ion reactions are, in principle, available on any hybrid instrument such as quadrupole/TOF, LIT/Orbitrap, and LIT/FT-ICR, provided the instrument control software can execute an ion/ion reaction experiment.
13
Ion/Ion Reactions in Electrodynamic Ion Traps
1.2.3 Ion/Ion Reactions in Hybrid Instruments Various hybrid tandem mass spectrometers, which combine two or more distinct types of mass analyzers, have been developed to maximize analytical performance and functionality. From the standpoint of ion/ion reactions, the incorporation of an electrodynamic ion trap into a hybrid instrument allows for the physical separation of the three basic steps involved in an ion/ion reaction experiment, that is, ionization, ion/ion reaction, and mass analysis of reaction products. The separation of these processes provides for the highest degree of flexibility and minimal compromises in the optimization of each step. To date, three major types of hybrid instruments have been described for ion/ion reaction studies using an electrodynamic ion trap as the reaction vessel. The three major types of hybrid instruments are: (i) quadrupole/TOF tandem mass spectrometer; (ii) Orbitrap*; and (iii) LIT /FT-ICR. The first type is a modified commercial quadrupole/TOF tandem mass spectrometer (QSTAR XL, Applied Biosystems/MDS Sciex, Concord, ON, Canada) [68] which consists of three quadrupoles (ion guide (Q0), mass filter (Q1), and Q2), and an orthogonal reflectron TOF mass analyzer, as shown schematically in Figure 1.5. Although, transmission mode ion/ion reactions [66,67] were demonstrated successfully later on this instrument, ion/ion reactions were implemented initially in the mutual storage OR
~
SK Q0
Q1
–’ve ESI/APCI +1500 V
GR
~
IQ1
+’ve nano–ESI
Cation injection
Aux. RF
TOF
Q2
IQ2 +18 V
IQ3 +20 V
+ 13 V +8 V
Anion injection Mutual storage Release to TOF
–2500 V
–14 V
–6 V
–8 V Rf barrier
0V
–8 V
0V
+8 V
–8 V
Rf barrier
Rf barrier
FIGURE 1.5 Schematic of a modified hybrid quadrupole/TOF mass spectrometer (QSTAR XL) and scan functions in a typical mutual storage mode ion/ion reaction (ions of opposite polarities are generated from a pulsed dual ion source). (Reproduced from Xia, Y.; Chrisman, P.A.; Erickson, D.E.; Liu, J.; Liang, X.R.; Londry, F.A.; Yang, M.J.; McLuckey, S.A., Anal. Chem. 2006, 78, 4146–4154. With permission from American Chemical Society.) * See Volume 4, Chapter 3: Theory and Practice of the Orbitrap™ Mass Analyzer by Alexander Makarov.
14
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
mode with the superposition of auxiliary RF signals on the two containment lenses of the Q2 quadrupole array. Compared to mass measurement with an electrodynamic ion trap, the use of a TOF mass analyzer provides superior mass analysis figure-of-merits in several respects; for example, a mass-resolving power of ca 8000 (M/ΔM FWHM) was obtained over a wide range of mass-to-charge ratios with a mass accuracy of ca 20 ppm for external calibration and 5 ppm for internal calibration [68]. Furthermore, an upper m/z limit of about 66,000 Th was observed, which is likely to be determined as much by the efficiency of the detector as it is by the ability of the LIT to store ions of such high m/z-value. A recent demonstration of bi-directional ion transfer between the three quadrupole arrays (that is, Q0, Q1, and Q2) has expanded greatly the MSn functionality of this platform, which allows for ready implementation of multi-step experiments including sequential ion/ion reactions [69]. Essentially all of the key ion/ ion reactions demonstrated in ion traps have been implemented readily on this instrument by use of a pulsed dual ion source. An example of a typical sequence of steps used in a mutual storage mode ion/ion reaction experiment with this instrument is shown in Figure 1.5. The second type, the LTQ Orbitrap XL, is the first commercial hybrid instrument designed with ion/ion reaction capabilities as a result of the efforts of Thermo Scientific to accommodate ETD experiments. The LTQ Orbitrap XL is comprised of an LIT, serving as both an ion/ion reaction vessel and mass analyzer, an Orbitrap [70,71] for high resolution mass analysis, and a ‘C-trap’ for ion transfer from the LIT to the Orbitrap; the ‘C-trap’ is an ion trapping device used to couple the LIT with the Orbitrap. The ion/ion ETD reaction is implemented in mutual storage mode with the positive analyte ions being produced via ESI from the front end of the instrument while the negative ions are formed by a chemical ionization source located at the rear end of the instrument. However, the first ETD experiment demonstrated on a hybrid LTQ Orbitrap used an earlier model of the LTQ Orbitrap XL that lacked the chemical ionization module [72,73]. The key hardware modifications were the application of an auxiliary RF trapping signal to the end lenses of the LIT for mutual ion storage and the adaptation of a pulsed dual ESI source to generate both positive analyte ions and negative ETD reagent ions [74] from the inlet of the instrument [75]. The ion/ ion reaction took place in the LIT and products were transferred via the ‘C-trap’ to the Orbitrap for mass analysis, which provided a mass-resolving power of ca 60,000 and a mass accuracy within 2 ppm. This instrument was used to demonstrate protein identification with high confidence due to both high sequence coverage (from the ion/ ion reaction) and high mass measurement accuracy (from the mass analyzer) [75]. A recent implementation of ETD reactions on the third type of hybrid instrument, a hybrid LIT/FT-ICR instrument [76], presents another example of the value in bringing high mass accuracy and mass-resolving power to the measurement of ion/ion reaction products. The hybrid instrument used a hexapole LIT as the reaction vessel by the superposition of auxiliary RF signals to the end lenses to effect mutual storage ion/ion reactions, the products of which were sent directly to the adjacent FT-ICR for mass analysis. The arrangement of the positive and negative ions sources is very similar to that for the Bruker Daltonics HCTultra™ post-translational modification (PTM) ion trap mass spectrometer [42] described in the previous section, which introduces positive ions formed by ESI from the front end of the instrument and
Ion/Ion Reactions in Electrodynamic Ion Traps
15
admits the negative ions from a negative chemical ionization source through the side of an octopole ion guide. The advantages associated with the coupling of an FT-ICR with a LIT for the ion/ion reaction were clear, as indicated by the resolving power of 60,000 and average mass accuracy of 1.36 ppm observed in the analysis of a small (about 10 kDa) protein by ETD with the use of a relatively low magnetic field strength (3T) cryo-magnet [76]. In summary, with the separation of the various components of an ion/ion reaction experiment, hybrid instruments provide potentially extremely powerful platforms for the gas phase analysis of ion–ion reaction products with high sensitivity, mass resolution, and mass accuracy.
1.3 METHODOLOGIES/APPLICATIONS Mass spectrometry applications that employ ion/ion chemistry have expanded rapidly in the past decade, particularly in biomolecule analysis. Ion/ion proton transfer reactions, for example, have been used widely for complex mixture analysis [12,26,77,78], ion concentration and purification [79–82], precursor ion charge-state manipulation [82–86], simplification of product ion mass spectra [76,81,83,88–93], etc. Charge inversion by ion/ion reactions enable conversion of ions of one polarity to the other and, with two sequential charge inversion reactions, are capable of increasing the net charge of an ion in the mass spectrometer [94–97]. In addition, metal–ion transfer between metal-containing reagents and biomolecules has been demonstrated both for polypeptides and for oligonucleotides [35,98–102]. Subsequent dissociation of the metal-adducted ions has been shown to be able to provide structural information that complements that derived from the dissociation of ions from the same analytes devoid of metals. As a unique dissociation tool using ion/ion chemistry, ETD plays a critical role in providing sequence information complementary to that obtained from other techniques and finds increasing use in peptide and protein characterization [32,33,103–111].
1.3.1 Charge State Manipulation: Proton Transfer 1.3.1.1 Macromolecule Mixture Analysis The analysis of macromolecules by mass spectrometry was facilitated by new ionization methods such as ESI. The multiple-charging phenomenon associated with ESI, however, often leads to a mass spectrum of limited value due to the possibility of severe peak overlap when complex mixtures are analyzed. This problem can be ameliorated to a large extent by use of an ion/ion proton transfer reaction to reduce the number of charges on the ions, which enhances significantly the informing power [112] of the approach by distributing the charge states over a wider range of mass space (provided that such a mass range can be accommodated) and reducing the number of peaks in the spectrum. When the charge states of all analyte ions are reduced largely either to + 1 or −1, the analyte molecular weights can be obtained directly from the mass spectrum, which resembles one obtained with Matrix-assisted laser desorption/ionization. Application of this approach to the analysis of protein mixtures was demonstrated in the early studies of ion/ion reactions using both an ion trap and an atmospheric pressure reactor external to a mass spectrometer [12,77].
16
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Analogous applications to nucleic acid mixture analysis have been described also [26]. The charge-squared dependence of the ion/ion reaction kinetics is of particular importance in this application as it gives rise to a desirable situation where analyte ions of various initial charge states can be reduced to largely singly-charged states with relatively little differential neutralization within the same reaction time window. Furthermore, the minimal degree of complex formation and negligible fragmentation observed commonly in proton transfer reactions contribute to the utility of this approach, which would otherwise add another factor of complexity to the analysis. 1.3.1.2 Precursor Ion Charge State Manipulation The charge state of an ion plays a central role in its chemistry, particularly in its gas phase dissociation [83,87–91,113]. For example, a general tendency has been noted in the fragmentation of protein ions that a high-charge state ion fragments preferentially at the N-terminal side of proline residues while the low-charge state gives products predominantly from C-terminal aspartic acid cleavages [83–85,87]. The chargedependent dissociation behavior of macromolecules makes desirable the ability to manipulate precursor ion charge states in bio-ion primary structure elucidation. A number of approaches are available for charge manipulation of multiply-charged positive ions, including the tuning of solution and electrospray interface conditions [114], and the use of gas-phase ion/molecule proton transfer reactions with strong bases [115–118]; however, ion/ion reactions in an electrodynamic ion trap provide the most comprehensive means for charge reduction of multiply-charged precursor ions. The high exothermicity associated with ion/ion reactions allows for an arbitrary degree of charge state reduction (including reduction to charge states not formed directly via ESI) with high efficiency through flexible control of the ion number densities in the ion trap and a well-defined reaction time window. The ability to accelerate selectively ions of particular mass-to-charge values in an electrodynamic ion trap can be used to inhibit selectively the ion/ion reaction rate of a specific ion, which allows for a gas-phase ion concentration and purification technique called ‘ion parking’ [79]. Ion parking can be effected by use of a low amplitude supplementary RF signal corresponding to the secular frequency of the ion of interest to accelerate the ion to a small degree so that the relative velocity between the two reactant ions is increased and their spatial overlap is decreased, leading to a reduced reaction rate. This phenomenon allows for gas-phase purification of a protein from a protein mixture simultaneously with concentration of the overall signal from highercharge states into a parked lower-charge state that can be interrogated further via tandem mass spectrometry (MS/MS) [81]. A demonstration of gas-phase ion concentration using ion parking is illustrated in Figure 1.6, where a distribution of charge states of bovine serum albumin is concentrated largely into a single-charge state. Application of this technique to protein analysis and unknown protein identification in a complex mixture derived from a whole cell lysate faction of Escherichia coli was described as an early example [80]. In addition to single frequency ion parking for ions within a narrow m/z-range, ion/ion reactions can be inhibited simultaneously for ions over a wider range, a phenomenon called ‘parallel ion parking,’ which can be effected by use of a tailored waveform [119], a single frequency alternating current (AC) signal of relatively high amplitude (a technique termed ‘high amplitude low
17
Ion/Ion Reactions in Electrodynamic Ion Traps (a)
(b)
Abundance
7500 500
[M+47H]47+ [M+38H]38+ [M+54H]54+ 1000
1500
(c)
2500
[M+25H]25+
12500 500
2000
1000
1500
2000
3000
3500
[M+22H]22+
4000
[M+17H]17+
2500
3000
3500
4000
2500
3000
3500
4000
[M+34H]34+
Abundance
150000
500
1000
1500
2000
m/z
FIGURE 1.6 Gas-phase concentration of bovine serum albumin. Mass spectra derived: (a) from native ESI; (b) after partial proton transfer reactions and in the absence of ion parking mode; and (c) after partial proton transfer reactions with ion parking mode. (Reproduced from Reid, G.E.; Wells, J.M.; Badman, E.R.; McLuckey, S.A., Int. J. Mass Spectrom. 2003, 222, 243–258. With permission from Elsevier.)
frequency’ (HALF) parallel ion parking) [120], or a dipolar DC potential across the end-cap electrodes of a 3D ion trap [121,122]. Examples of applications using parallel ion parking include the inhibition of sequential ion/ion reactions involving first generation products in an ETD experiment [119] and simultaneous gas-phase ion concentration for all protein components in a protein mixture [120]. 1.3.1.3 Simplification of Product Ion Mass Spectra A product ion mass spectrum from the dissociation of multiply-charged precursor ions consists generally of a mixture of fragment ions with various charge states up to that of the precursor ion. There is a general tendency for product ions derived from highly-charged precursor ions to be concentrated in a narrow mass-to-charge region that surrounds the precursor ion, leading to a mass spectrum of limited value with significant complexity and congestion due to severe peak overlap. Ion/ion proton transfer reactions have proven to be very useful for the simplification of such mass spectra by reducing product ions to largely singly-charged ions and separating the overlapped peaks of different masses and charges but similar mass-to-charge ratios. This approach
18
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a)
Pre-ion/ion rxn [M+19H]19+ (x3)
Abundance
400 300 200 100 500
700
900
1100
1300 m/z
1500
1700
1900
(b)
Post-ion/ion rxn 50
b88+ b89+– + V/A b97
Abundance
40
b86+ D/D
30 20 10
9000
b103+ V/T
2+ b104+– b110+ [M+2H]
b111+ y129+ L/G y2382+ L/G
b100+ A/Q
10000
11000
12000
13000
[M+3H]3+
y137+ V/T
y143+– y151+
15000
[M+H]+
y154+ D/D
y140+ A/Q 14000
y152+ V/A
16000
17000
y238+ y228+ V/G W/P
y174+ D/G
2+
7000 9000
b130+– y136+
11000
13000
15000
17000 19000 m/z
21000
23000
25000
27000
FIGURE 1.7 Product ion mass spectra of [M + 19H]19 + porcine elastase derived from: (a) CID of [M + 19H]19 + ; and (b) post-ion/ion proton transfer reactions. (Reproduced from Hogan, J.M.; McLuckey, S.A., J. Mass Spectrom. 2003, 38, 245–256. With permission from John Wiley & Sons Limited.)
has been used in a variety of ‘top-down’ studies of model proteins [83,86,89–91,113], and in the identification of a priori unknown proteins in complex mixtures using database searches against in silico product ion mass spectra predicted from whole proteins [80,81]. An example of product ion manipulation using ion/ion proton transfer reactions is shown in Figure 1.7, where isolated porcine elastase [M + 19H]19 + ions were subjected to collision-induced dissociation (CID) (Figure 1.7a) followed by an ion/ion reaction to reduce the entire product ion population of multiply-charged ions to largely singly-charged ions (Figure 1.7b). The congestion of the product ion mass spectrum in Figure 1.7a precluded confident interpretation. However, simplification of the product ion spectrum with ion/ion reactions allowed fragment ions to be assigned without charge state ambiguity. ‘Top-down’ protein identification via either CID or ETD followed by the examination of low mass product ions has been demonstrated also on instruments with moderate upper mass-to-charge limits (m/z = 2000–4000 Th) after all the product ions were reduced to largely singly-charged states [92,93,123]. It is noteworthy that in the latter case, the reagent ions for ETD and the reagent ions for subsequent proton transfer were both derived from the same neutral precursor species [76].
Ion/Ion Reactions in Electrodynamic Ion Traps
19
1.3.2 Charge Inversion Sequential proton transfer between multiply-charged ions of one polarity and a singlycharged ion of opposite polarity often leads to stepwise charge reduction until complete ion neutralization. However, multiple proton transfer within a single encounter can produce ions of opposite polarity, a phenomenon referred to as ‘charge inversion’ [94–97]. Charge inversion enables ion formation in one polarity and analysis in the opposite polarity mode. This capability is an example of how ion/ion reactions allow for the separation of the ionization step from the ion interrogation step in that ions can be formed in one polarity, converted to the opposite polarity, and subjected to dissociation for structure determination. An example of this type of process was demonstrated in a recent ETD study of a phosphopeptide in a simple peptide mixture. Hardly any positive ion signal was observed from the phosphopeptide under positive ESI conditions; however, deprotonated phosphopeptide, which was formed in high abundance via negative ESI, yielded abundant doubly-protonated phosphopeptide ions when the mixture solution was sprayed in the negative mode followed by charge inversion of the deprotonated phosphopeptide [97]. Both positive-to-negative and negative-to-positive charge inversion reactions have been effected by use of a variety of reagents including dendrimers, proteins, and small organic compounds, with dendrimers being the most widely used. In addition to polarity switching via a single charge inversion, two consecutive charge inversion reactions can result in a net increase in ion charge. Charge increase is desirable in many instances; for example, ions of higher charge often give better detector response and allow measurement to be made with higher resolution (for example, with an FT-ICR, for which mass resolution is inversely proportional to the m/z-value of the ion). Also, the ability to access a range of charge states opens the opportunity for the study of a wide range of chemistries. In a typical experiment for charge increase using positive ions as an example, the singly-charged analyte cation is allowed initially to react with multiply-charged reagent ions of opposite polarity, which leads to the charge inversion of the analyte ion to a negative ion (Equation 1.1). In the second charge inversion step, the net effect of charge increase can be achieved when the charge inverted ion, that is, the negative analyte ion, from the previous step reacts with another multiply-charged reagent ion of different polarity (Equation 1.2).
[ M + H]+ + [ N − nH]n− → [ M − H]− + [ N − (n − 2) H]( n− 2)−
(1.1)
[ M − H]− + [R + mH]m+ → [ M + 2H]2+ + [R + (m − 3)H]( m−3)+
(1.2)
1.3.3 Metal–Ion Transfer The ability to insert selectively metal cations into macromolecule ions in the gas phase is attractive both for the study of the intrinsic interactions of metal ions with macromolecules and for maximizing the structural information available from tandem mass spectrometry. For example, singly-charged, singly-sodiated peptide cations have been shown to fragment next to the C-terminal residue [124], whereas the protonated version fragments at various locations along the peptide backbone. The insertion
20
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Abundance
(a)
[M+2H]2+
3e4
[M+3H]3+
2e4 1e4
500 (b) Abundance
[M–2H+2ca]2+
3e4
850 m/z
1200
[M+3H]3+
2e4
[M+H+Ag]2+
1e4
[M+2H]2+
500
[M+2Ag]2+
850 m/z
1200
Abundance
(c) 3e4 2e4 1e4 500
[M+3H]3+ [M+Ni]2+ [M+2H]2+
850 m/z
[M–2H+2Ni]2+ 1200
FIGURE 1.8 Metal-ion transfer via cation switching ion/ion reactions of Trp-11 neurotensin with: (a) calcium acetate anions; (b) silver nitrate anions; and (c) nickel acetate anions. (Reproduced from Newton, K.A.; McLuckey, S.A., J. Am. Chem. Soc. 2003, 125, 12404–12405. With permission from American Chemical Society.)
of various metal ions into peptide or protein ions has been accomplished in the gas phase by cation-switching reactions involving multiply-protonated polypeptides and metal-containing anions, as illustrated in Figure 1.8, where calcium, silver, and nickelcontaining trp-11 neurotensin were formed in the gas phase [98]. The selective removal of cations, either protons or metal ions, from peptide or protein ions may be desirable also in some cases and a number of examples of ion/ion reactions for this purpose have been discussed [99,100]. In addition to metal transfer into polypeptide ions, transition metal ion insertion into oligodeoxynucleotide anions has been effected in the gas phase via ion/ion reactions of transition metal complex cations with multiply-charged oligodeoxynucleotide anions. Depending upon the metal, ligands, and reactant ion charge states, metal transfer to the oligodeoxynucleotide anion competes more or less effectively with other processes, such as electron transfer and cation/anion complex formation [102]. Gas-phase metal-ion transfer to macromolecules via ion/ion reactions
Ion/Ion Reactions in Electrodynamic Ion Traps
21
enables the separation of the ionization process from the formation of metal-containing macromolecule ions, which allows the optimization of both processes individually and the production of metal-containing species not formed readily in solution.
1.3.4 Electron Transfer Dissociation (ETD) Although the first bio-ion/ion electron transfer reaction involved multiply-deprotonated oligonucleotides in reaction with xenon cations [125,126] (later work was extended to those involving multiply-deprotonated peptides [127]), extensive applications of ETD have been focused largely on the reaction of multiply-protonated polypeptides with radical anions, due to the unique sequence information provided by this reactive combination. When an electron is transferred from an anion to a multiply-charged polypeptide cation, an open-shell radical cation is created, which fragments subsequently to produce c- and z-type sequence ions from backbone N−Cα bond cleavages, a phenomenon highly analogous to ECD [128,129]. Compared to traditional activation methods such as CID, ETD tends to be less selective than CID in terms of the range of structurally-informative channels that contribute to the product ion mass spectrum and, as a result, tends to provide more extensive sequence information. In addition, ETD tends not to cleave bonds to PTM that are relatively labile under CID conditions. Both features of ETD have been noted in the first ETD study on multiply-protonated polypeptides, as shown in Figure 1.9, in which a methyl-esterified phosphopeptide generated from a tryptic digest of human nuclear proteins was subjected to dissociation by both CID and ETD [32]. The CID product ion mass spectrum in Figure 1.9a is dominated by product ions corresponding to neutral losses of phosphoric acid and either water or methanol and yields little, if any, sequence ion information, whereas complete sequence coverage is achieved in the ETD spectrum, Figure 1.9b, with 13 out of 14 possible c/z-type ions observed. The propensity of ETD to retain the PTM information is demonstrated also in Figure 1.9b where no fragment ions are observed corresponding to the loss of phosphoric acid. Although Figure 1.9 illustrates an extreme case where essentially no sequence information can be acquired from CID, there is a consensus that, in general, ETD gives complementary sequence information to that provided by other activation methods, such as CID, in either modified or unmodified peptides [67,106,130,131]. This conclusion has been supported widely by many studies, as exemplified by a recent report that described a rapid alternating transmission mode ETD and CID for the characterization of polypeptides ions, including a phosphopeptide, a glycopeptide, and an unmodified tryptic peptide [67]. In the glycopeptide (Figure 1.10), for instance, ETD (Figure 1.10a) generates information about both peptide sequence and locations of the glycosylation sites, whereas CID (Figure 1.10b) provides information about the glycan structure. Motivated by the complementary nature of ETD, combined techniques involving ETD, for example CID coupled with ETD, have been used for peptide unknown protein database search with improved performance [106,132]. In addition to phosphorylation and glycosylation, disulfide linkage formation is another PTM that can be characterized more readily by ETD than by excitation with either photons or low-energy collisions, the latter of which tend not to cleave disulfide bonds present in multiply-protonated polypeptides [133,134]. The rupture
22
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
144 300
b
E
CID
R E R pS L pS 1234 1091 935 768 655 488 332 189
y
(M+3H–H3PO4–MeOH)+++
100 Relative abundance
467 580 747 903 1046 1234
R
(M+3H–2H3PO4)+++
Int: 2.08 × 108
(M+3H–H3PO4–H2O)+++ (M+3H–H3PO4)+++
50
0 (b)
c
161 317 E R
Relative abundance
z
1234 1075 919 752 639 (M+3H)+++
14 10 6 2
484 597 764 920 1063 1234 R E R pS L pS
z1 y1 200
c2
z2
ETD
472 316 173
(M+2H)++
c6
c4 z6 c3
z3
a2 400
z5
z4
z6** 600
(M+1H)+
Int: 2.19 × 105
m/z
c7 y6
800
z7 a7 1000
1200
FIGURE 1.9 Comparison of mass spectra derived from: (a) CID; and (b) ETD of a phosphopeptide. (Reproduced from Syka, J.E.P.; Coon, J.J.; Schroeder, M.J.; Shabanowitz, J.; Hunt, D.F., Proc. Natl. Acad. Sci. USA. 2004, 101, 9528–9533. With permission from National Academy of Sciences.)
of disulfide linkages using ETD is particularly useful in the characterization of polypeptides with internal disulfide linkages due to the fact that the sequence information within the cyclic structure formed by the disulfide bond is, in most cases, accessible only when the ring structure is opened [135]. In addition to the primary structure information obtained from backbone cleavage, ETD allows also for a ready differentiation of some isomeric structures of the protein side-chain, for example, isoaspartic acid and aspartic acid residues, because its characteristic dissociation of N−Cα bonds leads to the formation of diagnostic ions [136]. Besides the structural information provided for polypeptides, ETD has been shown to be useful for the characterization of glycerophosphocholine lipids [137] and may also play an important role in quantitation of peptides and proteins. The variety of applications related to ETD is expected to increase rapidly as its compatibility with high throughput experiments is
c3
z3
c4
600
c5
c6
c7 c8 c9 1100
c10
c11 c12
c13
2+
1600
z4 8
9
Man
Man
GlcNAc
GlcNAc
2100
2600
Activation on 2+: 48.60 kHz, 1.0 Vpp z6 z14 z12 z10 z 5 z7 z z13 z11 z
Xyl
Fuc
Man
SKP A Q G Y G Y L G V F N NSK
3100
0 700
800
[M-FucXylMan3+3H]3+ 900
1000
1100 m/z
1200
-Man
1300
-Man
1400
-Xyl 1500
SKP A Q G Y G Y L G V F N NSK [M-Xyl+3H]3+ [M-Fuc+3H]3+ Fuc GlcNAc [M-Man+3H]3+ [M-FucXylMan GlcNAc+2H]2+ 3 2+ [M-FucXylMan3GlcNAc2+2H] GlcNAc [M-XylMan3GlcNAc+2H]2+ 4.0E3 [M-FucXyl+3H]3+ 3+ 2+ Xyl Man Man [M-FucXylMan3+2H] [M-XylMan+3H]3+ [M-FucXylMan2+2H]2+ [M-FucMan+3H]3+ Man 2+ [M-FucXylMan+3H]3+ [M-FucXylMan+2H]2+ [M-Fuc+2H] -GlcNAc 2.0E3 -Man -GlcNAc [M-FucXyl+2H]2+ [M-FucXylMan2+3H]3+
6.0E3
0 100
40
80
120
160
3+ x 22.9
Neutral losses
Abundance (arb. unit)
FIGURE 1.10 Product ion mass spectra of a tryptic lectin glycopeptide derived from: (a) transmission mode ETD of the triply-charged peptide with azobenzene radical anions; and (b) beam-type CID of the triply-charged peptide. (Reproduced from Han, H.; Xia, Y.; Yang, M.; McLuckey, S.A., Anal. Chem. 2008, 80, 3492–3497. With permission from American Chemical Society.)
(b)
(a)
Ion/Ion Reactions in Electrodynamic Ion Traps 23
24
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
developed. For example, implementation of ETD on the chromatographic timescale has been demonstrated already [138], and has been found to be dependent upon the high rates of reaction that can be achieved in the LIT. Because of the importance of ETD, a variety of efforts have been made for its characterization and to improve its performance. A major complication in an ETD experiment is the existence of various competing channels, which include proton transfer, side-chain losses, electron transfer without dissociation (ET, no D), and electron transfer leading to backbone cleavages (that is, ETD), with only the latter being desirable in terms of structural characterization [139]. The competing channels are affected by many factors including, but not limited to, the identity of the reagent radical anion, bath gas temperature, the cation charge identity and its location on a multiply-charged polypeptide, the peptide size, and the charge states of the polypeptide [34,131,139–142]. Recent studies have shown that an ETD reagent anion with favorable Franck–Condon factors and a relatively low electron affinity for the neutral form of the reagent tends to give higher electron transfer relative to proton transfer [141]. When the cation charge identities are considered, similar degrees of electron transfer were observed for multiply-protonated peptide ions containing protonated histidine, arginine, or lysine, which were much higher than that for peptides having fixed-charge sites. Among the four types of cation-charge sites, protonated histidine showed the highest degree of ET, no D, while no apparent intact electron transfer products were observed for peptides with protonated arginine or lysine. All cation types showed side-chain losses with arginine yielding the greatest fraction and lysine giving the smallest [142]. A general trend with increased partitioning into the ET channel was observed when ETD was performed on a polypeptide with higher charge, which was accompanied also by the decreased contribution from side-chains losses and increased partitioning into the ETD channel [139]. These results highlight the importance of using as high a charge state as possible in an ETD experiment to maximize sequence information. As a result, attention has been directed to the development of various methods to increase the charges on the peptides such as the manipulation of the solution composition of tryptic digests [143], use of other enzymatic digestions, such as endoproteinase Lys-C digestion [132], and microwave-assisted D-cleavages [144], to produce peptides of larger size which are expected to produce high charges from ESI, etc. It is noteworthy that not only the charge state but also the size of the polypeptide affects significantly the partitioning of the ETD channel [131,139], with larger size correlating inversely with the fraction of ETD relative to the sequence-uninformative channels. The ET, no D channel is not informative in the sense that no sequence ions are produced directly. However, the ET, no D product has been proposed to consist of c/z-fragments remaining bound via noncovalent interactions [128] or comprising of an intact protein ion with a covalent bond weakened significantly by the electron attachment [145]. As a result, a variety of approaches have been developed to dissociate ET, no D products to improve the production of the sequence ions [66,67,146,147].
1.4 CONCLUSIONS The robust nature of ion/ion chemistry (for example, fast kinetics due to the long range Coulombic attraction and the universal nature of ion/ion reactions because of high
Ion/Ion Reactions in Electrodynamic Ion Traps
25
exothermicities) enables implementation of ion/ion reactions under a variety of conditions; however, the combination of ion/ion reactions with electrodynamic ion traps has many advantages. The use of an electrodynamic ion trap as an ion/ion reaction vessel allows for the separation of the major components of an MS experiment, that is, ionization, reaction, and mass analysis. The MSn functionality of an ion trap enables consecutive ion/ion reactions between multiple mass selection steps and allows for the exploration of different ion/ion reaction phenomena within a single experiment. The ease with which an LIT can be coupled with other mass spectrometric components makes possible the implementation of ion/ion reactions with virtually any form of mass analyzer. In addition to the commonly-used mutual storage mode in LITs, ion/ion reactions can be implemented in transmission mode, which requires minimal hardware modification and provides higher instrument duty cycles. The development of various new ion inlets such as pulsed dual and triple ion sources delivers high versatility to an instrument for the exploration of the high dimensionality of the ion/ion chemistry. Proton transfer ion/ion reactions have found various applications in electrodynamic ion traps including biopolymer mixture analysis by charge reduction of multiply-charged components generated by ESI to largely singly-charged ions. A similar strategy has been applied also to mixtures of product ions generated in a tandem mass spectrometric experiment to facilitate spectral interpretation. The well-defined ion frequencies established by ion storage in quadrupolar ion traps combined with ion/ion chemistry allow for ready charge reduction of a precursor ion to any arbitrary extent and gas-phase ion purification and concentration by use of ‘ion parking’ or ‘parallel ion parking.’ Charge inversion together with both metal-ion insertion and removal provide means to convert ions from one type to another. As an important dissociation tool, ETD provides sequence information, complementary to that provided by traditional CID methods, and tends to retain PTM information during the dissociation. The unique information provided by ETD makes it a valuable tool in the study of molecular biology. The recent introduction of commercially-available instruments capable of executing ion/ion reactions in ion traps will accelerate undoubtedly the development and applications of ion/ion reactions.
REFERENCES
1. Smith, D.; Adams, N.G.; Church, M.J. Mutual neutralization rates of ionospherically important ions. Planet Space Sci. 1976, 24, 697–703. 2. Mahan, B.H. Advances in Chemical Physics, Vol. 23, pp. 1–40. Wiley, New York, 1973. 3. Bates, D.R. Ion-ion reactions in an ambient gas. In Advances in Atomic and Molecular Physics, eds. D.R. Bates and B. Bederson, Vol. 20, pp. 1–37. Academic Press, Orlando, 1985. 4. Flannery, M.R. Theory of ion-ion recombination. Philos. T. Roy. Soc. A 1982, 304, 447–497. 5. Thomson, J.J.; Rutherford, E. On the passage of the electricity through gases exposed to Röntgen rays. Philos. Mag. 1896, 42, 392–407. 6. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; Whitehouse, C.M. Electrospray ionizationprinciples and practice. Mass Spectrom. Rev. 1990, 9, 37–70. 7. Smith, R.D.; Loo, J.A.; Edmonds, C.G.; Barinaga, C.J.; Udseth, H.R. New developments in biochemical mass spectrometry-electrospray ionization. Anal. Chem. 1990, 62, 882–899. 8. Gaskell, S.J. Electrospray: Principles and practice. J. Mass Spectrom. 1997, 32, 677–688.
26
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
9. Cole, R.B., Ed. Electrospray Ionization Mass Spectrometry, Fundamentals, Instrumentation, and Applications, Wiley, New York, 1997. 10. Loo, R.R.O.; Udseth, H.R.; Smith, R.D. Evidence of charge inversion in the reaction of singly charged anions with multiply charged macro-ions. J. Phys. Chem. 1991, 95, 6412–6415. 11. Loo, R.R.O.; Udseth, H.R.; Smith, R.D. A new approach for the study of gas-phase ion-ion reactions using electrospray ionization. J. Am. Soc. Mass Spectrom. 1992, 3, 695–705. 12. Scalf, M.; Westphall, M.S.; Krause, J.; Kaufman, S.L.; Smith, L.M. Controlling charge states of large ions. Science 1999, 283, 194–197. 13. Scalf, M.; Westphall, M.S.; Smith, L.M. Charge reduction electrospray mass spectrometry. Anal. Chem. 2000, 72, 52–60. 14. Ebeling, D.D.; Westphall, M.S.; Scalf, M.; Smith, L.M. Corona discharge in charge reduction electrospray mass spectrometry. Anal. Chem. 2000, 72, 5158–5161. 15. Frey, B.L.; Lin, Y.; Westphall, M.S.; Smith, L.M. Controlling gas-phase reactions for efficient charge reduction electrospray mass spectrometry of intact proteins. J. Am. Soc. Mass Spectrom. 2005, 16, 1876–1887. 16. McLuckey, S.A.; Stephenson, J.L. Ion/ion chemistry of high mass multiply charged ions. Mass Spectrom. Rev. 1998, 17, 369–407. 17. Pitteri, S.J.; McLuckey, S.A. Recent developments in the ion/ion chemistry of high-mass multiply charged ions. Mass Spectrom. Rev. 2005, 24, 931–958. 18. Xia, Y.; McLuckey, S.A. Evolution of instrumentation for the study of gas-phase ion/ion chemistry via mass spectrometry. J. Am. Soc. Mass Spectrom. 2008, 19, 173–189. 19. Mather, R.E.; Todd, J.F.J. The quadrupole ion store (Quistor) 7. Simultaneous positivenegative ion mass spectrometry. Int. J. Mass Spectrom. Ion Phys. 1980, 33, 159–165. 20. Williams, J.D.; Cooks, R.G. Reduction of space-charging in the quadrupole ion-trap by sequential injection and simultaneous storage of positively and negatively charged ions. Rapid Commun. Mass Spectrom. 1993, 7, 380–382. 21. Louris, J.N.; Brodbelt-Lustig, J.S.; Cooks, R.G.; Glish, G.L.; Van Berkel, G.J.; McLuckey, S.A. Ion isolation and sequential stages of mass-spectrometry in a quadrupole ion trap mass-spectrometer. Int. J. Mass Spectrom. Ion Processes 1990, 96, 117–137. 22. Johnson, J.V.; Yost, R.A.; Kelley, P.E.; Bradford, D.C. Tandem-in-space and tandem-intime mass-spectrometry-triple quadrupoles and quadrupole ion traps. Anal. Chem. 1990, 62, 2162–2172. 23. Stafford, G.C.; Kelley, P.E.; Syka, J.E.P.; Reynolds, W.E.; Todd, J.F.J. Recent improvements in and analytical applications of advanced ion trap technology. Int. J. Mass Spectrom. Ion Processes 1984, 60, 85–98. 24. Stephenson, J.L.; McLuckey, S.A. Ion/ion reactions in the gas phase: Proton transfer reactions involving multiply-charged proteins. J. Am. Chem. Soc. 1996, 118, 7390–7397. 25. March, R.E.; Londry, F.A. Theory of quadrupole mass spectrometry. In Practical Aspects of Ion Trap Mass Spectrometry, eds. R.E. March and J.F.J. Todd, Vol. 1, Chapter 2, pp. 25–48. CRC Press, Boca Raton, 1995. 26. McLuckey, S.A.; Wu, J.; Bundy, J.L.; Stephenson, J.L.; Hurst, G.B. Oligonucleotide mixture, analysis via electrospray and ion/ion reactions in a quadrupole ion trap. Anal. Chem. 2002, 74, 976–984. 27. Stephenson, J.L.; McLuckey, S.A. Anion effects on storage and resonance ejection of high mass-to-charge cations in quadrupole ion trap mass spectrometry. Anal. Chem. 1997, 69, 3760–3766. 28. Stephenson, J.L.; McLuckey, S.A. Adaptation of the Paul Trap for study of the reaction of multiply charged cations with singly charged anions. Int. J. Mass Spectrom. Ion Processes 1997, 162, 89–106. 29. McLuckey, S.A.; Glish, G.L.; Asano, K.G.; Grant, B.C. Atmospheric sampling glowdischarge ionization source for the determination of trace organic-compounds in ambient air. Anal. Chem. 1988, 60, 2220–2227.
Ion/Ion Reactions in Electrodynamic Ion Traps
27
30. Reid, G.E.; Wells, J.M.; Badman, E.R.; McLuckey, S.A. Performance of a quadrupole ion trap mass spectrometer adapted for ion/ion reaction studies. Int. J. Mass Spectrom. 2003, 222, 243–258. 31. Kaiser, R.E.; Cooks, R.G.; Stafford, G.C.; Syka, J.E.P.; Hemberger, P.H. Operation of a quadrupole ion trap mass-spectrometer to achieve high mass charge ratios. Int. J. Mass Spectrom. Ion Processes 1991, 106, 79–115. 32. Syka, J.E.P.; Coon, J.J.; Schroeder, M.J.; Shabanowitz, J.; Hunt, D.F. Peptide and protein sequence analysis by electron transfer dissociation mass spectrometry. Proc. Natl. Acad. Sci. USA 2004, 101, 9528–9533. 33. Pitteri, S.J.; Chrisman, P.A.; Hogan, J.M.; McLuckey, S.A. Electron transfer ion/ion reactions in a three-dimensional quadrupole ion trap: Reactions of doubly and triply protonated peptides with SO2.–. Anal. Chem. 2005, 77, 1831–1839. 34. Pitteri, S.J.; Chrisman, P.A.; McLuckey, S.A. Electron transfer ion/ion reactions of doubly protonated peptides: Effect of elevated bath gas temperature. Anal. Chem. 2005, 77, 5662–5669. 35. Payne, A.H.; Glish, G.L. Gas-phase ion/ion interactions between peptides or proteins and iron ions in a quadrupole ion trap. Int. J. Mass Spectrom. 2001, 204, 47–54. 36. Herron, W.J.; Goeringer, D.E.; McLuckey, S.A. Ion-ion reactions in the gas-phase: Proton-transfer reactions of protonated pyridine with multiply-charged oligonucleotide anions. J. Am. Soc. Mass Spectrom. 1995, 6, 529–532. 37. Creaser, C.S. Chemical ionization in ion trap mass spectrometry. In Practical Aspects of Ion Trap Mass Spectrometry, eds. R.E. March and J.F.J. Todd, Vol. III, Chapter 7, pp. 239–253. CRC Press, Boca Raton, FL, 1995. 38. Berberich, D.W.; Yost, R.A. Negative chemical-ionization in quadrupole ion-trap massspectrometry-effects of applied voltages and reaction-times. J. Am. Soc. Mass Spectrom. 1994, 5, 757–764. 39. Wells, J.M.; Chrisman, P.A.; McLuckey, S.A. “Dueling” ESI: Instrumentation to study ion/ion reactions of electrospray-generated cations and anions. J. Am. Soc. Mass Spectrom. 2002, 13, 614–622. 40. Wells, J.M.; Chrisman, P.A.; McLuckey, S.A. Formation and characterization of proteinprotein complexes in vacuo. J. Am. Chem. Soc. 2003, 125, 7238–7249. 41. Badman, E.R.; Chrisman, P.A.; McLuckey, S.A. A quadrupole ion trap mass spectrometer with three independent ion sources for the study of gas-phase ion/ion reactions. Anal. Chem. 2002, 74, 6237–6243. 42. Tang, N.; Goodley, P.C.; Horn, D.; Miller, C.; Hoerth, P.; Mohsin, S. Improving protein identification using a combination of collision induced dissociation and electron transfer dissociation MS/MS, Proc. 54th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, 2006. 43. Senko, M.W.; Hendrickson, C.L.; Emmett, M.R.; Shi, S.D.H.; Marshall, A.G. External accumulation of ions for enhanced electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. J. Am. Soc. Mass Spectrom. 1997, 8, 970–976. 44. Sannes-Lowery, K.; Griffey, R.H.; Kruppa, G.H.; Speir, J.P.; Hofstadler, S.A. Multipole storage assisted dissociation, a novel in-source dissociation technique for electrospray ionization generated ions. Rapid Commun. Mass Spectrom. 1998, 12, 1957–1961. 45. Belov, M.E.; Nikolaev, E.N.; Anderson, G.A.; Udseth, H.R.; Conrads, T.P.; Veenstra, T.D.; Masselon, C.D.; Gorshkov, M.V.; Smith, R.D. Design and performance of an ESI interface for selective external ion accumulation coupled to a Fourier transform ion cyclotron mass spectrometer. Anal. Chem. 2001, 73, 253–261. 46. Campbell, J.M.; Collings, B.A.; Douglas, D.J. A new linear ion trap time-of-flight system with tandem mass spectrometry capabilities. Rapid Commun. Mass Spectrom. 1998, 12, 1463–1474.
28
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
47. Collings, B.A.; Campbell, J.M.; Mao, D.M.; Douglas, D.J. A combined linear ion trap time-of-flight system with improved performance and MSn capabilities. Rapid Commun. Mass Spectrom. 2001, 15, 1777–1795. 48. Ijames, C.F. A proposed two dimensional quadrupole/electrostatic ion trap time-of-flight mass spectrometer. Proc. 44th ASMS Conference on Mass Spectrometry and Allied Topics, Portland, Oregon, 1996. 49. Dolnikowski, G.G.; Kristo, M.J.; Enke, C.G.; Watson, J.T. Ion trapping technique for ion molecule reaction studies in the center quadrupole of a triple quadrupole mass spectrometer. Int. J. Mass Spectrom. Ion Processes 1988, 82, 1–15. 50. Cha, B.C.; Blades, M.; Douglas, D.J. An interface with a linear quadrupole ion guide for an electrospray-ion trap mass spectrometer system. Anal. Chem. 2000, 72, 5647–5654. 51. Schwartz, J.C.; Senko, M.W.; Syka, J.E.P. A two dimensional quadrupole ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 2002, 13, 659–669. 52. Hager, J.W. A new linear ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 2002, 16, 512–526. 53. Hager, J.W.; Londry, F.A. Method for providing barrier fields at the entrance and exit end of a mass spectrometer, U.S. Patent 2005, 7,227,130, B2. 54. Xia, Y.; Wu, J.; McLuckey, S.A.; Londry, F.A.; Hager, J.W. Mutual storage mode ion/ion reactions in a hybrid linear ion trap. J. Am. Soc. Mass Spectrom. 2005, 16, 71–81. 55. Londry, F.A.; Hager, J.W. Mass selective axial ion ejection from a linear quadrupole ion trap. J. Am. Soc. Mass Spectrom. 2003, 14, 1130–1147. 56. Wu, J.; Hager, J.W.; Xia, Y.; Londry, F.A.; McLuckey, S.A. Positive ion transmission mode ion/ion reactions in a hybrid linear ion trap. Anal. Chem. 2004, 76, 5006–5015. 57. Hirabayashi, A.; Sakairi, M.; Koizumi, H. Sonic spray mass spectrometry. Anal. Chem. 1995, 67, 2878–2882. 58. Hirabayashi, A.; Sakairi, M.; Koizumi, H. Sonic spray ionization method for atmosphericpressure ionization mass-spectrometry. Anal. Chem. 1994, 66, 4557–4559. 59. Takats, Z.; Nanita, S.C.; Cooks, R.G.; Schlosser, G.; Vekey, K. Amino acid clusters formed by sonic spray ionization. Anal. Chem. 2003, 75, 1514–1523. 60. Xia, Y.; Liang, X.R.; McLuckey, S.A. Sonic spray as a dual polarity ion source for ion/ ion reactions. Anal. Chem. 2005, 77, 3683–3689. 61. Xia, Y.; Liang, X.R.; McLuckey, S.A. Pulsed dual electrospray ionization for ion/ion reactions. J. Am. Soc. Mass Spectrom. 2005, 16, 1750–1756. 62. Liang, X.R.; Xia, Y.; McLuckey, S.A. Alternately pulsed nanoelectrospray ionization/ atmospheric pressure chemical ionization for ion/ion reactions in an electrodynamic ion trap. Anal. Chem. 2006, 78, 3208–3212. 63. Liang, X.R.; Han, H.L.; Xia, Y.; McLuckey, S.A. A pulsed triple ionization source for sequential ion/ion reactions in an electrodynamic ion trap. J. Am. Soc. Mass Spectrom. 2007, 18, 369–376. 64. Liang, X.R.; Hager, J.W.; McLuckey, S.A. Transmission mode ion/ion electron-transfer dissociation in a linear ion trap. Anal. Chem. 2007, 79, 3363–3370. 65. Liang, X.R.; McLuckey, S.A. Transmission mode ion/ion proton transfer reactions in a linear ion trap. J. Am. Soc. Mass Spectrom. 2007, 18, 882–890. 66. Xia, Y.; Han, H.; McLuckey, S.A. Activation of intact electron-transfer products of polypeptides and proteins in cation transmission mode ion/ion reactions. Anal. Chem. 2008, 80, 1111–1117. 67. Han, H.; Xia, Y.; Yang, M.; McLuckey, S.A. Rapidly alternating transmission mode electron-transfer dissociation and collisional activation for the characterization of polypeptide ions. Anal. Chem. 2008, 80, 3492–3497. 68. Xia, Y.; Chrisman, P.A.; Erickson, D.E.; Liu, J.; Liang, X.R.; Londry, F.A.; Yang, M.J.; McLuckey, S.A. Implementation of ion/ion reactions in a quadrupole/time-of-flight tandem mass spectrometer. Anal. Chem. 2006, 78, 4146–4154.
Ion/Ion Reactions in Electrodynamic Ion Traps
29
69. Xia, Y.; Thomson, B.A.; McLuckey, S.A. Bi-directional ion transfer between quadrupole arrays: MSn ion/ion reaction experiments on a quadrupole/time-of-flight tandem mass spectrometer. Anal. Chem. 2007, 79, 8199–8206. 70. Hu, Q.Z.; Noll, R.J.; Li, H.Y.; Makarov, A.; Hardman, M.; Cooks, R.G. The Orbitrap: A new mass spectrometer. J. Mass Spectrom. 2005, 40, 430–443. 71. Scigelova, M.; Makarov, A. Orbitrap mass analyzer-overview and applications in proteomics. Proteomics 2006, 6, 16–21. 72. Erickson, B. Linear ion trap/Orbitrap mass spectrometer. Anal. Chem. 2006, 78, 2089–2089. 73. Makarov, A.; Denisov, E.; Kholomeev, A.; Baischun, W.; Lange, O.; Strupat, K.; Horning, S. Performance evaluation of a hybrid linear ion trap/Orbitrap mass spectrometer. Anal. Chem. 2006, 78, 2113–2120. 74. Huang, T.Y.; Emory, J.F.; O’Hair, R.A.J.; McLuckey, S.A. Electron transfer reagent anion formation via electrospray ionization and collision-induced dissociation. Anal. Chem. 2006, 78, 7387–7391. 75. McAlister, G.C.; Phanstiel, D.; Good, D.M.; Berggren, W.T.; Coon, J.J. Implementation of electron-transfer dissociation on a hybrid linear ion trap-orbitrap mass spectrometer. Anal. Chem. 2007, 79, 3525–3534. 76. Kaplan, D.A.; Hartmer, R.; Speir, J.P.; Stoermer, C.; Gumerov, D.; Easterling, M.L.; Brekenfeld, A.; Kim, T.; Laukien, F.; Park, M.A. Electron transfer dissociation in the hexapole collision cell of a hybrid quadrupole-hexapole Fourier transform ion cyclotron resonance mass spectrometer. Rapid Commun. Mass Spectrom. 2008, 22, 271–278. 77. Stephenson, J.L.; McLuckey, S.A. Ion/ion proton transfer reactions for protein mixture analysis. Anal. Chem. 1996, 68, 4026–4032. 78. VerBerkmoes, N.C.; Strader, M.B.; Smiley, R.D.; Howell, E.E.; Hurst, G.B.; Hettich, R.L.; Stephenson, J.L. Intact protein analysis for site-directed mutagenesis overexpression products: Plasmid-encoded R67 dihydrofolate reductase. Anal. Biochem. 2002, 305, 68–81. 79. McLuckey, S.A.; Reid, G.E.; Wells, J.M. Ion parking during ion/ion reactions in electrodynamic ion traps. Anal. Chem. 2002, 74, 336–346. 80. Reid, G.E.; Shang, H.; Hogan, J.M.; Lee, G.U.; McLuckey, S.A. Gas-phase concentration, purification, and identification of whole proteins from complex mixtures. J. Am. Chem. Soc. 2002, 124, 7353–7362. 81. Amunugama, R.; Hogan, J.M.; Newton, K.A.; McLuckey, S.A. Whole protein dissociation in a quadrupole ion trap: Identification of an a priori unknown modified protein. Anal. Chem. 2004, 76, 720–727. 82. He, M.; Reid, G.E.; Shang, H.; Lee, G.U.; McLuckey, S.A. Dissociation of multiple protein ion charge states following a single gas-phase purification and concentration procedure. Anal. Chem. 2002, 74, 4653–4661. 83. Wells, J.M.; Stephenson, J.L.; McLuckey, S.A. Charge dependence of protonated insulin decompositions. Int. J. Mass Spectrom. 2000, 203, A1–A9. 84. Reid, G.E.; Wu, J.; Chrisman, P.A.; Wells, J.M.; McLuckey, S.A. Charge-state-dependent sequence analysis of protonated ubiquitin ions via ion trap tandem mass spectrometry. Anal. Chem. 2001, 73, 3274–3281. 85. Hogan, J.M.; McLuckey, S.A. Charge state dependent collision-induced dissociation of native and reduced porcine elastase. J. Mass Spectrom. 2003, 38, 245–256. 86. Pitteri, S.J.; Chrisman, P.A.; Badman, E.R.; McLuckey, S.A. Charge-state dependent dissociation of a trypsin/inhibitor complex via ion trap collisional activation. Int. J. Mass Spectrom. 2006, 253, 147–155. 87. Mekecha, T.T.; Amunugama, R.; McLuckey, S.A. Ion trap collision-induced dissociation of human hemoglobin α-chain cations. J. Am. Soc. Mass Spectrom. 2006, 17, 923–931.
30
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
88. Cargile, B.J.; McLuckey, S.A.; Stephenson, J.L. Identification of bacteriophage MS2 coat protein from E-coli lysates via ion trap collisional activation of intact protein ions. Anal. Chem. 2001, 73, 1277–1285. 89. Newton, K.A.; Chrisman, P.A.; Reid, G.E.; Wells, J.M.; McLuckey, S.A. Gaseous apomyoblobin ion dissociation in a quadrupole ion trap: [M + 2H]2 + –[M + 21H]21 + . Int. J. Mass Spectrom. 2001, 212, 359–376. 90. Wells, J.M.; Reid, G.E.; Engel, B.J.; Pan, P.; McLuckey, S.A. Dissociation reactions of gaseous ferro-, ferri-, and apo-cytochrome c ions. J. Am. Soc. Mass Spectrom. 2001, 12, 873–876. 91. Engel, B.J.; Pan, P.; Reid, G.E.; Wells, J.M.; McLuckey, S.A. Charge state dependent fragmentation of gaseous protein ions in a quadrupole ion trap: Bovine ferri-, ferro-, and apo-cytochrome c. Int. J. Mass Spectrom. 2002, 219, 171–187. 92. Coon, J.J.; Ueberheide, B.; Syka, J.E.P.; Dryhurst, D.D.; Ausio, J.; Shabanowitz, J.; Hunt, D.F. Protein identification using sequential ion/ion reactions and tandem mass spectrometry. Proc. Natl. Acad. Sci. USA 2005, 102, 9463–9468. 93. Bowers, J.J.; Liu, J.; Gunawardena, H.P.; McLuckey, S.A. Protein identification via ion trap collision-induced dissociation and examination of low mass product ions. J. Mass Spectrom. 2008, 43, 23–34. 94. He, M.; Emory, J.F.; McLuckey, S.A. Reagent anions for charge inversion of polypeptide/ protein cations in the gas phase. Anal. Chem. 2005, 77, 3173–3182. 95. He, M.; McLuckey, S.A. Two ion/ion charge inversion steps to form a doubly protonated peptide from a singly protonated peptide in the gas phase. J. Am. Chem. Soc. 2003, 125, 7756–7757. 96. He, M.; McLuckey, S.A. Increasing the negative charge of a macroanion in the gas phase via sequential charge inversion reactions. Anal. Chem. 2004, 76, 4189–4192. 97. Gunawardena, H.P.; Emory, J.F.; McLuckey, S.A. Phosphopeptide anion characterization via sequential charge inversion and electron-transfer dissociation. Anal. Chem. 2006, 78, 3788–3793. 98. Newton, K.A.; McLuckey, S.A. Gas-phase peptide/protein cationizing agent switching via ion/ion reactions. J. Am. Chem. Soc. 2003, 125, 12404–12405. 99. Newton, K.A.; McLuckey, S.A. Generation and manipulation of sodium cationized peptides in the gas phase. J. Am. Soc. Mass Spectrom. 2004, 15, 607–615. 100. Newton, K.A.; He, M.; Amunugama, R.; McLuckey, S.A. Selective cation removal from gaseous polypeptide ions: Proton vs. sodium ion abstraction via ion/ion reactions. Phys. Chem. Chem. Phys. 2004, 6, 2710–2717. 101. Hodges, B.D.M.; Liang, X.; McLuckey, S.A. Generation of di-lithiated peptide ions from multiply protonated peptides via ion/ion reactions. Int. J. Mass Spectrom. 2007, 267, 183–189. 102. Barlow, C.K.; Hodges, B.D.M.; Xia, Y.; O’Hair, R.A.J.; McLuckey, S.A. Gas-phase ion/ ion reactions of transition metal complex cations with multiply charged oligodeoxynucleotide anions. J. Am. Soc. Mass Spectrom. 2008, 19, 281–293. 103. Chi, A.; Huttenhower, C.; Geer, L.Y.; Coon, J.J.; Syka, J.E.P.; Bai, D.L.; Shabanowitz, J.; Burke, D.J.; Troyanskaya, O.G.; Hunt, D.F. Analysis of phosphorylation sites on proteins from Saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proc. Natl. Acad. Sci. USA 2007, 104, 2193–2198. 104. Khidekel, N.; Ficarro, S.B.; Clark, P.M.; Bryan, M.C.; Swaney, D.L.; Rexach, J.E.; Sun, Y.E.; Coon, J.J.; Peters, E.C.; Hsieh-Wilson, L.C. Probing the dynamics of O-GlcNAc glycosylation in the brain using quantitative proteomics. Nature Chem. Biol. 2007, 3, 339–348. 105. Catalina, M.I.; Koeleman, C.A.M.; Deelder, A.M.; Wuhrer, M. Electron transfer dissociation of N-glycopeptides: Loss of the entire N-glycosylated asparagine side chain. Rapid Commun. Mass Spectrom. 2007, 21, 1053–1061.
Ion/Ion Reactions in Electrodynamic Ion Traps
31
106. Molina, H.; Horn, D.M.; Tang, N.; Mathivanan, S.; Pandey, A. Global proteomic profiling of phosphopeptides using electron transfer dissociation tandem mass spectrometry. Proc. Natl. Acad. Sci. USA 2007, 104, 2199–2204. 107. Wuhrer, M.; Stam, J.C.; van de Geijn, F.E.; Koeleman, C.A.M.; Verrips, C.T.; Dolhain, R.J.E.M.; Hokke, C.H.; Deelder, A.M. Glycosylation profiling of immunoglobulin G (IgG) subclasses from human serum. Proteomics 2007, 7, 4070–4081. 108. Srikanth, R.; Wilson, J.; Bridgewater, J.D.; Numbers, J.R.; Lim, J.; Olbris, M.R.; Kettani, A.; Vachet, R.W. Improved sequencing of oxidized cysteine and methionine containing peptides using electron transfer dissociation. J. Am. Soc. Mass Spectrom. 2007, 18, 1499–1506. 109. Zhang, Q.B.; Tang, N.; Brock, J.W.C.; Mottaz, H.M.; Ames, J.M.; Baynes, J.W.; Smith, R.D.; Metz, T.O. Enrichment and analysis of nonenzymatically glycated peptides: Boronate affinity chromatography coupled with electron-transfer dissociation mass spectrometry. J. Proteome Res. 2007, 6, 2323–2330. 110. Zhang, Q.B.; Tang, N.; Schepmoes, A.A.; Phillips, L.S.; Smith, R.D.; Metz, T.O. Proteomic profiling of nonenzymatically glycated proteins in human plasma and erythrocyte membranes. J. Proteome Res. 2008, 7, 2025–2032. 111. Bunger, M.K.; Cargile, B.J.; Ngunjiri, A.; Bundy, J.L.; Stephenson, J.L. Automated proteomics of E. coli via top-down electron-transfer dissociation mass spectrometry. Anal. Chem. 2008, 80, 1459–1467. 112. Liu, J.; Chrisman, P.A.; Erickson, D.E.; McLuckey, S.A. Relative information content and top-down proteomics by mass spectrometry: Utility of ion/ion proton-transfer reactions in electrospray-based approaches. Anal. Chem. 2007, 79, 1073–1081. 113. Watson, D.J.; McLuckey, S.A. Charge state dependent ion trap collision-induced dissociation of reduced bovine and porcine trypsin cations. Int. J. Mass Spectrom. 2006, 255, 53–64. 114. Muddiman, D.C.; Cheng, X.H.; Udseth, H.R.; Smith, R.D. Charge-state reduction with improved signal intensity of oligonucleotides in electrospray ionization mass spectrometry. J. Am. Soc. Mass Spectrom. 1996, 7, 697–706. 115. McLuckey, S.A.; Van Berkel, G.J.; Glish, G.L. Reactions of dimethylamine with multiply charged ions of cytochrome c. J. Am. Chem. Soc. 1990, 112, 5668–5670. 116. McLuckey, S.A.; Glish, G.L.; Van Berkel, G.J. Charge determination of product ions formed from collision-induced dissociation of multiply protonated molecules via ion molecule reactions. Anal. Chem. 1991, 63, 1971–1978. 117. Williams, E.R. Proton transfer reactivity of large multiply charged ions. J. Mass Spectrom. 1996, 31, 831–842. 118. Carr, S.R.; Cassady, C.J. Reactivity and gas-phase acidity determinations of small peptide ions consisting of 11 to 14 amino acid residues. J. Mass Spectrom. 1997, 32, 959–967. 119. Chrisman, P.A.; Pitteri, S.J.; McLuckey, S.A. Parallel ion parking: Improving conversion of parents to first-generation products in electron transfer dissociation. Anal. Chem. 2005, 77, 3411–3414. 120. Chrisman, P.A.; Pitteri, S.J.; McLuckey, S.A. Parallel ion parking of protein mixtures. Anal. Chem. 2006, 78, 310–316. 121. Grosshans, P.B.; Ostrander, C.M.; Walla, C.A. Methods and apparatus to control charge neutralization reactions in ion traps, U.S. Patent 2003, 6,570,151, B1. 122. Grosshans, P.B.; Ostrander, C.M.; Walla, C.A. Methods and apparatus to control charge neutralization reactions in ion traps, U.S. Patent 2004, 6,674,067, B2. 123. Chi, A.; Bai, D.L.; Geer, L.Y.; Shabanowitz, J.; Hunt, D.F. Analysis of intact proteins on a chromatographic time scale by electron transfer dissociation tandem mass spectrometry. Int. J. Mass Spectrom. 2007, 259, 197–203. 124. Lin, T.; Payne, A.H.; Glish, G.L. Dissociation pathways of alkali-cationized peptides: Opportunities for C-terminal peptide sequencing. J. Am. Soc. Mass Spectrom. 2001, 12, 497–504.
32
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
125. Herron, W.J.; Goeringer, D.E.; McLuckey, S.A. Gas-phase electron-transfer reactions from multiply-charged anions to rare-gas cations. J. Am. Chem. Soc. 1995, 117, 11555–11562. 126. McLuckey, S.A.; Stephenson, J.L.; O’Hair, R.A.J. Decompositions of odd- and evenelectron anions derived from deoxy-polyadenylates. J. Am. Soc. Mass Spectrom. 1997, 8, 148–154. 127. Coon, J.J.; Shabanowitz, J.; Hunt, D.F.; Syka, J.E.P. Electron transfer dissociation of peptide anions. J. Am. Soc. Mass Spectrom. 2005, 16, 880–882. 128. Zubarev, R.A.; Kelleher, N.L.; McLafferty, F.W. Electron capture dissociation of multiply charged protein cations. A nonergodic process. J. Am. Chem. Soc. 1998, 120, 3265–3266. 129. Zubarev, R.A. Reactions of polypeptide ions with electrons in the gas phase. Mass Spectrom. Rev. 2003, 22, 57–77. 130. Zubarev, R.A.; Zubarev, A.R.; Savitski, M.M. Electron capture/transfer versus collisionally activated/induced dissociations: Solo or duet? J. Am. Soc. Mass Spectrom. 2008, 19, 753–761. 131. Good, D.M.; Wirtala, M.; McAlister, G.C.; Coon, J.J. Performance characteristics of electron transfer dissociation mass spectrometry. Mol. Cell Proteomics 2007, 6, 1942–1951. 132. Wu, S.L.; Huehmer, A.F.R.; Hao, Z.Q.; Karger, B.L. On-line LC-MS approach combining collision-induced dissociation (CID), electron-transfer dissociation (ETD), and CID of an isolated charge-reduced species for the trace-level characterization of proteins with post-translational modifications. J. Proteome Res. 2007, 6, 4230–4244. 133. Chrisman, P.A.; Pitteri, S.J.; Hogan, J.M.; McLuckey, S.A. SO2– electron transfer ion/ion reactions with disulfide linked polypeptide ions. J. Am. Soc. Mass Spectrom. 2005, 16, 1020–1030. 134. Gunawardena, H.P.; Gorenstein, L.; Erickson, D.E.; Xia, Y.; McLuckey, S.A. Electron transfer dissociation of multiply protonated and fixed charge disulfide linked polypeptides. Int. J. Mass Spectrom. 2007, 265, 130–138. 135. Stephenson, J.L.; Cargile, B.J.; McLuckey, S.A. Ion trap collisional activation of disulfide linkage intact and reduced multiply protonated polypeptides. Rapid Commun. Mass Spectrom. 1999, 13, 2040–2048. 136. O’Connor, P.B.; Cournoyer, J.J.; Pitteri, S.J.; Chrisman, P.A.; McLuckey, S.A. Differentiation of aspartic and isoaspartic acids using electron transfer dissociation. J. Am. Soc. Mass Spectrom. 2006, 17, 15–19. 137. Liang, X.L.R.; Liu, J.; Leblanc, Y.; Covey, T.; Ptak, A.C.; Brenna, J.T.; McLuckey, S.A. Electron transfer dissociation of doubly sodiated glycerophosphocholine lipids. J. Am. Soc. Mass Spectrom. 2007, 18, 1783–1788. 138. Udeshi, N.D.; Shabanowitz, J.; Hunt, D.F.; Rose, K.L. Analysis of proteins and peptides on a chromatographic timescale by electron-transfer dissociation MS. FEBS J. 2007, 274, 6269–6276. 139. Liu, J.; Huang, T.Y.; McLuckey, S.A. Charge state dependence of proton transfer versus electron transfer in a gas-phase ion/ion electron transfer dissociation process on tryptic peptides. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1, 2008. 140. Coon, J.J.; Syka, J.E.P.; Schwartz, J.C.; Shabanowitz, J.; Hunt, D.F. Anion dependence in the partitioning between proton and electron transfer in ion/ion reactions. Int. J. Mass Spectrom. 2004, 236, 33–42. 141. Gunawardena, H.P.; He, M.; Chrisman, P.A.; Pitteri, S.J.; Hogan, J.M.; Hodges, B.D.M.; McLuckey, S.A. Electron transfer versus proton transfer in gas-phase ion/ion reactions of polyprotonated peptides. J. Am. Chem. Soc. 2005, 127, 12627–12639. 142. Xia, Y.; Gunawardena, H.P.; Erickson, D.E.; McLuckey, S.A. Effects of cation chargesite identity and position on electron-transfer dissociation of polypeptide cations. J. Am. Chem. Soc. 2007, 129, 12232–12243.
Ion/Ion Reactions in Electrodynamic Ion Traps
33
143. Kjeldsen, F.; Giessing, A.M.B.; Ingrell, C.R.; Jensen, O.N. Peptide sequencing and characterization of post-translational modifications by enhanced ion-charging and liquid chromatography electron-transfer dissociation tandem mass spectrometry. Anal. Chem. 2007, 79, 9243–9252. 144. Hauser, N.J.; Han, H.L.; McLuckey, S.A.; Basile, F. Electron transfer dissociation of peptides generated by microwave D-cleavage digestion of proteins. J. Proteome Res. 2008, 7, 1867–1872. 145. Turecek, F. N-Cα bond dissociation energies and kinetics in amide and peptide radicals. Is the dissociation a non-ergodic process? J. Am. Chem. Soc. 2003, 125, 5954–5963. 146. Swaney, D.L.; McAlister, G.C.; Wirtala, M.; Schwartz, J.C.; Syka, J.E.P.; Coon, J.J. Supplemental activation method for high-efficiency electron-transfer dissociation of doubly protonated peptide precursors. Anal. Chem. 2007, 79, 477–485. 147. Han, H.L.; Xia, Y.; McLuckey, S.A. Beam-type collisional activation of polypeptide cations that survive ion/ion electron transfer. Rapid Commun. Mass Spectrom. 2007, 21, 1567–1573.
Hydrogen/ 2 Gas-Phase Deuterium Exchange in Quadrupole-Ion Traps Joseph E. Chipuk and Jennifer S. Brodbelt Contents 2.1 Introduction..................................................................................................... 36 2.1.1 Historical Perspective.......................................................................... 36 2.1.2 Theory of Gas-phase Hydrogen/Deuterium (H/D) Exchange Experiments......................................................................................... 37 2.1.2.1 Deuterating Agents............................................................... 38 2.1.2.2 Proposed Mechanisms.......................................................... 38 2.2 Practical Aspects of Gas-phase Hydrogen/Deuterium (H/D) Exchange........40 2.2.1 Motivation for Gas-phase Hydrogen/Deuterium (H/D) exchange Experiments........................................................................ 41 2.2.2 Instrumentation for Gas-Phase Hydrogen/Deuterium (H/D) exchange Experiments........................................................................ 42 2.2.2.1 Ion Trapping for Gas-phase Hydrogen/Deuterium (H/D) Exchange.................................................................... 42 2.2.2.2 Reagent Inlet Systems........................................................... 42 2.2.3 Methods...............................................................................................44 2.2.3.1 Typical Reaction Conditions.................................................44 2.2.3.2 Mass Spectral Interpretation................................................. 45 2.2.3.3 Reaction Kinetics.................................................................. 45 2.3 Current Areas of Research..............................................................................46 2.3.1 Small Molecules.................................................................................. 47 2.3.1.1 Fundamental Studies of Model Compounds......................... 47 2.3.1.2 Isomer Differentiation........................................................... 49 2.3.2 Peptides and Proteins........................................................................... 50 2.3.2.1 Model Peptides...................................................................... 51 2.3.2.2 Proteins................................................................................. 51 2.3.3 Nucleosides, Nucleotides, and Oligonucleotides................................. 54 2.4 Conclusions...................................................................................................... 55 Acknowledgment...................................................................................................... 55 References................................................................................................................. 56
35
36
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
2.1 INTRODUCTION Ever since the first report of using hydrogen/deuterium (H/D) exchange reactions in the gas phase to count ‘active’ hydrogen atoms in organic ions [1], the potential utility of the H/D exchange method for probing structural aspects of ions has been recognized by a succession of various research groups. The H/D exchange method involves allowing a mass-selected set of ions to undergo reactions with a deuterated reagent in which labile, accessible hydrogen atoms of the analyte ions (typically surface-exposed hydrogen atoms attached to nitrogen, oxygen, or sulfur atoms) may exchange for deuterium atoms, thus causing easily monitored step-wise mass shifts in the mass spectrum. Numerous studies have shown that different active sites within ions exhibit different rates of H/D exchange, and the rates and extents of H/D exchange depend on both the intrinsic basicities (or acidities) of the reagent and analyte and the accessibilities of active hydrogen atoms. Investigations of H/D exchange reactions are a natural fit for quadrupole ion trap mass spectrometers because the kinetics of reactions can be monitored accurately by variation of the ion storage time, deuterated reagents can be introduced easily and in controlled fashion to the ion trap via several alternative means, and targeted analyte ions can be mass selected readily and isolated prior to reaction. As illustrated in this chapter, applications of H/D exchange in quadrupole-ion traps have ranged from those involving small organic molecules, especially involving comparisons of isomers, to larger biological molecules for which conformational effects play a significant role.
2.1.1 Historical Perspective The origin of gas-phase H/D exchange reactions can be traced to the pioneering experiments by Hunt and co-workers who investigated gas-phase chemical ionization (CI) using D2O as a reagent gas [1]. These studies differed from modern gasphase H/D-exchange reactions in two important ways: (1) they involved the reaction of ‘ionized D2O’ (meaning the radical cation D2O + • generated by electron ionization (EI) of D2O, the CI product ion D3O +, or the CI product ion clusters D + (D2O)n +) with neutral analytes rather than the reaction of ionized analytes with a neutral deuterating agent; and (2) the CI conditions employed were more energetic than those for reactions between neutral deuterating agents and ions generated by soft ionization methods such as electrospray ionization (ESI). Nevertheless, the groundbreaking revelation came not in the efficacy of ionized D2O as an exceptional CI reagent, but rather the appearance of the isotopically-modified mass spectra. In addition to the ionization of the neutral analyte, the observed mass spectra contained mass-shifted peaks associated with the incorporation of a deuterium atom for each of the ‘active or acidic’ hydrogen atoms, establishing a new landmark gas-phase ion/molecule reaction. Furthermore, while small amounts of deuterium exchange were observed also for ketones, aldehydes, and esters, no appreciable exchange was noted for other unsaturated organics such as benzene and stilbene. Thus, the reaction of ionized D2O was determined to be selective for particular hydrogen atoms and was deemed, therefore, to be useful in determining the types of functional groups present in a molecule known to contain specific heteroatoms.
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
37
The initial CI studies performed by Hunt and co-workers inspired Beauchamp and co-workers to investigate the reactions of protonated benzene and its substituted derivatives with neutral D2O in an ion cyclotron mass spectrometer [2]. Interestingly, the efficacy of the H/D exchange reaction was found to be dependent on both the protonation of the aromatic ring and the identity of any aromatic substituents, because none of the compounds containing electron-withdrawing groups were observed to undergo exchange. Furthermore, the kinetics and extents of the reaction were shown to differ for structural isomers such as o- and m-difluorobenzenes. At first glance, the results appeared to be contradictory to those reported previously by Hunt [1] in that benzene and other aromatic molecules were observed to form isotope exchange products. However, the key difference between the two experiments was determined to be, not surprisingly, the identities of the pair of reactants in each case; this critical distinction explained the variation in results observed. Ultimately, Hunt and co-workers continued their research into isotope exchange under CI conditions by exploring the differences in exchange between various deuterated reagent gases (D2O, C2H5OD, and ND3) [3]. Dramatic differences in the CI mass spectra were observed as the deuterating agent/CI reagent gas was varied. In some cases, analytes that were observed previously to exchange very little with a particular deuterating agent were observed to exchange all of the available hydrogen atoms with alternate agents. In contrast, other ion/CI-reagent pairings produced no additional exchange regardless of the deuterating agent/CI reagent gas used. Furthermore, in addition to positive-mode CI using protonated species, CI in the negative ion mode was investigated also and similar isotope exchange reactions were observed. The importance of this investigation became clear during the interpretation of the experimental results. Foremost, the efficiency of the exchange reaction could be ascribed to the relative difference in basicities or acidities of the reagent and analyte, since the extent of deuterium incorporation was shown to decrease as the proton affinity difference between the reactant and analyte increased. Furthermore, Hunt and co-workers were the first to propose formally the formation of an ion–molecule complex as a necessary condition for H/D exchange; they argued that exchange must occur during the lifetime of this complex. Years of subsequent research have continued to affirm this hypothesis. Finally, the authors speculated that the observed isotopeexchange reaction was dependent on temperature, especially as it relates to internal energy of the ions investigated. Collectively, these concepts became the foundation for continuing research into the mechanism of the gas-phase H/D exchange reaction and spurred many other groups to investigate the utility of the gas-phase H/D exchange reaction as an analytical tool.
2.1.2 Theory of Gas-phase Hydrogen/Deuterium (H/D) Exchange Experiments Gas-phase H/D exchange reactions constitute a specific type of ion/molecule reaction. The vast majority of these reactions involve gas-phase ions containing acidic hydrogen atoms bound to oxygen, nitrogen, or sulfur atoms (that is, acidic hydrogen atoms in alcohols, phenols, carboxylic acids, amines, amides, or thiols). However, specific cases of H/D exchange of non-labile hydrogen atoms (that is, those bound to atoms such as
38
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
carbon) have been reported also [4–6]. In these instances, subtle nuances in the local electronic environment of the carbon atom produced unusually acidic hydrogen atoms that were amenable to exchange. Again, unlike CI, the ion participant in the ion/molecule reaction in gas-phase H/D exchange is the analyte under study while the neutral molecule is one of many deuterating agents. 2.1.2.1 Deuterating Agents Popular choices for deuterating agents for gas-phase H/D exchange include D2O, CH3OD, CD3OD, ND3, DI, and D2S, although more exotic reagents such as CF3CH2OD, C6H5CH2OD, and C6D5OD have been used also. Historically, the choice of deuterating agent has been based primarily on its gas-phase acidity or basicity relative to that of the analyte of interest, although practical factors such as the vapor pressure of the deuterating agent and corrosivity of the agent toward instrumentation components may influence the decision also. In the gas phase, the acidity of a molecule is defined as the molar Gibbs energy, ΔG, required to dissociate it heterolytically into a proton and an anion [7]. H/D exchange reactions involve both analyte ions and deuterating agents that contain relatively acidic hydrogen or deuterium atoms. Likewise, the gas-phase basicity is the Gibbs energy of the associated protonation reaction, and the H/D exchange reaction involves relatively basic oxygen, nitrogen, or sulfur atoms that accept the exchanged deuteron. Table 2.1 lists the gas-phase acidities and basicities of several common deuterating agents. 2.1.2.2 Proposed Mechanisms Gas-phase H/D exchange is a complicated process that may depend on a number of factors including the relative acidity and basicity of the analyte and deuterating agent, the proximity of the charge site to the potential exchange site, the intramolecular interaction of the labile hydrogen atoms with other basic sites, the internal energy of the analyte ion, and the frequency of collisions and energetics of ion–molecule complexation. Therefore, different reaction conditions (for example, different deuterating agents, reagent pressures, ion temperatures) can promote different levels of H/D
TABLE 2.1 Gas-Phase Acidity and Basicity of Common Deuterating Agents Deuterating Agent Hydroiodic acid (DI) Deuterium sulfide (D2S) Deuterated ethanol (C2D5OD) Deuterated methanol (CD3OD) Deuterium oxide (D2O) Deuterated ammonia (ND3)
Gas-Phase Acidity (kJ mol −1)
Gas-Phase Basicity (kJ mol −1)
1315 1469 1561 1598 1636 1690
601 674 746 725 660 819
Source: All data are taken from Hunter, E.P.; Lias, S.G. NIST Chemistry WebBook. National Institute of Standards and Technology: Gaithersburg, MD, 1998.
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
39
exchange. Furthermore, different mechanisms are favored depending on the reactivity and gas-phase conformations of the reactants. The most fundamental principle of gas-phase H/D exchange involves the difference in basicity (or acidity) of the analyte and deuterating agent. In general, H/D exchange is favored when either the gas-phase acidity of the analyte and deuterating agent or gas-phase basicity of the analyte and deuterating agent are similar (that is, differing by less than 84 kJ mol−1). In these cases, the proton and deuteron transfer can occur via a relatively thermoneutral reaction. Indeed, this general statement has held true for many small molecule H/D exchange studies [7,9], but not for others involving amino acids and peptides, where exchange has been shown to occur when the difference in gas-phase basicities is greater than 84 kJ mol−1 [10]. Ultimately, the difference in gas-phase acidity (or basicity) of the reaction participants is only one key aspect of the H/D exchange process. Another factor, which has a major impact on gas-phase H/D exchange, is the proximity of the charge site to the labile hydrogen atoms. In some cases, the mechanism involved in the H/D exchange requires the two reactive sites to be relatively close in order to facilitate ion–molecule complexation and subsequent exchange. The relay mechanism shown in Figure 2.1 is an important example. This mechanism requires the formation of a stable ion–molecule complex in which a deuterating reagent bridges the gap between a charge site and a labile hydrogen atom, thereby allowing indirect interaction of the two remote sites. In this case, the charge site serves as the deuterium acceptor in the exchange process, while the labile hydrogen is transferred to the deuterating agent.
O
O
P
HO
O
Nucleobase
O
O H
D
O
H
D
O
O
D
O
P
HO
Nucleobase
O
O
P O
O O
D
O
Nucleobase
O
H
O
O
D
O
O
P
HO
O
HO
Nucleobase
O
D
D
O O
H
D
FIGURE 2.1 H/D exchange via the relay mechanism. The reaction is shown for a generic mononucleotide anion reacting with D2O. (Reproduced from Chipuk, J.E.; Brodbelt, J.S. J. Am. Soc. Mass Spectrom. 2007, 18, 724–736. With permission from Elsevier.)
40
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V M+H or M–H O
R
O
D
C
O O
H
D
M+H or M–H
M+H or M–H
R
-
D
C
O O
H
D
R
O
D
O
H
C
O
D
FIGURE 2.2 H/D exchange via a Flip-Flop mechanism. The reaction is shown for a generic carboxylic acid reacting with deuterium oxide (D2O).
In contrast, the flip-flop mechanism shown in Figure 2.2 is independent of the charge site and involves the interaction of the deuterating reagent directly with the labile hydrogen atom. For this route, the ion–molecule complex forms a pseudo-ring structure by the attraction of partially-charged participants, and exchange occurs as one of the potential outcomes of its disassembly. In many cases, the inherent gas-phase acidity or basicity of a deuterating agent results in the reagent favoring particular exchange mechanisms regardless of the ion involved. D2O is one example, as the relatively low acidity and basicity result in D2O reacting typically via the relay mechanism shown in Figure 2.1. Another example is deuterated ammonia, ND3. In this case, the basicity of the deuterating agent favors the abstraction of a proton from positively-charged ions, such as the N-terminus of peptides, in what would be a nominally-endothermic process. This transition results in the formation of an ammonium ion [ND3H] + and is typically accompanied by simultaneous solvation of the resultant ion. Subsequent shuttling of protons via tautomerism or other resonance may then result in conditions favorable for the transfer of one of the deuterium atoms to the analyte, resulting effectively in H/D exchange. The gas-phase chemistry of ND3 is well suited to this onium mechanism, making reactions of ND3 with positive ions among the most efficient. In contrast, the low gas-phase basicity of D2O alone would be likely to preclude this type of exchange. While the aforementioned exchange mechanisms have received the most attention, the debate over the mechanism of gas-phase H/D exchange reactions continues. While many groups have concluded that specific results seem to be correlated to a particular mechanism, it is likely that in many other cases multiple mechanisms are contributing simultaneously.
2.2 PRACTICAL ASPECTS OF GAS-PHASE Hydrogen/ Deuterium (H/D) EXCHANGE Gas-phase H/D experiments are carried out in a number of ways and utilize a variety of instrumentation configurations. In many instances, the results of the experiment are determined by the conditions utilized, and particular attention must be paid to the methodology employed. Furthermore, interpretation of the exchange data often requires additional insight when compared to typical mass spectral analysis, and rigorous molecular modeling is often a companion effort. Collectively, these complexities necessitate extra discussion of the process of conducting gas-phase H/D exchange experiments in ion-trapping instruments.
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
41
2.2.1 Motivation for Gas-phase Hydrogen/Deuterium (H/D) Exchange Experiments Gas-phase H/D exchange experiments are carried out typically with one of three major goals in mind: to elucidate the gas-phase structure of a particular ion; to differentiate between structurally-isomeric ions; or to assess the reactivity of ions in the absence of solvent. Many of these experiments are complemented by extensive computational modeling of the ion and/or ion–molecule complex as a means of validating the experimental results and, thereby, furthering the understanding of the mechanism of the experiment. The search for structural information of ions in the gas phase is relevant for all ions but, in recent years, the search has been targeted especially at large biomolecules where secondary structures are complex and critical for function in solution. For example, these experiments may be aimed at elucidating the folding and unfolding of proteins or probing the tertiary structure of oligonucleotides such as DNA quadruplexes. In doing so, these results are often complementary to those obtained by other experimental techniques such as ion-mobility measurements. Ultimately, the motivation for all of these experiments is to understand the readily accessible gas-phase conformations, how they interconvert, and how they may be correlated with structures and behaviors of molecules in the native solution state. In the case of isomeric differentiation, gas-phase H/D exchange serves to complement other techniques such collisionally-induced dissociation (CID), infrared multi-photon dissociation (IRMPD), or ultra-violet photon dissociation (UVPD). For small molecules, these dissociation techniques may at times prove to be inadequate to differentiate isomeric or enantiomeric species, and thus gas-phase H/D exchange may be an alternate way to determine structural differences, provided that the isomers contain exchangeable acidic hydrogen atoms. For oligomeric species such as peptides and oligonucleotides, differences in H/D exchange patterns may assist in determining the sequence of amino acids or nucleotides. While gasphase H/D exchange is not likely to replace more powerful bottom-up or top-down sequencing techniques, it can be used in specific cases as a secondary means of qualitative identification. Gas-phase H/D exchange is also useful for chemists interested in probing fundamental aspects of ion reactivity and thermochemistry. Reactions involving neutral deuterated molecules of varying acidity allow an intricate method for assessing the tendency of ions to react via gas-phase acid–base reactions. In some cases, these fundamental experiments produce results that are contrary to what is believed to take place in solution, again highlighting the unique chemistry that can take place in the absence of a ubiquitous solvent. In addition, reactions can be performed, on occasion, in relatively-controlled conditions where both the kinetic and potential energies of the ions can be calculated and, in some cases, manipulated via additional blackbody heating, off-resonance excitation, or direct activation via interaction with electromagnetic radiation, such as with an IR laser. Ultimately, these types of experiments are performed to increase the fundamental knowledge base of gas-phase ion chemistry.
42
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
2.2.2 Instrumentation for Gas-phase Hydrogen/Deuterium (H/D) Exchange Experiments In the gas-phase H/D exchange process, the lifetime of the ion–molecule complex is critical, as results from numerous experiments have demonstrated that reactions proceed at varying rates. It is, therefore, not surprising that the bulk of the gas-phase H/D exchange studies have been performed using either a quadrupole ion trap or a Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer, because these two techniques are so well suited to trapping ions for variable periods to allow kinetic analysis. While the two methodologies have many fundamental similarities, they also have several inherent differences, including the temperatures of ions, the number of collisions that ions undergo with the reagent or buffer gas (due to the dramatic difference in operating pressures), and the total reaction time. These factors may influence the results of the exchange reactions. The following sections will focus primarily on conducting H/D exchange reactions in each of two-dimensional and three-dimensional quadrupole-ion traps. 2.2.2.1 Ion Trapping for Gas-phase Hydrogen/Deuterium (H/D) Exchange While the majority of H/D exchange reactions reported in the literature have been conducted in FT-ICR mass spectrometers, numerous studies have been reported that utilize quadrupole-ion traps. Amongst these, the vast majority have utilized threedimensional hyperbolic quadrupoles, such as Thermo Finnigan LCQ [11–14] or Hitachi 3DQ mass spectrometer [5,6,15–17]. The less frequent use of two-dimensional ion traps, such as a linear ion trap (LIT) [18–21], is likely due to the simple fact that there are fewer research groups that possess the instrumentation and have adapted it for H/D exchange studies. Other ion-trap configurations, such as a cylindrical ion trap (CIT) or rectilinear ion trap (RIT), should also be amenable to conducting H/D-exchange reactions, since these instruments have been reported to be useful for conducting other ion/molecule reactions [22]. In theory, all ion traps with sufficient resolution should be amenable to conducting gas-phase H/D exchange reactions, provided that they can be integrated with an appropriate ion source and maintain proper operating pressure when integrated with a deuterating reagent inlet system. Ion introduction for gas-phase H/D exchange is performed typically at atmospheric pressure, and most often through the use of ESI. However, some examples that utilize Matrix-assisted laser desorption/ionization (MALDI) [23] can be found in the literature and gas-phase isotope exchange studies also utilize CI [1,24]. ESI lends itself particularly well to gas-phase H/D exchange experiments because protonation and deprotonation often involve relatively basic groups, such as amines, or relatively acidic hydrogen atoms such as those found in phosphates. In many cases, these sites either contain or are in close proximity to more than one labile hydrogen atom (for example, terminal phosphates in oligonucleotides). 2.2.2.2 Reagent Inlet Systems Undertaking H/D exchange reactions in an ion trap mass spectrometer requires a means of introducing a deuterating agent. Several popular methods include introduction of deuterating agent through the helium bath gas line [11–14], through a leak
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
43
valve directly to the ion trap or to the ion trap vacuum manifold [5,6,15–17], via direct and continuous infusion to the ion trap chamber [18–21], or in some cases through a pulsed-valve system [23]. Introduction of neutral reagent through the helium bath gas line of a Thermo Finnigan LCQ was described first by Gronert [25]. In this system, a measured flow of liquid deuterating agent is added via a syringe pump to a measured flow of helium. The rapid vaporization that takes place at the syringe tip allows for a range of mixing ratios of the reagent and the helium gas. While most of the total flow, including the deuterating agent, is directed to a flow meter, a small portion is drawn into the ion trap. In many cases, the internal down-stepping regulation of helium is bypassed and helium is, instead, delivered directly to the trap at reduced pressures. This adjustment reduces the dead volume of the introduction pathway and speeds up the response of the system to changes in the flow rate, and hence concentration, of the deuterating agent. This type of reagent inlet system allows for measurable and finely tunable deuterating reagent pressures into the ion trap with an estimated uncertainty of ca 20%, most of which is caused by the uncertainty in measuring the helium pressure itself [25]. The primary benefit is the ability to add small amounts of deuterating agent in a reproducible manner, which leads ultimately to the ability to conduct kinetic reaction measurements at a series of known reagent concentrations. The primary downside of this approach is contamination of the helium bath gas line with the deuterating reagent. While this contamination may be of lesser concern when the instrumentation is utilized exclusively for ion/molecule reactions such as H/D exchange, normal operation of the instrument under standard analytical conditions may be prone to interferences caused by the presence of reactive contaminants. Introduction of the neutral deuterating reagent through a leak valve is also an effective way to implement gas-phase H/D exchange [5,6,14–17]. In this approach, a liquid or gaseous reagent stored in a secondary vessel (for example, a glass finger tube for liquids or a lecture bottle for gases) is subjected to the vacuum of the mass spectrometer through an adjustable needle valve. When the reagent is a volatile or moderatelyvolatile liquid at room temperature, such as D2O or CD3OD, the reduced pressure will increase the vapor pressure of the liquid, and gas-phase neutral deuterating agent will be transferred to the ion trap via the pressure differential. Naturally, gaseous reagents such as D2S and ND3 are transferred easily in a similar manner. In some cases, the deuterating agent is leaked directly into the interior volume of the ion trap via a gas line plumbed through one of the hyperbolic end-cap electrodes of the trap. In other cases, the reagent is leaked into the vacuum manifold where it fills the evacuated space and, eventually, equilibrates with the helium bath gas. Introduction of deuterating agent via a leak valve has the advantages of maintaining the integrity of the helium transfer line and providing easily sufficient reactant to ensure that the concentration of the deuterating reagent is not the limiting factor in the exchange reaction. The primary disadvantage of this introduction method, relative to the helium gas line method, is the inability to control precisely the addition of small amounts of deuterating agent. In these instances, there is nearly always a stoichiometric excess of the deuterated reactant and exchange reactions that are concentration dependent and have very fast exchange rates become difficult to model kinetically. Therefore, the leak valve method may favor comparison of the extent of H/D exchange among different species over rigorous modeling of the reaction kinetics for rapidly exchanging systems.
44
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
A method for direct infusion of a deuterating agent to a RIT has been described also [18–21]. This procedure is similar to the leak valve approach except that a 0.9 mm nozzle is placed 4 mm radially from the center of the ion trap and is used to deliver the deuterating agent directly to the interior of the ion trap via the formation of a free jet. While H/D exchange is believed to occur in the free-jet expansion, it is assumed to be more extensive in the surrounding trapping environment as continuous infusion builds up a background pressure of deuterating agent. Calculations supporting this assumption suggest that there is a greater number density of deuterating agent molecules in the background compared to the free-jet area. In addition to these continuous introduction schemes, instruments have been constructed that utilize a pulsed-valve method for introducing the deuterating agent [23]. While this method is similar to the leak valve method, it offers the advantage of allowing the ion trap to ‘reset’ after each mass analysis scan and, therefore, to allow accumulation of ions within the ion trap before they are subjected to the neutral deuterating agent. In this way, all ions have a similar opportunity for exchange regardless of whether they were ionized and accumulated at the beginning of the ionization period or toward the end. The primary disadvantage of this method is that high-pressure pulses reduce the reaction stability produced by equilibrating the ion trap to the deuterating agent. The inherent variability in the pulsing system may change significantly the concentration of the deuterating agent on a scan-to-scan basis. In so doing, the variability of the exchange results may increase, especially as concentrations are reduced and when the kinetics of rapidly-exchanging systems are being investigated.
2.2.3 Methods As mentioned previously, gas-phase H/D exchange reactions are a form of ion/molecule reaction in which an analyte ion reacts with a neutral deuterated molecule within the confines of an ion-trap or ion-drift tube. The product of the reaction is a covalent exchange of hydrogen for deuterium in what amounts to a double replacement reaction that is expected to proceed through a stable ion–molecule complex. Like other organic reactions, the extent of gas-phase H/D exchange reactions depends on the number density of molecules of each reactant species present and the time during which they are allowed to react. 2.2.3.1 Typical Reaction Conditions The pressure of the ion-trapping system is critical to the rate, and, ultimately, to the extent, of any gas-phase H/D exchange reaction. Deuterating agents are introduced typically to the trap at pressures between 0.1 and 3.0 mTorr with higher pressures favoring more rapid exchange but, potentially, compromising mass spectral resolution. These pressures are approximately four orders of magnitude larger than those used in an FT-ICR cell and thus, have led several researchers to conduct H/D exchange experiments in parallel on both types of instruments [13,14,26]. Reaction times typically range from milliseconds to 10 s, although exchange times up to 40 s have been accomplished [13]. Furthermore, when a pulsed deuterating agent introduction system is not used, the actual reaction time is approximately the sum of the ion accumulation time, any isolation time, and any exchange time, since the deuterating
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
45
agent is present within the ion trap throughout the entire analysis. Exchange times of less than 10 s are far shorter than those used typically in FT-ICR studies. Therefore, the two mass analyzers provide somewhat complementary information in that a quadrupole ion trap can monitor the reaction of high concentrations of both ions and deuterating agents for a relatively short period of time, while FT-ICR mass analyzers monitor typically much smaller reactant concentrations but for much longer time periods. Aside from the reagent pressure and reaction times, H/D exchange experiments are sensitive to contamination from hydrogen sources such as water or methanol. Because these ions provide a means of exchange of deuterated species back to unexchanged ones, a conditioning procedure is carried out normally, whereby the ion trap is exposed to the gas-phase deuterating agent for up to one hour prior to analysis, so as to remove any unwanted contamination in the ion trap. As with other analyses, it is often the case that samples are analyzed multiple times across different days to assess reproducibility. This practice is especially important when making comparisons between reaction rates of isomeric species, because many reactions are very sensitive to the deuterating reagent pressure and this pressure may be difficult to control on a day-to-day basis. Therefore, whenever possible, series of analyses are run consecutively on a given day to control against pressure variation. 2.2.3.2 Mass Spectral Interpretation The evolution of a typical gas-phase H/D exchange reaction can be monitored by varying the reaction time while keeping constant the accumulation time, isolation time, and pressure of the deuterating agent. An example of such an evolution is shown in Figure 2.3. As the reaction time increases, the relative intensity of the isolated analyte ion, m/z 306, decreases as labile hydrogen atoms are exchanged for deuterium atoms. The relative intensities of m/z 307, 308, and 309 increase initially and then decrease, while that of m/z 310 increases until, after 10 s, the exchange of all four labile hydrogen atoms by deuterium atoms is complete. The ion of m/z 310 in Figure 2.3(d) has four deuterium atoms and represents the complete H/D exchange. The ion of m/z 311 does not represent additional exchange because it is the 13C isotopic analog of the peak of m/z 310. Indeed, contributions to a particular ion population m/z from non-exchanged isotopes, such as 13C, need to be corrected prior to performing any detailed analysis of the relative peak intensities. 2.2.3.3 Reaction Kinetics Many gas-phase H/D exchange experiments are performed to assess not only the extent to which the reaction takes place, but also the rate at which the reaction proceeds. Besides providing valuable insight into gas-phase ion reactivity, reaction rates are investigated because they often provide additional information that can be used to differentiate isomeric species or to give indications about the secondary structure of larger biomolecules in the gas phase. To perform a kinetic analysis, a series of experiments is performed at various exchange times ranging over several orders of magnitude (for example, 0 ms to 10 s). After correcting the peak intensities for non-exchanged isotopic contributions such as 13C, the series of data is fit typically to a system of coupled ordinary differential equations using programs such as KinFit [27] to determine the rate constants.
46
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a)
(b) 25,000
15,000
20,000 Intensity
15,000
*306
10,000
Intensity
D0
D2
10,000 *306
5000
5000 0 304 305 306 307 308 309 310 311 312 313 314 m/z
(c)
(d)
10,000
5000
30,000
D3
*306
D4
25,000 Intensity
Intensity
0 304 305 306 307 308 309 310 311 312 313 314 m/z
20,000 15,000 10,000
*306
5000 0 304 305 306 307 308 309 310 311 312 313 314 m/z
0 304 305 306 307 308 309 310 311 312 313 314 m/z
FIGURE 2.3 Representative mass spectra for the reaction of deprotonated 2-deoxy-5-cytidine monophosphate (5-dCMP) with D2O in a quadrupole-ion trap. Reaction times are (a) 0 ms, (b) 250 ms, (c) 2000 ms, and (d) 10,000 ms. Peaks are labeled as Dn, where n equals the number of incorporated deuterium atoms. (Reproduced from Chipuk, J.E.; Brodbelt, J.S. J. Am. Soc. Mass Spectrom. 2007, 18, 724–736. With permission from Elsevier.)
A plot of the calculated rate constant equation for each exchange along with the exchange data is prepared usually as an illustrative tool. An example of such a plot is shown in Figure 2.4.
2.3 CURRENT AREAS OF RESEARCH Initially, the most common application of H/D exchange was aimed at identification of the presence of particular acidic or basic sites in a molecule. This concept evolved quickly into investigations to distinguish between specific structural and stereoisomers of trapped ions. The invention of soft ionization techniques, such as MALDI and especially ESI, led to a completely different set of applications for gas-phase H/D exchange. The rapidity with which these new ionization techniques were introduced allowed the study of a wide range of biomolecules such as peptides, proteins, and oligonucleotides in the gas phase, and used the internal volume of the ion trap as a configurable chemical reactor. While some of the research in this area has utilized H/D exchange to differentiate between structural variations (that is, sequence of subunits), the majority of
47
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps 1 OH
0.8
OH
Relative abundance
OCH3
CH2OH
CH3OH O OH OH
D(0)
0.6
O
O
O
O
OH
O
OH
0.4
D(1)
0.2
0
D(2)
0
2
D(3) 4
D(4) 6
8
10
H/D exchange time (s)
FIGURE 2.4 Kinetic plot of the H/D exchange reactions of the deprotonated flavonoid, neohesperidin, fitted with KinFit. D(0) represents the initial precursor ion, and D(1), D(2), D(3), and D(4) represent ions incorporating from one to four deuterium atoms, respectively. (Reproduced from Zhang, J.; Brodbelt, J.S. J. Am. Chem. Soc. 2004, 126, 5906–5919. With permission from American Chemical Society.)
the work aims to investigate a much broader question, that of the conformation and secondary structures of these macromolecules in the gas phase.
2.3.1 Small Molecules Research in the area of small molecule gas-phase H/D exchange in ion traps has focused primarily on two fronts: (1) model systems to investigate the mechanisms of H/D exchange; and (2) isomer differentiation. 2.3.1.1 Fundamental Studies of Model Compounds A classic example of a mechanism-motivated H/D study involved the exchange behavior of various polyamines that formed singly-protonated, doubly-protonated, and sodium- or potassium-cationized species upon ESI [28]. The goal was to determine the impact of various factors such as ligand flexibility, size, basicity, and location of basic groups on the H/D-exchange reaction. It was shown that, under identical conditions, exchange with singly-protonated analytes was faster and more efficient than exchange with either sodium- or potassium-cationized complexes. These results were explained by either the loss of a low energy exchange pathway between the alkali metal-cationized species and the reagent gas, or by a structural change that was induced by the presence of a bulkier cation as compared to a proton. Furthermore, the H/D exchange of the doubly-protonated species was found to be much more efficient than for the singly-protonated species. Ultimately, it was concluded that gas-phase
Intensity
(c)
50,000 OH D(0) O 45,000 O 40,000 H 35,000 OH 30,000 25,000 H H 20,000 H 15,000 D(1) + 13C D(0) PA 10,000 5000 0 160 161 162 163 164 165 166 167 168 169 170 Mass (m/z) O D(1) + 13C D(0) OH 16,000 14,000 H D(0) 12,000 H 10,000 8000 H H 6000 OH O 13 4000 C D(1) TPA 2000 0 160 161 162 163 164 165 166 167 168 169 170 Mass (m/z)
(b)
Intensity
Intensity
(a)
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(d)
Intensity
48
D(2) 50,000 O OH 45,000 40,000 H 35,000 H 30,000 25,000 O 20,000 H 15,000 H OH D(0) 10,000 13 CD(2) D(1) IPA 5000 0 160 161 162 163 164 165 166 167 168 169 170 Mass (m/z) O OH 18,000 D(0) 16,000 H H 14,000 H 12,000 H 10,000 8000 D(1) + 13C D(0) H H 6000 4000 OH O 13 C D(1) 2000 NAPA 0 210 211 212 213 214 215 216 217 218 219 220 Mass (m/z)
FIGURE 2.5 H/D exchange mass spectra for deprotonated (a) phthalic acid, (b) isophthalic acid, (c) terephthalic acid, and (d) 2,6-naphthalic acid after 10 s exchange with D2O. (Reproduced from Chipuk, J.E.; Brodbelt, J.S. Int. J. Mass Spectrom. 2007, 267, 98–108. With permission from Elsevier.)
basicity, conformation of the ion, and interaction of the ion with the exchange reagent all play a role in the H/D exchange behavior. Another example of this type of study involved the H/D exchange reactions of deprotonated aromatic dicarboxylic acids [5]. In this case, the intent was to determine the difference in observed H/D exchange behavior as the relative position of the two carboxylic acid groups varied. Figure 2.5 summarizes the H/D exchange behavior and illustrates that the spatial relationship of the deprotonated sites has a tremendous impact on the extent of the H/D exchange. The exchange mass spectrum in Figure 2.5a illustrates the influence of intramolecular hydrogen bonding between the deprotonated site and proximate labile hydrogen atoms and how it may decrease the proclivity of these hydrogen atoms to exchange. The mass spectra in Figure 2.5c and d describe the opposite scenario, and illustrate that H/D exchange may occur via mechanisms that do not require the proximity of the labile hydrogen atoms to the deprotonated site, albeit at a much slower rate. Finally, the mass spectrum in Figure 2.5b shows the H/D exchange of traditionally non-labile hydrogen atoms. In this instance, the aromatic hydrogen located between the two carboxylic acid groups is unusually acidic and may allow the formation of carbanions that react subsequently with excess deuterating agent. While the differentiation between various phthalic acid isomers is not a chemistry landmark analytical application, the impact of the results extends beyond the differentiation of isomers and illustrates the intricacies of the gas-phase H/D exchange process. The conclusions reinforced previous studies that suggested that H/D exchange
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
49
behavior was dependent on gas-phase acidity, intramolecular hydrogen bonding, and intermolecular hydrogen bonding in the ion–molecule complex. In addition, it was suggested that multiple reaction mechanisms could occur. 2.3.1.2 Isomer Differentiation The utility of gas-phase H/D exchange as a method of isomer differentiation is exemplified by an in-depth study of the H/D exchange behavior of five isomeric flavonoid glycosides and their corresponding aglycons [15]. Although most flavonoids have a common phenyl-benzopyrone skeleton, they differ with respect to each other by the positions of hydroxyl and methoxy groups; the location, number, and identities of saccharides involved in glycosylation; and the intersaccharide linkage in the case of polysaccharides [15]. Figure 2.6 shows several H/D-exchange mass spectra obtained for five isomeric flavonoid monosaccharides of m/z 447. In Figure 2.6a and d, H/D exchange was very minimal, while in Figure 2.6c and e, the exchange was quite extensive. Obviously, the relative position and location of the labile hydrogen atoms was important to the exchange behavior. Moreover, unlike other molecules where a particular site is favored for protonation or deprotonation, the flavonoids contain many seemingly equivalent sites for deprotonation during ESI. Through extensive experimentation and high-level computational calculations, it was concluded that the location of the most likely deprotonation site (that is, the most acidic hydrogen), the ability of the charge to migrate via resonance, the relative acidities of the various labile hydrogen and deuterium atoms, and the flexibility of the various moieties to adopt favorable gas-phase conformations all played a role in determining the H/D exchange behavior. It was, therefore, a summation of all of these factors that resulted in the H/D exchange being remarkably sensitive to the subtle structural differences within the flavonoid isomers. Another interesting example of isomer differentiation by gas-phase H/D exchange was reported for catechins [6]. In this case the low-energy nature of the H/D exchange process was utilized to explore the possibility of differentiating various stereoisomers of both galloylated and non-galloylated species. Interestingly, stereoisomerism was found to have little effect on the reaction kinetics of the non-galloylated catechins with deuterium oxide, while the galloylated species, which differed only in the chirality of one the carbon atoms on the flavonoid ring, had distinctive H/D exchange kinetics. Furthermore, all of the non-galloylated catechin isomers were observed to exchange aromatic or allylic non-labile hydrogen atoms. An accompanying computational investigation showed that the addition of the gallate moiety increased significantly the gas-phase acidity of the hydroxyl hydrogen atoms and, thus, disfavored dramatically the scrambling of these hydrogen atoms to non-labile aromatic sites; in contrast, the non-galloylated species were shown to be prone to this type of rearrangement. The differentiation of isomers by gas-phase H/D exchange continues to be explored. The most compelling experiments are aimed at distinguishing quickly and accurately compounds that either cannot be distinguished by higher energy dissociation methods, such as CID, or require lengthy chromatographic runs, as in liquid chromatographymass spectrometry (LC-MS) or gas chromatography-mass spectrometry (GC-MS) studies. These isomer differentiation studies tend to reveal information about the effects of gas-phase conformations and thermochemistry that can be used to understand the reactions of more complex molecules that contain similar functional groups.
50
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a)
Luteolin-4´-O-glucoside
(b)
CH2CH
OH
Intensity
OH
OH
448
OH OH O
O
HO
444
OH
OH OH
*447
452
456
444
460
OH
OH
HO
452
(d)
O
456
460
Astragalin
O
*447
OH
OH O
HO
OH OH
O
OH
452 456 Mass (m/z)
O
Intensity
Intensity
OH
448
O
OH
448
CH2OH
444
O
Mass (m/z)
Orientin
*447
O
O
Mass (m/z) (c)
OH
CH2OH
Intensity
*447
Luteolin-7-O-glucoside
O
OH
O CH2OH OH
O
OH
444
460
(e)
448
Quercitrin
452 456 Mass (m/z)
O
OH
460
OH OH O
Intensity
HO
OH
*447
O CH3
O
O OH OH OH
444
448
452 456 Mass (m/z)
460
FIGURE 2.6 H/D exchange of five deprotonated flavonoid monosaccharides (a) Luteolin4′-O-glucoside, (b) Luteolin-7-O-glucoside, (c) Orientin, (d) Astragalin, and (e) Quercitrin after 10 s reaction with D2O. (Reproduced from Zhang, J.; Brodbelt, J.S. J. Am. Chem. Soc. 2004, 126, 5906–5919. With permission from American Chemical Society.)
2.3.2 Peptides and Proteins Investigating the behavior of biomolecules such as peptides and proteins is one of the most important and challenging aspects of modern analytical chemistry. While solution-phase studies offer the advantage of being similar to the native biological environment, these experiments are often time consuming and sensitive to variations in the solution conditions. In contrast, gas-phase studies are more efficient and can
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
51
be performed in the absence of solvent interferences. Indeed, proteomics has become a commercialized endeavor and the sequencing of peptides and proteins via mass spectrometric techniques has become standard practice in many laboratories. While the sequence of a peptide or protein is valuable information, the secondary and tertiary structures of the biomolecules are known to be critical to their biological activity. Therefore, additional experiments such as gas-phase H/D exchange have been designed to investigate the higher-order structures of these biomolecular ions. Ultimately, these gas-phase studies are aimed at providing insight into the behavior of the biomolecules in their native, solution-phase environment. 2.3.2.1 Model Peptides Numerous H/D exchange studies of peptides have been reported. Initial work on glycine oligomers using a 3D quadrupole ion trap confirmed not only the previous findings obtained by FT-ICR mass spectrometry but included studies of the collision-induced product ions as well as methylated oligomers [11]. Studies of H/D exchange reactions have been conducted also in concert with ion mobility studies and high-level computations to probe the gas-phase H/D exchange mechanisms active in a family of singly-protonated pentapeptides [26]. While a salt bridge mechanism appeared initially to be responsible for the H/D exchange, the authors concluded that, in fact, this structure was only a kinetic intermediate and that exchange followed from a relay mechanism. Peptides often contain highly basic residues and these oligomers provide an exceptional opportunity for the study of the impact of multiple charge sites on the gas-phase structure. Bradykinin is one of the polypeptides studied most often in mass spectrometry, due primarily to its manageable mass (1060.21 g mol−1) and two highly basic arginine residues. Gas-phase H/D exchange reactions of bradykinin have been studied using a MALDI-QIT instrument [19], an ESI-QIT instrument [14,29–30], and an ESI-LIT-TOF instrument [19]. Figure 2.7 shows a series of H/D exchange mass spectra for bradykinin ions collected using CD3OD as a reagent gas in an LIT-TOF mass spectrometer. Among the most significant findings is the evidence that the triply-protonated species (that is, [M + 3H]3 + ) and the doubly-protonated species (that is, [M + 2H]2 + ) of bradykinin produced multiple and distinct gas-phase conformations, as indicated by the corresponding bi-modal H/D exchange mass spectra [19,29]. The dissociation of these ‘fast’ and ‘slow’ exchanging species was studied subsequently in both FT-ICR and quadrupole ion trap instruments [14]. Another important finding was that the H/D exchange proceeded through a relay mechanism when D2O or DI was used as the deuterating agent, but probably through an alternate mechanism when ND3 was used as the deuterating agent [19,29–30]. Furthermore, results for bradykinin revealed that multiply-charged species exchanged faster than did singly-charged species and exchange of sodium-cationized species was very limited. 2.3.2.2 Proteins Gas-phase H/D exchange of proteins in ion-trapping instruments is a very challenging experiment. One noticeable difference between these experiments and those conducted on smaller molecules is the resolution of labeled species that can be attained typically. This difference stems directly from the fundamental operating
52
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a) BK+2H+
(g) BK+H+ 0s
(b)
(h) 0s
(c)
(i) 1s
(d)
(j) 5s
(e)
(k) 20s
(f )
(l) 80s
530 532 534 536
538 540 1060 m/z
1065
1070
1075
FIGURE 2.7 Mass spectra of doubly (a–f) and singly (g–l) protonated bradykinin molecules with different storage times, listed to the right of the figure. (a) and (g) were recorded with 7.4 × 10 –3 Torr of nitrogen in the LIT chamber while all others were recorded with a 1.7 × 10 –3 Torr base pressure of nitrogen and 5.7 × 10 –3 Torr of CD3OD in the trap chamber. (Reproduced from Mao, D.; Douglas, D.J. J. Am. Soc. Mass Spectrom. 2003, 14, 85–94. With permission from Elsevier.)
considerations of ESI and the limitations of ion-trapping instruments. Unlike small molecule exchange, where ionization involves typically a single deprotonation or protonation reaction, ESI of proteins produces multiply-protonated and multiply-deprotonated species having much higher charge states. This phenomenon is extremely beneficial in that it extends effectively the mass range of the ion trap and allows species with high molecular weights to be investigated in the ion trap. Unfortunately, the shift in charge state is not accompanied by an increase in resolving power of the mass spectrometer. For example, H/D exchange of a singly-protonated molecule of m/z 1000 requires only that the mass analyzer be capable of differentiating between
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
53
ions of m/z 1000 and m/z 1001 in order to monitor the isotopic exchange. In contrast, when a larger species with molecular weight of 10,000 Da is studied in the + 10 charge state, the mass spectrometer must be able to differentiate between ions of m/z 1001.0 and ions of m/z 1001.1 in order to detect a single isotopic exchange. Naturally, the gas-phase H/D-exchange analysis becomes more difficult as the mass and the most abundant charge state increase. Therefore, the utility of H/D exchange of proteins in ion-trapping instruments with moderate resolving power is limited to establishing trends in mass spectral shifts. Nonetheless, the technique provides unique insight into gas-phase conformations of proteins. For example, denatured proteins often undergo more extensive H/D exchange than do native proteins because the denatured proteins are less folded, thus, allowing access to interior active hydrogen atoms that were hydrogen bonded previously in the interior of the folded conformation. Indeed, the differentiation between folded and unfolded protein states is the primary focus of protein H/D exchange in both the gas phase and in solution. Studies of protein gas-phase H/D exchange in a LIT-TOF system include reactions of myoglobin and apomyoglobin with D2O and CD3OD [18]; lysozyme with D2O [20]; and ubiquitin, cytochrome c, apomyoglobin, and β-lactoglobulin with D2O [21]. An example of the H/D exchange of ubiquitin is shown in Figure 2.8. (a)
Relative intensity
(b)
(c)
(d)
(e)
1200
1220
m/z
1240
1260
FIGURE 2.8 Mass spectra of ubiquitin + 7 ions produced from 50:50 methanol:water at pH 2.0 at (a) and (b) 0, (c) 0.5 s, (d) 1 s, and (e) 5 s of trapping. The pressures in the trap chamber were (a) 7 mTorr N2 and (b)–(e) 2 mTorr N2 and 5 mTorr D2O. (Reproduced from Wright, P.J.; Zhang, J.; Douglas, D.J, J. Am. Soc. Mass Spectrom. 2008, 19, 1906–1913. With permission from Elsevier.)
54
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Obviously, the resolution of the mass spectra is not sufficient to discern the exact number of H/D exchanges. This particular series of mass spectra were used to correlate observations in the gas phase with similar ones obtained in solution, indicating that ubiquitin may have some memory of the solution conformation even after it is transferred to the gas phase via ESI. Similar results have been reported for studies utilizing a 3D quadrupole ion trap and CD3OD as the deuterating reagent [12].
2.3.3 Nucleosides, Nucleotides, and Oligonucleotides Fewer gas-phase H/D exchange experiments have been conducted on nucleic acids. The first gas-phase H/D exchange study in a quadrupole ion trap involved the reaction of various nucleosides and nucleoside analogs with CH3OD and ND3 [24]. The impact of the gas-phase basicity of the deuterating agent on the observed exchange of singly protonated nucleosides was evaluated. Exchange mass spectra obtained in the positive ion mode confirmed that exchange with ND3 was more rapid and more extensive than with CH3OD; these results were rationalized based on the greater similarities in gas-phase basicities of the ND3 deuterating agent and nucleosides. However, it was reported also that variations in the position of the exchangeable hydrogen atoms relative to the suspected protonation site resulted in differences in exchange behavior, and that the presence of other non-participatory functional groups could impact upon the H/D exchange. A more recent study examined the H/D exchange of deoxyribose monophosphate nucleotides, the building blocks of oligonucleotides. However, unlike the nucleosides, the mononucleotides contain a highly acidic phosphate group, which can be deprotonated easily during ESI. The study was undertaken in the negative ion mode using a weakly acidic deuterating agent, D2O, that had been confirmed previously to promote exchange via a relay mechanism [16]. Variations in the exchange behaviors of the mononucleotide isomers confirmed that the reactions were dependent on both the identity of the nucleobase and the position of the phosphate moiety. It was concluded also that the distinction between mononucleotide isomers by H/D exchange pattern showed promise for distinguishing sequences of larger oligonucleotides. Energy-variation studies, conducted either by heating the ion trap or by activating the trapped ions with an IR laser, indicated that additional energy influenced the H/D exchange pattern for the mononucleotides [16]. The investigation of the deoxyribose monophosphate nucleotides was followed by a subsequent investigation involving the simplest oligonucleotides, the dinucleotides. In this study, the H/D exchange reactions of dinucleotides containing only purine bases (that are, dAA, dAG, dGA, and dGG) and their 5′-monophosphate analogs (that are, 5′P-dAA, 5′P-dAG, 5′P-dGA, and 5′PdGG) were conducted using D2O in a quadrupole ion trap [17]. Significant differences in the rates and extents of exchange were found when the 5′-hydroxyl group of the dinucleotides was replaced by a phosphate functionality. Illustrative mass spectra are shown in Figure 2.9. Extensive and nucleobase-dependent exchange occurred for the deprotonated 5′-monophosphate dinucleotides, whereas the dinucleotides all exhibited essentially the same limited exchange. This result was correlated with the proximity of the deprotonated phosphate site to the exchangeable hydrogen atoms and the proclivity for particular nucleobase pairs to participate in nucleobase-stacking
55
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps (a)
(b)
15,000
[5´P-dGG-H]
15,000
D(9)
Intensity
10,000
D(8)
5000
13C
D(7)
D(2)
5000 D(0)
[5´P-dAA-H]
D(4)
D(1)
13C
0 674 675 676 677 678 679 680 681 682 683 684 685 686
0 642 643 644 645 646 647 648 649 650 651 652
Mass (m/z)
Mass (m/z)
(c)
(d)
30,000
10,000
D(6)
[5´P-dAG-H]
D(3)
[5´P-dGA-H]
D(2)
Intensity
20,000 D(5)
10,000 D(4)
D(7) 13C
Intensity
Intensity
10,000
D(3)
D(4) D(1)
5000
D(5) D(7)
D(0)
D(8)
0 658 659 660 661 662 663 664 665 666 667 668 669
13C 0 658 659 660 661 662 663 664 665 666 667 668 669
Mass (m/z)
Mass (m/z)
FIGURE 2.9 H/D-exchange mass spectra for deprotonated (a) 5′P-dGG, (b) 5′P-dAA, (c) 5′P-Dag, and (d) 5′P-dGA after 10 s exchange with D2O. Peaks are annotated as D(n) where n is the number of exchanged hydrogen atoms. Peaks labeled as 13C can be attributed solely to the isotopic contribution of the other exchanged peaks. (Reproduced from Chipuk, J.E.; Brodbelt, J.S. Int. J. Mass Spectrom, 2009 (In Press). With permission from Elsevier.)
interaction. In addition, results for the isomeric 5′-monophosphates, 5′-dAG, and 5′-dGA, were remarkably different, indicating that the H/D-exchange reaction was sequence dependent.
2.4 CONCLUSIONS Gas-phase H/D exchange reactions offer a versatile tool for ion trap mass spectrometry. The ability of ion traps to store ions for extended periods of time and to operate at relatively high pressures makes them robust for extensive studies of H/D exchange on a variety of analytes of interest, ranging from small organic molecules to biopolymers. Elucidation of the fundamental factors that influence gas-phase H/D exchange reactions, as well as the mechanisms, continue to be active areas of research, and it is anticipated that applications of H/D exchange will continue to expand because of the complementary information obtained compared to traditional ion-activation methods, especially when probing conformational effects in larger biopolymers.
Acknowledgment Funding from the Robert A. Welch Foundation (F-1155) and the National Institutes of Health (RO1 GM65956) is gratefully acknowledged.
56
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
REFERENCES
1. Hunt, D.F.; McEwen, C.N.; Upham, R.A. Determination of active hydrogen in organic compounds by chemical ionization mass spectrometry. Anal. Chem. 1972, 44, 1292–1294. 2. Freiser, B.S.; Woodin, R.L.; Beauchamp, J.L. Sequential deuterium exchange reactions of protonated benzenes with D2O in the gas phase by ion cyclotron resonance spectroscopy. J. Am. Chem. Soc. 1975, 97, 6893–6894. 3. Hunt, D.F.; Dethi, K. Gas-phase ion/molecule isotope-exchange reactions: methodology for counting hydrogen atoms in specific organic structural environments by chemical ionization mass spectrometry. J. Am. Chem. Soc. 1980, 102, 6953–6963. 4. Reed, D.; Kass, S. Hydrogen-deuterium exchange at nonlabile sites: a new reaction facet with broad implications for structural and dynamic determinations. J. Am. Soc. Mass Spectrom. 2001, 12, 1163–1168. 5. Chipuk, J.E.; Brodbelt, J.S. Investigation of the gas-phase hydrogen/deuterium exchange behavior of aromatic dicarboxylic acids in a quadrupole ion trap. Int. J. Mass Spectrom. 2007, 267, 98–108. 6. Niemeyer, E.D.; Brodbelt, J.S. Isomeric differentiation of green tea catechins using gasphase hydrogen/deuterium exchange reactions. J. Am. Soc. Mass Spectrom. 2007, 18, 1749–1759. 7. DePuy, C.H. An introduction to the gas phase chemistry of anions. Int. J. Mass Spectrom. 2000, 200, 79–96. 8. Hunter, E.P.; Lias, S.G. Proton affinity data. In NIST Chemistry WebBook. NIST standard reference database No. 69; Mallard, W.G.; Linstrom, P.J., Eds.; National Institute of Standards and Technology: Gaithersburg, MD, November, 1998; http://webbook.nist.gov. 9. Ausloos, P.; Lias, S.G. Thermoneutral isotope exchange reactions of cations in the gas phase. J. Am. Chem. Soc. 1981, 103, 3641–3647. 10. Gard, E.; Grenn, M.K.; Bregar, J.; Lebrilla, C.B. Gas-phase hydrogen/deuterium exchange as a molecular probe for the interaction of methanol and protonated peptides. J. Am. Soc. Mass Spectrom. 1994, 5, 623–631. 11. Reid, G.E.; Simpson, R.J.; O’Hair, R.A.J. Probing the fragmentation reactions of protonated glycine oligomers via multistage mass spectrometry and gas-phase ion-molecule hydrogen/deuterium exchange. Int. J. Mass Spectrom. 1999, 191, 209–230. 12. Evans, S.E.; Lueck, N.; Marzluff, E.M. Gas-phase hydrogen/deuterium exchange of proteins in an ion trap mass spectrometer. Int. J. Mass Spectrom. 2003, 222, 175–187. 13. Hermann, K.; Wysocki, V.; Vorpagel, E.R. Computational investigation and hydrogen/deuterium exchange of the fixed charge derivative Tris(2,4,6-Trimethoxyphenyl) Phosphonium: implications of the aspartic acid cleavage mechanism. J. Am. Soc. Mass Spectrom. 2005, 16, 1067–1080. 14. Herrmann, K.A.; Kuppannan, K.; Wysocki, V.H. Fragmentation of doubly-protonated ion populations labeled by H/D exchange with CD3OD. Int. J. Mass Spectrom. 2006, 249–250, 93–105. 15. Zhang, J.; Brodbelt, J.S. Gas-phase hydrogen/deuterium exchange and conformations of deprotonated flavonoids and gas-phase acidities of flavonoids. J. Am. Chem. Soc. 2004, 126, 5906–5919. 16. Chipuk, J.E.; Brodbelt, J.S. Gas-phase hydrogen/deuterium exchange of 5′- and 3′-mononucleotides in a quadrupole ion trap: exploring the role of conformation and system energy. J. Am. Soc. Mass Spectrom. 2007, 18, 724–736. 17. Chipuk, J.E.; Brodbelt, J.S. Gas-phase hydrogen/deuterium exchange of dinucleotides and 5′-monophosphate dinucleotides in a quadrupole ion trap. Int. J. Mass Spectrom. 2009. In Press.
Gas-Phase Hydrogen/Deuterium Exchange in Quadrupole-Ion Traps
57
18. Mao, D.; Ding, C.; Douglas, D.J. Hydrogen/deuterium exchange of myoglobin ions in a linear quadrupole ion trap. Rapid Commun. Mass Spectrom. 2002, 16, 1941–1945. 19. Mao, D.; Douglas, D.J. H/D exchange of gas phase bradykinin ions in a linear quadrupole ion trap. J. Am. Soc. Mass Spectrom. 2003, 14, 85–94. 20. Mao, D.; Babu, K.R.; Chen, Y-C.; Douglas, D.J. Conformations of gas-phase lysozyme ions produced from two different solution conformations. Anal. Chem. 2003, 75, 1325–1330. 21. Wright, P.J.; Zhang, J.; Douglas, D.J. Conformations of gas-phase ions of ubiquitin, cytochrome c, apomyoglobin, and β-lactoglobulin produced from two different solution conformations. J. Am. Soc. Mass Spectrom. 2008, 19, 1906–1913. 22. Chen, H.; Xu, R.; Chen, H.; Cooks, R.G.; Ouyang, Z. Ion/molecule reactions in a miniature RIT mass spectrometer. J. Mass Spectrom. 2005, 40, 1403–1411. 23. Kaltashov, I.A.; Doroshenko, V.M.; Cotter, R.J. Gas phase hydrogen/deuterium exchange reactions of peptide ions in a quadrupole ion trap mass spectrometer. Proteins: Struct. Funct. Genet. 1997, 28, 53–58. 24. Felix, T.; Reyzer, M.; Brodbelt, J. Hydrogen/deuterium exchange of nucleoside analogs in a quadrupole ion trap mass spectrometer. Int. J. Mass Spectrom. 1999, 191, 161–170. 25. Gronert, S. Estimation of effective ion temperatures in a quadrupole ion trap. J. Am. Soc. Mass Spectrom. 1998, 9, 845–848. 26. Wyttenbach, T.; Paizs, B.; Barran, P.; Breci, L.; Liu, D.; Suhai, S.; Wysocki, V.H.; Bowers, M.T. The effect of the initial water of hydration on the energetics, structures, and H/D exchange mechanism of a family of pentapeptides: an experimental and theoretical study. J. Am. Chem. Soc. 2003, 125, 13768–13775. 27. Dearden, D.V. KinFit: kinetics fitting for coupled ordinary differential equations, version 2.0 http://chemwww.byu.edu/people/dvdearden/kinfit.htm (posted April 2003). 28. Reyzer, M.L.; Brodbelt, J.S. Gas-phase H/D exchange reactions of polyamine complexes: (M + H) + , (M + alkali metal + ), and (M + 2H)2 + . J. Am. Soc. Mass Spectrom. 2000, 11, 711–721. 29. Schaaff, T.G.; Stephenson, J.L., Jr.; McLuckey, S.A. The reactivity of gaseous ions of bradykinin and its analogues with hydro- and deuteroiodic Acid. J. Am. Chem. Soc. 1999, 121, 8907–8919. 30. Schaaff, T.G.; Stephenson, J.L., Jr.; McLuckey, S.A. Gas phase H/D exchange kinetics: DI versus D2O. J. Am. Soc. Mass Spectrom. 2000, 11, 167–171.
for Multi-Stage 3 Methods Ion Processing Involving Ion/Ion Chemistry in a Quadrupole Linear Ion Trap Graeme C. McAlister and Joshua J. Coon Contents 3.1 Introduction..................................................................................................... 59 3.2 Electron Transfer Dissociation (ETD) Coupled with Collision-Induced Dissociation (CID).............................................................64 3.3 Collision-Induced Dissociation (Cid) and Electron Transfer Dissociation (ETD) Coupled with Proton Transfer......................................... 67 3.4 Ion Attachment (IA) Coupled with Collision-Induced Dissociation (Cid)........................................................................................... 70 3.5 Conclusion....................................................................................................... 71 References................................................................................................................. 73
3.1 INTRODUCTION Perhaps one of the most influential concepts in protein mass spectrometry has been the notion of enzymatic protein digestion to render a collection of peptides of suitable size for conventional tandem mass spectrometry (collision-induced dissociation, CID) [1,2]. Doubtless, this methodology has enabled significant progress for global protein identification; however, many investigators now realize this approach has significant limitations [3]. Their conclusion is based upon the following observations. First, protein post-translational modifications (PTMs) on multi-domain proteins, and among components of protein–protein machines, work in concert; to determine their biological relevance, these patterns must be detected within the context of one another, that is to say, across the whole protein. Second, transcriptional editing processes are pervasive in higher eukaryotes and difficult to predict, even with a completely-sequenced genome. For example, three-quarters of all human proteins are expected to have at least one splice variant, that is, the gene is ‘read’ in multiple ways giving rise to multiple proteins [4–6]. Other transcription and translation events such as gene fusion, 59
60
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
PTR
CID nH+
ETD
Peptides and proteins
IA
FIGURE 3.1 Quadrupole ion trap mass spectrometers are capable of concatenating multiple, unique ion-processing methods during routine, robust, day-to-day operation; for example, a user can combine electron transfer dissociation (ETD), proton transfer reaction (PTR), collision-induced dissociation (CID), and ion attachment (IA) in any order in but a single scan. This ability highlights how trapped-ion instruments are capable both of mass analysis and of functioning as an ion reaction vessel.
where two separate genes are fused together to form a new unique gene, and single nucleotide polymorphisms, where two DNA molecules differ from each other by a single nucleotide, occur also. Thus, the use of short peptides as proxy markers for genes are inadequate and often misleading [7]. Today, ion trap mass spectrometers offer multiple ion-manipulation methodologies so that whole protein sequence analysis is increasingly achievable; such analysis is described as top-down proteomics [3,8–15]. With these ion-handling methods, ion trap mass spectrometers are well positioned to accelerate our ability to process effectively large species in the gas phase.* The basis of this approach is the implementation of multi-functional tools for systematic ion manipulation and processing, as depicted diagrammatically in Figure 3.1. Ion/ion chemical reactions, which include electron transfer (ET), proton transfer, and ion attachment (IA) (see below), represent one family of such tools. These technologies can be meshed with other more conventional ion trap processing methodologies such as ion isolation and CID. Together, these individual components form the foundation of, or comprise the basic toolset for, a versatile approach that promises to accelerate markedly the field of large molecule mass spectrometry. Tandem mass spectrometry (MS/MS), in its simplest and earliest implementations on quadrupole ion traps (QITs), employed sequential (tandem-in-time MS/MS) isolation and energetic (for example, collisional) activation steps [16–22]. Modern MS/ MS experiments can involve multiple stages of mass selectivity (MSn), which employ different dissociation methods. Depending upon precursor chemistry, sometimes * See Volume 5, Chapter 4: Chemical Derivatization and Multistage Tandem Mass Spectrometry for Protein Structural Characterization by Jennifer M. Froelich, Yali Lu, and Gavin E. Reid.
61
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
there is no adequate activation method; in these cases, it is often advisable to combine multiple activation methods into a single concerted MSn scan type. In this chapter we discuss these concatenated MSn scan functions, focusing particularly on those scan functions that facilitate large molecule analysis. Virtually all Quadrupole ion trap (QIT)-based MS/MS experiments comprise three basic steps: (i) precursor ion isolation; (ii) precursor ion manipulation (for example, collisions, reactions such as ion/ion or ion/molecule, etc.); and (iii) product ion mass/ charge ratio analysis. The appeal of MS/MS for protein sequence analysis has come, in part, because of its ability to provide primary sequence information, and its relative ease of use when compared to alternative wet chemistry-based methods like Edman degradation [23–26]. The utility of MS/MS can often be extended by implementing multiple iterations (MSn), in which subsequent activation steps are employed to manipulate selected product ions. Due to the iterative and expandable nature of QIT-based MSn analysis, many mass-selective and/or activating steps can be strung together in a single scan; for example, one experiment utilized 11 sequential activation steps to produce an MS12 spectrum [27]. The ideal ion activation strategy would, of course, generate sufficient information to identify the precursor ion in a single concerted step (that is, MS2). But, from the study of a variety of activation strategies, it has become apparent that their utility varies depending on size and charge state of the precursor ion and the presence of any labile functional groups; in some cases, one activation method is better suited for a particular subset of ions than another and, in some instances, there is no ideal activation method [28–32]. In these latter situations, coupling together multiple activation methods, to probe fully a precursor population can provide the necessary information. The scan type employed generally by ion traps relies on activation by collision, whereupon precursor ion kinetic energy is transferred to internal energy; however, as the efficacy of this process is proportional to the ratio of target (neutral) mass to projectile (ion) mass, CID is not effective for high mass ions. Gas-phase ion/ion manipulations that react cation precursors with an anionic reagent have become established in response to the development of electrospray ionization, which can produce multiply-charged ions from large polyatomic molecules, such as whole proteins [33]. The development of ion/ion chemistries from this multiple-charging capability is related directly to the fact that at least one ion/ion product must retain a charge if the product ion population is meant to be studied by mass spectrometry. The original descriptions of ion/ion chemistry focused on PTR and its application to both individual species of ions and entire ion populations [34–43]. From these original studies, the library of ion/ion reactions has expanded to include three main types:
1. Proton transfer reaction (PTR), whereby an ionic reagent either abstracts or donates a proton to a precursor [34–43]
[ M + 3H]3+ + A − → [ M + 2H]2+ + HA
(3.1);
62
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
2. ET, whereby an ionic reagent either abstracts or donates an electron to a precursor [44,45]
[ M + 3H]3+ + A −• → [ M + 3H]2+• + A
3. IA, whereby the ion reagent and precursor react to form a new complex [46–49]
[ M + 3H]3+ + A − → [ M + 3H + A]2+
(3.2);
(3.3).
Central to implementing MSn scans is the temporal flexibility allowed for by ion/ ion reactions. Following an ion/ion reaction, a completely different ion-manipulation step can be conducted nearly instantaneously, with the only required time being the few milliseconds necessary to prepare the electronics for whatever form of activation is to follow. For such experiments, ion/ion reactions offer a significant time advantage over ion/molecule reactions, which can require tens of seconds, or even minutes, between admission of the molecular reagent to the reaction chamber and to the completion of purging of that reagent so that the subsequent activation method may be initiated [50–60]. Technological advancement has been the primary contributor to this temporal flexibility associated with ion/ion reactions [61]. Ion/ion reactions require typically multiple ion sources for even MS2 experiments (that is, one for the cation precursor population and one for the anion reagent population), and often any additional MSn stages that require additional ion reagents may necessitate even more sources. Hence, the ability to concatenate multiple unique ion/ion activation methods during routine, robust, day-to-day operation has been made possible by the development of mass spectrometers that can accommodate simultaneously multiple ion sources. To meet these needs, researchers have developed schemes and instruments that allow ions to be injected into the QIT from multiple directions, for example, through each of an end-cap electrode and the ring electrode of a three-dimensional (3D) ion trap, in this way, two separate and distinct ion sources and ion pathways can be employed in generating and supplying the precursor ion and reagent ion populations (Figure 3.2a) [34,45,48,62–66]. Other approaches focus on adapting a single source region and ion pathway to accommodate both populations [43,67–73]. The latter scheme often requires that implementation of the sources be separated in time, that is, either the voltage applied to the sources or their placement in front of the inlet of the instrument varies during the scan cycle. In Figure 3.2b, two ESI sources are located at the MS inlet: one of the sources, a standard ESI source, is used to generate the reagent anions; the second source, a nano-spray static tip, is used to generate the precursor peptide and protein cations. The high-voltage power supplies for the two sources are triggered during the scan depending upon which ion population is needed. In a few cases, ion/ion reagents and precursors have been generated simultaneously via bipolar ionization, for example sonic spray ionization, SSI [74,75]. Ion traps themselves have been the focus of instrumentation projects. For example, linear ion traps were adapted to permit the imposition of radio-frequency (RF) voltages both
63
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry (a)
Ion trap analyzer Guard ring
Electrospray needle
Electron multiplier
Protein sample infusion
Conversion dynode
Gate lens
Application of high voltage DC pulser (Gate lens)
Positive ions Negative ions Needle valve
(b)
(c) A ESI + Curtain plate Orifice Skimmer
Fused silica capillary Capillary holder
Triggered ± HV
B
Back
Center
Cl – Back lens Precursor ions moved to front section +
D E
Triggered ± HV
Front
Front lens
C Nano-ESI tip
ASGDI sample containment vessel
F
0V –10 V +5 V
Anion injection
+ +
–
Ion/ion reaction
–
–
+5 V 0V
–
+ –
0V
End reaction and scan out –
+
–
0V
FIGURE 3.2 Multiple methods are used to generate and to inject reagent ions for conducting ion/ion reactions in a quadrupole ion trap. (a) Schematic of a quadrupole ion trap, which was adapted to permit the injection of reagent ions through the ring electrode (Reproduced from Stephenson, J.L.; McLuckey, S.A. Int. J. Mass Spectrom. Ion Processes. 1997, 162, 89–106. With permission from Elsevier.) (b) Method of injection of cations and reagent anions via dual atmospheric pressure ion emitters; this approach requires very little, if any, modification of the ion trap system (Reproduced from Xia, Y.; McLuckey, S.A. J. Am. Soc. Mass Spectrom. 2008, 19, 173–189. With permission from Elsevier.) (c) A schematic diagram illustrating the implementation of ion/ion reactions on a quadrupole linear ion trap system. Note reagent anions are generated on the right side of the diagram via a CI source. This quadrupole linear ion trap system has been adapted such that either it can mix selectively ions of opposite polarity or it can segregate ions of opposite polarity via a DC offset and charge-sign independent trapping. (Reproduced from Syka, J.E.P., Coon, J.J.; Schroeder, M.J.; Shabanowitz, J.; Hunt, D.F., Proc. Natl. Acad. Sci. 2004, 101, 9528–9533. With permission from the National Academy of Sciences.)
64
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
radially and axially thereby permitting charge-sign independent trapping, which in turn allows the device to function as an ion/ion reaction vessel* (Figure 3.2c) [45]. In this chapter, we present details of how the compilation of ion-manipulation methodologies in QIT mass spectrometers can eliminate many of the restrictions that limit currently large molecule mass spectrometry. Specifically, we highlight three MSn interrogation schemes that show genuine promise to enable whole protein analysis: (i) ETD coupled with CID; (ii) CID and ETD coupled with Proton Transfer; and (iii) IA coupled with CID.
3.2 ELECTRON TRANSFER DISSOCIATION (ETD) COUPLED WITH COLLISION-INDUCED DISSOCIATION (CID) Electron-based dissociation methods (ETD and its forerunner electron capture dissociation, ECD) are highly efficient at generating c- and z• -type fragment ions from peptide and whole protein precursor ions via a process that is essentially independent of peptide length, amino acid composition, and post-translation modification (PTM) state [28,45,76–81]. Precursor charge state and m/z-value are important factors, however, that can affect the dissociation efficiency of ETD methods [82–85]. The most widely-held explanation for why lowly-charged, high-m/z ions do not fragment efficiently by ETD is that there is a charge-to-residue threshold to efficient ETD, that is, a minimum charge density is necessary in order to achieve direct dissociation [85,86]. If a particular precursor ion falls below this threshold then, following cleavage of the backbone bond, the probability that the two product ions remain non-covalently bound increases. To obviate this problem more energy can be added to the post-reaction product, for example for ECD, analyte ions can be activated by infrared photons prior to ECD, a process that has been termed activated ion ECD (AI-ECD) [10,14,87–90]. The additional energy absorbed disrupts preferentially the non-covalent bonds and increases fragmentation efficiency. The QIT-based ion/ion analog of AI-ECD was described first by Swaney et al. [32]. There we described in detail how gentle CID of the non-dissociative ET products (ETnoD) could increase significantly the yield of c- and z• -type product ions; this fragmentation process is referred to as EtcaD in Ref. [32]. Optimization of the energy was essential for the preferential breakage of the non-covalent bonds because, if too much energy was imparted, then b- and y-type product ions were formed. Over a large data set, that is, liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis of a complex mixture, we showed how data-dependent activation of the ETnoD products by these optimal CID parameters increased the overall number of c- and z• -type ions produced, thereby ETD efficiency was increased as well (Figure 3.3). Since this original description of the ETD/CID MS/MS scan type (ETcaD), it has been implemented on commercial instruments, applied toward biologically-relevant samples, and expanded upon [28,85,91,92]. For example, Han et al. [92] coupled ETD with beam-type CID on their QTRAP hybrid triple quadrupole/linear ion trap instrument (Figure 3.4). An advantage of both this implementation, and beam-type CID in general, is that the spectra do not suffer from the low mass cut-off that is * See Volume 5, Chapter 1: Ion/Ion Reactions in Electrodynamic Ion Traps by Jian Liu and Scott A. McLuckey.
ETD, 532 m/z
300
400
z+3
1000
1200
z 11
+
1400
×8.5
(M+2H)++
(M+2H)++
1000
z+13
z+12
900
z+8
ET-CID,532 m/z
300
z5 c + 5
400
z3
200
400
Intensity: 1.0×105 z3
z4 600
I G S E I S S LT L E E A R
(e)
200
Intensity: 1.0×105 z2
VV D IV D TF R
z6
500
z+5 700
800 m/z
z7
400
600
800
1000 m/z
×4
1200
(Precursor)
1400
1600
1800
c20
z+20 2000
×8
200
z4 400
Intensity: 1.0×104
ET-CID,966 m/z
600
z5
z6
Y6
800
z7 z8
1000 m/z
z9
1200
1400
(Precursor) z ×4 z 12 11 z 13 z14 z13
(M+2H)++
1600
z16
1800
2000
(M+2H)++ z+ c+20 20 c19
1400
z+13
(M+2H)++
1000
(M+2H)++
FIGURE 3.3 Comparison of ETD (a–c) and ETcaD (d–f) product ion mass spectra (single scan) for three tryptic peptides. Each product ion mass spectrum was acquired during a data-dependent analysis of a complex tryptic peptide mixture derived from Arabidopsis. (Reproduced from Swaney D.L.; McAlister, G.C.; Wirtala, M.; Schwartz, J.C.; Syka, J.E.P.; Coon, J.J., Anal. Chem. 2007, 79, 477–485. With permission from the American Chemical Society.)
200
Intensity: 1.0×104
ETD, 966 m/z
(M+2H)++
1200
z+12 c12
c13
z8 z+ 8 900
z+11
z+7
c11
800
z+10
z+6
1000
z9
(M+2H)++
(Precursor) ×2 z8
600 m/z
z4 ×1.5
(Precursor)
(M+2H)++
VG P P PA P S GG L P GT DN SD QA R
800 m/z
z+10
800
z+7
ET-CID,532 m/z
VG P P PA P S GG L P GT DN SD QA R
600
700
(Precursor) ×3
5
(M+2H)++
600 m/z
z
+
z+6
×5
(d)
(f )
400
500
z+4
(Precursor)
(M+2H)++
(M+2H)++
(c)
200
Intensity: 1.0×105
I G S E I S S LT L E E A R
ETD, 753 m/z
(b)
200
Intensity: 1.0×105
VV D IV D TF R
(a)
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry 65
66
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Nanospray emitter
Q0
IQ2 ~ IQ3 ~ Q2 Q1 Q3
Pulsed + HV Pulsed Reagent vapor – HV APCI emitter
FIGURE 3.4 Schematic diagram of the QTRAP hybrid triple quadrupole/linear ion trap instrument that has been modified to conduct ion/ion reactions in Q2. (Reproduced from Han, H.L.; Xia, Y.; McLuckey, S.A., Rapid Commun. Mass Spectrom. 2007, 21, 1567–1573. With permission from John Wiley & Sons, Inc.)
characteristic of ion trap CID. Also, multiple ETnoD product ions can be activated simultaneously instead of sequentially, as would be required with ion trap CID. An analog to coupling ETD with CID, which uses CID as the primary dissociation mechanism, is the addition of a PTR activation step prior to CID. Like ETD, CID activation is also charge dependent. But unlike ETD, where fragmentation efficiency increases with increasing charge density, CID fragmentation efficiency tends to be higher for lowly-charged precursor ions [93–95]. Unfortunately, during typical proteomics experiments, ionization mechanisms determine almost exclusively precursor charge state, and larger precursors tend to ionize through the acquisition of more charges. To some degree, these trends obstruct the interrogation of all large precursor ions via CID. The inclusion of a PTR activation step prior to CID can help mitigate this problem. As noted earlier, PTR involves anionic reagents that abstract protons from cationic precursors [34–43]. In the context of this experiment, a PTR step helps by stripping away charges from a precursor, that is too highly-charged to be dissociated effectively via CID, and places it in a charge state where it is more amenable to fragmentation [96]. An interesting variant of this experiment involves performing ion parking, pioneered by McLuckey et al. [96], during the PTR ion/ion reaction. Ion parking is the quenching of an ion/ion reaction at a specific product ion mixture instead of letting those products react further with additional reagent ions. Ion parking can take two forms: (i) separation of the reagents and products in space via segregation of the ion clouds; and (ii) excitation of the product ions via the application of a supplemental waveform, which prevents the oppositely-charged ions from forming an orbiting complex, a prerequisite to an ion/ion reaction [13,97–102]. In either case, the inclusion of a parking step allows all of the precursor ions to be converted efficiently from one or multiple charge states to product ions of a lower-charge state. As noted earlier, the ideal mass spectrometer-based activation method would require only a single step. However, as we have learned more about ion activation methods, both their abilities and their limitations, we have arrived at the conclusion that no single ion activation method is best suited for interrogating every possible precursor ion. Hence, the experiments discussed here highlight the utility of mixing multiple activation methods in a single scan as a viable alternative; when CID is combined with ETD, ETD-like dissociation is observed wherein rich product ion mass spectra permit identification of precursor ions that have low charge states and m/z-values.
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
67
3.3 COLLISION-INDUCED DISSOCIATION (CID) AND ELECTRON TRANSFER DISSOCIATION (ETD) COUPLED WITH PROTON TRANSFER (PTR) By analogy with the successful coupling of ETD with CID, wherein the second activation step augmented the first activation step by converting first generation product ions into new species that are more informative of sequence, CID, and ETD have been coupled to PTR [36,77]. QITs are capable of high-sensitivity analysis and fast duty cycles (spectra s−1) at unit mass-resolving power. This limited resolution generally means that only singly and doubly-charged ions can be identified. Operation of QITs at resolutions sufficient to determine fragment ion charges greater than two is possible, but generally is considered impractical as the scan rates and sensitivity are substantially reduced [103]. Principally as a result of this limitation, MS/MS analysis on QITs has been restricted to lowly-charged precursor ions (for example, in bottom-up proteomics), as these populations tend to produce lowly-charged product ions that are ideal for QIT analysis. Proton transfer (PTR), one of the ion/ion reactions detailed above, has been developed primarily as a means of simplifying mixtures of highly-charged ions [34–43]. Higher-charged precursor ions will react faster than lower-charge precursor ions as reaction rates for an ion/ion reactions increase by the square of the charge state of the participating ions [35,104,105]. Therefore, PTR activation of a mixture of highlycharged ions will result in the ion population being concentrated into lower-charge states. The utility then of PTR, as it applies to interrogation of larger highly-charged precursor peptides and proteins, is that following the initial activation and dissociation (MS2) of the precursor cation, the product population can be simplified and concentrated en masse into lower-charge state ions by a subsequent PTR activation step. Initially, this multi-stage activation scheme employed CID as the prime activation step (MS2) [36]. These experiments, pioneered by Stephenson and McLuckey, were carried out on a Finnigan 3D QIT, which was modified to allow for injection of PTR reagent anions, which were generated via glow discharge of the sampled headspace of perfluoro-1,3-dimethyl-cyclohexane (PDCH), through an additional hole in the ring electrode. Cation precursors, generated via electrospray, were injected through a hole in one of the end-cap electrodes. Highlighted in Figure 3.5 is the interrogation of melittin via CID/PTR. The mass spectrum in Figure 3.5a is the post-CID MS/MS spectrum of the [M + 4H]4 + melittin ion, and Figure 3.5b displays the same ion population after being subjected to a PTR reaction for 100 ms. Following PTR activation, the b- and y-type ions are concentrated in their lower-charge states. In this example, the coupling of CID with PTR allowed for the successful interrogation of a higher-charged precursor species by removing charge from the multiply-protonated CID product ions. ET dissociation is coupled easily also with a subsequent PTR step. Even though dissociation via ETD is inherently destructive to precursor charges, in that typically one charge is neutralized for every backbone bond that is broken, in general, this procedure does not lead inevitably to a reduction of the aggregate charge of the ion population to the point that would permit isotopic resolution for identification. For highly-charged precursor ions, multiple ion/ion reactions are required in order to achieve the desired level of charge reduction; however, multiple ETD reactions will result in multiple breakages of the peptide backbone bonds which, in turn, produce
68
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
12000
Intensity (arbitrary units)
(a)
2+ y13
8000
[M+4H–2NH3]4+
[M+4H]4+
3+ y13
4000
3+ y24
2+; b+ y21 13 3+ y17
b5+
+
b12
0 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 m/z (b) G I G A V L K V L T T G L P A L I S W I K R K R Q Q–NH2
Intensity (arbitrary units)
1600
+ y13
1200 2+ y13
800
400
b+8 b+7
+ b10
b+9
+
+ b + b12 13 y11 + + y10 y12 b+ 11
+ y14
[M+H–NH3]+ + y16
+ y15
[M+H–2NH3]+ + y19 + y17 y+ y+ + 21 + 24 + y18 y20 y22
0 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 m/z
FIGURE 3.5 (a) Pre-ion/ion reaction (PTR) MS/MS spectrum of the [M + 4H]4 + parent ion of melittin. (b) Post-ion/ion reaction (PTR) MS3 spectrum of the [M + 4H]4 + parent ion of melittin. (Reproduced from Stephenson, J.L.; McLuckey, S.A. Anal. Chem. 1998, 70, 3533–3544. With permission from the American Chemical Society.)
internal fragments that will convolute substantially the product ion mass spectrum. Hence, two types of ion/ion reactions are required; one to break the backbone bonds and one to strip away the excess charges. The first reported use of the ETD/PTR scan type involved reactions between cation precursors with radical fluoranthene anions (ETD), followed by reactions of cationic ETD product ions with benzoic acid anions (PTR) (Figure 3.6) [77]. These
69
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry (a) 15 ms ETD
Int: 1.2 × 104
15 ms ETD/50 ms PTR
Int: 9.0 × 103
15 ms ETD/100 ms PTR
Int: 4.6 × 103
(b)
(c)
(d) 15 ms ETD/150 ms PTR c2 c3 c4 c7 c5 z8 z3 z4 z5 z6 z7 c6 400
600
800
z17++ c17++ c8
Int: 3.4 × 103
c8
c17/z17 c10 z13 c15/z15 z15 c9 c c z9 z10 z11c11z12 12 c13 z1414 c16
1000
1200 m/z
1400
1600
1800
2000
(e) M Q I F V KT LT G K T I TL E V E S S D T ID N V K S K IQ D K E G I P P D Q Q R L I F A G KQ L E D G R T L S D Y N I Q K E S T LH L V L R L R G G
FIGURE 3.6 Product ion mass spectrum of ubiquitin generated by sequential ion/ion reactions. (a) Whole protein dissociation (ubiquitin + 13, having 13 residues, that is, 13 amino-acids, m/z 659) after reacting for 15 ms with the radical anion of fluoranthene. Note production of several hundred highly-charged unresolved c- and z-type product ions. (b–d), The subsequent reaction of these product ions with even-electron anions of benzoic acid for 50 (b), 100 (c), and 150 (d) ms. Note the gradual degradation of multiply-charged products, leaving predominately doubly and singly-charged fragments after 150 ms. (e) The resulting sequence coverage considering only singly-charged product ions. Each mass spectrum is the average of 50 spectra (30-s acquisition), and the relative ion abundance is indicated on the ordinate. (Reproduced from Coon, J.J.; Ueberheibe, B.; Syka, J.E.P.; Dryhurst, D.D.; Ausio, J.; Shabanowitz, J.; Hunt, D.F., Proc. Natl. Acad. Sci. 2005, 102, 9463–9468.)
experiments were conducted on a modified Finnigan LTQ mass spectrometer. The instrument was adapted to accommodate a chemical ionization (CI) source on the far side of the linear ion trap relative to the standard atmospheric pressure source, and the electronics supplying the linear ion trap were modified to allow for the
70
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
superimposition of an RF voltage on the end lenses, which allows for charge-sign independent trapping. Both fluoranthene (for ETD) and benzoic acid (for PTR) were volatilized and ionized in the added CI source (that is, one source was used to generate both reagent anions); depending upon which ion/ion activation method was being employed, the competing ion/ion reagent was removed by the application of a selective waveform. In Figure 3.6, four mass spectra are presented, each one utilizing the same ETD reaction parameters but for progressively longer PTR times (0–100 ms, in Figure 3.6A through D, respectively). As both sets of authors noted in these papers, by deconvoluting the charge states of the product ions using an additional PTR step, an experimenter can extend significantly the size range of precursor ions that are compatible with QIT instruments [36,77]. The propensity of researchers to perform bottom-up proteomics experiments involving smaller precursor peptides is not the result of a lack of interest in interrogating longer precursor peptides and whole proteins. Since the advent of these scan types, which involve ion dissociation followed by PTR, they have been employed in studies to investigate biological problems such as histone PTM state and to identify intact proteins [106,107]. Identification of PTM motifs and the importance of alternative splicing on biological systems require the analysis of at least large peptides and, preferably, intact proteins.
3.4 ION ATTACHMENT (IA) COUPLED WITH COLLISION-INDUCED DISSOCIATION (CID) In the two previously-described experiments, the second activation step augmented the first activation step by converting the first generation product ions into new species that were more informative of sequence; for example, PTR was coupled to CID to convert the CID product ions, which were produced from the interrogation of high-charge state precursors, into forms that were better suited for QIT analysis. But for both the MS2 and MS3 scan types, the dominant fragmentation pathways were the same, that is, the inclusion of PTR did not result in any additional backbone bonds being broken. Coupling an IA reaction to CID however, is done for tangential reasons, that is, IA is included before the CID step to bias the CID process toward forming different product ions [49]. The goal is still to produce sequence-informative ions that would not have been possible using only an MS2 scan type but, in this case, the production of sequenceinformative ions is achieved by biasing the CID process away from generating one set of product ions and toward the generation of a completely different set. To date there has been very little published about IA reactions. A few metal-based anions, phosphorus hexafluoride, and I¯ have all been reported to form complexes with peptide cations [37,108,109]. Yet upon CID, these complexes tend to separate easily, which implies that they are simply coming together to form a long-lived intermediate of the PTR pathway. Glish and Payne reported the first IA work detailing how activation with FeCO −2 resulted in peptide fragment ions that contained the anion reagent, which implies that the anion forms bond with the peptide that are stronger than the peptide’s own intramolecular backbone bonds [48]. Also, Gunawardena et al. showed that a reaction between AuCl −2 and peptide cations resulted in selective cleavage of disulfide bonds [46].
71
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
O H3N+
N H
R+
N H
O
+
NH3
O
H N
2.
O ÖH
O
OH
3.
O
R
HN N
O O
N O O
NH O N –2 O + Re O N OH
H
–
Re
OH
ÖH O
OH
OH
+
N
O
O
H
4. O
+
N H
R+
O
O
H N
H3N+
O –2 Re OH O Ö
O – Re O
1.
O O
OH
5. O
R+
+
R +
O NH + 3
+
Re– N
H2O
NH3
N
N
O
O
O
+ H2O
O
OH
FIGURE 3.7 Proposed mechanism for the attachment of rhenate to a doubly-protonated p eptide. (Reproduced from McAlister, G.C.; Kiessel, S.E.B.; Coon, J.J., Int. J. Mass Spectrom. 2008, 276, 149–152. With permission from Elsevier).
Recently, we demonstrated that reactions between rhenate Re O3− and multiplycharged peptide cations resulted in a new compound that, upon CID, retained the metal oxide (Figure 3.7) [49]. These experiments were conducted on a modified Finnigan LTQ mass spectrometer, which was adapted to accommodate a CI source. The Re O3− was generated initially by leaking atmospheric air into the CI source region, and, later, by leaking in high-purity 18O2. In both cases, molecular oxygen reacted with the rhenium filament of the CI source to generate the anion reagent. Of particular interest to us, the presence of rhenate alters the preferred CID fragmentation pathways (Figure 3.8). Upon CID, the singly-charged synthetic peptide RAAAKAAAK produces the b8+ ion; however, following the addition of a Re O3− ion to the doubly-charged species via IA, CID produces fragment ions derived from four different backbone bond cleavages. One exciting possible application of this MS3 scan type, which involves an IA reaction followed by CID, is site-specific gas phase disassembly of peptide and protein cations. In this application, anion reagents would bind to the peptide or protein with a high degree of site or motif specificity. Then during the following dissociation step, the bound anion reagent would direct backbone cleavage. In this manner, a large protein could be broken apart systematically and interrogated in much smaller chunks, which are more readily sequenced by the instrument. The anions presented in this section, which are capable of forming covalently-bound complexes with cation precursors, represent a solid first step; however, they are not at the level of functionality described above.
3.5 CONCLUSION The perfect ion activation technique would produce a sequence-informative spectrum in a single concerted step (that is, MS2) regardless of peptide length, amino acid
72
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a) [M+H]+
b8+H2O [M+H]+
(Precursor)
R A A A KA A A K
b8
[M+H–H2O]+
(b) [M+2H+ReO3–H2O]+
a4+ReO
[M+2H+ReO3–3H2O]+ RAAAKAAAK b4+ReO a8+ReO a7+ReO b +ReO b8+ReO 7
a3+ReO 300
400
500
600
700 m/z
800
900
1000
FIGURE 3.8 Collision-induced dissociation (CID) of the singly-protonated peptide RAAAKAAAK (a) and the same peptide following attachment of the rhenate anion (b). Note the preferred dissociation pathways change. (Reproduced from McAlister, G.C.; Kiessel, S.E.B.; Coon, J.J., Int. J. Mass Spectrom. 2008, 276, 149–152. With permission from Elsevier.)
composition, or presence of PTMs. Unfortunately, there are niches of peptide and protein precursors, delineated by size, charge, amino acid composition, and PTM state, for which no ideal activation method exists. As a means of accommodating this diverse set of precursors, selected activation methods can be brought together in multiple-stage activation schemes (MSn) that are tailored sufficiently to the precursor ion chemistry to allow for characterization of the sequence. In this vein, ion/ion reactions allow for a variety of ion manipulations that are otherwise not possible, are coupled easily with CID, and with each other in discrete reaction sequences. The three scan types described here are by no means a complete listing of all the possible MSn scan types which contain ion/ion reactions. For example, McLuckey and coworkers have published extensively on charge inversion reactions and their utility in MSn scans as a means of dissociating precursor polarity from the ionization process; for example, precursors can be ionized in negative mode but activated in a positive charge state following charge inversion [46,73,110–112]. There have been studies that have examined the affect of CID on specific ETD fragment ions [113]. For every ion/ ion reaction presented here there are numerous different reagent analogs, each possessing a unique set of drawbacks and benefits, which can and in many cases have been used. As trapped-ion mass spectrometers capable of ion/ion reactions have become more prevalent and accessible to the scientific community, the body of knowledge covering ion/ion chemistries has grown accordingly. This rapid increase in knowledge has been transformative in the sense that these instruments have undergone a transition from simple mass spectrometers into gas-phase chemical reactors of extraordinary
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
73
performance. For example, such gas-phase ion-trapping devices have evolved to the point where multi-stage reactions can be performed in milliseconds, reaction parameters such as time and reagent population size can be manipulated easily, and single product ion detection has been realized. In touching lightly on these attributes of gas-phase ion-trapping devices, this chapter has described merely the first stages of this exciting transition of enormous potential.
REFERENCES
1. Aebersold, R.; Mann, M. Mass spectrometry-based proteomics. Nature 2003, 422, 198–207. 2. McCormack, A.L.; Schieltz, D.M.; Goode, B.; Yang, S.; Barnes, G.; Drubin, D.; Yates, J.R. Direct analysis and identification of proteins in mixtures by LC/MS/MS and database searching at the low-femtomole level. Anal. Chem. 1997, 69, 767–776. 3. Du, Y.; Meng, F.Y.; Patrie, S.M.; Miller, L.M.; Kelleher, N.L. Improved molecular weight-based processing of intact proteins for interrogation by quadrupole-enhanced FT MS/MS. J. Proteome Res. 2004, 3, 801–806. 4. Johnson, J.M.; Castle, J.; Garrett-Engele, P.; Kan, Z.Y.; Loerch, P.M.; Armour, C.D.; Santos, R.; Schadt, E.E.; Stoughton, R.; Shoemaker, D.D. Genome-wide survey of human alternative pre-mRNA splicing with exon junction microarrays. Science 2003, 302, 2141–2144. 5. Resch, A.; Xing, Y.; Modrek, B.; Gorlick, M.; Riley, R.; Lee, C. Assessing the impact of alternative splicing on domain interactions in the human proteome. J. Proteome Res. 2004, 3, 76–83. 6. Stamm, S.; Ben-Ari, S.; Rafalska, I.; Tang, Y.S.; Zhang, Z.Y.; Toiber, D.; Thanaraj, T.A.; Soreq, H. Function of alternative splicing. Gene 2005, 344, 1–20. 7. Godovac-Zimmermann, J.; Kleiner, O.; Brown, L.R.; Drukier, A.K. Perspectives in spicing up proteomics with splicing. Proteomics 2005, 5, 699–709. 8. Amunugama, R.; Hogan, J.M.; Newton, K.A.; McLuckey, S.A. Whole protein dissociation in a quadrupole ion trap: Identification of an a priori unknown modified protein. Anal. Chem. 2004, 76, 720–727. 9. Ge, Y.; ElNaggar, M.; Sze, S.K.; Bin Oh, H.; Begley, T.P.; McLafferty, F.W.; Boshoff, H.; Barry, C.E. Top down characterization of secreted proteins from Mycobacterium tuberculosis by electron capture dissociation mass spectrometry. J. Am. Soc. Mass Spectrom. 2003, 14, 253–261. 10. Ge, Y.; Lawhorn, B.G.; ElNaggar, M.; Strauss, E.; Park, J.H.; Begley, T.P.; McLafferty, F.W. Top down characterization of larger proteins (45 kDa) by electron capture dissociation mass spectrometry. J. Am. Chem. Soc. 2002, 124, 672–678. 11. Ge, Y.; Lawhorn, B.G.; Elnaggar, M.; Sze, S.K.; Begley, T.P.; McLafferty, F.W. Detection of four oxidation sites in viral prolyl-4-hydroxylase by top-down mass spectrometry. Protein Science 2003, 12, 2320–2326. 12. Hogan, J.M.; Pitteri, S.J.; McLuckey, S.A. Phosphorylation site identification via ion trap tandem mass spectrometry of whole protein and peptide ions: Bovine alpha-crystallin A chain. Anal. Chem. 2003, 75, 6509–6516. 13. Reid, G.E.; Shang, H.; Hogan, J.M.; Lee, G.U.; McLuckey, S.A. Gas-phase concentration, purification, and identification of whole proteins from complex mixtures. J. Am. Chem. Soc. 2002, 124, 7353–7362. 14. Sze, S.K.; Ge, Y.; Oh, H.; McLafferty, F.W. Top-down mass spectrometry of a 29-kDa protein for characterization of any posttranslational modification to within one residue. Proc. Natl. Acad. Sci. 2002, 99, 1774–1779.
74
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
15. Forbes, A.J.; Patrie, S.M.; Taylor, G.K.; Kim, Y.B.; Jiang, L.H.; Kelleher, N.L. Targeted analysis and discovery of posttranslational modifications in proteins from methanogenic archaea by top-down MS. Proc. Natl. Acad. Sci. 2004, 101, 2678–2683. 16. Brodbelt, J.S.; Kenttamaa, H.I.; Cooks, R.G. Energy-resolved collisional activation of dimethyl phosphonate and dimethyl phosphite ions in a quadrupole ion trap and a triple quadrupole mass-spectrometer. Org. Mass Spectrom. 1988, 23, 6–9. 17. Louris, J.N.; Brodbelt, J.S.; Cooks, R.G. Photodissociation in a quadrupole ion trap mass-spectrometer using a fiber optic interface. Int. J. Mass Spectrom. Ion Processes 1987, 75, 345–352. 18. Louris, J.N.; Cooks, R.G.; Syka, J.E.P.; Kelley, P.E.; Stafford, G.C.; Todd, J.F.J. Instrumentation, applications, and energy deposition in quadrupole ion-trap tandem mass spectrometry. Anal. Chem. 1987, 59, 1677–1685. 19. McLuckey, S.A.; Glish, G.L.; Kelley, P.E. Collision-activated dissociation of negative-ions in an ion trap mass-spectrometer. Anal. Chem. 1987, 59, 1670–1674. 20. Stafford, G.C.; Kelley, P.E.; Syka, J.E.P.; Reynolds, W.E.; Todd, J.F.J. Recent improvements in and analytical applications of advanced ion trap technology. Int. J. Mass Spectrom. Ion Processes 1984, 60, 85–98. 21. Strife, R.J.; Kelley, P.E.; Weber-Grabau, M. Tandem mass spectrometry of prostaglandins: A comparison of an ion trap and a reversed geometry sector instrument. Rapid Commun. Mass Spectrom. 1988, 2, 105–109. 22. Syka, J.E.P.; Louris, J.N.; Kelley, P.E.; Stafford, G.C.; Reynolds, W.E. Method of operating ion trap detector in MS/MS mode. U.S. Patent 1988, 4,736,101. 23. Sadygov, R.G.; Cociorva, D.; Yates, J.R. Large-scale database searching using tandem mass spectra: Looking up the answer in the back of the book. Nature Methods 2004, 1, 195–202. 24. LeDuc, R.D.; Taylor, G.K.; Kim, Y.B.; Januszyk, T.E.; Bynum, L.H.; Sola, J.V.; Garavelli, J.S.; Kelleher, N.L. ProSight PTM: An integrated environment for protein identification and characterization by top-down mass spectrometry. Nucleic Acids Research 2004, 32, W340–W345. 25. Perkins, D.N.; Pappin, D.J.C.; Creasy, D.M.; Cottrell, J.S. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999, 20, 3551–3567. 26. Eng, J.K.; McCormack, A.L.; Yates, J.R. An approach to correlate tandem mass-spectral data of peptides with amino-acid-sequences in a protein database. J. Am. Soc. Mass Spectrom. 1994, 5, 976–989. 27. Louris, J.N.; Brodbeltlustig, J.S.; Cooks, R.G.; Glish, G.L.; Vanberkel, G.J.; McLuckey, S.A. Ion isolation and sequential stages of mass-spectrometry in a quadrupole ion trap mass-spectrometer. Int. J. Mass Spectrom. Ion Processes 1990, 96, 117–137. 28. Molina, H.; Matthiesen, R.; Kandasamy, K.; Pandey, A. Comprehensive comparison of collision induced dissociation and electron transfer dissociation. Anal. Chem. 2008, 80, 4825–4835. 29. Zubarev, R.A.; Zubarev, A.R.; Savitski, M.M. Electron capture/transfer versus collisionally activated/induced dissociations: Solo or duet? J. Am. Soc. Mass Spectrom. 2008, 19, 753–761. 30. Good, D.M.; Wirtala, M.; McAlister, G.C.; Coon, J.J. Performance characteristics of electron transfer dissociation mass spectrometry. Mol. Cell. Proteomics 2007, 6, 1942–1951. 31. Swaney, D.L.; McAlister, G.C.; Coon, J.J. Decision tree-driven tandem mass spectrometry for shotgun proteomics. Nature Methods 2008, 5, 959–964. 32. Swaney, D.L.; McAlister, G.C.; Wirtala, M.; Schwartz, J.C.; Syka, J.E.P.; Coon, J.J. Supplemental activation method for high-efficiency electron-transfer dissociation of doubly protonated peptide precursors. Anal. Chem. 2007, 79, 477–485.
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
75
33. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; Whitehouse, C.M. Electrospray ionization for mass-spectrometry of large biomolecules. Science 1989, 246, 64–71. 34. Stephenson, J.L.; McLuckey, S.A. Adaptation of the Paul Trap for study of the reaction of multiply charged cations with singly charged anions. Int. J. Mass Spectrom. Ion Processes 1997, 162, 89–106. 35. Stephenson, J.L.; McLuckey, S.A. Ion/ion reactions in the gas phase: Proton transfer reactions involving multiply-charged proteins. J. Am. Chem. Soc. 1996, 118 (31), 7390–7397. 36. Stephenson, J.L.; McLuckey, S.A. Simplification of product ion spectra derived from multiply charged parent ions via ion/ion chemistry. Anal. Chem. 1998, 70, 3533–3544. 37. Stephenson, J.L.; Van Berkel, G.J.; McLuckey, S.A. Ion-ion proton transfer reactions of bio-ions involving noncovalent interactions: Holomyoglobin. J. Am. Soc. Mass Spectrom. 1997, 8, 637–644. 38. Frey, B.L.; Lin, Y.; Westphall, M.S.; Smith, L.M. Controlling gas-phase reactions for efficient charge reduction electrospray mass spectrometry of intact proteins. J. Am. Soc. Mass Spectrom. 2005, 16, 1876–1887. 39. Ebeling, D.D.; Westphall, M.S.; Scalf, M.; Smith, L.M. Corona discharge in charge reduction electrospray mass spectrometry. Anal. Chem. 2000, 72, 5158–5161. 40. Scalf, M.; Westphall, M.S.; Smith, L.M. Charge reduction electrospray mass spectrometry. Anal. Chem. 2000, 72 (1), 52–60. 41. Scalf, M.; Westphall, M.S.; Krause, J.; Kaufman, S.L.; Smith, L.M. Controlling charge states of large ions. Science 1999, 283, 194–197. 42. Loo, R.R.O.; Udseth, H.R.; Smith, R.D. A new approach for the study of gas-phase ion-ion reactions using electrospray ionization. J. Am. Soc. Mass Spectrom. 1992, 3, 695–705. 43. Loo, R.R.O.; Udseth, H.R.; Smith, R.D. Evidence of charge inversion in the reaction of singly charged anions with multiply charged macroions. J. Phys. Chem. 1991, 95, 6412–6415. 44. Coon, J.J.; Shabanowitz, J.; Hunt, D.F.; Syka, J.E.P. Electron transfer dissociation of peptide anions. J. Am. Soc. Mass Spectrom. 2005, 16, 880–882. 45. Syka, J.E.P.; Coon, J.J.; Schroeder, M.J.; Shabanowitz, J.; Hunt, D.F. Peptide and protein sequence analysis by electron transfer dissociation mass spectrometry. Proc. Natl. Acad. Sci. 2004, 101, 9528–9533. 46. Gunawardena, H.P.; Emory, J.F.; McLuckey, S.A. Phosphopeptide anion characterization via sequential charge inversion and electron-transfer dissociation. Anal. Chem. 2006, 78, 3788–3793. 47. Wells, J.M.; Chrisman, P.A.; McLuckey, S.A. Formation of protein-protein complexes in vacuo. J. Am. Chem. Soc. 2001, 123, 12428–12429. 48. Payne, A.H.; Glish, G.L. Gas-phase ion/ion interactions between peptides or proteins and iron ions in a quadrupole ion trap. Int. J. Mass Spectrom. 2001, 204, 47–54. 49. McAlister, G.C.; Kiessel, S.E.B.; Coon, J.J. In vacuo formation of peptide-metal coordination complexes. Int. J. Mass Spectrom. 2008, 276, 149–152. 50. Pyatkivskyy, Y.; Ryzhov, V. Coupling of ion-molecule reactions with liquid chromatography on a quadrupole ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 2008, 22, 1288–1294. 51. Jarvis, M.J.Y.; Koyanagi, G.K.; Zhao, X.; Covey, T.R.; Bohme, D.K. Scrubbing ions with molecules: Kinetic studies of chemical noise reduction in mass spectrometry using ion-molecule reactions with dimethyl disulfide. Anal. Chem. 2007, 79, 4006–4012. 52. Shoukry, M.M.; Shehata, M.R.; Hamza, M.S.A.; van Eldik, R. Equilibrium, kinetic and solvent effect studies on the reactions of [Ru-III(Hedta)(H2O)] with thiols. Dalton Transactions 2005, 24, 3921–3926.
76
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
53. Custer, T.G.; Kato, S.; Bierbaum, V.M.; Howard, C.J.; Morrison, G.C. Gas-phase kinetics and mechanism of the reactions of protonated hydrazine with carbonyl compounds. Gas-phase hydrazone formation: Kinetics and mechanism. J. Am. Chem. Soc. 2004, 126, 2744–2754. 54. Gronert, S.; Huang, R.; Li, K.H. Gas phase derivatization in peptide analysis I: The utility of trimethyl borate in identifying phosphorylation sites. Int. J. Mass Spectrom. 2004, 231, 179–187. 55. Gao, H.; Petzold, C.J.; Leavell, M.D.; Leary, J.A. Investigation of ion/molecule reactions as a quantification method for phosphorylated positional isomers: An FT-ICR approach. J. Am. Soc. Mass Spectrom. 2003, 14, 916–924. 56. Milman, B.L. Cluster ions of diquat and paraquat in electrospray ionization mass spectra and their collision-induced dissociation spectra. Rapid Commun. Mass Spectrom. 2003, 17, 1344–1349. 57. Leavell, M.D.; Leary, J.A. Probing isomeric differences of phosphorylated carbohydrates through the use of ion/molecule reactions and FT-ICR MS. J. Am. Soc. Mass Spectrom. 2003, 14, 323–331. 58. Gronert, S.; O’Hair, R.A.J. Gas phase reactions of trimethyl borate with phosphates and their non-covalent complexes. J. Am. Soc. Mass Spectrom. 2002, 13, 1088–1098. 59. Leavell, M.D.; Kruppa, G.H.; Leary, J.A. Analysis of phosphate position in hexose monosaccharides using ion-molecule reactions and SORI-CID on an FT-ICR mass spectrometer. Anal. Chem. 2002, 74, 2608–2611. 60. Green, M.K.; Lebrilla, C.B. Ion-molecule reactions as probes of gas-phase structures of peptides and proteins. Mass Spectrom. Rev. 1997, 16, 53–71. 61. Xia, Y.; McLuckey, S.A. Evolution of instrumentation for the study of gas-phase ion/ion chemistry via mass spectrometry. J. Am. Soc. Mass Spectrom. 2008, 19, 173–189. 62. Hogan, J.M.; Pitteri, S.J.; Chrisman, P.A.; McLuckey, S.A. Complementary structural information from a tryptic N-linked glycopeptide via electron transfer ion/ion reactions and collision-induced dissociation. J. Proteome Res. 2005, 4, 628–632. 63. Herron, W.J.; Goeringer, D.E.; McLuckey, S.A. Product ion charge state determination via ion/ion proton transfer reactions. Anal. Chem. 1996, 68, 257–262. 64. Herron, W.J.; Goeringer, D.E.; McLuckey, S.A. Gas-phase electron-transfer reactions from multiply-charged anions to rare-gas cations. J. Am. Chem. Soc. 1995, 117, 11555–11562. 65. Herron, W.J.; Goeringer, D.E.; McLuckey, S.A. Ion-ion reactions in the gas-phase-protontransfer reactions of protonated pyridine with multiply-charged oligonucleotide anions. J. Am. Soc. Mass Spectrom. 1995, 6, 529–532. 66. Berberich, D.W.; Yost, R.A. Negative chemical-ionization in quadrupole ion-trap massspectrometry-effects of applied voltages and reaction-times. J. Am. Soc. Mass Spectrom. 1994, 5, 757–764. 67. McAlister, G.C.; Phanstiel, D.; Good, D.M.; Berggren, W.T.; Coon, J.J. Implementation of electron-transfer dissociation on a hybrid linear ion trap-orbitrap mass spectrometer. Anal. Chem. 2007, 79, 3525–3534. 68. Williams, D.K.; McAlister, G.C.; Good, D.M.; Coon, J.J.; Muddiman, D.C. Dual electrospray ion source for electron-transfer dissociation on a hybrid linear ion traporbitrap mass spectrometer. Anal. Chem. 2007, 79, 7916–7919. 69. Gunawardena, H.P.; McLuckey, S.A. Synthesis of multi-unit protein hetero-complexes in the gas phase via ion-ion chemistry. J. Mass Spectrom. 2004, 39, 630–638. 70. Wells, J.M.; Chrisman, P.A.; McLuckey, S.A. Formation and characterization of proteinprotein complexes in vacuo. J. Am. Chem. Soc. 2003, 125, 7238–7249. 71. Badman, E.R.; Chrisman, P.A.; McLuckey, S.A. A quadrupole ion trap mass spectrometer with three independent ion sources for the study of gas-phase ion/ion reactions. Anal. Chem. 2002, 74, 6237–6243.
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
77
72. Wells, J.M.; Chrisman, P.A.; McLuckey, S.A. “Dueling” ESI: Instrumentation to study ion/ion reactions of electrospray-generated cations and anions. J. Am. Soc. Mass Spectrom. 2002, 13, 614–622. 73. Xia, Y.; Liang, X.R.; McLuckey, S.A. Pulsed dual electrospray ionization for ion/ion reactions. J. Am. Soc. Mass Spectrom. 2005, 16, 1750–1756. 74. Xia, Y.; Liang, X.R.; McLuckey, S.A. Sonic spray as a dual polarity ion source for ion/ ion reactions. Anal. Chem. 2005, 77, 3683–3689. 75. Takats, Z.; Nanita, S.C.; Cooks, R.G.; Schlosser, G.; Vekey, K. Amino acid clusters formed by sonic spray ionization. Anal. Chem. 2003, 75, 1514–1523. 76. Coon, J.J.; Syka, J.E.P.; Shabanowitz, J.; Hunt, D.F. Tandem mass spectrometry for peptide and protein sequence analysis. Biotechniques 2005, 38, 519–523. 77. Coon, J.J.; Ueberheide, B.; Syka, J.E.P.; Dryhurst, D.D.; Ausio, J.; Shabanowitz, J.; Hunt, D.F. Protein identification using sequential ion/ion reactions and tandem mass spectrometry. Proc. Natl. Acad. Sci. 2005, 102, 9463–9468. 78. Hauser, N.J.; Han, H.L.; McLuckey, S.A.; Basile, F. Electron transfer dissociation of peptides generated by microwave D-cleavage digestion of proteins. J. Proteome Res. 2008, 7, 1867–1872. 79. Gunawardena, H.P.; Gorenstein, L.; Erickson, D.E.; Xia, Y.; McLuckey, S.A. Electron transfer dissociation of multiply protonated and fixed charge disulfide linked polypeptides. Int. J. Mass Spectrom. 2007, 265, 130–138. 80. Liang, X.R.; Hager, J.W.; McLuckey, S.A. Transmission mode ion/ion electron-transfer dissociation in a linear ion trap. Anal. Chem. 2007, 79, 3363–3370. 81. Huang, T.Y.; Emory, J.F.; O’Hair, R.A.J.; McLuckey, S.A. Electron-transfer reagent anion formation via electrospray ionization and collision-induced dissociation. Anal. Chem. 2006, 78, 7387–7391. 82. Iavarone, A.T.; Paech, K.; Williams, E.R. Effects of charge state and cationizing agent on the electron capture dissociation of a peptide. Anal. Chem. 2004, 76, 2231–2238. 83. Pitteri, S.J.; Chrisman, P.A.; McLuckey, S.A. Electron-transfer ion/ion reactions of doubly protonated peptides: Effect of elevated bath gas temperature. Anal. Chem. 2005, 77, 5662–5669. 84. Zubarev, R.A.; Horn, D.M.; Fridriksson, E.K.; Kelleher, N.L.; Kruger, N.A.; Lewis, M.A.; Carpenter, B.K.; McLafferty, F.W. Electron capture dissociation for structural characterization of multiply charged protein cations. Anal. Chem. 2000, 72, 563–573. 85. Good, D.M.; Wirtala, M.; McAlister, G.C.; Coon, J.J. Performance characteristics of electron transfer dissociation mass spectrometry. Mol. Cell. Proteomics 2007, 6, 1942–1951. 86. Olsen, J.V.; Haselmann, K.F.; Nielsen, M.L.; Budnik, B.A.; Nielsen, P.E.; Zubarev, R.A. Comparison of electron capture dissociation and collisionally activated dissociation of polycations of peptide nucleic acids. Rapid Commun. Mass Spectrom. 2001, 15, 969–974. 87. Hakansson, K.; Chalmers, M.J.; Quinn, J.P.; McFarland, M.A.; Hendrickson, C.L.; Marshall, A.G. Combined electron capture and infrared multiphoton dissociation for multistage MS/MS in a Fourier transform ion cyclotron resonance mass spectrometer. Anal. Chem. 2003, 75, 3256–3262. 88. Sze, S.K.; Ge, Y.; Oh, H.B.; McLafferty, F.W. Plasma electron capture characterization of large dissociation for the proteins by top down mass spectrometry. Anal. Chem. 2003, 75, 1599–1603. 89. Shi, S.D.H.; Hemling, M.E.; Carr, S.A.; Horn, D.M.; Lindh, I.; McLafferty, F.W. Phosphopeptide/phosphoprotein mapping by electron capture dissociation mass spectrometry. Anal. Chem. 2001, 73, 19–22. 90. Horn, D.M.; Ge, Y.; McLafferty, F.W. Activated ion electron capture dissociation for mass spectral sequencing of larger (42 kDa) proteins. Anal. Chem. 2000, 72, 4778–4784.
78
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
91. Xia, Y.; Han, H.; McLuckey, S.A. Activation of intact electron-transfer products of polypeptides and proteins in cation transmission mode ion/ion reactions. Anal. Chem. 2008, 80, 1111–1117. 92. Han, H.L.; Xia, Y.; McLuckey, S.A. Beam-type collisional activation of polypeptide cations that survive ion/ion electron transfer. Rapid Commun. Mass Spectrom. 2007, 21, 1567–1573. 93. Watson, D.J.; McLuckey, S.A. Charge state dependent ion trap collision-induced dissociation of reduced bovine and porcine trypsin cations. Int. J. Mass Spectrom. 2006, 255, 53–64. 94. Pitteri, S.J.; Chrisman, P.A.; Badman, E.R.; McLuckey, S.A. Charge-state dependent dissociation of a trypsin/inhibitor complex via ion trap collisional activation. Int. J. Mass Spectrom. 2006, 253, 147–155. 95. Wells, J.M.; McLuckey, S.A. Collision-induced dissociation (CID) of peptides and proteins. Biol. Mass Spectrom. 2005, 402, 148–185. 96. Liu, J.; Chrisman, P.A.; Erickson, D.E.; McLuckey, S.A. Relative information content and top-down proteomics by mass spectrometry: Utility of ion/ion proton-transfer reactions in electrospray-based approaches. Anal. Chem. 2007, 79, 1073–1081. 97. Chrisman, P.A.; Pitteri, S.J.; McLuckey, S.A. Parallel ion parking of protein mixtures. Anal. Chem. 2006, 78, 310–316. 98. Chrisman, P.A.; Pitteri, S.J.; McLuckey, S.A. Parallel ion parking: Improving conversion of parents to first-generation products in electron transfer dissociation. Anal. Chem. 2005, 77, 3411–3414. 99. Goeringer, D.E.; Asano, K.G.; McLuckey, S.A.; Hoekman, D.; Stiller, S.W. Filtered noise field signals for mass-selective accumulation of externally formed ions in a quadrupole ion-trap. Anal. Chem. 1994, 66, 313–318. 100. Grosshans, P.B.; Ostrander, C.M.; Walla, C.A. Methods and apparatus to control charge neutralization reactions in ion traps. U.S. Patent 2003, 6,570,151. 101. Grosshans, P.B.; Ostrander, C.M.; Walla, C.A. Methods and apparatus to control charge neutralization reactions in ion traps. U.S. Patent 2004, 6,674,067. 102. McLuckey, S.A.; Reid, G.E.; Wells, J.M. Ion parking during ion/ion reactions in electrodynamic ion traps. Anal. Chem. 2002, 74, 336–346. 103. Schwartz, J.C.; Syka, J.E.P.; Jardine, I. High-resolution on a quadrupole ion trap massspectrometer. J. Am. Soc. Mass Spectrom. 1991, 2, 198–204. 104. Stephenson, J.L.; McLuckey, S.A. Ion/ion proton transfer reactions for protein mixture analysis. Anal. Chem. 1996, 68, 4026–4032. 105. McLuckey, S.A.; Stephenson, J.L.; Goeringer, D.E. Gas-phase bio-ion/ion reactions: Charge transfer and ion pairing. Abstr. Pap. Am. Chem. Soc. 1996, 212, 17.I 106. Phanstiel, D.; Brumbaugh, J.; Berggren, W.T.; Conard, K.; Feng, X.; Levenstein, M.E.; McAlister, G.C.; Thomson, J.A.; Coon, J.J. Mass spectrometry identifies and quantifies 74 unique histone H4 isoforms in differentiating human embryonic stem cells. Proc. Natl. Acad. Sci. 2008, 105, 4093–4098. 107. Bowers, J.J.; Liu, J.; Gunawardena, H.P.; McLuckey, S.A. Protein identification via iontrap collision-induced dissociation and examination of low-mass product ions. J. Mass Spectrom. 2008, 43, 23–34. 108. Newton, K.A.; Amunugama, R.; McLuckey, S.A. Gas-phase ion/ion reactions of multiply protonated polypeptides with metal containing anions. J. Phys. Chem. A 2005, 109, 3608–3616. 109. Newton, K.A.; McLuckey, S.A. Generation and manipulation of sodium cationized peptides in the gas phase. J. Am. Soc. Mass Spectrom. 2004, 15 (4), 607–615. 110. He, M.; Emory, J.F.; McLuckey, S.A. Reagent anions for charge inversion of polypeptide/protein cations in the gas phase. Anal. Chem. 2005, 77, 3173–3182.
Methods for Multi-Stage Ion Processing Involving Ion/Ion Chemistry
79
111. He, M.; McLuckey, S.A. Increasing the negative charge of a macroanion in the gas phase via sequential charge inversion reactions. Anal. Chem. 2004, 76, 4189–4192. 112. He, M.; McLuckey, S.A. Two ion/ion charge inversion steps to form a doubly protonated peptide from a singly protonated peptide in the gas phase. J. Am. Chem. Soc. 2003, 125, 7756–7757. 113. Han, H.L.; Xia, Y.; McLuckey, S.A. Ion trap collisional activation of c and z(center dot) ions formed via gas-phase ion/ion electron-transfer dissociation. J. Proteome Res. 2007, 6, 3062–3069.
Part II Ion Conformation and Structure
Derivatization 4 Chemical and Multistage Tandem Mass Spectrometry for Protein Structural Characterization Jennifer M. Froelich, Yali Lu, and Gavin E. Reid Contents 4.1 Introduction.....................................................................................................84 4.2 Protein Identification and Characterization..................................................... 85 4.2.1 Analytical Strategies to Overcome the Mixture Complexity and Dynamic Range Limitations Associated with Proteome Analysis............................................................................... 85 4.2.1.1 Protein and Peptide Separation............................................. 85 4.2.1.2 Protein and Peptide Depletion.............................................. 85 4.2.1.3 Protein and Peptide Enrichment and ‘Targeted’ Analysis................................................................................. 86 4.2.2 Challenges Associated with the Application of Tandem Mass Spectrometry Strategies for Peptide Sequence Analysis and Characterization.................................................................................. 87 4.2.2.1 Chemical Derivatization Strategies to Direct the Fragmentation Reactions of Protonated Peptides Toward the Formation of ‘Sequence’ Product Ions.............. 88 4.2.2.2 Chemical Derivatization Strategies to Direct the Fragmentation Reactions of Protonated Peptides Toward the Formation of Diagnostic ‘Non-Sequence’ Product Ions.......................................................................... 91 4.3 Quantitative Analysis of Protein Expression................................................... 93 4.3.1 Two-Dimensional Differential Gel Electrophoresis............................ 93 4.3.2 Label-Free Quantitative Analysis........................................................94 4.3.3 Stable-Isotope Labeling Quantitative Analysis...................................94 4.4 Protein Structure, Protein Folding, and Protein–Protein Interactions.......... 101 4.4.1 Cross-linking Strategies Employing Affinity Tags or Stable Isotope Labels.................................................................................... 102 83
84
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
4.4.2 Cross-linking Strategies Employing ‘Solution-phase’ Cleavage Sites.................................................................................... 104 4.4.3 Cross-linking Strategies Employing ‘Gas-phase’ Cleavage Sites................................................................................... 104 4.5 Concluding Remarks..................................................................................... 109 References............................................................................................................... 109
4.1 INTRODUCTION Major goals within the field of proteomics are to identify, to characterize, and to quantify changes in protein expression, as well as to characterize protein–protein interactions, either at a particular time throughout the cell cycle or in response to a particular type of stimulation (for example, disease). The outcome of this research should enable a more complete understanding of the processes which control normal cellular function and the changes in cell regulation that lead to the onset and progression of disease. Due to its speed, sensitivity, and specificity, mass spectrometry (MS) has become one of the leading technologies in the field of proteomics [1]. The development of soft ionization techniques such as electrospray ionization (ESI) [2] and matrix-assisted laser desorption ionization (MALDI) [3], which have enabled large biological molecules to be ionized with minimal fragmentation, as well as significant advances in the bioinformatics tools which are employed for data analysis [4–7], have contributed to the overall success of MS-based proteomics research. Although a variety of instrumentation platforms are currently available, the quadrupole ion trap mass spectrometer offers the significant advantage of being able to perform multistage tandem mass spectrometry (MS n) to obtain detailed structural information for an ion of interest. Thus, aside from their relatively low cost and high sensitivity, the MS n capabilities of the quadrupole ion trap make this type of mass spectrometer particularly well-suited for the identification and structural characterization of biological molecules. The ‘bottom-up’ or ‘shotgun’ tandem mass spectrometric (MS/MS) approach has emerged as one of the dominant methods employed for protein identification, characterization, and quantitative analysis and for the characterization of protein–protein interactions. A typical bottom-up approach involves the enzymatic digestion of either unresolved protein mixtures, or individual proteins that have been resolved by electrophoretic or chromatographic methods. The resultant peptide mixture is fractionated using one or two-dimensional capillary liquid chromatography (LC) and introduced to the mass spectrometer via ESI or MALDI [8,9]. To obtain detailed structural information regarding the amino acid sequence of the peptide, or to identify and to localize modification sites to particular amino acid residues located within the peptide sequence, individual protonated precursor ions are isolated automatically and subjected to MS/MS analysis [10,11]. The identification of each peptide, and the protein from which it was derived, is achieved subsequently by de novo sequencing [10] or by database search algorithms which correlate the experimental product ion mass spectra with
Chemical Derivatization and Multistage Tandem Mass Spectrometry
85
product ion mass spectra generated theoretically for peptides of the same mass contained within a known protein sequence database [4–7]. While the bottom-up approach has proven successful, there are several drawbacks, which limit its comprehensive application. This chapter will provide an overview of strategies that have been developed to address these limitations, with particular emphasis on the role of chemical derivatization strategies for enhancing the capabilities of multistage tandem mass spectrometric methods for targeted protein identification, characterization, and quantitative analysis, and for the characterization of protein–protein interactions.
4.2 PROTEIN IDENTIFICATION AND CHARACTERIZATION 4.2.1 Analytical Strategies to Overcome the Mixture Complexity and Dynamic Range Limitations Associated with Proteome Analysis The increase in sample mixture complexity resulting from proteolytic digestion, and the dynamic range associated with the proteome, present formidable challenges for protein identification and characterization. To address these challenges, a range of analytical strategies, involving the extensive separation, depletion, enrichment, or ‘targeted’ analysis of proteins, or their proteolytically-derived peptides, have been developed recently. 4.2.1.1 Protein and Peptide Separation Off or on-line one or two-dimensional chromatography methods are employed routinely to fractionate extensively intact protein mixtures [12], or proteolyticallyderived peptide mixtures originating from purified proteins [13], prior to their analysis by MS. Numerous other strategies have been used in conjunction with chromatographic separation to increase the number of unique peptide ions selected for analysis by MS/MS, particularly for those present at low abundance. Dynamic exclusion is one such approach, whereby the m/z-values of precursor ions selected previously for fragmentation are placed automatically into an exclusion list for a defined period of time to prevent their re-selection [14–16]. An iterative survey scan approach has been described also, which subjects peptide mixtures to multiple replicate LC-MS/MS analyses [14–16]. In each individual analysis, precursor ions are selected for fragmentation from a narrow m/z-window rather than from the entire m/z-range. In addition, Wang and Li have described recently a strategy by which the m/z-values of peptide ions that are identified positively in an initial LC-MS/MS run are placed into an exclusion list in subsequent runs of the same peptide mixture to prevent these peptides from being re-selected for fragmentation [17]. 4.2.1.2 Protein and Peptide Depletion The identification of low abundance proteins present in complex biological samples (for example, human serum or cerebrospinal fluid) can be improved also by removing
86
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
proteins which are present at high abundance. Traditionally, the removal of such proteins has been achieved using dye–ligand affinity chromatography [18] or antibody-based methods such as immunoaffinity chromatography [19,20]. In an alternative approach, combinatorial ligand library beads have been employed recently to increase the concentration of low abundance proteins while effectively reducing the concentration of high abundance proteins [21–23]. In this approach, a protein mixture is exposed to beads which have been synthesized with a library of diverse ligands. Each individual protein or peptide present within the sample will bind to the ligand exhibiting the strongest intermolecular interaction. Those proteins that are present at high abundance will continue to bind to the beads until a saturation limit is reached. However, low abundance proteins, which are present at concentrations below the saturation limits of the beads, will be bound extensively. Using this approach, the dynamic range of protein concentrations present within a complex biological sample can be reduced significantly. 4.2.1.3 Protein and Peptide Enrichment and ‘Targeted’ Analysis In efforts to decrease sample mixture complexity and to improve dynamic range, numerous ‘targeted’ approaches have been described, which analyze only a subset of the peptides contained within a proteolytically-derived peptide mixture. For example, affinity capture methods have been employed extensively for the enrichment of peptides containing specific post-translational modifications or selected amino acid residues. Immobilized metal-ion affinity chromatography (IMAC) incorporating Fe3 + , Ga3 + , or Al3 + , has been used to isolate phosphorylated peptides [24,25], while the enrichment of histidine-containing peptides has been achieved using IMAC columns loaded with Cu2 + [26]. In an analogous approach, organomercurial agarose beads have been employed to isolate cysteine-containing peptides from a tryptic digest of yeast cell lysates [27]. Metal–oxide affinity chromatography (MOAC) methods utilizing titanium dioxide (TiO2), zirconium dioxide (ZrO2), and aluminum oxide (Al2O3), have been used also for the highly selective enrichment of phosphopeptides as an alternative strategy to IMAC [28–30]. The enrichment of glycosylated peptides prior to mass spectrometric analysis has been achieved using lectin affinity chromatography [31]. As an alternative ‘targeted’ approach, peptide subsets may be enriched via the chemical derivatization of specific functional groups within a peptide (that is, amino acid side chains) followed by isolation using affinity capture, covalent capture, or chromatographic strategies [14,32–48]. Specific examples include the biotinylation of cysteine residues within a peptide and subsequent enrichment using streptavidinaffinity chromatography [14,32,33], or the introduction of a quaternary amine tag to the side chain of cysteine residues followed by their isolation using strong cation exchange (SCX) chromatography [38]. Thiol-specific covalent resins have been employed to enrich for cysteine-containing peptides [41–44]. For example, Wang et al. have used a cysteine covalent capture strategy to characterize the mouse brain proteome [44]. It was reported that, in conjunction with global tryptic digestion, a total of 7792 proteins were identified following LC-MS/MS analysis using a linear quadrupole ion trap with 1564 proteins identified exclusively from the sample which had been subjected to cysteine peptide enrichment.
Chemical Derivatization and Multistage Tandem Mass Spectrometry
87
4.2.2 Challenges Associated with the Application of Tandem Mass Spectrometry Strategies for Peptide Sequence Analysis and Characterization Another challenge facing the identification and characterization of protonated peptide ions is the ability to form a series of b- and y-type ‘sequence’ product ions, via fragmentation of the peptide amide bonds during low-energy collision-induced dissociation (CID) tandem mass spectrometry that are required typically for subsequent database analysis and protein identification. It is well documented that the fragmentation reactions of protonated peptide ions are influenced strongly by both the amino acid composition and charge state of the precursor ion (that is, proton mobility) [49]. For example, the formation of dominant b- or y-type sequence ions resulting from enhanced cleavage at the amide bond N-terminal to proline residues is observed frequently under ‘mobile’ proton conditions (that is, when the total number of ionizing protons exceeds the combined number of arginine, lysine, and histidine residues) [50]. Under ‘non-mobile’ proton conditions, where ionizing protons are ‘sequestered’ at the side chains of basic amino acids (for example, arginine, lysine, or histidine), the formation of dominant b- or y-type sequence ions resulting from enhanced fragmentation at the amide bond C-terminal to aspartic acid residues are observed commonly [51]. Under certain conditions of proton mobility, the presence of several common post-translational or process-induced protein modifications may result in the formation of dominant ‘non-sequence’ neutral-loss product ions, via fragmentations occurring at the modified amino acid side chain [52]. Examples include the neutral loss of CH3SOH (64 Da) from methionine sulfoxide-containing peptides [53], the neutral loss of alkyl sulfenic acid (RSOH) from S-alkyl cysteine sulfoxide-containing peptides [54–57], the neutral loss of H3PO4 (98 Da) from phosphoserine or phosphothreoninecontaining peptides [58–61], the neutral loss of SO3 (80 Da) from O-sulfonoserine, O-sulfonothreonine, or thiosulfate(–S–SO3H) containing peptides [62–64], and the loss of a glycan moiety from O-linked N-acetylgalactosamine-containing peptides [65,66]. The formation of these non-sequence neutral-loss product ions in high relative abundance may limit the amount of sequence information that is available for unambiguous identification of a peptide, or for localization of the modification site to a particular amino acid residue within the peptide sequence. Despite this risk of losing sequence information, the observation of diagnostic non-sequence side chain neutral-loss product ions may also be beneficial, by enabling the identification of these modified peptides in the gas-phase, thereby reducing the sample mixture complexity. Furthermore, observation of these diagnostic non-sequence product ions could potentially improve the specificity of database search analysis strategies by enabling searches against only a subset of the peptides contained within a protein sequence database. When required, additional sequence information can be obtained by automatically subjecting the initial neutral-loss product ion to MS3 analysis in the quadrupole ion trap [25,28,67–69]. Typically, however, only a fraction of the peptide precursor ions containing these modifications will give rise to the diagnostic product ion of interest at sufficiently-high abundance to permit MS3 examination, due to the
88
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
strong proton mobility dependence associated with the mechanisms responsible for these losses [53,57,61]. Thus, it may be desirable to direct the fragmentation reactions of protonated peptides toward the formation of diagnostic product ions, through the use of chemical derivatization approaches. 4.2.2.1 Chemical Derivatization Strategies to Direct the Fragmentation Reactions of Protonated Peptides Toward the Formation of ‘Sequence’ Product Ions In an effort to improve peptide identification and characterization by de novo sequencing or database search algorithm strategies, numerous chemical derivatization approaches have been developed to direct the fragmentation reactions of peptides toward the formation of a series of b- and/or y-type sequence product ions. One such approach involves chemical derivatization of the side chains of basic amino acid residues within a peptide to alter proton affinity [70–73]. For example, the reagents acetylacetone [70] and malondialdehyde [71] have been employed to modify chemically guanidino groups on the side chain of arginine residues in an effort to decrease the proton affinity of these sites. By decreasing the proton affinity of the arginine side chain, an ionizing proton is less likely to be sequestered and is available, therefore, to initiate cleavage of the amide bonds along the peptide backbone. It has been demonstrated that quadrupole ion trap CID-MS/MS of these chemically-modified arginine-containing peptides results in an increased number and intensity of band y-type sequence product ions compared to their non-derivatized counterparts, thereby improving peptide identification [71]. In a similar approach, the ε-amino group of C-terminal lysine residues has been converted to an imidazole derivative via chemical modification with 2-methoxy-4,5-dihydro-1H-imidazole [72,73]. In contrast to producing both b- and y-type ions, the increased proton affinity resulting from this chemical modification results predominantly in the formation of a series of y-type ions, yielding a simplified CID-product ion mass spectrum for interpretation, particularly for de novo peptide sequencing. The chemical derivatization of peptide N- or C-termini to incorporate a fixed positive or negative charge is yet another approach that has been employed extensively to direct peptide ion fragmentation toward the formation of a desired series of sequence product ions [74–84]. For example, it has been shown that the chemical derivatization of peptide N-termini with S-pentafluorophenyl [N-tris(2,4,6-trimethoxyphenyl) phosphonium]acetate bromide (TMPP-AcSC6F5 bromide) [75,76] (Scheme 4.1) or with [tris(2,4,6-trimethoxyphenyl)phosphonium]acetic acid N-hydroxysuccinimide ester (TMPP-Ac-OSu) [77] to form [tris(2,4,6-trimethoxyphenyl)-phosphonium] acetyl (TMPP-Ac) derivatives directs the gas-phase CID fragmentation reactions of protonated peptides toward the formation of a series of a- or b-type sequence product ions. A representative example is shown in Figure 4.1a and b for the doublycharged ([M + 2H]2 + ) underivatized and doubly-charged ([M + + H]2 + ) TMPP-Ac derivatized precursor ions of the tryptic peptide PHPFHFFVYK, which has been subjected to low-energy CID-MS/MS in a quadrupole ion trap [77]. In contrast to the underivatized peptide, N-terminal derivatization to form the quaternary phosphonium ion results predominantly in the formation of a b-type ion series, and, to a lesser extent, a-type ions. Adamczyk et al. have demonstrated that N-terminal b- and
Chemical Derivatization and Multistage Tandem Mass Spectrometry
89
OCH3 +
H3CO
–
P–CH2–CO–SC6F5Br + NH2–Peptide–COOH OCH3
3
DMAP, 15´ Room temperature
OCH3 +
H3CO
P–CH2–CO–NH–Peptide–COOH OCH3
3
(TMPP-Ac-Peptide)
SCHEME 4.1 Reaction of S-pentafluorophenyl [N-tris(2,4,6-trimethoxyphenyl)phosphonium] acetate bromide (TMPP-AcSC6F5 bromide) with the N-terminus of a peptide to form the [tris(2,4,6-trimethoxyphenyl)phosphonium]acetyl (TMPP-Ac) derivative. (Reproduced from Sadagopan, N.; Watson, J.T., J. Am. Soc. Mass Spectrom. 2000, 11, 107–119. With permission from the American Society for Mass Spectrometry, published by Elsevier Inc.)
a-type ions are formed predominately following CID-MS3 analysis of the TMPP-Ac containing b-type ions produced in the first stage of tandem mass spectrometry analysis [77], which can be particularly useful for the analysis of larger peptides, for which incomplete sequence information is obtained frequently following CID-MS/ MS alone. Recently, fixed-charge chemical derivatization of peptide N-termini with (N-succinimidyloxycarbonylmethyl)tris(2,4,6-trimethoxyphenyl)phosponium acetate to form an acetylphosphonium derivative has been used to increase the sequence coverage obtained by electron capture dissociation (ECD) of o-phosphorylated and o-glycosylated peptides [79]. Peptide N-termini have been modified chemically also with 4-sulfophenyl isothiocyanate and various other sulfonic acid derivatives to incorporate a fixed negative charge [81–84]. The introduction of a fixed negative charge on the N-terminus of the peptide, and the localization of an ionizing proton on the side chain of arginine or lysine residues contained within the peptide sequence, results in the formation of a neutral peptide molecule. To analyze the peptide via MS, a second ionizing proton is, therefore, required. Because the arginine or lysine side chains are already protonated, the second ionizing proton is able to move along the peptide backbone and initiate cleavage of the amide bonds. Using this approach, a single series of y-type sequence product ions are generated. Numerous chemical derivatization strategies have been developed also to direct the fragmentation reactions of protonated peptides toward the selective formation of single characteristic sequence-type product ions [85–87]. For example, Summerfield and coworkers have demonstrated that N-terminal derivatization, with phenylisothiocyanate to form the corresponding phenylthiocarbamoyl (PTC) derivative, results in exclusive fragmentation of the amide bond between the first two amino
90
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
y8
100
60
20 0 200 (b)
y82+ b82+ y4
y2 400
y5
600
PFHFVY
b2 PFHFFV
40
[y9–17]2+
Relative abundance
80
y6 b7 y7 b 8
800 m/z
1000
1200
1400
b82+
100
b72+
0 500
700
[b9+18]2+
a7
b5
b2 b92+ a9
2+
2+
a6 2+
20
b62+
2+
40
a82+
60
a52+
Relative abundance
80
900
b4 a4 1000
b5
m/z
b6 1300
b7
b8
1500
1700
1900
FIGURE 4.1 Quadrupole ion trap CID-product ion mass spectra of (a) the doubly-charged, [M + 2H]2 + , tryptic peptide PHPFHFFVYK (m/z 660), and (b) its doubly-charged, [M+ + H]2 + , [tris(2,4,6-trimethoxyphenyl)phosphonium]acetyl (TMPP-Ac) derivative (m/z 946). (Reproduced from Adamczyk, M.; Gebler, J.C.; Wu, J., Rapid Commun. Mass Spectrom. 1999, 13, 1413–1422. With permission from John Wiley & Sons.)
acid residues under mobile proton conditions to generate a b1 ion and the complementary yn–1 ion [85,86]. The incorporation of a fixed-charge TMPP-Ac tag to the N-terminus of aspartic acid-containing peptides has been employed also to promote enhanced cleavage C-terminal to aspartic acid residues [87]. The specific information obtained, regarding the presence and location of an aspartic acid residue within a peptide sequence, has been shown to improve the specificity of database search analysis strategies employed for protein identification [88].
Chemical Derivatization and Multistage Tandem Mass Spectrometry
91
4.2.2.2 Chemical Derivatization Strategies to Direct the Fragmentation Reactions of Protonated Peptides Toward the Formation of Diagnostic ‘Non-sequence’ Product Ions Recently, a fixed-charge chemical derivatization strategy, termed Selective Extraction of Labeled Entities by Charge derivatization and Tandem mass spectrometry (SELECT), has been developed to direct the fragmentation reactions of peptide ions toward the formation of diagnostic non-sequence product ions, so as to improve the capabilities of tandem mass spectrometry for selective peptide identification and characterization [89–91]. In an initial demonstration of this approach, the side chains of methionine residues within a peptide or protein were alkylated with phenacylbromide (BrCH2COC6H5) to yield a fixed-charge sulfonium ion (Scheme 4.2) [89,90]. Under low-energy CID-MS/MS conditions, fragmentation of these sulfonium ioncontaining peptides was directed exclusively toward the site of the fixed-charge resulting in the formation of a single diagnostic product ion, independent of the proton mobility of the precursor ion. A representative example is shown in Figure 4.2 for a tryptic digest of yeast enolase, derivatized with a 1000-fold molar excess of phenacylbromide, subjected to ESI, and then examined by multistage tandem mass spectrometry (MS/MS and MS3) in a linear quadrupole ion trap. The full scan MS spectrum is shown in Figure 4.2a to demonstrate the mixture complexity associated with the sample. The CID product ion mass spectra obtained for the doubly- and triply-charged precursor ions, [M + + H]2 + and [M + + 2H]3 + , respectively, of the side chain phenacyl sulfonium ion fixed-charge derivative of the methionine-containing peptide AAQDSFAAGWGVMVSHR are shown in Figure 4.2b and c, respectively. The observation of a single diagnostic product ion corresponding to the neutral loss of phenacyl methyl sulfide (CH3SCH2COC6H5, 166 Da) enables methioninecontaining peptides to be selectively ‘enriched’ in the gas-phase from within the complex proteolytically-derived peptide mixture. To obtain further structural information regarding the amino acid sequence of the peptide, the initial neutral loss product ions from Figure 4.2b and c were then isolated and subjected to further dissociation via MS3 (Figure 4.2d and e).
H3C
O C
N H
H 3C
S CH2
(i) BrCH2COC6H5
CH2
(ii) MS
CH
C O
[M]
H N
O C
+ S
C
CH2
O
CH2 N H
CH
C
H N
(iii) CID –CH3SCH2COC6H5
[M +nH –CH3SCH2COC6H5](n+1)+
O [M+nH+CH2COC6H5](n+1)+
SCHEME 4.2 Fixed-charge derivatization and selective gas-phase fragmentation of methionine fixed-charge sulfonium ion derivatized peptides. (Reproduced from Amunugama, M.; Roberts, K.D.; Reid, G.E., J. Am. Soc. Mass Spectrom. 2006, 17, 1631–1642. With permission from the American Society for Mass Spectrometry, published by Elsevier Inc.)
92
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a)
369.1
% Relative abundance
100
553.0 379.1 392.5 490.8
400
[M++2H]3+ 637.0 709.1
600
A A Q D S FA A G W G V M V S H R
[M++H]2+ 955.5
815.7
800
(b)
m/z
1200
1400
(c) [M++H–CH3SCH2COC6H5]2+
% Relative abundance
100
[M++2H–CH3SCH2COC6H5]3+
100
[M++H–2H2O]2+ [M++H]2+
[M++2H]3+
400 600 800 1000 1200 1400 1600 1800 (d)
200
400
600
800
1000
(e)
100
% Relative abundance
1000
–H2O
y153+
100
y142+ y152+
y153+ y6 y7 y2
y3 y4
y5
y112+ 2+ y12 2+ y9 –H O y132+ 2
y13 y9 b14 b15 y y8 b 10 y11 b16 10 b12
400 600 800 1000 1200 1400 1600 1800 m/z
b3 y2 b4 b5
200
400
b6
b7 b8
600 m/z
b9
800
b10
1000
FIGURE 4.2 Linear quadrupole ion trap multistage tandem mass spectrometry (MS/MS and MS3) analysis of a reduced and S-carboxyamidomethylated tryptic digest of yeast enolase following sulfonium ion derivatization with phenacylbromide. The mass spectrum obtained by ESI is shown in panel (a). The CID product ion mass spectra of the [M + + H]2 + (m/z 955.5) and [M + + 2H]3 + (m/z 637.0) precursor ions of the methionine side chain phenacyl sulfonium ion fixed-charge derivative of AAQDSFAAGWGVMVSHR are shown in panels (b) and (c), respectively. Panels (d) and (e) show the CID-MS3 product ion mass spectra of the [M + + H –CH3SCH 2COC6H5]2 + (m/z 872.5) and [M + + 2H–CH3SCH 2COC6H5]3 + (m/z 581.7) product ions, respectively. Key: * = –NH 3.
More recently, this strategy has been applied to the selective identification and characterization of cysteine-containing peptides, through the introduction of a fixed-charge on the side chain of cysteine residues via chemical derivatization with the reagent (3-[N-bromoacetamido]propyl)-methylphenacylsulfonium bromide (BAPMPS) [92]. Regardless of the amino acid targeted, the SELECT approach addresses issues associated with the identification and characterization of
Chemical Derivatization and Multistage Tandem Mass Spectrometry
93
protonated peptides by reducing the sample mixture complexity in the gas-phase, thereby improving the dynamic range. Furthermore, this approach has the potential to improve the specificity of current database search analysis strategies, by enabling searches against only a subset of the peptides contained within a protein sequence database (that is, those containing methionine or cysteine). A number of approaches described in the literature have demonstrated also that selective dissociation can be achieved by first generating a radical site within a peptide or protein [93–96]. For example, Ly and Julien described recently an approach whereby reactive tyrosine residues within individual proteins were converted to 3-iodotyrosine under natively-folded conditions [95]. The modified tyrosinecontaining proteins were ionized via ESI, introduced to a linear quadrupole ion trap and, subsequently, subjected to UV photodissociation which resulted in the formation of a radical site on the aromatic ring of the modified tyrosine residues. Following re-isolation and low-energy CID, radical-directed selective cleavage adjacent to the tyrosine residues was observed, resulting in the formation of a-type sequence product ions. Similar results were obtained also for proteins containing exposed histidine residues, which were shown also to be susceptible to iodination. The information generated via this site-specific fragmentation could be used potentially to reduce the computational time associated with database search analysis strategies.
4.3 QUANTITATIVE ANALYSIS OF PROTEIN EXPRESSION In addition to protein identification and characterization, another major goal of proteomics research is to quantify protein expression levels. However, MS is not inherently quantitative. Thus, the intensity of a peptide ion introduced to the mass spectrometer via ESI or MALDI does not reflect necessarily the amount of peptide present in the sample, due to the strong dependence of ionization on the physical and chemical nature of the analyte. To overcome this challenge, numerous quantitative analysis strategies have been developed to measure the differences in protein abundances between two different cellular states of a biological system (for example, normal and diseased cells).
4.3.1 Two-Dimensional Differential Gel Electrophoresis Differential quantitative analysis has been performed previously at the protein level using two-dimensional differential gel electrophoresis (2D DIGE) [97–99]. In this approach, individual protein populations are labeled covalently with structurally similar, but spectrally distinct, fluorophores. The protein populations are then combined and separated by 2D polyacrylamide gel electrophoresis (PAGE). Protein quantitation is achieved via imaging of the gel using different fluorescence excitation wavelengths. To determine the identity of those proteins whose cellular expression levels are either up or down-regulated, gel spots are individually excised, subjected to enzymatic in-gel digestion, and the resultant peptide mixture is analyzed subsequently by MS. Although this technique overcomes many of the disadvantages associated with protein quantitation via traditional 2D PAGE, it still suffers from limited dynamic range (104), as well as a limited ability to resolve proteins with extremes
94
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
of molecular weight and pI (where pI is the pH at which a molecule carries zero net electrical charge). In addition, accurate quantitative analysis is precluded when two or more proteins are present in the same gel spot.
4.3.2 Label-Free Quantitative Analysis The quantitative analysis of protein expression levels from different protein populations has been achieved also using ‘label-free’ MS-based approaches [100–103]. In the label-free approach, control and experimental samples are digested enzymatically and analyzed individually by LC-MS and MS/MS. Protein abundances are then determined by either summing the extracted ion chromatographic peak areas [100,103], by summing the peptide identification scores obtained from database analysis [101], or by summing the MS/MS spectral counts [102,103], for all the peptide ions identified from a single protein. However, in order to achieve accurate quantitation, highly-reproducible LC-MS analysis is required to minimize shifts in retention time and fluctuations in ion signal intensity. Furthermore, protein quantitation may be precluded when either peptides are present at low abundance or isobaric peptide ions elute at the same retention time.
4.3.3 Stable-Isotope Labeling Quantitative Analysis Differential quantitative analysis has been achieved also via the incorporation of differential stable isotope labels between control and experimental samples. To date, the majority of stable isotope-labeling methods have involved either in vivo metabolic labeling [104–106] or in vitro chemical derivatization [32,33,35,41–43,106–108]. In vivo metabolic labeling approaches such as SILAC (Stable Isotope Labeling by Amino acids in Cell culture) [105] incorporate a differential stable isotope label by growing one population of cells in normal media and a second population of cells in media enriched with an isotopically-encoded amino acid. Following protein extraction, the two protein populations are combined and digested enzymatically. Protein expression levels are determined subsequently by MS analysis via comparison of the relative abundances of intact peptide precursor ions derived from the ‘light’ and ‘heavy’ isotopically-labeled samples. In vitro chemical derivatization approaches employed for differential quantitative analysis either label all peptides within a proteolytically-derived peptide mixture (that is, at the N- or C-terminus), or target specific amino acid side chains or posttranslational modifications. A general overview of commonly-employed chemical derivatization strategies is shown in Figure 4.3. A more detailed discussion of each of these chemical derivatization approaches can be found in a review article by Julka and Regnier [106] and the references cited therein. Although numerous chemical derivatization and quantitative analysis strategies have now been described, one of the earliest of these involved use of the isotope-coded affinity tag (ICAT) reagent [32]. The first generation ICAT reagent, designed by Gygi et al., consisted of an iodoacetyl thiol-specific reactive group, a biotin tag, and an oxyethylene linker region which contained either eight hydrogen atoms (light ICAT reagent) or eight deuterium atoms (heavy ICAT reagent) [32]. In this approach, all cysteine residues from
(55) R–CHO
F
Biotin-SS
(17) NO2
SCI
N HN I RNHCOCH 2 OH SH CH3 CH2 CH2 H2NCHCONHCHCONHCHNHCH2CONHCH2CONHCHCONHCHCONHCHCONHCH2COOH CH2 CH2 CH2 COOH CH2 CH2 NH2 (68) (22,57,58) 18 CH3OH O Labeling (68)CH3OH (116) OMe (108,109,110) NH2 HN N CH3O NH2
N R
(88,89) (100) O CH2CHCONH2
FIGURE 4.3 Generic summary of the stable isotope labeling strategies employed currently for MS-based relative protein quantitation. All reactions shown to occur on the amino terminus apply also to the ε-amino group of lysine residues. The numbers in parentheses indicate the references cited in the review article by Julka and Regnier. (Reproduced from Julka, S.; Regnier, F., J. Proteome Res. 2004, 3, 350–363. With permission from the American Chemical Society.)
(51) O2N
(54) RNCS NO2
(13) RCOOOCR
(11,14,18) R–CO–OR
O (16,26,27) RNHCOCH2I
(102)
Chemical Derivatization and Multistage Tandem Mass Spectrometry 95
96
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
control and experimental samples are labeled with the light and heavy ICAT reagent, respectively. After labeling, the two protein populations are combined, subjected to enzymatic digestion, and then the ICAT-labeled peptides are isolated from nonlabeled peptides using avidin affinity chromatography. Relative protein expression levels are then determined by measuring the peak ion signal intensity ratios of peptide pairs in a mass scan, while peptide identification is achieved by MS/MS analysis of individual peptide precursor ions. Using the ICAT approach, protein quantitation and a reduction in sample mixture complexity are achieved simultaneously. Despite the initial success of this approach, a number of disadvantages associated with the first generation ICAT reagent have been noted. For example, the presence of deuterium atoms in the linker region of the heavy ICAT reagent may result in the chromatographic separation of light and heavy ICAT-labeled peptides during reversed-phase chromatography, thereby precluding their accurate quantitation. To ensure that light and heavy ICAT-labeled peptides co-elute, a second generation ICAT reagent was designed to include 13C rather than 2H in the linker region [33]. In addition, fragmentation of the bulky biotin tag has been observed frequently during CID-tandem mass spectrometric analysis, which complicates interpretation of the resultant product ion mass spectra for peptide identification. Thus, an acid-cleavable group, which connects the biotin moiety with the thiol-specific isotope tag, was also incorporated into the second generation ICAT reagent to cleave the biotin moiety from modified peptides [33]. In contrast to the first generation ICAT reagent, the small size of the remaining tag results in minimal fragmentation following CIDtandem mass spectrometry. As an alternative to the solution-phase ICAT approach, a number of solid-phase isotope-labeling strategies have been developed [41–43]. Solid-phase covalent capture methods enable more stringent wash conditions to be employed in an effort to remove non-specifically bound peptides. In addition, peptide labeling and isolation can be achieved in a single step. Zhou et al. have described a method for solid-phase stableisotope labeling of cysteine-containing peptides from Saccharomyces cerevisiae using controlled-pore glass beads containing an o-nitrobenzyl-based photocleavable linker, a stable isotope tag incorporating either seven hydrogen or seven deuterium atoms, and a thiol-specific iodoacetyl group [41]. Proteins from control and experimental samples were proteolyzed individually and then cysteine-containing peptides were captured by d0- or d7-beads, respectively. The beads were then combined, washed, and cleaved photolytically to release the differentially-labeled cysteine-containing peptides. Following LC-MS/MS analysis with a quadrupole ion trap, it was determined that more proteins could be identified and quantified using the solid-phase isotope-labeling approach than when the samples were prepared using the first generation solution-phase ICAT reagent. A similar solid-phase isotope-labeling approach termed acid-labile isotope-coded extractants (ALICE) has also been developed by Qiu et al. [42]. In this approach, cysteine-containing peptides are captured using a non-biological polymer which has been modified chemically to contain a maleimido thiol-reactive group and an acid-labile linker in both heavy and light isotope-encoded forms. For the majority of stable-isotope labeling approaches, including those described above, quantitative analysis is achieved in a mass scan. Limitations are encountered,
Chemical Derivatization and Multistage Tandem Mass Spectrometry
97
however, when one or both of the differentially-labeled peptide ions are present at or below the level of chemical noise in the mass spectrum, thereby precluding quantitative analysis. Furthermore, accurate quantitation may be precluded when the m/z values of differentially-labeled peptide ions overlap with non-labeled or other labeled components present in the peptide mixture. To address these challenges, several differential quantitative analysis strategies have been described recently whereby quantitation is performed by tandem mass spectrometry, rather than by a mass scan [109–115]. The iTRAQ approach, which utilizes a commercially-available multiplexed set of reagents to quantitate relative expression levels for multiple protein populations, is one example [112,113]. The initial reagent employed in this approach consists of a reporter group, a balance group, and an amine specific peptide reactive group (Figure 4.4a). Differential stable isotope labels are incorporated into the balance and reporter groups of four iTRAQ reagents in such a way that the tag generated upon reaction with a peptide has the same overall mass ( + 145.1 Da) (Figure 4.4b). To use this approach for quantitative analysis, peptide mixtures are labeled individually with one member of the multiplexed set after which the labeled peptide mixtures are combined and subjected to mass spectrometric analysis (Figure 4.4c). In a mass scan, identical tagged peptides from each of the four samples are present at the same m/z value, therefore the sensitivity in a mass scan is maximized. Upon CID-tandem mass spectrometric analysis of the peptide precursor ions, the balance group is lost as a neutral, while the reporter group retains a charge to generate low m/z product ions at m/z 114, 115, 116, or 117, which are used subsequently for quantitative analysis. The b- and y-type ions that may be generated also during the MS/MS experiment remain isobaric and can be used to identify the sequence of the labeled peptide. Compared to conventional MS-based approaches, increased sensitivity and greater specificity is achieved due to the reduction in chemical noise associated with the MS/MS technique. To increase the number of protein populations, which can be analyzed via this approach, an 8-plex version of the iTRAQ strategy was introduced recently [114]. In a recent publication, Li and Zeng have described a similar tandem mass spectrometric-based quantitative analysis strategy termed Cleavable Isobaric Labeled Affinity Tag or CILAT [115]. Essentially a hybrid of the ICAT and iTRAQ approaches, the CILAT reagent includes an isobaric tag consisting of a reporter group and a balance group. The reagent incorporates also a biotin moiety for the enrichment of modified peptides via avidin affinity chromatography and an acid cleavable linker to remove the biotin moiety prior to MS/MS analysis. The thiol group employed in this reagent is used to modify tyrosine residues within peptides that have been converted to ortho-quinone via oxidation with tyrosinase. Although quantitative measurement strategies based on tandem mass spectrometric analysis overcome the limitations associated with normal mass scanning approaches, the product ions of low m/z required for quantitation are often found to lie below the low mass cut-off introduced during MS/MS in quadrupole ion trap mass spectrometers, thereby precluding their use for quantitative analysis in this type of instrumentation. Thus, to date, the majority of proteomic experiments utilizing these reagents has been performed using quadrupole time-of-flight (QTOF), timeof-flight/time-of-flight (TOF/TOF) and, to a lesser extent, hybrid triple-quadrupole/
98
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a)
(b)
Isobaric tag Total mass = 145
H N
Amine specific peptide reactive group (NHS)
Reporter group mass 114 – 117 (Retain charge)
N
O
Pepti d e O
N
O N N
N O O
Balance group Mass 31–28 (neutral loss)
m/z 114 (+1)
13
m/z 115 (+2)
13
m/z 116 (+3)
13
m/z 117 (+4)
13
13
C
C
C2
O (+3) O (+2)
18
C2 C2
18
15
N
15
N
13
(+1)
C
(+0)
(c) 114
31
NHS + p e p t id e
115 30
NHS + p ep t id e
116 117
29 28
NHS + p ep t id e
114 114 31 -NH-p e p tid e MS/MS 115 30 -NH-p e p tid e 116 29 -NH-p e p tide MS 117 28 -NH-p e p ti d e
Mix
115 116
b P
E
P
T
I
D
E y
117
NHS + p e p t id e Reporter-Balance-Peptide INTACT –4 Samples identical m/z
-Peptide fragments EQUAL -Reporter ions DIFFERENT
FIGURE 4.4 (a) Structure of the iTRAQ reagent that consists of a reporter group, a mass balance group, and a peptide reactive group. The overall mass of the reporter and balance components of the molecule is kept constant (145.1 Da) using differential isotopic enrichment with 13C, 15N, and 18O atoms. (b) Upon reaction with a peptide, the tag forms an amide linkage to any peptide amine (N-terminus or ε-amino group of lysine). When subjected to CID, fragmentation of the tag amide bond results in the loss of the balance group as a neutral species, while the charge is retained by the reporter group fragment. The numbers in parentheses indicate the number of enriched centers in either the reporter group or balance group of the molecule. (c) A mixture of four identical peptides, each labeled with one member of the multiplex set, appears as a single, unresolved precursor ion in a mass scan (identical m/z). Following CID, the four reporter group ions appear as distinct masses (m/z 114–117). All other sequence-informative fragment ions (that is, b- and y-type ions) remain isobaric. (Reproduced from Ross, P.L.; Huang, Y.N.; Marchese, J.N.; Williamson, B.; Parker, K.; Hattan, S.; Khainovski, N.; Pillai, S.; Dey, S.; Daniels, S.; Purkayastha, S.; Juhasz, P.; Martin, S.; Bartlet-Jones, M.; He, F.; Jacobson, A.; Pappin, D.J., Mol. Cell. Proteomics. 2004, 3, 1154– 1169. With permission of the American Society for Biochemistry and Molecular Biology.)
linear ion trap mass spectrometers [116–120]. However, a new ‘High Amplitude Short Time Excitation’ (HASTE) dissociation technique (an analogous method is termed Pulsed Q Collision-Induced Dissociation (PQD) in the commercially-available ion trap mass spectrometer platforms available from Thermo Scientific), has been implemented recently which enables the product ions of low m/z, excluded normally from product ion mass spectra, to be observed [121]. Using the iTRAQ approach coupled with PQD-MS/MS on a Thermo linear quadrupole ion trap, Meany et al. were able to quantify successfully carbonylated proteins enriched from rat skeletal
Chemical Derivatization and Multistage Tandem Mass Spectrometry
99
muscle mitochondria [122]. The PQD-MS/MS product ion mass spectrum of the triply-protonated peptide IEGTPLEAMQKK, derived from the carbonylated protein NADH dehydrogenase 1 beta sub-complex 3, is shown in Figure 4.5a as a representative example [122]. An expansion of the low m/z region of the product ion mass spectrum depicted in Figure 4.5a shows the iTRAQ reporter ions at m/z 114–117 (a)
563.3
100 90 Relative abundance
80 70 60
291.3
50 40 30 20
617.4 447.9 674.4 527.3
894.6 823.6 750.5
145.0 116.1
10 0
200
(b)
400
Control
100
600 m/z 1 115.1
2 116.1
90
800
1000
1200
3 117.1
80 70 60 50 40 30 20 10 0
113
114
115
116
117
118
119
FIGURE 4.5 (a) Linear quadrupole ion trap Pulsed Q Collision-Induced Dissociation (PQD)-tandem mass spectrometric analysis of the [M + 3H]3 + precursor ion of the peptide IEGTPLEAMQKK derived from the carbonylated protein NADH dehydrogenase (ubiquinone) 1 beta sub-complex 3. (b) An expansion of the low m/z region of the product ion mass spectrum in panel A shows the relative abundance of the iTRAQ reporter ions at m/z 114–117. The peptide showed increased abundance in the samples that had been labeled with biotin hydrazide prior to avidin enrichment (iTRAQ reporter ions m/z 115–117) as compared to the control sample that had not been labeled with biotin hydrazide prior to avidin purification (iTRAQ reporter ion m/z 114). (Reproduced from Meany, D.L.; Xie, H.; Thompson, L.V.; Arriaga, E.A.; Griffin, T.J. Proteomics. 2007, 7, 1150–1163. With permission from Wiley-VCH Verlag GmbH.)
100
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
which were used for quantitative analysis (Figure 4.5b). Griffin et al. also compared PQD-MS/MS in a linear quadupole ion trap mass spectrometer with MS/MS in a QTOF mass spectrometer for the quantitative analysis of iTRAQ-labeled peptides derived from a standard yeast lysate mixture [123]. It was determined that similar quantitative accuracy could be achieved using both instrumentation platforms, although optimization of the amplitude of the collision energy was required for efficient fragmentation by PQD. Protein kinases, extracted from cells that had been incubated with various drugs, have been quantified also using the iTRAQ approach and PQD-MS/MS in a linear quadrupole-Orbitrap mass spectrometer [124]. It has been demonstrated also that the iTRAQ tandem mass spectrometric quantitative analysis strategy can be used in conjunction with the quadrupole ion trap by performing multiple stages of mass analysis (that is, MS3) [125]. For example, chemical derivatization with the iTRAQ reagent not only labels the N-terminus of a peptide, but the lysine side chain also. Thus, tryptic peptides with a modified lysine residue present at the C-terminus will produce a y1 product ion at m/z 291 following CID-tandem mass spectrometry. To generate the low m/z iTRAQ reporter ions required for quantitation, the y1 product ion is isolated and subjected to datadependent CID-MS3. Using this approach, peptide identification is achieved in the MS/MS scan, while quantitation is achieved via MS3. Regardless of the instrumentation platform employed for analysis, each of the quantitative tandem mass spectrometric derivatization strategies requires an ionizing proton to initiate cleavage of the stable isotope-containing label in order to generate the low m/z reporter ions required for quantitation. Therefore, the fragmentation reactions associated with these strategies are expected to be highly dependent upon the proton mobility of the precursor ion, such that the characteristic isotopically-encoded low mass reporter ions may often be observed at sufficient levels in only a sub-set of the total peptide ions selected for MS/MS to enable their use for quantification. Another limitation is that the desired fragmentation pathway giving rise to the low m/z reporter ions of interest is typically only one of many dissociation channels, including those resulting in the formation of b- and y-type sequence product ions, thereby ‘diluting’ the mass spectrum and limiting the dynamic range for quantitative analysis. To overcome some of these challenges, the SELECT approach described in Section 4.2.2.2 above has been applied to quantitative analysis, via the incorporation of light and heavy isotopically-encoded labels into the fixed-charge chemical derivatization reagents. In the initial report by Reid et al., methionine-containing peptides were alkylated with either ‘light’ 1H5-phenacylbromide or ‘heavy’ 2H5-phenacylbromide to form a fixed-charge sulfonium ion on the side chain of methionine residues [89]. The light and heavy phenacyl sulfonium ion fixed-charge-containing peptides were combined and subjected subsequently to LC-MS/MS analysis. The relative abundances of the neutral loss product ions generated by CID-MS/MS, formed via the loss of CH3SCH2COC6H5 (166 Da) and CH3SCH2COC62H5 (171 Da), respectively, were used for subsequent quantitative analysis. In contrast to the tandem mass spectrometric quantitative analysis strategies described above, the neutral loss product ions required for quantitation via the SELECT approach are formed independently of the proton mobility of the precursor ion, such that these product ions are observed under all experimental conditions. Furthermore, formation of the neutral loss product
Chemical Derivatization and Multistage Tandem Mass Spectrometry
101
ions at a m/z value close to that of the selected precursor ion circumvents the lowmass cut-off limitation of the ion trap. Characterization of the identified and quantified methionine peptides may be achieved readily by subjecting the common neutral loss product ion formed from the light or heavy labeled peptides to MS3 analysis. However, one of the potential disadvantages to this initial approach is that some, albeit limited, chromatographic separation of the 1H5 and 2H5 forms of the fixed-charge sulfonium ion derivatives was observed following reversed-phase chromatography of these peptides. Thus, to ensure that the labeled peptides co-elute during reversedphase chromatographic separation, and to improve the capabilities of this approach, more recent studies have employed 12C6-phenacylbromide and 13C6-phenacylbromide derivatives for fixed-charge chemical derivatization of methionine-containing peptides [126]. Although not investigated to date, light and heavy isotopically-encoded labels could also be incorporated readily into the previously-described alkylating reagent (3-[N-bromoacetamido]propyl)-methylphenacylsulfonium bromide [92] to enable the selective tandem mass spectrometric quantitative analysis of cysteinecontaining peptides.
4.4 PROTEIN STRUCTURE, PROTEIN FOLDING, AND PROTEIN–PROTEIN INTERACTIONS Due to their high sensitivity and rapid analysis capabilities, numerous mass spectrometric approaches have been developed also to study protein structures and the dynamics of protein folding as well as to characterize protein–protein interactions. For example, hydrogen–deuterium exchange combined with MS has been employed extensively and successfully to probe solvent-accessible regions of proteins and to examine protein folding [127]. A variety of covalent-labeling approaches have been used also to modify oxidatively the side chains of solvent-accessible amino acid residues followed by mass spectrometric analysis to determine the site(s) of modification. Examples include the use of hydroxyl radical probes generated by (1) highenergy synchrotron radiolysis of water [128], (2) UV irradiation of hydrogen peroxide [129], (3) modification using electrochemical oxidation [130], and (4) direct hydrogen peroxide oxidation [131]. Numerous chemical derivatization reagents including diethylpyrocarbonate, butanedione, and phenylboronic acid have been used also to modify specific amino acid side chains [132–135]. Chemical cross-linking combined with MS is also a promising approach for studying the low-resolution structure of protein topology as well as for probing protein–protein interactions [136]. Cross-linkers can link covalently interacting regions either within a single protein or between individual subunits of multi-protein complexes. Most importantly, non-covalent protein–protein interactions, which may be transient or dependent on specific physiological conditions, can be captured into long-lived covalent complexes [137]. In a typical bottom-up approach, cross-linked proteins or protein complexes are separated typically by one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and subjected subsequently to enzymatic in-gel digestion, chromatographic fractionation, and mass spectrometric analysis [138]. Ultimately, the assignment of distance constraints, within a single protein or protein
102
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
complex, can be employed to provide information regarding the protein–protein interactions and three-dimensional structures of the proteins or protein complex. However, despite the relative simplicity of cross-linking approaches, the identification of cross-linked products by MS is not straightforward due to the high complexity of the peptide mixtures generated via enzymatic digestion. The well-known term, “looking for a needle in a haystack” is used often to describe the challenges associated with the identification of cross-linked peptides from within a large number of unmodified peptides. In addition, numerous types of cross-linked peptides can be produced as a result of the chemical cross-linking reaction, which can complicate further peptide analysis. For example, dead-end modified peptides (Type 0) are formed when one of the reactive groups of the cross-linker has reacted with a protein while the second reactive group has been hydrolyzed. A type 0 modification does not provide any information related to the distance constraints between two amino acid residues, but may indicate reactive groups exposed on the protein surface. Intra-molecular cross-linked peptides (Type 1) are formed when both reactive groups of the cross-linker have reacted with two amino acid residues on the same peptide chain. Type 1 cross-linked peptides can be used to map the low-resolution three-dimensional structure of proteins. Finally, inter-molecular cross-linked peptides (Type 2) are formed when the reactive groups of the cross-linker have reacted with two amino acid residues on two different peptide chains. When the cross-linked peptides are from two proteins within a protein complex, information about the interacting sites can be obtained. To date, a number of chemical modification techniques have been developed to address the challenges associated with the identification of cross-linked peptides, including the enrichment of cross-linker-containing species by specific affinity tags, the introduction of discriminating properties such as differential isotope labels, or by the introduction of specific ‘solution-phase’ or ‘gas-phase’ cleavage sites, either within the cross-linking reagent itself or within the cross-linked peptides.
4.4.1 Cross-Linking Strategies Employing Affinity Tags or Stable Isotope Labels Specific affinity tags have been incorporated into the cross-linking reagent to enrich for cross-link-containing species prior to mass spectrometric analysis [139–143] (Figure 4.6a). Most commonly, a biotin functional group is incorporated into the cross-linker, followed by affinity purification of the cross-linked peptides using either an avidin affinity column or avidin beads. Differential stable isotope labeling strategies have been employed also to detect cross-linked peptides [144–151]. Using this approach, isotopic labels are incorporated within the protein or peptide, or within the cross-linker itself, to produce a distinctive mass shift or isotopic pattern following mass spectrometric analysis (Figure 4.6b). One approach which incorporates a stable isotope label within the polypeptide chain involves the introduction of two 18O atoms to each of the C-terminal carboxyl groups during proteolytic digestion in 18O enriched water [145]. Thus, inter-molecular cross-linked peptides are distinguished readily by a characteristic mass shift of 8 Da compared with peptides formed upon proteolysis in natural abundance water. However, it is not possible to distinguish
Chemical Derivatization and Multistage Tandem Mass Spectrometry (a)
Affinity labeled cross-linkers
Stable isotope labeled cross-linkers
S
HN O
(b)
103
N H
“Light” “Heavy”
(c) Solution cleavable cross-linkers S
“Light”
S
“Heavy”
FIGURE 4.6 General representation of (a) affinity labeled, (b) stable isotope labeled, and (c) solution cleavable cross-linking strategies used to identify cross-linked peptides.
between dead-end cross-linked peptides, intra-molecular cross-linked peptides, and unmodified peptides, as they will all exhibit the same mass increment of 4 Da following the introduction of two 18O atoms to their C-terminal carboxyl groups during proteolytic digestion in 18O enriched water. Another approach, described by Chen et al., involves reductive di-methylation of primary amino groups within a protein, followed by enzymatic hydrolysis and derivatization of the newly-formed N-termini with a 1:1 (w/w) mixture of 2,4-dinitrofluorobenzene-d 0 /d3 [146]. Due to the incorporation of two dinitrophenyl groups, inter-molecular cross-linked peptides are distinguished by a characteristic 1:2:1 isotope pattern in the mass spectra from the 1:1 isotope pattern of other cross-linked types and non-modified peptides. Also, for visualizing inter-molecular cross-linked peptides, a mixed isotope cross-linking (MIX) strategy was designed by mixing 1:1 uniformly 15N-labeled and unlabeled proteins to form a mixture of homodimers [147]. Molecular ions formed from crosslinked peptides of intermolecular origin are observed as a triplet or quadruplet of labeled peaks containing [15N/15N]/[15N/14N]/[14N/15N]/[14N/14N], while all other peptide species are observed as a doublet of [15N]/[14N] labeled peaks. While the introduction of stable isotope labels on either the protein or the peptide enables the ready identification of inter-molecular cross-linked peptides, intramolecular cross-linked peptides cannot be distinguished by this method. Thus, isotopic labels have been incorporated directly into the cross-linking reagent. It has been demonstrated that by reacting with 1:1 (w/w) mixtures of stable isotope-labeled and non-labeled cross-linking reagents, cross-linked peptides can be detected readily by their distinctive 1:1 isotope pattern [148,149]. Isotope-labeled cross-linkers have been combined with other strategies for further identification of cross-link types [150,151].
104
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
4.4.2 Cross-Linking Strategies Employing ‘Solution-phase’ Cleavage Sites Although isotope-labeling technologies have been employed widely to assist in the identification of cross-linked peptides, this approach still suffers from a number of limitations. For example, in a mass spectrum of high complexity, the characteristic isotopic pattern of cross-linked peptides will be masked either because of the small mass difference between isotopes, or because of the low relative abundances of the cross-linked products. Moreover, the use of only one isotope labeling process is not capable of identifying all the cross-linked types at once. Therefore, to improve the ability to identify and to sequence cross-linked peptides, a range of cleavable cross-linking reagents has been developed. The cleavage reaction can be performed in solution, by hydrolysis [151], or by the use of reducing agents in the case of disulfide containing cross-linking reagents [152], (Figure 4.6c). The thiol-cleavable cross-linking reagent 3,3′-dithio-bis(succinimidylpropionate) (DTSSP) has been applied to a number of proteins and protein complexes [152–154]. After reduction of the cross-linker disulfide bond, intermolecular cross-linked peptides give rise to two separated components, each containing a reduced linker. Intra-molecular cross-links yield two reduced-linker halves on the same peptide, whereas dead-end types only have one present. Different cross-link types are distinguished by their characteristic mass shifts before and after reduction. Another example involves cleavable cross-linkers combined with isotope-coding strategies [151]. The isotopically-labeled cleavable cross-linker is applied with its unlabeled counterpart to provide additional discriminating information because of the distinctive isotope-pairs thus, formed, allowing unambiguous identification of cross-linked products. However, the application of cleavable cross-linkers by a chemical reaction relies on mass spectrometric analysis before and after the cleavage reaction takes place. Thus, the identification of cross-linked products might be complicated when the corresponding signals of reduced cross-link species either are not observed, or they overlap with non-cross-linked peptides.
4.4.3 Cross-Linking Strategies Employing ‘Gas-phase’ Cleavage Sites To simplify the problem, tandem mass spectrometric methods employing low energy CID have been applied recently to the cleavage and identification of crosslink containing peptides in the gas-phase. The cleavage site can be incorporated on the side chain of the cross-linking reagent, resulting in the formation of a characteristic stable product ion or neutral loss upon MS/MS, while maintaining the cross-linked peptide linkage (Figure 4.7a). For example, Back et al. have described the use of a bifunctional lysine reactive cross-linker, N-benzyliminodiacetoylhydroxysuccinimid (BID), which yields a stable benzyl cation under low energy CID conditions, to identify successfully inter- and intra-molecular cross-linked peptides [155]. However, the low m/z-ratio of the benzyl cation can limit the application of BID when quadrupole ion trap instruments are used due to the low mass cut-off inherent to this instrument. In addition, the benzyl cation is observed only as a dominant product from certain precursor ion charge states of the protonated peptide. Bruce and co-workers have introduced a novel cross-linker strategy, termed protein
Chemical Derivatization and Multistage Tandem Mass Spectrometry Gas-phase cleavable cross-linkers
MS/MS
(a)
105
m/z
MS/MS
(b)
m/z
FIGURE 4.7 Tandem mass spectrometric strategies for the targeted identification of crosslinked peptides. Gas-phase cleavable cross-linking reagents can be designed such that the cleavage site is incorporated either (a) into the side chain of the cross-linking reagent or (b) directly into the cross-linker spacer chain.
interaction reporter (PIR), which incorporates an affinity tag, a hydrophilic group, a photocleavable group, and low-energy CID-cleavable bonds [143,156]. Upon tandem mass spectrometry, PIR intermolecular cross-linked peptides fragment at two cleavage sites within the linker, giving rise to a reporter ion and two separated peptides each with an additional fixed mass. However, the spacer-arm chain length of nearly 43 Å for the PIR makes the use of this reagent less informative in determining the specific interaction distances for protein–protein interactions. Alternatively, the cleavage site may be incorporated directly into the cross-linker spacer chain, resulting in cleavage of the cross-link upon MS/MS (Figure 4.7b). The incorporation of a single gas-phase cleavable bond within the linker region enables the use of MSn to determine the amino acid sequence of each peptide following the initial cleavage reaction. For example, Soderblom and Goshe have developed recently a set of single site gas-phase cleavable cross-linking reagents that can be fragmented selectively in the source region of the mass spectrometer (Figure 4.8) [157]. In this approach, the acquisition of a full mass scan is followed by in-source collision-induced dissociation (ISCID) due to the application of an additional potential offset between the skimmer lens and the multipole region of the mass spectrometer. Utilization of ISCID results in cleavage at the aspartyl-prolyl bond within the cross-linking reagent, which is possibly mediated by proton transfer from the aspartyl side chain to the basic amine of the adjacent prolyl residue. Thus, the ISCID mass spectral scan results in the formation of two peptide product ions, each containing a unique modification corresponding to the remaining portion of the cross-linking reagent. The precursor ions
106
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V SuDP Bisuccinimidyl-succinamyl-aspartyl-proline OH
O O
O O
N
N
N O
O
O
H
O O
N O
O
11.2 Å SuDPG Bisuccinimidyl-succinamyl-aspartyl-prolyl-glycine O O N O
OH
O O
N
N O
O
O
H
O
O N
N H
O
O
15.1 Å
FIGURE 4.8 Structures of the collision-induced dissociative cross-linking reagents bisuccinimidyl-succinamyl-aspartyl-proline (SuDP) and bisuccinimidyl-succinamyl-aspartylproline-glycine (SuDPG). Each reagent incorporates a single gas-phase cleavable bond within the linker region for selective fragmentation in the source region of the mass spectrometer. The calculated distance between each of the reactive sites is indicated for each reagent. (Reproduced from Soderblom, E.J.; Goshe, M.B., Anal. Chem. 2006, 78, 8059–8068. With permission from the American Chemical Society.)
from the ISCID mass spectral scan are subjected subsequently to tandem mass spectrometry in order to identify the peptides and to localize the modification introduced by the cross-linking reagent to specific lysine residues within the peptide sequence. Regardless of whether the cleavage site is incorporated on the cross-linker side chain or in the cross-linker spacer chain, it would be desirable for selective fragmentation at these sites to occur prior to cleavage along the peptide backbone. However, the mechanisms responsible for the gas-phase fragmentation reactions that give rise to the product ions of interest within protonated peptide ions (including cross-linked protonated peptide ions) are dependent typically on both the charge state and the amino acid composition of the peptide (that is, proton mobility) [49]. Thus the characteristic product ions required for cross-link identification may be observed only from a subset of the total cross-linked peptide ions that are subjected to dissociation. To improve the identification of cross-linked peptides, and to develop a ‘targeted’ multistage tandem mass spectrometric-based approach for the identification and characterization of protein–protein interactions, a fixed-charge sulfonium ioncontaining amine reactive cross-linking reagent, S-methyl 5,5′-thiodipentanoylhydroxysuccinimide, was synthesized recently, characterized and applied initially to the
Chemical Derivatization and Multistage Tandem Mass Spectrometry
107
analysis of the cross-linked products formed by reaction with various model peptides (Scheme 4.3) [158]. Under low-energy CID conditions, peptide ions containing this cross-linker undergo fragmentation exclusively at the C–S bond adjacent to the fixedcharge, independent of the charge state and amino acid composition of the precursor ion. A representative example of this cross-linking strategy is shown in Figure 4.9 for the homo-dimer of cross-linked neurotensin (pELYENKPRRPYIL) subjected to multistage tandem mass spectrometric analysis (MS/MS and MS3) in a quadrupole ion trap. The full scan mass spectrum obtained following the cross-linking reaction is shown in Figure 4.9a; the m/z-value is indicated for the homo-dimeric intermolecular cross-linked peptide ([2M + 4H + (I–S)]5 + ) (where (I–S) indicates the presence of the cross-link). CID of the [2M + 4H + (I–S)]5 + precursor ion resulted in the formation of two characteristic product ions, ([M + 2H + I]3 + and [M + 2H + S]2 + ), each containing unique modifications (I = + 83 Da; S = + 130 Da, (Scheme 4.3)) corresponding to the portion of the cross-link remaining on the peptide side-chain following cleavage of the sulfonium ion (Figure 4.9b). As shown in Scheme 4.3, formation of these characteristic product ions occurs via cleavage within the ionic cross-linker, which results in separation of the two peptide chains. These initial product ions formed from the MS/MS experiment were subjected to fragmentation via MS3 to obtain further structural information required for the identification of the peptide and the modification site(s) (Figure 4.9c and d). Although not shown here, O N
O
O S
O
O O
N
O
O H2N-Peptide (1)
H2N-Peptide (2)
O Peptide(1)
O
HN
S
Peptide(2)
NH
MS/MS O
O Peptide(1)
N H M+I
S
NH Peptide(2) M+S
SCHEME 4.3 Structure of the ionic cross-linking reagent S-methyl 5,5′-thiodipentanoylhydroxysuccinimide, an intermolecular cross-linked peptide product formed by reaction with this cross-linker, and the selective gas-phase fragmentation reaction of the intermolecular cross-linked peptides.
108
100
400
600
100
% Relative abundance % Relative abundance
O
NH
p ELYENKPRRPYIL
1000
[M+H]+
1200
1400
1600
1800
2000
1200
1400
1600
1800
2000
[M+2H+S]2+ 902.5 [2M+4H+(I–S)]5+
600
800
1000
y9†3+
100
[b12†+H2O]3+
ρ ELYENKPRRPYIL + NH O
†2+ y11†3+ †3+ a12 –H2O y9 2+
y8†3+
200 (d)
S +
[M+2H+I]3+ 586.1
400 (c)
800
MS
NH
O
[2M+4H+(I–S)]5+ [M+2H]2+
% Relative abundance
(b)
p ELYENKPRRPYIL
[M+3H]3+
% Relative abundance
(a)
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
300
y7
400
100
500
y10†*2+ y10†2+
600
b11†*2+
700
y11†2+
b12†*2+
b6†
800
900
1000
[b12†+H2O]2+ y9‡2+ b9‡2+ a9‡2+ ‡2+ y72+y4 y8
400
600
ρ ELYENKPRRPYIL
b12‡2+ * y7
‡2+ y5 a ‡2+ ‡ b13 12 b11‡2+ ‡2+ b6 y11 y7 a11‡2+
800
1000
NH
b8‡
y8‡* b ‡* 9
1200 m/z
O
S
b11‡*
b9‡
1400
1600
1800
2000
FIGURE 4.9 (a) An electrospray ionization mass scan of cross-linked neurotensin (pELYENKPRRPYIL) following reaction with the ionic cross-linking reagent S-methyl 5,5′-thiodipentanoylhydroxy-succinimide. (b) CID product ion mass spectrum of the [2M + 4H + (I–S)]5 + precursor ion, m/z 712.4, of neurotensin containing an intermolecular peptide cross-link. CID-MS3 product ion mass spectra of the (c) [M + 2H + I]3 + and (d) [M + 2H + S]2 + product ions, from panel B. † indicates product ions containing an I-type modification and ‡ indicates product ions containing an S-type modification (see Scheme 4.3).
Chemical Derivatization and Multistage Tandem Mass Spectrometry
109
dead-end cross-linked peptides, present in either their hydrolyzed or unhydrolyzed forms, are identified readily based on the observation of a characteristic neutral loss. Further characterization of these dead-end cross-linked peptides can be achieved, therefore, by isolating the initial neutral loss product ion then subjecting it to MS3. Tandem mass spectrometric analysis of intra-molecular cross-linked peptides in the ion trap results in initial cleavage of the cross-linker without changing the m/z of the precursor ion, which results in immediate further fragmentation of the peptide backbone to yield b- and y-type ions. Thus, detailed structural information can be obtained by direct analysis of the MS/MS product ion spectrum. The use of this ionic cross-linking reagent in conjunction with multistage tandem mass spectrometry therefore enables the different types of peptide cross-links to be readily distinguished. Furthermore, the solubility of the ionic cross-linker and its stability under aqueous conditions suggests that it holds great promise for future studies aimed at the structural analysis of large proteins or protein complexes. As an alternative to the gas-phase cleavage of cross-linking reagents via CID, infrared multiphoton dissociation (IRMPD) can be employed also to facilitate the identification of cross-linked peptides. Recently, Gardner et al. have developed a new IR chromogenic cross-linker (IRCX) that incorporates a phosphate functional group into the cross-linking reagent; the phosphate group has a strong infrared absorption at 10.6 µm [159].* Upon infrared irradiation at 10.6 µm, all peptides that contain this chromogenic cross-linker undergo photodissociation and are distinguished from non-modified peptides by a decrease in ion abundance. The IRCX-containing peptides identified in this manner can be interrogated further by IRMPD, CID, or by both methods. This new approach establishes, therefore, another promising chemical cross-linking pathway for the structural analysis of biological assemblies.
4.5 CONCLUDING REMARKS Chemical derivatization methods provide a useful additional tool for protein structural analysis, particularly when coupled with the multistage tandem mass spectrometric capabilities of modern ion trap mass spectrometers. The objective of this chapter was to provide a brief overview of the chemical derivatization strategies that are employed currently to address the challenges associated with protein identification, characterization, and quantitative analysis as well as for the characterization of protein–protein interactions.
REFERENCES
1. Aebersold, R.; Mann, M. Mass spectrometry-based proteomics. Nature 2003, 422, 198–207. 2. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; Whitehouse, C.M. Electrospray ionization for mass spectrometry of large biomolecules. Science 1989, 246, 64–71. 3. Karas, M.; Hillencamp, F. Laser desorption ionization of proteins with molecular masses exceeding 10,000 Daltons. Anal. Chem. 1988, 60, 2299–2301.
* See also Volume 4, Chapter 20, for further discussion of the strong infrared absorption at 10.6 µm of the phosphate group.
110
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
4. Eng, J.K.; McCormack, A.L.; Yates, J.R. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J. Am. Soc. Mass Spectrom. 1994, 5, 976–989. 5. Perkins, D.N.; Pappin, D.J.; Creasy, D.M.; Cottrell, J.S. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999, 20, 3551–3567. 6. MacCoss, M.J.; Wu, C.C.; Yates, J.R. Probability-based validation of protein identifications using a modified SEQUEST algorithm. Anal. Chem. 2002, 74, 5593–5599. 7. Creasy, D.M.; Cottrell, J.S. Error tolerant searching of uninterpreted tandem mass spectrometry data. Proteomics 2002, 2, 1426–1434. 8. Zugaro, L.M.; Reid, G.E.; Ji, H.; Eddes, J.S.; Murphy, A.C; Burgess, A.W.; Simpson, R.J. Characterization of rat brain stathmin isoforms by two-dimensional gel electrophoresis-matrix assisted laser desorption/ionization and electrospray ionization-ion trap mass spectrometry. Electrophoresis 1998, 19, 867–876. 9. Simpson, R.J.; Connelly, L.M.; Eddes, J.S.; Pereira, J.J.; Moritz, R.L.; Reid, G.E. Proteomic analysis of the human colon carcinoma cell line (LIM 1215): development of a membrane protein database. Electrophoresis 2000, 21, 1707–1732. 10. Hunt, D.F.; Yates, J.R.; Shabanowitz, J.; Winston, S.; Hauer, C.R. Protein sequencing by tandem mass spectrometry. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 6233–6237. 11. Steen, H.; Mann, M. The ABC’s (and XYZ’s) of peptide sequencing. Nature Rev. Mol. Cell Biol. 2004, 5, 699–711. 12. Apffel, A. Multidimensional Chromatography of Intact Proteins. In Purifying Proteins for Proteomics A Laboratory Manual; 1st Ed. Simpson, R.J., Ed. Cold Spring Harbor Press: New York, 2004; pp. 75–100. 13. Washburn, M.P.; Walters, D.; Yates, J.R. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 2001, 19, 242–247. 14. Spahr, C.S.; Susin, S.A.; Bures, E.J.; Robinson, J.H.; Davis, M.T.; McGinley, M.D.; Kroemer, G.; Patterson, S.D. Simplification of complex peptide mixtures for proteomic analysis: reversible biotinylation of cysteinyl peptides. Electrophoresis 2000, 21, 1635–1650. 15. Spahr, C.S.; Davis, M.T.; McGinley, M.D.; Robinson, J.H.; Bures, E.J.; Beierle, J.; Mort, J.; Courchesne, P.L.; Chen, K.; Wahl, R.C.; Yu, W.; Luethy, R.; Patterson, S.D. Towards defining the urinary proteome using liquid chromatography-tandem mass spectrometry I. Profiling an unfractionated tryptic digest. Proteomics 2001, 1, 93–107. 16. Davis, M.T.; Spahr, C.S.; McGinley, M.D.; Robinson, J.H.; Bures, E.J.; Beierle, J.; Mort, J.; Yu, W.; Luethy, R.; Patterson, S.D. Towards defining the urinary proteome using liquid chromatography-tandem mass spectrometry II. Limitations of complex mixture analyses. Proteomics 2001, 1, 108–117. 17. Wang, N.; Li, L. Exploring the precursor ion exclusion feature of liquid chromatography -electrospray ionization quadrupole time-of-flight mass spectrometry for improving protein identification in shotgun proteome analysis. Anal. Chem. 2008, 80, 4696–4710. 18. Ahmed, N.; Barker, G.; Oliva, K.; Garfin, D.; Talmadge, K.; Georgiou, H.; Quinn, M.; Rice, G. An approach to remove albumin for the proteomic analysis of low abundance biomarkers in human serum. Proteomics 2003, 3, 1980–1987. 19. Maccarrone, G.; Mifay, D.; Birg, I.; Rosenhagen, M.; Holsboer, F.; Grimm, R.; Bailey, J.; Zolotarjova, N.; Turck, C.W. Mining the human cerebrospinal fluid proteome by immunodepletion and shotgun mass spectrometry. Electrophoresis 2004, 25, 2402–2412. 20. Björhall, K.; Miliotis, T.; Davidsson, P. Comparison of different depletion strategies for improved resolution in proteomic analysis of human serum samples. Proteomics 2005, 5, 307–317.
Chemical Derivatization and Multistage Tandem Mass Spectrometry
111
21. Thulasiraman, V.; Lin, S.; Gheorghiu, L.; Lathrop, J.; Lomas, L.; Hammond, D.; Boschetti, E. Reduction of the concentration of proteins in biological fluids using a library of combinatorial ligands. Electrophoresis 2005, 26, 3561–3571. 22. Guerrier, L.; Claverol, S.; Fortis, F.; Rinalducci, S.; Timperio, A.M.; Antonioli, P.; Jandrot-Perrus, M.; Boschetti, E.; Righetti, P.G. Exploring the platelet proteome via combinatorial, hexapeptide ligand libraries. J. Proteome Res. 2007, 6, 4290–4303. 23. Shores, K.S.; Udugamasooriya, D.G.; Kodadek, T.; Knapp, D.R. Use of peptide analogue diversity library beads for increased depth of proteomic analysis: application to cerebrospinal fluid. J. Proteome Res. 2008, 7, 1922–1931. 24. Ndassa, Y.M.; Orsi, C.; Marto, J.A. Improved immobilized metal affinity chromatography for large-scale phosphoproteomics applications. J. Proteome Res. 2006, 5, 2789–2799. 25. Lee, J.; Xu, Y.; Chen, Y.; Sprung, R.; Chan Kim, S.; Xie, S.; Zhao, Y. Mitochondrial phosphoproteome revealed by an improved IMAC method and MS/MS/MS. Mol. Cell. Proteomics 2007, 6, 669–676. 26. Ren, D.; Penner, N.A.; Slentz, B.E.; Mirzaei, H.; Regnier, F.E. Evaluating immobilized metal affinity chromatography for the selection of histidine-containing peptides in comparative proteomics. J. Proteome Res. 2003, 2, 321–329. 27. Raftery, M.J. Enrichment by organomercurial agarose and identification of cyscontaining peptides from yeast cell lysates. Anal. Chem. 2008, 80, 3334–3341. 28. Olsen, J.V.; Blagoev, B.; Gnad, F.; Macek, B.; Kumar, C.; Mortensen, P.; Mann, M. Global, in vivo, and site-specific phosphorylation dynamics in signaling networks. Cell 2006, 127, 635–648. 29. Kweon, H.K.; Håkansson, K. Selective zirconium dioxide-based enrichment of phosphorylated peptides for mass spectrometric analysis. Anal. Chem. 2006, 78, 1743–1749. 30. Li, Y.; Xu, X.; Qi, D.; Deng, C.; Yang, P.; Zhang, X. Novel Fe3O4@TiO2 core-shell microspheres for selective enrichment of phosphopeptides in phosphoproteome analysis. J. Proteome Res. 2008, 7, 2526–2538. 31. Qiu, R.Q.; Regnier, F.E. Use of multidimensional lectin affinity chromatography in differential glycoproteomics. Anal. Chem. 2005, 77, 2802–2809. 32. Gygi, S.P.; Rist, B.; Gerber, S.A; Turecek, F.; Gelb, M.H.; Aebersold, R. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 1999, 17, 994–999. 33. Hansen, K.C.; Schmitt-Ulms, G.; Chalkley, R.J.; Hirsch, J.; Baldwin, M.A.; Burlingame, A.L. Mass spectrometric analysis of protein mixtures at low levels using cleavable 13Cisotope-coded affinity tag and multidimensional chromatography. Mol. Cell. Proteomics 2003, 2, 299–314. 34. Bernhard, O.K.; Kapp, E.A.; Simpson, R.J. Enhanced analysis of the mouse plasma proteome using cysteine-containing tryptic glycopeptides. J. Proteome Res. 2007, 6, 987–995. 35. Qian, W.-J.; Goshe, M.B.; Camp II, D.G.; Yu, L.-R.; Tang, K.; Smith, R.D. Phosphoprotein isotope-coded solid-phase tag approach for enrichment and quantitative analysis of phosphopeptides from complex mixtures. Anal. Chem. 2003, 75, 5441–5450. 36. Jalili, P.R.; Ball, H.L. Novel reversible biotinylated probe for the selective enrichment of phosphorylated peptides from complex mixtures. J. Am. Soc. Mass Spectrom. 2008, 19, 741–750. 37. Jalili, P.R.; Sharma, D.; Ball, H.L. Enhancement of ionization efficiency and selective enrichment of phosphorylated peptides from complex protein mixtures using a reversible poly-histidine tag. J. Am. Soc. Mass Spectrom. 2007, 18, 1007–1017. 38. Ren, D.; Julka, S.; Inerowicz, H.D.; Regnier, F.E. Enrichment of cysteine-containing peptides from tryptic digests using a quaternary amine tag. Anal. Chem. 2004, 76, 4522–4530.
112
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
39. Brittain, S.M.; Ficarro, S.B.; Brock, A.; Peters, E.C. Enrichment and analysis of peptide subsets using fluorous affinity tags and mass spectrometry. Nat. Biotechnol. 2005, 23, 463–468. 40. Weinberger, S.R.; Viner, R.I.; Ho, P. Tagless extraction-retentate chromatography: a new global protein digestion strategy for monitoring differential protein expression. Electrophoresis 2002, 23, 3182–3192. 41. Zhou, H.; Ranish, J.A.; Watts, J.D.; Aebersold, R. Quantitative proteome analysis by solid-phase isotope tagging and mass spectrometry. Nat. Biotechnol. 2002, 20, 512–515. 42. Qiu, Y.; Sousa, E.A.; Hewick, R.M.; Wang, J.H. Acid-labile isotope-coded extractants: a class of reagents for quantitative mass spectrometric analysis of complex protein mixtures. Anal. Chem. 2002, 74, 4969–4979. 43. Shi, Y.; Xiang, R.; Horváth, C.; Wilkins, J.A. Quantitative analysis of membrane proteins from breast cancer cell lines and MCF7 using multistep solid phase mass tagging and 2D LC/MS. J. Proteome Res. 2005, 4, 1427–1433. 44. Wang, H.; Qian, W.-J.; Chin, M.H.; Petyuk, V.A.; Barry, R.C.; Liu, T.; Gritsenko, M.A.; Mottaz, H.M.; Moore, R.J.; Camp II, D.G.; Khan, A.H.; Smith, D.J.; Smith, R.D. Characterization of the mouse brain proteome using global proteomic analysis complemented with cysteinyl-peptide enrichment. J. Proteome Res. 2006, 5, 361–369. 45. Lansdell, T.A.; Tepe, J.J. Isolation of phosphopeptides using solid phase enrichment. Tet. Lett. 2004, 45, 91–93. 46. Shen, M.; Guo, L.; Wallace, A.; Fitzner, J.; Eisenman, J.; Jacobson E.; Johnson, R.S. Isolation and isotope labeling of cysteine- and methionine-containing tryptic peptides: application to the study of cell surface proteolysis. Mol. Cell. Proteomics 2003, 2, 315–324. 47. Gevaert, K.; Ghesquière, B.; Staes, A.; Martens, L.; Van Damme, J.; Thomas, G.R.; Vandekerckhove, J. Reversible labeling of cysteine-containing peptides allows their specific chromatographic isolation for non-gel proteome studies. Proteomics 2004, 4, 897–908. 48. Gevaert, K.; Pinxteren, J.; Demol, H.; Hugelier, K.; Staes, A.; Van Damme, J.; Martens, L.; Vandekerckhove, J. Four stage liquid chromatographic selection of methionyl peptides for peptide-centric proteome analysis: the proteome of human multipotent adult progenitor cells. J. Proteome Res. 2006, 5, 1415–1428. 49. Kapp, E.A.; Schütz, F.; Reid, G.E.; Eddes, J.S.; Moritz, R.L.; O’Hair, R.A.J.; Speed, T.P.; Simpson, R.J. Mining a tandem mass spectrometry database to determine the trends and global factors influencing peptide fragmentation. Anal. Chem. 2003, 75, 6251–6264. 50. Breci, L.A.; Tabb, D.L.; Yates III, J.R.,; Wysocki, V.H. Cleavage N-terminal to proline: analysis of a database of peptide tandem mass spectra. Anal. Chem. 2003, 75, 1963–1971. 51. Tsaprailis, G.; Somogyi, A.; Nikolaev, E.N.; Wysocki, V.H. Refining the model for selective cleavage at acidic residues in arginine-containing protonated peptides. Int. J. Mass Spectrom. 2000, 195/196, 467–479. 52. Hung, C.-W.; Schlosser, A.; Wei, J.; Lehmann, W.D. Collision-induced reporter fragmentations for identification of covalently modified peptides. Anal. Bioanal. Chem. 2007, 389, 1003–1016. 53. Reid, G.E.; Roberts, K.D.; Kapp, E.A.; Simpson, R.J. Statistical and mechanistic approaches to understanding the gas-phase fragmentation behavior of methionine sulfoxide containing peptides. J. Proteome Res. 2004, 3, 751–759. 54. Steen, H.; Mann, M. Similarity between condensed phase and gas phase chemistry: fragmentation of peptides containing oxidized cysteine residues and its implications for proteomics. J. Am. Soc. Mass Spectrom. 2001, 12, 228–232.
Chemical Derivatization and Multistage Tandem Mass Spectrometry
113
55. Yagüe, J.; Núñez, A.; Boix, M.; Esteller, M.; Alfonso, P.; Casal, J.I. Oxidation of carboxyamidomethyl cysteine may add complexity to protein identification. Proteomics 2005, 5, 2761–2768. 56. Chowdhury, S.M.; Munske, G.R.; Ronald, R.C.; Bruce, J.E. Evaluation of low energy CID and ECD fragmentation behavior of mono-oxidized thio-ether bonds in peptides. J. Am. Soc. Mass Spectrom. 2007, 18, 493–501. 57. Froelich, J.M.; Reid, G.E. Mechanisms for the proton mobility dependent gas-phase fragmentation reactions of S-alkyl cysteine sulfoxide-containing peptide ions. J. Am. Soc. Mass Spectrom. 2007, 18, 1690–1705. 58. Annan, R.S.; Carr, S.A. Phosphopeptide analysis by matrix-assisted laser desorption time-of-flight mass spectrometry. Anal. Chem. 1996, 68, 3413–3421. 59. DeGnore, J.P.; Qin, J. Fragmentation of phosphopeptides in an ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 1998, 9, 1175–1188. 60. Tholey, A.; Reed, J.; Lehmann, W.D. Electrospray tandem mass spectrometric studies of phosphopeptides and phosphopeptide analogues. J. Mass Spectrom. 1999, 34, 117–123. 61. Palumbo, A.M.; Tepe, J.J.; Reid, G.E. Mechanistic insights into the multistage gas-phase fragmentation behavior of phosphoserine- and phosphothreonine-containing peptides. J. Proteome Res. 2008, 7, 771–779. 62. Medzihradszky, K.F.; Darula, Z.; Perlson, E.; Fainzilber, M.; Chalkley, R.J.; Ball, H.; Greenbaum, D.; Bogyo, M.; Tyson, D.R.; Bradshaw, R.A.; Burlingame, A.L. O-Sulfonation of serine and threonine: mass spectrometric detection and characterization of a new posttranslational modification in diverse proteins throughout the eukaryotes. Mol. Cell. Proteomics 2004, 3, 429–440. 63. Medzihradszky, K.F.; Guan, S.; Maltby, D.A.; Burlingame, A.L. Sulfopeptide fragmentation in electron-capture and electron-transfer dissociation. J. Am. Soc. Mass Spectrom. 2007, 18, 1617–1624. 64. Raftery, M.J. Selective detection of thiosulfate-containing peptides using tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2005, 19, 674–682. 65. Hanisch, F.G.; Green, B.N.; Bateman, R.; Peter-Katalinic, J. Localization of O-glycosylation sites of MUC1 tandem repeats by QTOF ESI mass spectrometry. J. Mass Spectrom. 1998, 33, 358–362. 66. Rademaker, G.J.; Pergantis, S.A.; Blok-Tip, L.; Langridge, J.I.; Kleen, A.; ThomasOates, J. Mass spectrometric determination of the sites of O-glycan attachment with low picomolar sensitivity. Anal. Biochem. 1998, 257, 149–160. 67. Chang, E.J.; Archambault, V.; McLachlin, D.T.; Krutchinsky, A.N.; Chait, B.T. Analysis of protein phosphorylation by hypothesis-driven multi-stage mass spectrometry. Anal. Chem. 2004, 76, 4472–4483. 68. Schroeder, M.J.; Shabanowitz, J.; Schwartz, J.C.; Hunt, D.F.; Coon, J.J. A neutral loss activation method for improved phosphopeptide sequence analysis by quadrupole ion trap mass spectrometry. Anal. Chem. 2004, 76, 3590–3598. 69. Wolschin, F.; Lehmann, U.; Glinski, M.; Weckwerth, W. An integrated strategy for identification and relative quantification of site-specific protein phosphorylation using liquid chromatography coupled to MS2/MS3. Rapid Commun. Mass Spectrom. 2005, 19, 3626–3632. 70. Dikler, S.; Kelly, J.W.; Russell, D.H. Improving mass spectrometric sequencing of arginine-containing peptides by derivatization with acetylacetone. J. Mass Spectrom. 1997, 32, 1337–1349. 71. Foettinger, A.; Leitner, A.; Linder, W. Derivatisation of arginine residues with malondialdehyde for the analysis of peptides and protein digests by LC-ESI-MS/MS. J. Mass Spectrom. 2006, 41, 623–632.
114
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
72. Peters, E.C.; Horn, D.M.; Tully, D.C.; Brock, A. A novel multifunctional labeling reagent for enhanced protein characterization with mass spectrometry. Rapid Commun. Mass. Spectrom. 2001, 15, 2387–2392. 73. Conrotto, P.; Hellman, U. Lys Tag: An easy and robust chemical modification for improved de novo sequencing with a matrix-assisted laser desorption/ionization tandem time-of-flight mass spectrometer. Rapid Commun. Mass. Spectrom. 2008, 22, 1823–1833. 74. Roth, K.D.W.; Huang, Z.-H.; Sadagopan, N.; Watson, J.T. Charge derivatization of peptides for analysis by mass spectrometry. Mass Spectrom. Rev. 1998, 17, 255–274. 75. Huang, Z.-H.; Wu, J.; Roth, K.D.W.; Yang, Y.; Gage, D.A.; Watson, J.T. A picomolescale method for charge derivatization of peptides for sequence analysis by mass spectrometry. Anal. Chem. 1997, 69, 137–144. 76. Sadagopan, N.; Watson, J.T. Investigation of the tris(trimethoxyphenyl)phosphonium acetyl charged derivatives of peptides by electrospray ionization mass spectrometry and tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 2000, 11, 107–119. 77. Adamczyk, M.; Gebler, J.C.; Wu, J. Charge derivatization of peptides to simplify their sequencing with an ion trap mass spectrometer. Rapid Commun. Mass. Spectrom. 1999, 13, 1413–1422. 78. Kuyama, H.; Sonomura, K.; Shima, K.; Nishimura, O.; Tsunasawa, S. An improved method for de novo sequencing of arginine-containing, Nα-tris(2,4,6-trimethoxyphenyl)phosphonium-acetylated peptides. Rapid Commun. Mass. Spectrom. 2008, 22, 2063–2072. 79. Chamot-Rooke, J.; van der Rest, G.; Dalleu, A.; Bay, S.; Lemoine, J. The combination of electron capture dissociation and fixed charge derivatization increases sequence coverage for O-glycosylated and O-phosphorylated peptides. J. Am. Soc. Mass Spectrom. 2007, 18, 1405–1413. 80. Müenchbach, M.; Quadroni, M.; Miotto, G.; James, P. Quantitation and facilitated de novo sequencing of proteins by isotopic N-terminal labeling of peptides with a fragmentation-directing moiety. Anal. Chem. 2000, 72, 4047–4057. 81. Lee, Y.H.; Kim, M.-S.; Choie, W.-S.; Min, H.-K.; Lee, S.-W. Highly informative proteome analysis by combining improved N-terminal sulfonation for de novo peptide sequencing and online capillary reverse-phase liquid chromatography/tandem mass spectrometry. Proteomics 2004, 4, 1684–1694. 82. Keough, T.; Youngquist, R.S.; Lacey, M.P. A method for high-sensitivity peptide sequencing using postsource decay matrix-assisted laser desorption ionization mass spectrometry. Proc. Natl. Acad. Sci. USA. 1999, 96, 7131–7136. 83. Bauer, M.D.; Sun, Y.; Keough, T.; Lacey, M.P. Sequencing of sulfonic acid derivatized peptides by electrospray mass spectrometry. Rapid Commun. Mass. Spectrom. 2000, 14, 924–929. 84. Keough, T.; Lacey, M.P.; Fieno, A.M.; Grant, R.A.; Sun, Y.; Bauer, M.D.; Begley, K.B. Tandem mass spectrometry methods for definitive protein identification in proteomics research. Electrophoresis 2000, 21, 2252–2265. 85. Summerfield, S.G.; Bolgar, M.S.; Gaskell, S.J. Promotion and stabilization of b1 ions in peptide phenylthiocarbamoyl derivatives: analogies with condensed-phase chemistry. J. Mass Spectrom. 1997, 32, 225–231. 86. Summerfield, S.G.; Steen, H.; O’Malley, M.; Gaskell, S.J. Phenyl thiocarbamoyl and related derivatives of peptides: Edman chemistry in the gas phase. Int. J. Mass Spectrom. 1999, 188, 95–103. 87. Gu, C.; Tsaprailis, G.; Breci, L.; Wysocki, V.H. Selective gas-phase cleavage at the peptide bond C-terminal to aspartic acid in fixed-charge derivatives of Asp-containing peptides. Anal. Chem. 2000, 72, 5804–5813.
Chemical Derivatization and Multistage Tandem Mass Spectrometry
115
88. Sullivan, A.G.; Brancia, F.L.; Tyldesley, R.; Bateman, R.; Sidhu, K.; Hubbard, S.J.; Oliver, S.G.; Gaskell, S.J. The exploitation of selective cleavage of singly protonated peptide ions adjacent to aspartic acid residues using a quadrupole orthogonal time-offlight mass spectrometer equipped with a matrix-assisted laser desorption/ionization source. Int. J. Mass Spectrom. 2001, 210/211, 665–676. 89. Reid, G.E.; Roberts, K.D.; Simpson, R.J.; O’Hair, R.A.J. Selective identification and quantitative analysis of methionine-containing peptides by charge derivatization and tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 2005, 16, 1131–1150. 90. Amunugama, M.; Roberts, K.D.; Reid, G.E. Mechanisms for the selective gas-phase fragmentation reactions of methionine side chain fixed charge sulfonium ion containing peptides. J. Am. Soc. Mass Spectrom. 2006, 17, 1631–1642. 91. Sierakowski, J.; Amunugama, M.; Roberts, K.D.; Reid, G.E. Substituent effects on the gas-phase fragmentation reactions of sulfonium ion containing peptides. Rapid Commun. Mass. Spectrom. 2007, 21, 1230–1238. 92. Roberts, K.D.; Reid, G.E. Leaving group effects on the selectivity of the gas-phase fragmentation reactions of side chain fixed-charge-containing peptide ions. J. Mass Spectrom. 2007, 42, 187–198. 93. Chu, I.K.; Rodriguez, C.F.; Hopkinson, A.C.; Siu, K.W.M.; Lau, T.-C. Formation of molecular radical cations of enkephalin derivatives via collision-induced dissociation of electrospray-generated copper(II) complex ions of amines and peptides. J. Am. Soc. Mass Spectrom. 2001, 12, 1114–1119. 94. Yin, H.; Chacon, A.; Porter, N.A.; Yin, H.; Masterson, D.S. Free radical-induced sitespecific peptide cleavage in the gas phase: low-energy collision-induced dissociation in ESI- and MALDI mass spectrometry. J. Am. Soc. Mass Spectrom. 2007, 18, 807–816. 95. Ly, T.; Julian, R.R. Residue-specific radical-directed dissociation of whole proteins in the gas-phase. J. Am. Chem. Soc. 2008, 130, 351–358. 96. Diedrich, J.K.; Julian, R.R. Site specific radical directed dissociation of peptides at phosphorylated residues. J. Am. Chem. Soc. 2008, 130, 12212–12213. 97. Unlu, M.; Morgan, M.E.; Minden, J.S. Differential gel electrophoresis: a single gel method for detecting changes in protein extracts. Electrophoresis 1997, 18, 2071–2077. 98. Tonge, R.; Shaw, J.; Middleton, B.; Rowlinson, R.; Rayner, S.; Young, J.; Pognan, F.; Hawkins, E.; Currie, I.; Davison, M. Validation and development of fluorescence twodimensional differential gel electrophoresis proteomics technology. Proteomics 2001, 1, 377–396. 99. Zhou, G.; Li, H.; DeCamp, D.; Chen, S.; Shu, H.; Gong, Y.; Flaig, M.; Gillespie, J.W.; Hu, N.; Taylor, P.R.; Emmert-Buck, M.R.; Liotta, L.A.; Petricoin III, E.F.; Zhao, Y. 2D Differential in-gel electrophoresis for the identification of esophageal scans cell cancerspecific protein markers. Mol. Cell. Proteomics 2002, 1, 117–124. 100. Chelius, D.; Zhang, T.; Wang, G.; Shen, R.-F. Global protein identification and quantification technology using two-dimensional liquid chromatography nanospray mass spectrometry. Anal. Chem. 2003, 75, 6658–6665. 101. Colinge, J.; Chiappe, D.; Lagache, S.; Moniatte, M.; Bougueleret, L. Differential proteomics via probabilistic peptide identification scores. Anal. Chem. 2005, 77, 596–606. 102. Liu, H.; Sadygov, R.G.; Yates III, J.R. A model for random sampling and estimation of relative protein abundance in shotgun proteomics. Anal. Chem. 2004, 76, 4193–4201. 103. Old, W.M.; Meyer-Arendt, K.; Aveline-Wolf, L.; Pierce, K.G.; Mendoza, A.; Sevinsky, J.R.; Resing, K.A.; Ahn, N.G. Comparison of label-free methods for quantifying human proteins by shotgun proteomics. Mol. Cell. Proteomics 2005, 4, 1487–1502. 104. Oda, Y.; Huang, K.; Cross, F.R.; Cowburn, D.; Chait, B.T. Accurate quantitation of protein expression and site-specific phosphorylation. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 6591–6596.
116
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
105. Ong, S.-E.; Blagoev, B.; Kratchmarova, I.; Bach, Kristensen, D.; Steen, H.; Pandey, A.; Mann, M. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell. Proteomics 2002, 1, 376–386. 106. Julka, S.; Regnier, F. Quantification in proteomics through stable isotope coding: a review. J. Proteome Res. 2004, 3, 350–363. 107. Kuhn, K.; Prinz, T.; Schafer, J.; Baumann, C.; Scharfke, M.; Kienle, S.; Schwarz, J.; Steiner, S.; Hamon, C. Protein sequence tags: a novel solution for comparative proteomics. Proteomics 2005, 5, 2364–2368. 108. Simons, B.L.; Wang, G.; Shen, R.-F.; Knepper, M.A. In vacuo isotope coded alkylation technique (IVICAT); an N-terminal stable isotopic label for quantitative liquid chromatography/mass spectrometry proteomics. Rapid Commun. Mass Spectrom. 2006, 20, 2463–2477. 109. Shi, Y.; Yao, X. Oxygen isotopic substitution of peptidyl phosphates for modificationspecific mass spectrometry. Anal. Chem. 2007, 79, 8454–8462. 110. Thompson, A.; Schäfer, J.; Kuhn, K.; Kienle, S.; Schwarz, J.; Schmidt, G.; Neumann, T.; Hamon, C. Tandem mass tags: a novel quantification strategy for comparative analysis of complex protein mixtures by MS/MS. Anal. Chem. 2003, 75, 1895–1904. 111. Dayon, L.; Hainard, A.; Licker, V.; Turck, N.; Kuhn, K.; Hochstrasser, D.F.; Burkhard, P.R.; Sanchez, J.-C. Relative quantification of proteins in human cerebrospinal fluids by MS/MS using 6-plex isobaric tags. Anal. Chem. 2008, 80, 2921–2931. 112. Ross, P.L.; Huang, Y.N.; Marchese, J.N.; Williamson, B.; Parker, K.; Hattan, S.; Khainovski, N.; Pillai, S.; Dey, S.; Daniels, S.; Purkayastha, S.; Juhasz, P.; Martin, S.; Bartlet-Jones, M.; He, F.; Jacobson, A.; Pappin, D.J. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol. Cell. Proteomics 2004, 3, 1154–1169. 113. Sachon, E.; Mohammed, S.; Bache, N.; Jenson, O.N. Phosphopeptide quantitation using amine-reactive tagging reagents and tandem mass spectrometry: application to proteins isolated by gel electrophoresis. Rapid Commun. Mass. Spectrom. 2006, 20, 1127–1134. 114. Choe, L.; D’Ascenzo, M.; Relkin, N.R.; Pappin D.; Ross, P.; Williamson, B.; Guertin, S.; Pribil, P.; Lee, K.H. 8-Plex quantitation of changes in cerebrospinal fluid protein expression in subjects undergoing intravenous immunoglobulin treatment for Alzeimer’s disease. Proteomics 2007, 7, 3651–3660. 115. Li, S.; Zeng, D. CILAT – A new reagent for quantitative proteomics. Chem. Commun. 2007, 21, 2181–2183. 116. Jones, A.M.E.; Bennett, M.H.; Mansfield, J.W.; Grant, M. Analysis of the defence phosphoproteome of Arabidopsis thaliana using differential mass tagging. Proteomics 2006, 6, 4155–4165. 117. Champion, P.A.D.; Stanley, S.A.; Champion, M.M.; Brown, E.J.; Cox, J.S. C-Terminal signal sequence promotes virulence secretion in mycobacterium tuberculosis. Science 2006, 313, 1632–1636. 118. Corvey, C.; Koetter, P.; Beckhaus, T.; Hack, J.; Hofmann, S.; Hampel, M.; Stein, T.; Karas, M.; Entian, K.-D. Carbon source-dependent assembly of the Snf1p kinase complex in Candida albicans. J. Biol. Chem. 2005, 280, 25323–25330. 119. Williamson, B.L.; Marchese, J.; Morrice, N.A. Automated identification and quantification of protein phosphorylation sites by LC-MS on a hybrid triple quadrupole linear ion trap mass spectrometer. Mol. Cell. Proteomics 2006, 5, 337–346. 120. Wolf-Yadlin, A.; Hautaniemi, S.; Lauffenburger, D.A.; White, F.M. Multiple reaction monitoring for robust quantitative proteomic analysis of cellular signaling networks. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5860–5865.
Chemical Derivatization and Multistage Tandem Mass Spectrometry
117
121. Cunningham, C.; Glish, G.L. High amplitude short time excitation: a method to form and detect low mass product ions in a quadrupole ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 2006, 17, 81–84. 122. Meany, D.L.; Xie, H.; Thompson, L.V.; Arriaga, E.A.; Griffin, T.J. Identification of carbonylated proteins from enriched rat skeletal muscle mitochondria using affinity chromatography-stable isotope labeling and tandem mass spectrometry. Proteomics 2007, 7, 1150–1163. 123. Griffin, T.J.; Xie, H.; Bandhakavi, S.; Popko, J.; Mohan, A.; Carlis, J.V.; Higgins, L. iTRAQ reagent-based quantitative proteomic analysis on a linear ion trap mass spectrometer. J. Proteome Res. 2007, 6, 4200–4209. 124. Bantscheff, M.; Eberhard, D.; Abraham, Y.; Bastuck, S.; Boesche, M.; Hobson, S.; Mathieson, T.; Perrin, J.; Raida, M.; Rau, C.; Reader, V.; Sweetman, G.; Bauer, A.; Bouwmeester, T.; Hopf, C.; Kruse, U.; Neubauer, G.; Ramsden, N.; Rick, J.; Kuster, B.; Drewes, G. Quantitative chemical proteomics reveals mechanisms of action of clinical ABL kinase inhibitors. Nat. Biotechnol. 2007, 25, 1035–1044. 125. Chang, B.; Ünlü, M.; Clauser, K.; Carr, S.A. iTRAQ-IT: Implementation of iTRAQ quantitation tags on ion trap instruments via MS3. Proc. 53rd ASMS Conference on Mass Spectrometry and Allied Topics, San Antonio, Texas, 2005. 126. Froelich, J.M.; Kaplinghat, S.; Reid, G.E. Automated neutral loss and data dependent energy resolved “pseudo MS3” for the targeted identification, characterization and quantitative analysis of methionine-containing peptides. Eur. J. Mass Spectrom. 2008, 14, 219–229. 127. Englander, S.W. Hydrogen exchange and mass spectrometry: a historical perspective. J. Am. Soc. Mass Spectrom. 2006, 17, 1481–1489. 128. Kiselar, J.G.; Maleknia, S.D.; Sullivan, M.; Downard, K.M.; Chance, M.R. Hydroxyl radical probe of protein surfaces using synchrotron X-ray radiolysis and mass spectrometry. Int. J. Radiat. Biol. 2002, 78, 101–114. 129. Sharp, J.S.; Becker, J.M.; Hettich, R.L. Analysis of protein solvent accessible surfaces by photochemical oxidation and mass spectrometry. Anal. Chem. 2004, 76, 672–683. 130. McClintock, C.; Kertesz, V.; Hettich, R.L. Development of an electrochemical oxidation method for probing higher order protein structure with mass spectrometry. Anal. Chem. 2008, 80, 3304–3317. 131. Carruthers, N.J.; Stemmer, P.M. Methionine oxidation in the calmodulin-binding domain of calcineurin disrupts calmodulin binding and calcineurin activation. Biochemistry 2008, 47, 3085–3095. 132. Azim-Zadeh, O.; Hillebrecht, A.; Linne, U.; Marahiel, M.A.; Klebe, G.; Lingelbach, K.; Nyalwidhe, J. Use of biotin derivatives to probe conformational changes in proteins. J. Biol. Chem. 2007, 282, 21609–21617. 133. Ladner, C.L.; Turner, R.J.; Edwards, R.A. Development of indole chemistry to label tryptophan residues in protein for determination of tryptophan surface accessibility. Protein Sci. 2007, 16, 1204–1213. 134. Leitner, A.; Linder, W. Functional probing of arginine residues in proteins using mass spectrometry and an arginine-specific covalent tagging concept. Anal. Chem. 2005, 77, 4481–4488. 135. Mendoza, V.L.; Vachet, R.W. Protein surface mapping using diethylpyrocarbonate with mass spectrometric detection. Anal. Chem. 2008, 80, 2895–2904. 136. Sinz, A. Chemical cross-linking and mass spectrometry to map three-dimensional protein structures and protein-protein interactions. Mass Spectrom. Rev. 2006, 25, 663–682. 137. Trakselis, M.A.; Alley, S.C.; Ishmael, F.T. Identification and mapping of protein-protein interactions by a combination of cross-linking, cleavage, and proteomics. Bioconj. Chem. 2005, 16, 741–750.
118
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
138. Young, M.M.; Tang, N.; Hempel, J.C.; Oshiro, C.M.; Taylor, E.W.; Kuntz, I.D.; Gibson, B.W.; Dollinger, D. High throughput protein fold identification by using experimental constraints derived from intramolecular cross-links and mass spectrometry. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 5802–5806. 139. Alley, S.C.; Ishmael, F.T.; Jones, A.D.; Benkovic, S.J. Mapping protein-protein interactions in the bacteriophage T4 DNA polymerase holoenzyme using a novel trifunctional photo-cross-linking and affinity reagent. J. Am. Chem. Soc. 2000, 122, 6126–6127. 140. Trester-Zedlitz, M.; Kamada, K.; Burley, S.K.; Fenyo, D.; Chait, B.T.; Muir, T.W. A modular cross-linking approach for exploring protein iInteractions. J. Am. Chem. Soc. 2003, 125, 2416–2425. 141. Sinz, A.; Kalkhof, S.; Ihling, C. Mapping protein interfaces by a trifunctional crosslinker combined with MALDI-TOF and ESI-FTICR mass spectrometry. J. Am. Soc. Mass Spectrom. 2005, 16, 1921–1931. 142. Ahrends, R.; Kosinski, J.; Kirsch, D.; Manelyte, L.; Giron-Monzon, L.; Hummerich, L.; Schulz, O.; Spengler, B.; Friedhoff, P. Identifying an interaction site between MutH and the C-terminal domain of MutL by crosslinking, affinity purification, chemical coding and mass spectrometry. Nucl. Acids Res. 2006, 34, 3169–3180. 143. Chowdhury, S.M.; Munske, G.R.; Tang, X.; Bruce, J.E. Collisionally activated dissociation and electron capture dissociation of several mass spectrometry-identifiable chemical cross-linkers. Anal. Chem. 2006, 78, 8183–8193. 144. Reynolds, K.J.; Yao, X.; Fenselau, C. Proteolytic 18O labeling for comparative proteomics: evaluation of endoprotease Glu-C as the catalytic agent. J. Proteome Res. 2002, 1, 27–33. 145. Back, J.W.; Notenboom, V.; de Koning, L.J.; Muijsers, A.O.; Sixma, T.K.; de Koster, C.G.; de Jong, L. Identification of cross-linked peptides for protein interaction studies using mass spectrometry and 18O labeling. Anal. Chem. 2002, 74, 4417–4422. 146. Chen, X.; Chen, Y.H.; Anderson, V.E. Protein cross-links: universal isolation and characterization by isotopic derivatization and electrospray ionization mass spectrometry. Anal. Biochem. 1999, 273, 192–203. 147. Taverner, T.; Hall, N.E.; O’Hair, R.A.J.; Simpson, R.J. Characterization of an antagonist interleukin-6 dimer by stable isotope labeling, cross-linking, and mass spectrometry. J. Biol. Chem. 2002, 277, 46487–46492. 148. Müller, D.R.; Schindler, P.; Towbin, H.; Wirth, U.; Voshol, H.; Hoving, S.; Steinmetz, M.O. Isotope-tagged cross-linking reagents. A new tool in mass spectrometric protein interaction analysis. Anal. Chem. 2001, 73, 1927–1934. 149. Pearson, K.M.; Pannell, L.K.; Fales, H.M. Intramolecular cross-linking experiments on cytochrome c and ribonuclease A using an isotope multiplet method. Rapid Commun. Mass Spectrom. 2002, 16, 149–159. 150. Seebacher, J.; Mallick, P.; Zhang, N.; Eddes, J.S.; Aebersold, R.; Gelb, M.H. Protein cross-linking analysis using mass spectrometry, isotope-coded cross-linkers, and integrated computational data processing. J. Proteome Res. 2006, 5, 2270–2282. 151. Petrotchenko, E.V.; Olkhovik, V.K.; Borchers, C.H. Coded cleavable cross-linker for studying protein-protein interaction and protein complexes. Mol. Cell. Proteomics 2005, 4, 1167–1179. 152. Bennett, K.L.; Kussmann, M.; Björk, P.; Godzwon, M.; Mikkelsen, M.; Sørensen, P.; Roepstorff, P. Chemical cross-linking with thiol-cleavable reagents combined with differential mass spectrometric peptide mapping—a novel approach to assess intermolecular protein contacts. Protein Sci. 2000, 9, 1503–1518. 153. Back, J.W.; Sanz, M.A.; de Jong, L.; de Koning, L.J.; Nijtmans, L.G.J.; de Koster, C.G.; Grivell, L.A.; van der Speck, H.; Muijsers, A.O. A structure for the yeast prohibitin complex: structure prediction and evidence from chemical crosslinking and mass spectrometry. Protein Sci. 2002, 11, 2471–2478.
Chemical Derivatization and Multistage Tandem Mass Spectrometry
119
154. Peterson, J.J.; Young, M.M.; Takemoto, L.J. Probing α-crystallin structure using chemical cross-linkers and mass spectrometry. Molecul. Vision 2004, 10, 857–866. 155. Back, J.W.; Hartog, A.F.; Dekker, H.L.; Muijsers, A.O.; de Koning, L.J.; de Jong, L. A new cross-linker for mass spectrometric analysis of the quaternary structure of protein complexes. J. Am. Soc. Mass Spectrom. 2001, 12, 222–227. 156. Tang, X.; Munske, G.R.; Siems, W.F.; Bruce, J.E. Mass spectrometry identifiable cross-linking strategy for studying protein-protein interactions. Anal. Chem. 2005, 77, 311–318. 157. Soderblom, E.J.; Goshe, M.B. Collision-induced dissociative chemical cross-linking reagents and methodology: applications to protein structural characterization using tandem mass spectrometry analysis. Anal. Chem. 2006, 78, 8059–8068. 158. Lu, Y.; Tanasova, M.; Borhan, B.; Reid, G.E. An ionic reagent for controlling the gas-phase fragmentation reactions of cross-linked peptides. Anal. Chem. 2008, 80, 9279–9287. 159. Gardner, M.W.; Vasicek, L.A.; Shabbir, S.; Anslyn, E.V.; Brodbelt, J.S. Chromogenic cross-linker for the characterization of protein structure by infrared multiphoton dissociation mass spectrometry. Anal. Chem. 2008, 80, 4807–4819.
Transform Ion 5 Fourier Cyclotron Resonance Mass Spectrometry in the Analysis of Peptides and Proteins Helen J. Cooper Contents 5.1 Fourier Transform Ion Cyclotron Resonance (FT-ICR)................................ 122 5.1.1 Principles of Fourier Transform Ion Cyclotron Resonance (FT-ICR)............................................................................................ 123 5.1.1.1 Ion Motion........................................................................... 123 5.1.1.2 Excitation and Detection..................................................... 125 5.1.2 Instrumentation.................................................................................. 126 5.1.2.1 Magnet................................................................................ 126 5.1.2.2 Ionization............................................................................ 127 5.1.2.3 Ion Transfer......................................................................... 128 5.1.2.4 Ion Cyclotron Resonance (ICR) Cell.................................. 128 5.1.3 Features of Fourier Transform Ion Cyclotron Resonance (FT-ICR) Mass Spectrometry............................................................ 128 5.1.3.1 Mass Accuracy.................................................................... 128 5.1.3.2 Resolving Power................................................................. 129 5.1.3.3 Sensitivity............................................................................ 130 5.2 Tandem Mass Spectrometry (MS/MS) in Fourier Transform Ion Cyclotron Resonance (FT-ICR)..................................................................... 130 5.2.1 Precursor Ion Isolation....................................................................... 131 5.2.2 Sustained Off-Resonance Irradiation Collision-Induced Dissociation (SORI-CID).................................................................. 132 5.2.2.1 Principles of Sustained Off-Resonance Irradiation Collision-Induced Dissociation (SORI-CID)...................... 132 5.2.2.2 Sustained Off-Resonance Irradiation Collision-Induced Dissociation (SORI-CID) of Peptides and Proteins.............133
121
122
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
5.2.3 Infrared Multiphoton Dissociation (IRMPD).................................... 133 5.2.3.1 Principles of Infrared Multiphoton Dissociation (IRMPD)............................................................................. 133 5.2.3.2 Infrared Multiphoton Dissociation (IRMPD) of Peptides and Proteins.......................................................... 133 5.2.4 Blackbody Infrared Radiative Dissociation (BIRD)......................... 134 5.2.4.1 Principles of Blackbody Infrared Radiative Dissociation (BIRD)........................................................... 134 5.2.4.2 Blackbody Infrared Radiative Dissociation (BIRD) of Peptides and Proteins.......................................................... 134 5.2.5 Electron Capture Dissociation (ECD)............................................... 135 5.2.5.1 Principles of Electron Capture Dissociation (ECD)........... 135 5.2.5.2 Electron Capture Dissociation (ECD) of Peptides and Proteins............................................................................... 135 5.2.5.3 Activated Ion Electron Capture Dissociation (AI-ECD)........................................................................... 137 5.3 Hybrid Fourier Transform Ion Cyclotron Resonance (FT-ICR) Instruments.................................................................................................... 138 5.3.1 Quadrupole-Fourier Transform Ion Cyclotron Resonance (FT-ICR)............................................................................................ 138 5.3.2 Linear Ion Trap-Fourier Transform Ion Cyclotron Resonance (FT-ICR)............................................................................................ 139 5.4 Applications of Fourier Transform Ion Cyclotron Resonance (FT-ICR) in Proteomics................................................................................................. 139 5.4.1 ‘Bottom-Up’ Approaches................................................................... 139 5.4.1.1 Peptide Mass Fingerprinting............................................... 139 5.4.1.2 Peptide Sequencing............................................................. 140 5.4.2 ‘Top-Down’ Approaches.................................................................... 143 5.5 Summary....................................................................................................... 144 References............................................................................................................... 145
5.1 FOURIER TRANSFORM ION CYCLOTRON RESONANCE (FT-ICR) Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry [1,2] has developed due to the fact that a charged particle in a uniform magnetic field will undergo cyclotron motion, that is, will describe a circular path, perpendicular to the direction of the magnetic field, and the frequency of that motion is inversely proportional to its mass-to-charge ratio. The technique of FT-ICR mass spectrometry offers the highest resolution and mass accuracy of all mass analyzers, making it ideal for the characterization of peptides and proteins. This chapter provides an overview of FT-ICR mass spectrometry and its applications in structural characterization of peptides and proteins. The principles of FT-ICR (ion motion, excitation/detection, and instrumental considerations) are discussed together with the features of FT-ICR that make it so suitable for peptide/protein analysis. Tandem mass spectrometry (MS/MS) techniques (sustained off-resonance irradiation collision-induced dissociation (SORI-CID), infrared
123
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
multiphoton dissociation (IRMPD), black body infrared radiative dissociation (BIRD), and electron capture dissociation (ECD) for the sequencing of peptides/proteins are described; see Section 5.2 for an explanation of these acronyms. The new generation of hybrid FT-ICR instruments are reviewed. Finally, the chapter includes a discussion of the applications of FT-ICR in ‘bottom-up’ and “top-down” proteomics.
5.1.1 Principles of Fourier Transform Ion Cyclotron Resonance (FT-ICR) 5.1.1.1 Ion Motion As described above, an ion moving in the presence of a uniform magnetic field, B, experiences the Lorentz force: F = ma = qvB
(5.1),
where m is the mass of the ion, a is the acceleration experienced by the ion, q is the charge of the ion, and v its velocity. The Lorentz force is directed perpendicular to the direction of the ion’s velocity and to the magnetic field. Consequently, in the absence of any collisions, the ion’s trajectory is curved into a circle of radius, r, that is, the ion acquires radial velocity in the xy plane (see Figure 5.1). The angular acceleration of the ion is given by: a=
2 v xy r
(5.2),
z x B Lorentz force
y
B
v
ω
FIGURE 5.1 Ion cyclotron motion. The ion describes a circular path perpendicular to the direction of the magnetic field.
124
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
where vxy is the velocity of the ion in the xy plane. The angular frequency of the ion, ω, is: ω=
v xy r
(5.3).
Rearranging Equations 5.1 through 5.3, we get the cyclotron equation ω=
qB m
(5.4).
The cyclotron equation shows that the frequency at which an ion undergoes cyclotron motion is inversely proportional to its mass-to-charge ratio. Thus, when the cyclotron frequency is measured, m/z may be calculated. Although ions are trapped in the plane perpendicular to the magnetic field, they are not trapped in the direction parallel to the magnetic field. It is necessary, therefore, to apply a trapping potential to two electrodes at either end of the ion cyclotron resonance (ICR) cell. The electric field that is produced is three-dimensional and non-linear because the trapping electrodes are finite in size. Two additional types of ion motion are observed as a consequence of the electrostatic field (see Figure 5.2). The ion will move back and forth parallel to the direction of the magnetic field; this motion is known as trapping oscillation. The frequency of trapping oscillation depends on the size and shape of the ICR cell. The trapping potential also has a radial component that produces an outward electrical force on the ion (cf the inwarddirected Lorentz force). The radial force on the ion is F=
qVtrap α r d2
(5.5),
where Vtrap is the trapping potential, α is a constant determined by the ICR cell geometry, and d is the distance between the trapping electrodes. As a consequence, the ion will undergo magnetron motion. The center of the cyclotron motion will describe a circular path perpendicular to the direction of the magnetic field. Magnetron motion and trapping oscillation frequencies are much smaller than the cyclotron frequency and are not detected typically in FT-ICR. z x
Cyclotron motion
y
Magnetron motion
Trapping oscillation
FIGURE 5.2 The three modes of motion undertaken by an ion trapped in an ICR cell: cyclotron motion, magnetron motion, and trapping oscillation.
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
125
As described, the radial electric force opposes the Lorentz force. Thus the overall force acting on the ion is now:
Force = mω2r = qωrB −
qVtrap α r d2
(5.6).
Rearranging Equation 5.6 gives a quadratic equation in ω that can be solved to give the reduced cyclotron frequency
ω=
ωc + 2
( ω2 ) − ω2 c
2
2 z
(5.7),
where ωc is the cyclotron frequency and ωz is the trapping oscillation frequency. As a result, the frequency-to-m/z calibration equation [3] is non-linear. It is necessary when calibrating, therefore, to use at least two m/z-values. Ideally, multiple m/z-values should be used as cyclotron motion is also affected by Coulombic repulsion between ions. 5.1.1.2 Excitation and Detection The goal of FT-ICR mass spectrometry is to measure the cyclotron frequency, by detecting the image current as a packet of ions passes by a detector plate, and thus calculate m/z. This goal cannot be achieved, however, simply by trapping ions in an ICR cell. First, ions of the same m/z-value are not necessarily in phase with each other. On injection to the ICR cell, an ion can start its cyclotron motion at any point on the cyclotron path. The net image current induced by two ions ‘opposite’ each other, that is, 180° out of phase, will be zero. Second, the cyclotron motion needs to be at a radius sufficiently large to permit detection. For example, at room temperature, a singly-charged ion of mass 1000 Da has a cyclotron radius of 0.08 mm in a 9.4 T magnetic field [2]. The diameter of the ICR cell is of the order of 10 cm. In order to detect ions trapped in an ICR cell, it is necessary first to induce coherent motion, that is, ions of the same m/z need to be traveling together in a single ion packet at a radius close to that of the ICR cell. Coherent motion is achieved by exciting the ions
Excite
Detect
Detect
Excite
FIGURE 5.3 Excitation of an ion trapped in an ICR cell.
126
0
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
100
200
300 400 500 Time (ms)
600
700
800
Fourier transform
150 200 Frequency (kHz)
100
Frequency to m/z calibration
12+ 600
800
250
7+ 1000
1200 1400 m/z
1600
1800
2000
FIGURE 5.4 FT-ICR analysis of electrosprayed ubiquitin. Detection of the image current results in the time-domain signal (top). Fourier transformation converts the time-domain signal to a frequency spectrum (middle) that is calibrated to give the mass spectrum (bottom).
via application of an oscillating electric field at, or near, the cyclotron frequency of the ions. A radiofrequency potential containing frequencies that span the desired m/z-range is applied to two of the ICR electrodes, see Figure 5.3. Ions orbiting at these frequencies absorb energy; their kinetic energy is increased and, hence, their cyclotron radius. The cyclotron radius following excitation is independent of m/z; all ions of a given m/z-range are excited to the same radius without any mass discrimination effects. Post-excitation, a second pair of electrode is used to detect the image current as the clouds of ions cycle around the cell. The resulting time-domain signal comprises the superimposed signals from each of the ion packets. A Fourier transform is applied to convert the time-domain signal into a frequency spectrum, that is, to ‘pull out’ each frequency in the complex signal [4]. A frequency-to-m/z calibration, based on Equations 5.4 and 5.7, is performed resulting in a mass spectrum as depicted in Figure 5.4. Figure 5.4 shows the FT-ICR analysis of electrosprayed ubiquitin, a small protein of ca 8.5 kDa. The mass spectrum reveals the presence of charge states + 7 through + 12, that is, peaks corresponding to [M + 7H]7 + through [M + 12H]12 + .
5.1.2 Instrumentation 5.1.2.1 Magnet A permanent magnet, the homogeneity of which is crucial for satisfactory performance, is central to the FT-ICR mass spectrometer. Commercial instruments featuring
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
127
superconducting magnets of field strength 3–15 T are available. Several parameters improve with magnetic field strength; these include mass resolving power, signalto-noise ratio, dynamic range, and mass accuracy [5]. There is also a concomitant increase in price! To address the problem of stray magnetic field, the magnet may be shielded either passively, for example, enclosed within steel, or actively. Active shielding involves a second set of coils outside the main magnet having a field in the opposing direction, thereby canceling the external magnetic field. 5.1.2.2 Ionization Most FT-ICR applications involving the analysis of peptides and proteins utilize electrospray ionization (ESI) [6]. It is possible also to couple matrix-assisted laser desorption/ionization (MALDI) with FT-ICR mass spectrometers, and these instruments are available commercially, however, they will not be discussed in detail here. 5.1.2.2.1 Electrospray Ionization (ESI) The techniques of ESI and MALDI have revolutionized biological mass spectrometry by enabling the generation of intact either protonated or deprotonated molecules from peptides, proteins, oligonucleotides, and sugars. So great has been their impact that the scientists behind their development, John Fenn (ESI) and Koichi Tanaka (MALDI), were awarded the Nobel Prize for Chemistry in 2002. Prior to the introduction of these techniques, it was not possible to investigate biological molecules greater than ca 2–3 kDa. In ESI [7], a solution containing the analyte flows through a capillary at the mouth of the mass spectrometer. The solution comprises typically an aqueous/organic mix such as 1:1 water:acetonitrile. In positive electrospray, that is, the generation of cations, the solution is acidified with formic or acetic acid (≤ 2%). Generation of anions (negative electrospray) involves typically addition of a base to the solution, for example, ammonium hydroxide. In experiments where it is necessary to preserve non-covalent interactions, for example, when analyzing a peptide:protein complex, the analyte may be sprayed from a solution containing a volatile buffer such as ammonium acetate at a concentration of 5–10 mM. It is very important to note, when considering the ESI of peptides and proteins, that either a high salt concentration or a solution containing detergent will affect adversely the ESI of the analyte. A potential is applied between the tip of the capillary and the entrance to the mass spectrometer, the result of which is dispersal of the solution into fine, charged droplets. Note that ESI is an atmospheric pressure interface. The charged droplets are swept toward the entrance to the mass spectrometer by the pressure gradient. As the droplets proceed, the solvent desorbs and the droplets decrease in size until the repulsion between the charges on the surface of the droplet is such that a Coulombic explosion occurs producing many smaller droplets. The process repeats itself until eventually gas-phase analyte ions remain. It is possible to enhance this process by applying low flows of nitrogen gas to encourage desorption of the solvent. The number of charges on the analyte ions depends on the nature of the analyte and of the solution. Positive electrospray of peptides and proteins tends to result in multiply-charged ions (see Figure 5.4), in which the charges are associated with basic amino acid residues (lysine, arginine, and histidine) and the N-terminus.
128
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Electrospray was first coupled with FT-ICR mass spectrometry in 1989 [8], and the two are very well-suited, particularly in the analysis of large biomolecules. ESI of a protein results in multiply-charged ions, many of which lie within the m/z-range of the FT-ICR mass spectrometer (ca 50–5000 Th). The high resolving power associated with FT-ICR allows the complex isotopic distributions that exist within each charge state to be scrutinized. 5.1.2.3 Ion Transfer As mentioned above, ESI is an atmospheric pressure process whereas the pressure in the ICR cell is of the order of 10−9 Torr. It is necessary, therefore, to transport the ions generated through a succession of differentially-pumped regions of lower pressure. In addition, the ESI source is situated usually outside of the magnetic field and the ions must be transported along the magnetic field lines. The ions produced by electrospray have both axial (parallel to the magnetic field) and radial (perpendicular to the magnetic field) velocity components. As the ions are transported, they experience a large magnetic field gradient. As the magnetic field strength increases, the radial velocity increases. Since kinetic energy must be conserved, the axial velocity decreases. Depending on the initial ratio of the axial to radial velocities, this phenomenon can lead to the magnetic mirror effect, in which the ions stop their forward motion and reverse without reaching the ICR cell. To avoid the magnetic mirror effect, either the ions are accelerated to a very high axial velocity by use of electrostatic lenses [9] or focusing devices such as radiofrequency (RF)-quadrupoles [10], hexapoles [11], or octopoles [12] are used to transport the ions. ESI is a continuous process. While the potential is applied between the capillary tip and the entrance to the mass spectrometer, ions are generated in a constant stream. However, the nature of the FT-ICR experiment requires discrete packets of ions that are first trapped in the ICR cell then excited and detected. To maximize the duty cycle, ions are accumulated externally before being transferred to the ICR cell [13]. Accumulation of ions can be achieved by use of storage hexapoles, octapoles, or ion traps. 5.1.2.4 Ion Cyclotron Resonance (ICR) Cell The principal component of an FT-ICR mass spectrometer is the ICR cell, located in the magnetic field in which the ions are trapped. The axis of the cell is aligned with the magnetic field. A number of different cell geometries are in use including cubic, cylindrical with end-caps, open cylindrical, and ‘matrix-shimmed’ [14]. The cylindrical design is particularly widespread. The effects of the cell geometry on ion motion are discussed above.
5.1.3 Features of Fourier Transform Ion Cyclotron Resonance (FT-ICR) Mass Spectrometry 5.1.3.1 Mass Accuracy FT-ICR mass spectrometry offers unprecedented mass accuracy. This feature is a consequence of the inherent accuracy of frequency measurement. Sub-ppm mass accuracies are obtained routinely. It is possible, up to ca m/z 400, to assign elemental composition (CcHhNnOoPpSs) based on mass alone. This approach has been applied to the fields of
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
129
petroleomics [15] and metabolomics [16]. Smith and co-workers have developed methods in which the mass accuracy associated with FT-ICR is applied to protein identification in proteomics [17,18]. The accurate mass tag (AMT), approach was applied to the analysis of Deinococcus radiodurans identifying over 60% of the predicted proteome [19]. The mass accuracy allows direct determination of small modifications in large intact protein ions, for example, the presence of a disulfide bond (−2 Da). In peptide analysis, it is possible to distinguish between amino acids glutamine and lysine (36 mDa). 5.1.3.2 Resolving Power FT-ICR mass spectrometry offers ultrahigh resolution. This feature is a result of the large numbers of cyclotron orbits during detection and the fact that cyclotron frequency is independent of ion velocity. Performance is not limited by the initial position, direction or speed of the ions, unlike mass spectrometers such as time-offlight or sector instruments. FT-ICR resolution can be defined as the full width of a spectral peak at half of the maximum peak height, that is, ∆m50% for mass spectra, or ∆ω50% for frequency domain spectra. The FT-ICR resolving power is defined as m/∆m50% or ω/∆ω50%. The first derivative of Equation 5.4 gives
m qB =− ∆ m50% m∆ω 50%
(5.8).
Provided that the peak width of the frequency domain spectral peak is independent of magnetic field, that is, post-excitation ion kinetic energy is constant, the FT-ICR resolving power increases linearly with increasing magnetic field. For a constant magnetic field, FT-ICR resolving power varies inversely with m/z. At the low pressure limit, that is, when the time-domain signal persists undamped throughout the acquisition period, the mass spectral peak width is independent of m/z. As m/z increases, the peak width remains constant, however, because ICR frequency varies inversely with m/z, the peaks are more closely spaced. The resolving power associated with FT-ICR has a number of advantages in the analysis of biomolecules. Unit mass resolution has been demonstrated for proteins up to 112 kDa [20], that is, it is possible to distinguish between isotopic peaks containing 13C, 15N, 18O, 34S, etc. The ability to distinguish between isotopic peaks allows direct determination of charge state. The nominal mass difference between isotopic ions is 1 Da, for example, one 12C versus one 13C. The difference in m/z-spacing between isotopic ion peaks corresponds to 1/z. When an ion is in the + 12 charge state, the spacing between the isotopic peaks in the mass spectrum will be 1/12 Th. When it is in the + 6 charge state, the spacing will be 1/6 Th. As protein ions increase in molecular weight, it becomes increasingly statistically likely that the monoisotopic peak will not be observed. Averagine has the molecular formula C4.9384H7.7583N1.3577O1.4773S0.0417, molecular mass 111.1254 Da, and is based on the statistical occurrence of common amino acids. By fitting the observed isotopic distribution against a model distribution based on averagine amino acid residues, it is possible to calculate the monoisotopic mass [21]. In addition to resolving isotopomers, it is possible to resolve isotopic fine structure, for example, to distinguish between two isotopic ions where one ion contains
130
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
two 13C atoms and one ion contains one 34S atom, that is, (Cc-213C2HhOoNnSs) versus (CcHhOoNnSs-134S) by FT-ICR mass spectrometry. Marshall and co-workers demonstrated resolution of isotopic fine structure resolution in a protein of 15.8 kDa [22]. In terms of peptide analysis, the resolving power of FT-ICR allows resolution of phosphorylated and sulfated peptides, PH versus S, a mass difference of 9.5 mDa [23]. In a truly elegant demonstration of resolving power, two peptides of nominal mass 904 Da, but differing in elemental composition by N4O versus S2H8, were baseline-resolved [24]. The actual difference in mass between these peptides was 0.45 mDa – less than the mass of an electron (0.54 mDa)! 5.1.3.3 Sensitivity The sensitivity of FT-ICR mass spectrometry is limited by image current detection. Approximately, 100 charges are required to generate a measurable frequency for a given value of m/z. ESI is advantageous as multiply-charged ions are generated, however, ESI is concentration-sensitive and the analysis of dilute samples can be problematic. Further problems arise because the generated ions need to be transported some distance through a pressure gradient starting at atmospheric pressure and ending at around 10−9 Torr. It is often the case that FT-ICR is interfaced with an on-line separation technique such as nano-liquid chromatography (LC) [25] when analyzing samples of low concentration and high complexity. This configuration concentrates the sample before it enters the mass spectrometer and has the additional advantage of on-line desalting. It should be noted that the presence of salt is common in biological samples and interferes with the electrospray process. Reproducible detection of 100 amol (1 µL of a 100 amol µL −1 solution loaded on column) and 300 amol (1 µL of a 300 amol µL −1 solution) of a single peptide in water and in artificial cerebrospinal fluid, respectively, have been demonstrated [11]. Ten amol (0.5 µL of a 20 amol µL −1 solution) of cytochrome C peptides in a six-protein mix have been detected by this method [26].
5.2 TANDEM MASS SPECTROMETRY (MS/MS) IN FOURIER TRANSFORM ION CYCLOTRON RESONANCE (FT-ICR) In order to gain sequence information about a peptide or protein, it is necessary to fragment the ion in a sequence-specific manner, that is, to cleave between each amino acid residue with a single cleavage per intact peptide or protein ion. A mixture of fragments, corresponding to all possible cleavages, will result and the sequence can be deduced from the mass differences between the fragments. The nomenclature for describing the fragments of peptide/protein ions was devised by Roepstorff et al. [27] and is illustrated in Figure 5.5. Fragmentation of peptide and protein ions in FT-ICR mass spectrometry may be induced by sustained off-resonance irradiation collision-induced dissociation (SORI-CID) [28], infrared multiphoton dissociation (IRMPD) [29,30], blackbody infrared radiative dissociation (BIRD) [31,32], surface-induced dissociation (SID) [33,34], and electron capture dissociation (ECD) [35,36]. These techniques are ‘true’ MS/MS techniques in which the precursor ion is isolated prior to fragmentation. Additional techniques in which ions are not isolated but fragmented before they
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
131
a b c
R1 H2N O
R3
O
H N R2
N H
O
H N O
OH R4
x y z
FIGURE 5.5 Nomenclature for describing the backbone product ions of peptide/protein ions. (Reproduced from Roepstorff, P.; Fohlman, J., Biol. Mass Spectrom. 1984, 11, 601.)
reach the ICR cell include multipole storage-assisted dissociation (MSAD) [37,38] and nozzle-skimmer dissociation [39].
5.2.1 Precursor Ion Isolation Isolation of precursor ions for FT-ICR MS/MS can take place either prior to the ions entering the ICR cell or after trapping of the ions in the cell, that is, FT-ICR MS/ MS can be spatial or temporal. Temporal MS/MS involves resonant excitation of the unwanted ions in the ICR cell; here, all but the required precursor ions are removed from the cell. This process is similar to that used to excite the ions to increased cyclotron orbits prior to image current detection. For the removal of ions, however, the ions are excited to a cyclotron radius greater than the dimensions of the ICR cell. Once the ions collide with the cell walls, they are ‘ejected’ from the trap. As described above, resonant excitation of ions prior to detection involves application to two of the ICR electrodes of a radiofrequency potential containing frequencies which span the desired m/z-range. In precursor isolation, the applied potential contains all frequencies except those corresponding to the ions to be isolated. An elegant feature of this approach is that it is possible to perform ‘double notch’ isolation where two (or more) ions of quite separate m/z can remain in the ICR cell. One method for temporal precursor isolation is ‘stored waveform inverse Fourier transform’ (SWIFT) [40]. In this method, the desired frequency domain profile (all frequencies except that of the ion of interest) is inversely Fourier transformed to a time domain waveform. This waveform is then applied to the ‘excite’ electrodes in the ICR cell and, thus, the precursor ions are isolated in the cell. An alternative technique for in-cell isolation is correlated sweep excitation (COSE) [41], also known as correlated harmonic excitation fields (CHEF) [42]. This method involves application of radiofrequency pulses to the ‘excite’ electrodes. The technique correlates the duration and frequency of the RF pulses with those appropriate to the ions to be isolated. Both SWIFT and COSE are capable of isolating single isotopomers in peptide and protein ions [43–45]. As described above, it is possible also to isolate precursor ions prior to entry to the ICR cell. Hybrid FT-ICR instruments, which are interfaced with front-end
132
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
resolving quadrupoles, were developed in the groups of Marshall [46,47] and Smith [48,49] and are now available commercially. These instruments allow mass-selective external accumulation of ions. Similarly, FT-ICR instruments interfaced with linear ion traps are available commercially also. In these instruments, ion isolation occurs solely at the front-end—there is no in-cell selection. Hybrid FT-ICR instruments are discussed in more detail below.
5.2.2 Sustained Off-Resonance Irradiation Collision-Induced Dissociation (SORI-CID) 5.2.2.1 Principles of Sustained Off-Resonance Irradiation Collision-Induced Dissociation (SORI-CID) Collision-induced dissociation (CID) [50] is the mainstay of MS/MS. The technique involves collisions between translationally-excited ions with inert gas atoms or molecules. Translational energy is converted to internal energy as a result of the inelastic collision. ‘Slow heating’ of the precursor ions to a higher Boltzmann temperature results in formation of product ions by the lowest energy pathway(s). In peptides and proteins, cleavage of the peptide bond occurs resulting in b and y fragments (see Figure 5.5). The first obstacle for FT-ICR-based CID, that the ions must be excited translationally, may be surmounted by resonant excitation of the ions (as in pre-detection excitation, or radial ejection, described above). However, the extent of excitation is limited by the magnetic field and the size of the trap. Excess excitation would result in ejection of the ions from the ICR cell. A second problem is that the products are formed off-axis because the precursor ions are increasing their cyclotron radius; this effect results in reduced resolving power and prevents any further fragmentation (MSn). SORI-CID, introduced by Gauthier and co-workers [28], is not beset by these problems. As the name suggests, ions are excited slightly off-resonance (500–2000 Hz). Such excitation results in acceleration and deceleration of the ions with a period equal to the difference between the excitation frequency and the ion cyclotron frequency. The periodic decrease in cyclotron radius means that ions are not ejected from the ICR cell. Prior to off-resonance excitation of the precursor ions, inert gas is leaked into the ICR cell. As the ions are excited, collisions with the gas result in conversion of translational energy to internal energy. Again, as a result of the periodic decrease in cyclotron radius, the product ions are formed close to the center of the cell, eliminating resolution issues. It is possible that the product ions have a cyclotron frequency equal to that of the applied excitation waveform. If this were the case, those product ions would be ejected from the ICR cell (resonant ejection). To avoid this occurrence, off-resonance excitation is performed in both directions, for example, ±500 Hz. A shortcoming of SORI-CID is that gas must be leaked into the ICR cell. However, high resolution FT-ICR measurements require cell low pressure. It is necessary, therefore, to introduce a delay in the experiment following the SORI-CID event. This delay allows the cell pressure to return to normal (ca 10−9 Torr). The delay is of the order of tens of seconds, making SORI-CID incompatible with on-line separation techniques.
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
133
5.2.2.2 Sustained Off-Resonance Irradiation Collision-Induced Dissociation (SORI-CID) of Peptides and Proteins SORI-CID is a ‘slow-heating’ fragmentation technique in which product ions are formed by the lowest energy pathways. For peptides and proteins, cleavage of the peptide bond occurs to give b and y ions. In addition, losses of small neutral molecules, such as water or ammonia, are observed frequently. Studies of SORI-CID of peptides and proteins have shown that the technique has high efficiency, selectivity, and resolving power [51]. The technique has been applied to peptides [43,45,52–54] and proteins including cytochome c (ca 12 kDa) [55], myoglobin (ca 17 kDa) [45,56], carbonic anhydrase (ca 29 kDa) [57], and monomeric M-CSF (ca 25 kDa) [58]. Multiplestage SORI-CID (up to MS4) has been reported [59]. In addition to providing primary sequence information, SORI-CID can be used to dissociate non-covalent protein complexes [60] and to remove salt from protein ions in the ICR cell [58].
5.2.3 Infrared Multiphoton Dissociation (IRMPD) 5.2.3.1 Principles of Infrared Multiphoton Dissociation (IRMPD) In IRMPD [29,30], ions are activated by irradiation with photons in the ICR cell. The photons are provided typically by a 10.6 µm CO2 laser. Precursor ions are heated slowly to their dissociation threshold and, as with SORI-CID, they fragment via the lowest energy pathways. The optimum irradiation time for peptides and proteins is 100–200 ms [29]. IRMPD offers a number of advantages over SORI-CID. The principal advantage is that the introduction of gas to the ICR cell is obviated. There is no requirement for a pump-down delay, hence the speed of analysis is greater and the method is compatible with on-line separation techniques [61]. Unlike SORI-CID, blind-spots in the MS/MS spectrum do not occur because there is no resonant excitation of the product ions. All product ions are formed on-axis so there is no loss of resolution and it is possible to undertake further stages of MS/MS. A consequence of the on-axis position is the potential for secondary fragmentation, that is, the product ions can be photon activated also. Fragmentation of product ions can complicate spectral interpretation. 5.2.3.2 Infrared Multiphoton Dissociation (IRMPD) of Peptides and Proteins IRMPD of peptide and protein ions results in cleavage of the peptide bond to give b and y product ions (see Figure 5.5). Greater product ion fragmentation is observed with IRMPD than with SORI-CID. This feature is a consequence of the on-axis nature of the IRMPD technique. Typically, peptide-bond cleavage is more extensive in IRMPD than in SORI-CID. The fact that IRMPD and SORI-CID produce similar product ions, despite being dissimilar activation methods, is indicative of rapid energy partitioning within the precursor ion. An example of an IRMPD mass spectrum of a peptide is given in Figure 5.6; b and y product ions dominate the mass spectrum, with loss of water and ammonia observed also. IRMPD has been applied to peptides and proteins including ubiquitin (ca 8.5 kDa), cytochrome C (ca 12 kDa), carbonic anhydrase (ca 29 kDa), glucokinase (ca 34 kDa),
134
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V ×3
×3
ISQAVHAAHAEINEAGR
[M+2H]2+
y3 y7–H2O y6 y4–H2O
y1 y2 200
b4
400
b8 y8 b10
y5
600
b7
y7
800
y9 y 10
b9
m/z
y12
y11
b14–NH3
b11
b14
b12 1000
1200
1400
1600
FIGURE 5.6 IRMPD FT-ICR mass spectrum obtained from [M + 2H]2 + ions of the ovalbumin tryptic peptide ISQAVHAAHAEINEAGR. Precursor ions were irradiated for 80 ms at 60% laser power.
protein A (ca 45 kDa), and serum albumin (ca 67 kDa) [29,62,63]. IRMPD has been applied also to the analysis of ubiquitinated and sumoylated proteins [64,65], and glycoproteins [66,67].
5.2.4 Blackbody Infrared Radiative Dissociation (BIRD) 5.2.4.1 Principles of Blackbody Infrared Radiative Dissociation (BIRD) In BIRD [32], ions are activated by absorption of infrared photons emitted by nearby materials. The vacuum chamber surrounding the ICR cell, when heated normally, emits infrared radiation. This thermal infrared radiation is absorbed by the precursor ions and they are heated close to the temperature of the chamber. As a result, the precursor ions fragment via the lowest energy pathways. Temperatures of up to 500 K are accessible by this method. The rate of dissociation of peptides and proteins at these temperatures is slow and it typically takes 10–1000 s to acquire a BIRD mass spectrum. In addition, long times are needed for the temperature of the ICR cell to equilibrate with that of the vacuum chamber. Unlike SORI-CID, the BIRD technique is neither troubled by the presence of blind-spots in the resulting mass spectra (there is no resonant excitation of product ions), nor are the product ions formed off-axis. 5.2.4.2 Blackbody Infrared Radiative Dissociation (BIRD) of Peptides and Proteins The long timescales associated with BIRD mean that it is not the technique of choice when sequencing peptides and proteins. Moreover, BIRD has been shown to be
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
135
inefficient at protein sequencing when compared with other methods such as CID or ECD [68]. Very low energy fragmentation channels are observed as a result of the slow activation over relatively long timescales. The lowest energy fragmentation channel is often that of extensive water loss. Loss of up to eight water molecules from ubiquitin precursor ions has been demonstrated [32]. Although BIRD is not ideal for peptide/protein sequencing, the fact that the temperature of the ions’ environment can be measured accurately means it is eminently suitable for determining rate constants, activation energies, and dissociation energies.
5.2.5 Electron Capture Dissociation (ECD) 5.2.5.1 Principles of Electron Capture Dissociation (ECD) ECD [35] involves irradiation of precursor ions trapped in the ICR cell with low energy (< 0.2 eV) electrons. In order for the product ions to be detected following an electron/ion reaction, the precursor ion must be multiply-charged. Clearly, the capture of an electron by a singly-charged cation would result in a net charge of zero. The electron capture cross-section increases with the square of the charge of the precursor ion, hence multiply-charged precursor ions are preferred. Happily, ESI generates multiply-charged ions. ECD is virtually exclusive to FT-ICR mass spectrometers. The method of ECD on other mass analyzers has been demonstrated [69] but is non-trivial due to the problems with trapping electrons. In the initial ECD experiments, electrons were provided by a conventional heated-filament situated behind the ICR cell. That configuration required long irradiation times (tens of seconds) due to a weak flux of low-energy electrons and inadequate spatial overlap between the electrons and the trapped ion cloud. An emission current of 1 mA produced by a filament source results in an electron beam cross-section of << 1 mm2, whereas the diameter of an ion cloud trapped in an ICR cell can be > 1 mm. The introduction of the heated dispenser cathode overcame this problem [70]. The emitting area of the cathode is > 1 mm2. The heated dispenser cathode has a lower surface temperature and results in more uniform electron energy because of the elimination of the voltage drop across the emitting surface. A consequence of this development is that electron irradiation times have been reduced vastly, to milliseconds, making the technique compatible with on-line separation techniques (see below). ECD is not an efficient process. At least one charge is neutralized as a result of electron capture. Furthermore, a given precursor ions cleaves into any one of a rather large number of possible product ions, with the result that a high abundance of precursor ions is required in order to detect a particular product ion. An ultimate limit of ca 30% conversion of precursor to product ions exists. 5.2.5.2 Electron Capture Dissociation (ECD) of Peptides and Proteins SORI-CID, IRMPD, and BIRD of peptides and proteins progress via the lowest energy pathway(s). Because the fragmentation patterns are non-random, for example, cleavage of the amide bond N-terminal to proline is favored, incomplete sequence coverage can result. ECD, however, is the result of radical ion chemistry and appears to be non-ergodic, that is, fragmentation occurs before energy is randomized throughout the precursor ion. Consequently, site-specific cleavage is not observed.
136
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Typically, the dominant product in an ECD mass spectrum corresponds to the charge-reduced radical cation [M + nH](n-1) + •, that is, the precursor ion that has captured an electron but has not fragmented into sequence-specific ions. Another common product is [M + (n−1)H](n−1) + , that is, the precursor ion that has captured an electron and has lost a hydrogen atom [71]. In terms of peptide backbone fragmentation, ECD results in cleavage of the N–Cα bond to give c and z• (or c• and z) product ions (see Figure 5.5). The mechanism by which N–Cα cleavage proceeds has been the subject of intense debate. Recent work by Turecek and co-workers [72] suggests that the mechanism is defined by the electronic state of the charge-reduced species; electron capture occurs via a cascade process in which many electronic states are sampled [73]. In the ground state charge-reduced species, cleavage proceeds via proton transfer to an internallysolvated amide carbonyl. In the excited electronic state, amide bonds acquire sufficient electron density as to render them superbasic. N–Cα cleavage then occurs via proton transfer from a remote (but accessible) site. As described above, there are no favored sites of cleavage within peptides, however cleavage at proline is rarely observed [74]. The proline side-chain is cyclic and observation of c/z product ions requires two bonds to be broken. An example of an ECD mass spectrum is shown in Figure 5.7. ECD offers a number of advantages over the ‘slow-heating’ techniques for peptide and protein analysis. The extent of cleavage throughout the peptide backbone tends to be greater resulting in longer peptide sequence tags [35]. For larger polypeptides and proteins (> 20 kDa), however, few backbone fragments are observed [75]. This effect arises because whereas ECD cleaves the peptide backbone, it does not disrupt non-covalent bonds. The folded protein ion, therefore, remains intact. Activated ion techniques are used to address this problem and are described in more detail below. A second advantage of ECD for peptide/protein analysis is that backbone fragments tend to retain post-translational modifications (PTMs) [76–80], such as phosphorylation and glycosylation. It is possible, therefore, to identify sites of modification directly. ‘Slow-heating’ techniques tend to result in loss of labile modifications, often [M+2H]2+
FESNFNTQATNR
z11 z2 z4 200
400
z5 600
z6
z7 800 m/z
z8
z9 z10
1000
1200
c11 [M+2H]+ 1400
FIGURE 5.7 ECD FT-ICR mass spectrum of [M + 2H]2 + ions of the lysozyme [52–63] tryptic peptide FESNFNTQATNR.
137
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
[M+2H] + [M+2H]2+ FQ p SEEQQQTEDELQDK
z5
z4
500
z8 z6
z12
z9 z10
z7
1000
m/z
c11 c12 z11
1500
c13
c14
c15
z13
2000
FIGURE 5.8 ECD FT-ICR mass spectrum of [M + 2H]2 + ions of the β-casein phosphopeptide FQpSEEQQQTEDELQDK; p denotes phosphorylation.
at the expense of sequence fragments. It is possible to confirm the presence of the modification but not necessarily the site. An example of an ECD mass spectrum of a phosphopeptide is shown in Figure 5.8. Although there are two possible sites of phosphorylation in this peptide (Ser3 and Thr9), it is possible to localize the modification to the serine residue based on the ECD fragmentation pattern. ECD results in a number of minor fragmentation channels in addition to the c/z• cleavage described above. Dissociation to a • and y product ions may also be observed [81]. ECD is unique in that it cleaves disulfide bonds [81]. Disulfide bonds are not cleaved following CID, and serve as a hindrance to interpretation of CID mass spectra as overlapping series of b and y fragments are observed. ECD results also in the direct loss of amino acid side-chains from the precursor rather than as the result of secondary fragmentation [82]. Peaks corresponding to amino acid side chain loss are similar in abundance to those of c and z• product ions, and can be used diagnostically to confirm the presence of particular amino acid residues. Hot ECD (HECD) [83] is a related technique in which higher energy electrons are utilized. HECD mass spectra are characterized by the presence of secondary product ions arising from the loss of amino acid side chains from z• ions. These secondary ions are useful in distinguishing the isomers leucine and isoleucine. Loss of •CH(CH3)2 (43 Da) is observed from z• ions containing an N-terminal leucine and loss of •CH2CH3 (29 Da) is observed from z• ions containing N-terminal isoleucine. 5.2.5.3 Activated Ion Electron Capture Dissociation (AI-ECD) As described above, ECD of large polypeptides and proteins (> 20 kDa) is characterized by low fragmentation efficiency. Abundant peaks corresponding to chargereduced species but few backbone product ions are observed. Electron capture cleaves
138
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
the peptide backbone but fails to disrupt non-covalent bonds. Similar behavior is seen for smaller tightly-folded species. Activated ion ECD [75,84,85], in which ions are heated prior to, during, or after ECD circumvents this problem. As a result of heating, the ions are unfolded enabling the sequence fragments to dissociate. Activation of ions can be achieved by infrared irradiation, blackbody irradiation, or collisional activation using a nitrogen gas pulse. An alternative approach is plasma ECD [86], in which electrons (0.1−15 eV) are collided with pulsed nitrogen gas prior to the trapping of ions in the ICR cell. The induced plasma conditions result in significant increase in ECD efficiency. A single plasma ECD mass spectrum of carbonic anhydrase (ca 29 kDa) showed peaks corresponding to cleavage of 183/253 N–Cα bonds cf 116/258 by activated ion ECD.
5.3 HYBRID FOURIER TRANSFORM ION CYCLOTRON RESONANCE (FT-ICR) INSTRUMENTS Since 2000, the field has moved increasingly toward hybrid FT-ICR instruments in which the FT-ICR is interfaced with a front-end mass analyzer. The groups of Marshall [46,47] and Smith [48,49] introduced the quadrupole-FT-ICR. That configuration is available commercially. The hybrid linear ion trap FT-ICR [87] was introduced commercially in 2003. Hybrid instruments offer greater versatility in terms of mass-selective external accumulation with the associated increase in sensitivity and dynamic range.
5.3.1 Quadrupole-Fourier Transform Ion Cyclotron Resonance (FT-ICR) The benefits of the accumulation of ions external to the ICR cell are described above. As ion detection and accumulation are separated physically, one packet of ions is being detected whilst the next packet is being accumulated. The duty cycle, that is, the fraction of time that ions are accumulated for detection, can therefore approach 100%. In addition, external accumulation of ions results in enhanced signal-to-noise ratio and mass resolving power [13]. A factor which limits the maximum achievable duty cycle is the time taken to purge the external trap of ions. Smith and co-workers [48] introduced a 10-cm long segmented accumulation quadrupole that could be purged completely of ions in 400 µs. Marshall and co-workers introduced an alternative approach, in which a direct current (DC) voltage is applied to angled wires positioned between adjacent rods of the accumulation octopole [88]. Further benefits can be realized by the implementation of mass-selective external accumulation. The dynamic range and sensitivity of the instrument are improved. Mass-selective external accumulation can be achieved by interfacing a quadrupole mass filter with the FT-ICR mass spectrometer. The quadrupole can be operated either in RF/DC mass filtering mode, in which one m/z region traverses the quadrupole, or in RF-only resonant dipolar excitation mode. The latter allows selective removal of multiple m/z peaks. For example, Smith and co-workers showed that this mode could be applied to remove the [M + 16H]16 + and [M + 14H]14 + ions of myoglobin from the charge-state envelope + 13 through + 18 [49].
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
139
5.3.2 Linear Ion Trap-Fourier Transform Ion Cyclotron Resonance (FT-ICR) An alternative approach was introduced by Hunt and co-workers [87]. Those researchers coupled a linear quadrupole ion trap, consisting of four rods of hyberbolic crosssection, with an FT-ICR mass spectrometer. The linear ion trap allows accumulation of larger populations of ions than does a standard three-dimensional (3D) ion trap. The hybrid linear ion trap-FT-ICR instrument enables simultaneous detection in both mass analyzers. This aspect is particularly advantageous for ‘data-dependent’ MS/ MS methods used in proteomics, and is discussed further below. The commercial version of this instrument features automated gain control that accumulates a fixed number of charges before delivery to the ICR cell. Because the ‘ideal’ ion density is attained in the cell, space-charge effects resulting in loss of mass resolution and mass accuracy, are eliminated.
5.4 APPLICATIONS OF FOURIER TRANSFORM ION CYCLOTRON RESONANCE (FT-ICR) IN PROTEOMICS Proteomics [89,90] is the study of the entire complement of proteins expressed by a cell or tissue type. The focus of a proteomics experiment, for example, might be identification of proteins that differ according to growth conditions or according to disease state. The aims are to identify and to characterize the maximum number of significant proteins. Proteomics experiments can be described as ‘bottom-up,’ in which proteins are digested with a protease and the resulting peptides are analyzed by mass spectrometry. Alternatively, a “top-down” approach, in which intact proteins are characterized, can be applied [53]. FT-ICR has found applications in both approaches, as discussed below.
5.4.1 ‘Bottom-Up’ Approaches ‘Bottom-up’ proteomics involves digestion of proteins with a protease, usually trypsin, and subsequent mass spectrometric analysis of the resulting peptides. The masses of the tryptic peptides are characteristic of the parent protein. Either a peptide mass fingerprinting approach can be employed, or peptide sequencing by MS/MS can be performed. Peptide sequencing involves generally separation of the peptide mixture by on-line LC. As the peptides elute, they are ionized by electrospray and analyzed by MS/MS. In both the peptide mass fingerprinting and peptide sequencing methods, the data are searched against a protein database. 5.4.1.1 Peptide mass fingerprinting In proteomics, peptide mass fingerprinting of the peptide mixture is undertaken frequently by ionization by MALDI followed by time-of-flight mass analysis. However, the high mass accuracy and resolving power of FT-ICR mass spectrometry can be exploited for this approach. It is possible to resolve virtually all peptide isobars differing by up to two amino acids, even those differing by the smallest mass difference of 3.4 mDa. Resolution of two peptides differing by 11 mDa in a complex mixture of 1000s of peptides has been demonstrated [91]. Clearly, isomers require MS/MS
140
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
while, in the case of leucine/isoleucine, HECD is required. Further confidence in the accuracy of mass measurement can be gained through use of a dual electrospray source [92,93], in which an internal mass calibrant is introduced via a separate electrospray emitter. Consequently, space-charge effects in the ICR cell are mitigated without the inherent problem of electrospray ion suppression. Peptide mass fingerprinting can be performed by use of MALDI-FT-ICR [94,95]. For example, Przybylski and co-workers applied MALDI-FT-ICR to the proteomic analysis of cryoglobulins from a hepatitis C patient [96], and to alveolar proteomics associated with proteinosis and cystic fibrosis [97]. Alternatively, LC can be coupled with ESI FT-ICR for peptide mass fingerprinting [94]. Among other applications, LC coupled with ESI-FT-ICR has been used in the proteomic analysis of Escherichia coli [98], the proteomic analysis of amniotic fluid [99], the identification of brain natriuritic peptide (BNP-32) in plasma following heart failure [100], and in the molecular differentiation of ischemic and valvular heart disease [101]. An alternative approach, which exploits the high mass accuracy of FT-ICR, is the use of accurate mass tags (AMT) [17–19]. The approach involves initial creation of a set of AMTs which act as biomarkers for their parent proteins. Potential mass tags are generated by LC along with MS/MS performed on a conventional ion trap instrument, and then validated by FT-ICR and LC retention time. This initial procedure is relatively time-consuming but, once the AMTs are generated, high-throughput experiments can be performed subsequently. The approach has been applied to the global analysis of the Deinococcus radiodurans proteome [19], and proteomics analysis of breast carcinoma cells [102]. 5.4.1.2 Peptide Sequencing Methods incorporating FT-ICR MS/MS have been applied also to bottom-up proteomic analyes. Hakansson et al. [66] applied ESI FT-ICR and IRMPD MS/MS to the analysis of glycoproteins isolated from human cerebrospinal fluid. Brock and co-workers [103] combined MALDI FT-ICR with SORI-CID. The throughput of this approach is hampered by the timescales associated with SORI-CID. Laskin and co-workers [104] compared approaches utilizing SORI-CID and SID coupled to ESI. The protein identification scores were comparable for the two techniques. SID has the advantage that no pump-down delay is needed and, therefore, more cycles of MS/ MS can be completed. 5.4.1.2.1 Liquid Chromatography (LC) Tandem Mass Spectrometry (MS/MS) The majority of peptide-sequencing proteomic experiments involve coupling of LC with MS/MS. Protein spots may be excised from a two-dimensional (2-D) gel and digested prior to reversed-phase LC-MS/MS. Alternatively, a whole cell lysate may be digested and separated by 1 or 2-D (strong cation exchange and reversed phase) on-line LC followed by MS/MS; this approach is known as the shotgun approach [105]. The hybrid linear ion trap FT-ICR instrument allows simultaneous collection of MS data in the ICR cell and CID MS/MS data in the linear ion trap [106]. A common workflow involves one FT-ICR survey MS scan and linear ion trap CID scans of the three most abundant ions. Dynamic exclusion prevents re-analysis of precursor ions. The timescale for an FT-ICR scan is ca 1s (100,000 resolution at m/z 400) whereas a linear ion trap CID event takes ca 300 ms. This method is demonstrated in Figure 5.9.
141
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (a) #1345 RT: 29.86 512.2562
714.8323
489.5394 432.2817
707.6076
735.4080
524.9059 674.3078 570.7399 625.3184 400
500
600
767.8837 789.4778
700 m/z
800
900
1000
(b) 513.16
#1346 RT: 29.87 Full ms2 524.91
657.30 488.56 648.26
383.28 374.38 200
300
400
500
666.27 713.84
600 m/z
730.28 700
800
900
1000
(c) 974.16 #1347 RT: 29.88 Full ms2 625.32
746.14
276.05 231.13 200
551.97 606.90 391.13 504.10 616.25 431.12
300
400
500
(d)
600 m/z
956.15 819.13
700
401.06 335.98 300
433.05 400
606.00 547.98 530.07 500
695.40
851.28
749.90 600 m/z
900
1000
863.85
676.78
Full ms2 707.61
200
800
777.83
630.81
#1348 RT: 29.89
859.12
700
800
945.33
900
1000
FIGURE 5.9 ‘Snapshot’ of an LC CID MS/MS analysis of a tryptic digest of a mixture of six proteins (bovine serum albumin, transferrin, cytochrome c, lysozyme, alcohol dehydrogenase, and β-galactosidase). RT = retention time. (a) FT-ICR survey scan (#1345); (b) Linear ion trap CID MS/MS of precursor m/z 524.9 (#1346); (c) Linear ion trap CID MS/MS of precursor m/z 625.3 (#1347); and (d) Linear ion trap CID MS/MS of precursor m/z 707.6 (#1348).
142
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Figure 5.9 shows a snapshot of an LC CID MS/MS analysis of a tryptic digest of a standard six-protein mixture. The top mass spectrum (scan # 1345 in this experiment) is an FT-ICR survey scan. In the subsequent scan (# 1346), CID MS/MS of the precursor ion m/z 524.9 is performed in the linear ion trap. That precursor ion is the most abundant ion in the survey scan that has not been subjected previously to MS/MS, that is, is not on the exclusion list. Scan # 1347 is the CID mass spectrum of precursor m/z 625.3, the second most abundant ion, not on the exclusion list, in the survey scan. The sequence ends with CID of precursor m/z 707.6, the third most abundant ion, not on the exclusion list, in the survey scan. The subsequent scan (not shown) is an FT-ICR survey scan. An alternative workflow favored by some researchers is one FT-ICR survey MS scan followed by CID in the linear ion trap of the 10 most-abundant ions [107]. These ‘parallel-processing’ approaches have been applied to a diverse range of studies including analysis of the chicken egg white proteome [108], the low molecular weight proteome of Halobacterium salinarum [109], the endocervical mucas proteome [110], sumoylation in Saccharomyces cerevisiae [111], and the tear fluid proteome [112]. It is also possible to combine on-line LC with ECD MS/MS. This approach was applied to the analysis of the protein Fc-ROR2 that was isolated from chondrocytes and digested with trypsin [113]. Analysis by LC ECD MS/MS cannot be undertaken in a parallel manner: ECD must take place in the ICR cell. The previous survey scan must, therefore, be completed prior to ECD. Consequently, the duty cycle is reduced. A further disadvantage of this approach is the inherent inefficiency of ECD (see above). It is necessary to accumulate more precursor ions for ECD, with a concomitant increase in experiment time. The accumulation time is of the order of seconds rather than the milliseconds required for accumulation for CID. Nevertheless, studies have shown that ECD results in longer peptide sequence tags than does CID, thus improving confidence in peptide assignment [114]. Approaches which combine LC with ECD and CID have been developed also and are discussed further below [115–118]. 5.4.1.2.2 Post-Translational Modification (PTM) Analysis Post-translational modification (PTM) of proteins plays a vital role in many biological processes. For example, phosphorylation is a key event in many signaling cascades, ubiquitination targets proteins for degradation, and glycosylation is involved in cell–cell recognition. Identifying and characterizing modified proteins is a major goal in proteomics. The ease with which this can be undertaken depends largely on the lability of the modification and its stoichiometry. ‘Parallel-processing’ methods in which high resolution, high mass accuracy FT-ICR survey scans are combined with lower specification CID scans (see above) have been applied to the study of ubiquitination and sumoylation of proteins [119–121]. Both of these modifications are relatively stable. ‘Parallel-processing’ methods have been applied also to global analyzes of the phosphoproteome [122], and hundreds of phosphoproteins have been identified. The Ascore algorithm, developed by Gygi and co-workers [123], can be applied to these data to determine site localization confidence. Phosphorylation of serine and threonine (but less so phosphotyrosine) are particularly labile modifications. CID of peptides containing either phosphoserine or phosphothreonine tends to result in loss of phosphoric acid (H3PO4,−98 Da) at the expense of peptide backbone
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
143
product ions. Observation of a peak corresponding to the neutral loss confirms the presence of the modification but often precludes its localization. Global phopshoproteome analyzes can be performed also by use of electron transfer dissociation (ETD) [124,125]. These analyzes utilized either 3D quadrupolar ion trap or linear ion trap instruments and are beyond the scope of this chapter. Targeted ECD approaches have been developed that exploit the advantages of ECD, in particular the retention of labile modifications on peptide backbone fragments, while minimizing the disadvantage of time scale. Experiments are performed on a hybrid linear ion trap FT-ICR mass spectrometer. Neutral losstriggered ECD (NL-ECD) [118] uses observation of a neutral loss as a trigger for an ECD event. The most abundant multiply-charged ion identified in the FT-ICR survey scan is subjected to CID in the linear ion trap in the subsequent scan. When a neutral loss peak, for example, H3PO4, −98 Da, is observed, the following scan will be ECD of the precursor ion (not the neutral loss peak). An alternative approach involving separate LC analyses has been developed. The first LC experiment involves CID of the eluting peptides. The purpose of this analysis is phosphopeptide discovery based on mass of the precursor and any sequence product ions observed. The m/z ratios of the putative phosphopeptides are added to an inclusion list and an LC-ECD analysis is performed. In this experiment, only those ions on the inclusion list are interrogated by ECD. This approach has been applied to the analysis of the protein Sprouty2 [126]. Fourteen sites of phosphorylation were identified of which 11 were novel. Zubarev and co-workers utilized a combined ECD CID approach for the bottom-up analysis of phosphorylation in human α-casein [127]. The method involved an FT-ICR survey scan followed by ECD and CID of the two most abundant precursor ions. These researchers identified a site of phosphorylation that, although known in the bovine form, had not been reported previously for human α-casein.
5.4.2 ‘Top-Down’ Approaches FT-ICR mass spectrometry has great potential for ‘top-down’ proteomics [128], that is, characterization of intact proteins. The high resolution and mass accuracy are well-suited to the analysis of large biomolecules. Moreover, these features allow direct and accurate mass measurement of multiply-charged product ions, that is, in top-down MS/MS. To date, top-down MS/MS has been applied to characterization of proteins and large polypeptides up to 60 kDa [129]. The top-down approach offers some advantages over the bottom-up approach for protein characterization. Because intact proteins are analyzed, 100% sequence coverage is achieved. The method is, therefore, particularly suited to PTM analysis. The highest mass molecule for which unit resolution was achieved was 112,508 Da [20] by use of a 9.4 T instrument. As FT-ICR magnetic field strength increases, unit resolution should be achieved for even higher mass molecules. Unit resolution enables modifications such as disulfide bridge formation (−2 Da) or deamination ( + 1 Da) to be identified. A disadvantage of the bottom-up approach is that any connectivity between modifications or mutations in the protein sequence is lost. For example, a protein with
144
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
mutation at amino acid Xxx may be phosphorylated uniquely at amino acid Zzz. Any such information is lost following proteolytic digestion. For example, Ge et al. [130], performed top-down analyses of proteins isolated from Mycobacterium tuberculosis. One protein assignment, which was based on the mass of the intact species, was found subsequently to be erroneous based on ECD MS/MS data. The MS/MS results showed that, in fact, the species was a truncated version of a different protein. It would not have been possible to demonstrate this distinction using a bottom-up approach. A potential drawback for top-down analyses is that, as the molecular weight of a protein increases, it becomes less likely that the monoisotopic peak will be observed. For proteins ≥ 15 kDa, the monoisotopic peak is << 1% abundance and cannot be detected [131,132]. It has been shown also that natural variation in 13C versus 14C abundance can shift the most abundant isotopic peak of carbonic anhydrase by 1 Da [131]. These phenomena place a limit on the accurate determination of protein molecular mass and, consequently, PTM assignment; for example, is a 2 Da shift due to either 13C 14C abundance and deamination, or disulfide cleavage? Marshall and co-workers showed that by isolating proteins from E. coli grown on 13C-depleted glucose and 15N-depleted ammonium sulfate, it is possible to reduce significantly the mass shift and width of the isotope distribution [133]. For 99.99% 12C and 99.99% 14N, the upper mass limit for a relative abundance of 1% of monoisotopic peak is 100 kDa. FT-ICR-based top-down proteomics has been applied to protein discovery in Methanococcus jannaschii [134], Shewanella oneidensis [135], and human HeLa cells [136]. Examples of applications of the top-down approach for targeted protein characterization include the analysis of post-translationally modified histones [137,138], determination of variations in the sequence of β-thymosin [139], and phosphorylation analysis of cardiac troponin I [140].
5.5 SUMMARY This chapter has provided an overview of the principles of FT-ICR mass spectrometry and its applications in the structural analysis of peptides and proteins. The features of FT-ICR mass spectrometry that make the technique so suitable for peptide/protein analysis are ultrahigh mass accuracy and resolution. The central element of FT-ICR is the cyclotron motion of an ion in a magnetic field. Measurement of the frequency of cyclotron motion enables calculation of the ion’s m/z ratio. In addition to cyclotron motion, an ion undergoes also trapping oscillation and magnetron motion when constrained by a trap of finite size. The FT-ICR mass spectrometer comprises an external ionization source and ion accumulation apparatus, the ICR cell and a superconducting magnet. Recent developments have lead to the introduction of hybrid instruments in which FT-ICR is interfaced with a different mass analyzer such as a quadrupole mass filter or a linear ion trap. To gain sequence information about a peptide, it is necessary to perform MS/MS. FT-ICR MS/MS techniques include SORI-CID, IRMPD, BIRD, and ECD. Finally, FT-ICR mass spectrometry has applications in both bottom-up and top-down proteomics.
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
145
REFERENCES
1. Comisarow, M.B.; Marshall, A.G. Fourier transform ion cyclotron resonance spectroscopy. Chem. Phys. Lett. 1974, 25, 282–283. 2. Marshall, A.G.; Hendrickson, C.L.; Jackson, G.S. Fourier transform ion cyclotron resonance mass spectrometry: A primer. Mass Spectrom. Rev. 1998, 17, 1–35. 3. Ledford, E.B., Jr.; Rempel, D.L.; Gross, M.L. Space charge effects in Fourier transform mass spectrometry. Mass calibration. Anal. Chem. 1984, 56, 2744–2748. 4. Marshall, A.G.; Verdun, F.R. Fourier Transforms in NMR, Optical, and Mass Spectrometry: A User’s Handbook. 1990, Amsterdam: Elsevier. 5. Marshall, A.G.; Guan, S. Advantages of high magnetic field for FT-ICR mass spectrometry. Rapid Commun. Mass Spectrom. 1996, 10, 1819–1823. 6. Hendrickson, C.L.; Emmett, M.R. Electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Annu. Rev. Phys. Chem. 1999, 50, 517–536. 7. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; Whitehouse, C.M. Electrospray ionization for mass spectrometry of large biomolecules. Science, 1989, 246, 64–71. 8. Henry, K.D.; Williams, E.R.; Wang, B-H.; McLafferty, F.W.; Shabanowitz, J.; Hunt, D.F. FT-MS of large molecules by ESI. Proc. Natl. Acad. Sci. 1989, 86, 9075–9078. 9. Kofel, P.; Allemann, M.; Kellerhaos, H.; Wancek, K.P. External generation of ions in ICR spectrometry. Int. J. Mass Spectrom. Ion Processes 1985, 65, 97–103. 10. Hunt, D.F.; Shabanowitz, J.; McIver, R.T.; Hunter, R.L.; Syka, J.E.P. Ionization and mass analysis of nonvolatile compounds by particle bombardment tandem-quadrupole FT-MS. Anal. Chem. 1985, 57, 765–768. 11. Quenzer, T.L.; Emmett, M.R.; Hendrickson, C.L.; Kelly, P.H.; Marshall, A.G. High sensitivty FT-ICR mass spectrometry for biological analysis with nano-LC and microelectrospray ionisation. Anal. Chem. 2000, 73, 1721–1725. 12. Marshall, A.G.; Guan, S. Ion optics for FT-ICR mass spectrometry. Nucl. Instr. Meth. Phys. Res. A 1995. 363, 397–405. 13. Senko, M.W.; Hendrickson, C.L.; Emmett, M.R.; Shi, S.D-H.; Marshall, A.G. External accumulation of ions for enhanced electrospray ionization FT-ICR mass spectrometry. J. Am. Soc. Mass Spectrom. 1997, 8, 970–976. 14. Guan, S.; Marshall, A.G. Ion traps for FT-ICR mass spectrometry-principles and design of geometric ad electric configurations. Int. J. Mass Spectrom. Ion Processes 1995, 146/147, 261–296. 15. Hughey, C.A.; Cooper, H.J.; Rodgers, R.P.; Marshall, A.G.; Robbins, W.K.; Qian, K. Resolution and identification of elemental compositions for more than 3000 crude acids in heavy petroleum by negative-ion microelectrospray high-field FT-ICR mass spectrometry. Energy & Fuels 2001, 15, 1505–1511. 16. Southam, A.D.; Payne, T.G.; Cooper, H.J.; Arvanitis, T.N.; Viant, M.R. Dynamic range and mass accuracy of wide-scan direct infusion nanoelectrospray FT-ICR mass spectrometry-based metabolomics increased by the spectral stitching method. Anal. Chem. 2007, 79, 4595–4602. 17. Conrads, T.P.; Anderson, G.A.; Veenstra, T.D.; Pasa-Tolic, L.; Smith, R.D. Utility of accurate mass tags for proteome-wide protein identification. Anal. Chem. 2000, 72, 3349–3354. 18. Smith, R.D.; Anderson, G.A.; Lipton, M.S.; Pasa-Tolic, L.; Shen, Y.F.; Conrads, T.P.; Veenstra, T.D.; Udseth, H.R. An accurate mass tag strategy for quantitative and highthroughput proteome measurements. Proteomics 2002, 2(5), 513–523. 19. Lipton, M.S.; Pasa-Tolic, L.; Anderson, G.A.; Anderson, D.J.; Auberry, D.L.; Battista, K.R.; Daly, M.J.; Fredrickson, J.; Hixson, K.K.; Kostandarithes, H.; Masselson, C.D.; Markillies, L.M.; Moore, R.J.; Romine, M.F.; Shen, Y.F.; Stritmatter, E.; Tolic, N.; Udseth,
146
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
H.R.; Venkateswaran, A.; Wong, L.K.; Zhao, R.; Smith, R.D. Global analysis of the Deinococcus radiodurans proteome by accurate mass tags. Proc. Natl. Acad. Sci. 2002, 99(17), 11049–11054. 20. Kelleher, N.L.; Senko, M.W.; Siegel, M.M.; McLafferty, F.W. Unit resolution massspectra of 112 kDa molecules with 3 Da accuracy. J. Am. Soc. Mass Spectrom. 1997, 8, 380–383. 21. Senko, M.W.; Beu, S.C.; McLafferty, F.W. Determination of monoisotopic masses and ion populations for large biomolecules from resolved isotopic distributions. J. Am. Soc. Mass Spectrom. 1995, 6, 229–233. 22. Shi, S.D-H.; Hendrickson, C.L.; Marshall, A.G. Counting individual sulfur atoms in a protein by ultrahigh resolution FT-ICR mass spectrometry: Experimental resolution of isotopic fine structure in proteins. Proc. Natl. Acad. Sci. 1998, 95, 11532–11537. 23. Bossio, R.E.; Marshall, A.G. Baseline resolution of isobaric phosphorylated and sulfated peptides and nucleotides by electrospray ionization FT-ICR MS: Another step toward mass spectrometry-based proteomics. Anal. Chem. 2002, 74, 1674–1679. 24. He, F.; Hendrickson, C.L.; Marshall, A.G. Baseline mass resolution of peptide isobars: A new record for molecular mass resolution. Anal. Chem. 2001, 73, 647–650. 25. Kennedy, R.T.; Jorgenson, J.W. Preparation and evaluation of packed capillary liquidchromatography columns with inner diameters from 20 μM to 50 μm. Anal. Chem. 1989, 61, 1128–1135. 26. Martin, S.E.; Shabanowitz, J.; Hunt, D.F.; Marto, J.A. Subfemtomole MS and MS/MS peptide sequence analysis using nano-HPLC micro-ESI FT-ICR mass spectrometry. Anal. Chem. 2000, 72, 4266–4274. 27. Roepstorff, P.; Fohlman, J. Proposal for a common nomenclature for sequence ions in mass spectra of peptides. Biol. Mass Spectrom. 1984, 11, 601. 28. Gauthier, J.W.; Trautman, T.R.; Jacobson, D.B. Sustained off-resonance irradiation for collision-activated dissociation involving Fourier transform mass spectrometry. Collision-activated dissociation technique that emulates infrared multiphoton dissociation. Anal. Chim. Acta. 1991, 246, 211–225. 29. Little, D.P.; Speir, J.P.; Senko, M.W.; O’Connor, P.B.; McLafferty, F.W. Infrared multiphoton dissociation of large multiply-charged ions for biomolecule sequencing. Anal. Chem. 1994, 66, 2809–2815. 30. Woodlin, R.L.; Bomse, D.S.; Beauchamp, J.L. Multiphoton dissociation of molecules with low power continuous wave infrared laser radiation. J. Am. Chem. Soc. 1978, 100, 3248–3250. 31. Dunbar, R.C.; McMahon, T.B. Activation of unimolecular reactions by ambient blackbody radiation. Science 1998, 279, 194–197. 32. Price, W.D.; Schnier, P.D.; Williams, E.R. Tandem mass spectrometry of large biomolecule ions by blackbody infrared radiative dissociation. Anal. Chem. 1996, 68, 859–866. 33. Chorush, R.A.; Little, D.P.; Beu, S.C.; Wood, T.D.; McLafferty, F.W. Surface induced dissociation of multiply protonated proteins. Anal. Chem. 1995, 67, 1042–1046. 34. McCormack, A.L.; Jones, J.L.; Wysocki, V.H. Surface-induced dissociation of multiply protonated peptides. J. Am. Soc. Mass Spectrom. 1992, 3, 859–862. 35. Zubarev, R.A.; Kelleher, N.L.; McLafferty, F.W. ECD of multiply charged protein cations. A non-ergodic process. J. Am. Chem. Soc. 1998, 120, 3265–3266. 36. Cooper, H.J.; Hakansson, K.; Marshall, A.G. The role of electron capture dissociation in biomolecular analysis. Mass Spectrom. Rev. 2005, 24, 201–222. 37. Hakansson, K.; Axelsson, J.; Palmblad, M.; Hakansson, P. Mechanistic studies of multipole storage assisted dissociation. J. Am. Soc. Mass Spectrom. 2000, 11, 210–217. 38. Sannes-Lowery, K.; Griffey, R.H.; Kruppa, G.H.; Speir, J.P.; Hofstadler, S.A. Multipole storage assisted dissociation, a novel in-source dissociation technique for electrospray ionization generated ions. Rapid Commun. Mass Spectrom. 1998, 12, 1957–1961.
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
147
39. Loo, J.A.; Edmonds, C.G.; Udseth, H.R.; Smith, R.D. Collisional activation and dissociation of large multiply charged proteins produced by electrospray ionisation. Anal. Chim. Acta. 1990, 241, 167–173. 40. Marshall, A.G.; Wang, T-C.L.; Ricca, T.L. Tailored excitation for FTICR MS. J. Am. Chem. Soc. 1985, 107, 7893–7897. 41. Heck, A.J.R.; de Koning, L.J.; Pinske, F.A.; Nibbering, N.M.M. Mass-specific selection of ions in Fourier transform ion cyclotron resonance mass spectrometry-unintentional offresonance excitation of selected ions. Rapid Commun. Mass Spectrom. 1991, 5, 406–414. 42. De Koning, L.J.; Nibbering, N.M.M.; van Orden, S.L.; Laukien, F.H. Mass selection of ions in an FT-ICR trap using correlated harmonic excitation fields (CHEF). Int. J. Mass Spectrom. Ion Processes 1997, 165, 209–219. 43. Heck, A.J.R.; Derrick, P.J. Ultrahigh mass accuracy in isotope selective collision-induced dissociation using correlated sweep excitation and sustained off resonance iradiation: A Fourier transform mass spectrometry case study on the [M + 2H]2+ ion of bradykinin. Anal. Chem. 1997, 69, 3603–3607. 44. Freitas, M.A.; Hendrickson, C.L.; Emmett, M.R.; Marshall, A.G. High-Field Fourier transform ion cyclotron resonance mass spectrometry for simultaneous trapping and gas phase hydrogen/deuterium exchange of peptides. J. Am. Soc. Mass Spectrom. 1998, 9, 1012–1019. 45. Heck, A.J.R.; Derrick, P.J. Selective fragmentation of single isotopic ions of proteins up to 17 kDa using 9.4 T FTICR. Eur. Mass Spectrom. 1998, 4, 181–188. 46. Hendrickson, C.L.; Quinn, J.P.; Emmett, M.R.; Marshall, A.G. Mass-selective external ion accumulation for Fourier transform ion cyclotron resonance mass spectrometry. 49th ASMS Conference on Mass Spectrometry and Allied Topics. May 27–31, 2001. Chicago, IL. 47. Patrie, S.M.; Charlebois, J.P.; Whipple, D.; Kelleher, N.L.; Hendrickson, C.L.; Quinn, J.P.; Marshall, A.G.; Mukhopadhyay, B. Construction of a hybrid quadrupole/Fourier transform ion cyclotron resonance mass spectrometer for versatile MS/MS above 10 kDa. J. Am. Soc. Mass Spectrom. 2004, 15(7), 1099–1108. 48. Belov, M.; Nikolaev, E.N.; Anderson, G.A.; Auberry, K.J.; Harkewicz, R.; Smith, R.D. Electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry using ion preselection and external accumulation for ultrahigh sensitivity. J. Am. Soc. Mass Spectrom. 2001, 12, 38–48. 49. Belov, M.E.; Nikolaev, E.N.; Anderson, G.A.; Udseth, H.R.; Conrads, T.P.; Veenstra, T.D.; Masselson, C.D.; Gorshkov, M.V.; Smith, R.D. Design and performance of an ESI interface for selective ion accumulation coupled to a Fourier transform ion cyclotron mass spectrometer. Anal. Chem. 2001, 73, 253–261. 50. Hayes, R.N.; Gross, M.L. Collision induced dissociation. Methods Enzymol. 1990, 193, 237–263. 51. Senko, M.W.; Speir, J.P.; McLafferty, F.W. Collisional activation of large multiply charged ions using Fourier transform mass spectrometry. Anal. Chem. 1994, 66, 2801–2808. 52. Froesch, M.; Bindila, L.; Zamfir, A.; Peter-Katalinic, J. Siaylation analysis of O-glycosylated siaylated peptides from urine of patients suffering from Schindler’s disease by FT-ICR mass spectrometry and SORI-CID. Rapid Commun. Mass Spectrom. 2003, 17(24), 2822–2832. 53. Kelleher, N.L.; Lin, H.Y.; Valaskovic, G.A.; Aaserud, D.J.; Fridriksson, E.K.; McLafferty, F.W. Top down versus bottom up protein characterization by tandem high-resolution mass spectrometry. J. Am. Chem. Soc. 1999, 121, 806–812. 54. Kelleher, N.L.; Nicewonger, R.B.; Begley, T.H.; McLafferty, F.W. Identification of modification sites in large biomolecules by stable isotope labelling and tandem high resolution mass spectrometry. J. Biol. Chem. 1997, 272, 32215–32220. 55. Wu, Q.; Van Orden, S.; Cheng, X.; Bakhtiar, R.; Smith, R.D. Characterization of cytochrome c variants with high-resolution FTICR mass spectrometry: Correlation of fragmentation and structure. Anal. Chem. 1995, 67, 2498–2509.
148
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
56. Wang, Y.; Shi, S.D-H.; Hendrickson, C.L.; Marshall, A.G. Mass-selective ion accumulation and fragmentation in a linear octopole ion trap external to a Fourier transform ion cyclotron resonance mass spectrometer. Int. J. Mass Spectrom. 2000, 198, 113–120. 57. O’Connor, P.B.; Speir, J.P.; Senko, M.W.; Little, D.P.; McLafferty, F.W. Tandem mass spectrometry of carbonic anhydrase. J. Mass Spectrom. 1995, 30, 88–93. 58. Maier, C.S.; Yan, X.; Harder, M.E.; Schimerlik, M.I.; Deinzer, M.L.; Pasa-Tolic, L.; Smith, R.D. ESI FTICR MS analysis of the recombinant human macrophage colony stimulating factor beta and derivatives. J. Am. Soc. Mass Spectrom. 1999, 11, 237–243. 59. Huang, Y.; Pasa-Tolic, L.; Guan, S.; Marshall, A.G. Collision-induced dissociation for mass spectrometric analysis of biopolymers: High resolution FT-ICR MS4. Anal. Chem. 1994, 66, 4385–4389. 60. Schwartz, B.L.; Bruce, J.E.; Anderson, G.A.; Hofstadler, S.A.; Rockwood, A.L.; Smith, R.D.; Chilkoti, A.; Stayton, P.S. Dissociation of tetrameric ions of noncovalent streptavidin complexes formed by electrospray ionisation. J. Am. Soc. Mass Spectrom. 1995, 6, 459–465. 61. Li, W.Q.; Hendrickson, C.L.; Emmett, M.R.; Marshall, A.G. Identification of intact proteins in mixtures by alternated capillary liquid chromatography electrospray ionisation and LC ESI infrared multiphoton dissociation FT-ICR mass spectrometry. Anal. Chem. 1999, 71, 4397–4402. 62. Aaserud, D.J.; Little, D.P.; O’Connor, P.B.; McLafferty, F.W. Distinguishing N-terminus and C-terminus ions for mass spectrometry sequencing of proteins without prior degradation. Rapid Commun. Mass Spectrom. 1995, 9, 871–876. 63. Dufresne, C.P.; Wood, T.D.; Hendrickson, C.L. High-resolution electrospray ionisation FT-MS with IRMPD of glucokinase from Bacillus stearothermophilus. J. Am. Soc. Mass Spectrom. 1998, 9(11), 1222–1225. 64. Cooper, H.J.; Heath, J.K.; Jaffray, E.; Hay, R.T.; Lam, T.T.; Marshall, A.G. Identification of sites of ubiquitination in proteins: A Fourier transform ion cyclotron resonance mass spectrometry approach. Anal. Chem. 2004, 76, 6982–6988. 65. Cooper, H.J.; Tatham, M.H.; Jaffray, E.; Heath, J.K.; Lam, T.T.; Marshall, A.G.; Hay, R.T. FT-ICR mass spectrometry for the analysis of SUMO modification: Identification of lysines in RanBP2 and SUMO targeted for modification during the E3 autoSUMOylation reaction. Anal. Chem. 2005, 77, 6310–6319. 66. Hakansson, K.; Emmett, M.R.; Marshall, A.G.; Davidsson, P.; Nilsson, C.L. Structural analysis of 2-D gel separated glycoproteins from human cerebrospinal fluid by tandem high-resolution mass spectrometry. J. Proteome Res. 2003, 2(6), 581–588. 67. Sihlborn, C.; Davidsson, P.; Emmett, M.R.; Marshall, A.G.; Nilsson, C.L. Glycoproteomics of cerebrospinal fluid in neurodegenerative disease. Int. J. Mass Spectrom. 2004, 234(1–3), 145–152. 68. Ge, Y.; Horn, D.M.; McLafferty, F.W. Blackbody infrared radiative dissociation of larger (42 kDa) multiply charged proteins. Int. J. Mass Spectrom. 2001, 210/211, 203–214. 69. Baba, T.; Black, D.; Glish, G.L. ECD coupled with a linear RF QIT-TOF mass spectrometer. 51st ASMS Conference on Mass Spectrometry and Allied Topics. June 8–12, 2003. Montreal, Canada. 70. Tsybin, Y.O.; Hakansson, P.; Budnik, B.A.; Haselmann, K.F.; Kjeldsen, F.; Gorshkov, M.; Zubarev, R.A. Improved low-energy electron injection systems for high rate electron capture dissociation in Fourier transform ion cyclotron resonance mass spectrometry. Rapid Commun. Mass Spectrom. 2001, 15, 1849–1854. 71. Breuker, K.; Oh, H.; Cerda, B.; Horn, D.M.; McLafferty, F.W. Hydrogen atom loss in ECD: A Fourier transform ion cyclotron resonance study with single isotopomeric ions. Eur. J. Mass Spectrom. 2002, 8, 177–180. 72. Syrstad, E.A.; Turecek, F. Towards a general mechanism of electron capture dissociation. J. Am. Soc. Mass Spectrom. 2005, 16(2), 208–224.
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
149
73. Sawicka, A.; Skurski, P.; Hudgins, R.R.; Simons, J. Model calculations relevant to disulfide bond cleavage via electron capture influenced by positively charged groups. J. Phys. Chem. B 2003, 107, 13505–13511. 74. Cooper, H.J.; Hudgins, R.R.; Hakansson, K.; Marshall, A.G. Secondary fragmentation of linear peptides in electron capture dissociation. Int. J. Mass Spectrom. 2003, 228, 723–728. 75. Horn, D.M.; Ge, Y.; McLafferty, F.W. Activated ion electron capture dissociation for mass spectral sequencing of larger (42 kDa) proteins. Anal. Chem. 2000, 73, 4778–4784. 76. Hakansson, K.; Cooper, H.J.; Emmett, M.R.; Costello, C.E.; Marshall, A.G.; Nilsson, C.L. ECD and IRMPD MS/MS of an N-glycosylated tryptic peptide to yield complementary sequence information. Anal. Chem. 2001, 73, 4530–4536. 77. Kelleher, N.L.; Zubarev, R.A.; Bush, K.; Furie, B.; Furie, B.C.; McLafferty, F.W.; Walsh, C.T. Localization of labile posttranslational modifications by ECD: The case of gammacarboxyglutamic acid. Anal. Chem. 1999, 71, 4250–4253. 78. Mirgorodskaya, E.; Roepstorff, P.; Zubarev, R.A. Localization of O-glycosylation sites in peptides by ECD in a Fourier transform mass spectrometer. Anal. Chem. 1999, 71, 4431–4436. 79. Shi, S.D-H.; Hemling, M.E.; Carr, S.A.; Horn, D.M.; Lindh, I.; McLafferty, F.W. Phosphopetide/phosphoprotein mapping by ECD mass spectrometry. Anal. Chem. 2001, 73, 19–22. 80. Stensballe, A.; Jensen, O.N.; Olsen, J.V.; Haselmann, K.F.; Zubarev, R.A. ECD of singly and multiply phosphorylated peptides. Rapid Commun. Mass Spectrom. 2000, 14, 1793–1800. 81. Zubarev, R.A.; Kruger, N.A.; Fridriksson, E.K.; Lewis, M.A.; Horn, D.M.; Carpenter, B.K.; McLafferty, F.W. ECD of gaseous multiply charged proteins is favored a disulfide bonds and other sites of high hydrogen atom affinity. J. Am. Chem. Soc. 1999, 121, 2857–2862. 82. Cooper, H.J.; Hudgins, R.R.; Hakansson, K.; Marshall, A.G. Characterization of amino acid side chain losses in ECD. J. Am. Soc. Mass Spectrom. 2002, 13, 241–249. 83. Kjeldsen, F.; Haselmann, K.F.; Budnik, B.A.; Jensen, F.; Zubarev, R.A. Dissociative capture of hot (3-13 eV_ electrons by polypeptide polycations: An efficient process accompanied by secondary fragmentation). Chem. Phys. Lett. 2002, 356, 201–206. 84. Ge, Y.; Lawhorn, B.G.; ElNaggar, M.; Strauss, E.; Park, J-H.; Begley, T.; McLafferty, F.W. Top down characterization of larger proteins by ECD mass spectrometry. J. Am. Chem. Soc. 2002, 124, 672–678. 85. Sze, S.K.; Ge, Y.; Oh, H.; McLafferty, F.W. Top-down characterization of a 29 kDa protein for characterization of any post-translational modification to within one residue. Proc. Natl. Acad. Sci. 2002, 99, 1774–1779. 86. Sze, S.K.; Ge, Y.; Oh, H.; McLafferty, F.W. Plasma ECD for the characterization of large proteins by top-down mass spectrometry. Anal. Chem. 2003, 75, 1599–1603. 87. Syka, J.E.P.; Marto, J.A.; Bai, D.L.; Horning, S.; Senko, M.W.; Schwartz, J.C.; Ueberheide, B.; Garcia, B.; Busby, S.; Muratore, T.; Shabanowitz, J.; Hunt, D.F. Novel linear quadrupole ion trap/FT mass spectrometer: Performance characterisation and use in the comparative analysis of histone H3 post-translational modifications. J. Proteome Res. 2004, 3, 621–626. 88. Wilcox, B.E.; Hendrickson, C.L.; Marshall, A.G. Improved ion extraction from a linear octopole ion trap: SIMION analysis and experimental demonstration. J. Am. Soc. Mass Spectrom. 2002, 13, 1304–1312. 89. Godovac-Zimmermann, J.; Brown, L. Perspectives for mass spectrometry and functional proteomics. Mass Spectrom. Rev. 2001, 20, 1–57. 90. Peng, J.; Gygi, S.P. Proteomics: The move to mixtures. J. Mass Spectrom. 2001, 36, 1083–1091.
150
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
91. He, F.; Emmett, M.R.; Hakansson, K.; Hendrickson, C.L.; Marshall, A.G. Theoretical and experimental prospects for protein identification based solely on accurate mass measurement. J. Proteome Res. 2004, 3, 61–67. 92. Nepomuceno, A.I.; Muddiman, D.C.; Bergen, H.R.; Craighead, J.R.; Burke, M.J.; Caskey, P.E.; Allan, J.A. Dual electrospray ionization source for confident generation of accurate mass tags using LC FT-ICR mass spectrometry. Anal. Chem. 2003, 75, 3411–3418. 93. Williams, D.K.; Hawkridge, A.M.; Muddiman, D.C. Sub-ppm mass measurement accuracy of intact proteins and product ions achieved using a dual electrospray ionisation quadrupole FT-ICR mass spectrometer. J. Am. Soc. Mass Spectrom. 2007, 18(1), 1–7. 94. Witt, M.; Fuchser, J.; Baykut, G. FT-ICR mass spectrometry with NanoLC/microelectrospray ionization and MALDI: Analytical performance in peptide mass fingerprint analysis. J. Am. Soc. Mass Spectrom. 2003, 14(6), 553–561. 95. Horn, D.M.; Peters, E.C.; Klock, H.; Meyers, A.; Brock, A. Improved protein identification using automated high mass measurement accuracy MALDI FT-ICR MS peptide mass fingerprinting. Int. J. Mass Spectrom. 2004, 238(2), 189–196. 96. Damoc, E.; Youhnovski, N.; Crettaz, D.; Tissot, J-D.; Przybylski, M. High resolution proteome analysis of cryoglobulins using FT-ICR mass spectrometry. Proteomics 2003, 3, 1425–1433. 97. Bai, Y.; Galetskiy, D.; Damoc, E.; Paschen, C.; Liu, Z.; Griese, M.; Liu, S.; Przybylski, M. High resolution mass spectrometric alveolar proteomics: Identification of surfactant protein SP-A and SP-D modifications in proteinosis and cystic fibrosis patients. Proteomics 2004, 4, 2300–2309. 98. Ihling, C.; Sinz, A. Proteome analysis of E. coli using HPLC and FT-ICR mass spectrometry. Proteomics 2005, 5, 2029–2042. 99. Nilsson, S.; Ramstrom, M.; Palmblad, M.; Axelsson, O.; Bergquist, J. Explorative study of the protein composition of amniotic fluid by LC ESI FT-ICR mass spectrometry. J. Proteome Res. 2004, 3, 884–889. 100. Hawkridge, A.M.; Heublein, D.M.; Bergen, H.R.; Cataliotti, A.; Burnett, J.C.; Muddiman, D.C. Quantitative mass spectral evidence for the absence of circulating brain natriuretic peptide (BNP-32) in severe human heart failure. Proc. Natl. Acad. Sci. 2005, 102(48), 17442–17447. 101. Baykut, D.; Grapow, M.; Bergquist, M.; Amirkhani, A.; Ivonin, I.A.; Reineke, D.; Grussenmeyer, T.; Hakansson, P.; Zerkowski, H.R.; Baykut, G.; Bergquist, J. Molecular differentiation of ischemic and valvular heart disease by liquid chromatography FT-ICR mass spectrometry. Eur. J. Med. Res. 2006, 11(6), 221–226. 102. Umar, A.; Luider, T.M.; Foekens, J.A.; Pasa-Tolic, L. Nano-LC FT-ICR MS improves proteome coverage attainable for similar to 3000 laser-microdissected breast carcinoma cells. Proteomics 2007, 7(2), 323–329. 103. Brock, A.; Horn, D.M.; Peters, E.C.; Shaw, C.M.; Ericson, C.; Phung, Q.T.; Salomon, A.R. An automated MALDI quadrupole FT-ICR mass spectrometer for ‘bottom-up’ proteomics. Anal. Chem. 2003, 75, 3419–3428. 104. Fernandez, F.M.; Wysocki, V.H.; Futrell, J.H.; Laskin, J. Protein identification via surfaced induced dissociation in an FT-ICR mass spectrometer and a patchwork sequencing approach. J. Am. Soc. Mass Spectrom. 2006, 17, 700–709. 105. McDonald, W.H.; Yates, J.R. Shotgun proteomics and biomarker discovery. Dis. Markers 2002, 18(2), 99–105. 106. Peterman, S.M.; Dufresne, C.P.; Horning, S. The use of a hybrid linear trap/FT-ICR mass spectrometer for on-line high reoslution/high mass accuracy bottom-up sequencing. J. Biomol. Tech. 2005, 16(2), 112–124. 107. Haas, W.; Faherty, B.K.; Gerber, S.A.; Elias, J.E.; Beausoleil, S.A.; Bakalarski, C.E.; Li, X.; Villen, J.; Gygi, S.P. Optimization and use of peptide mass measurement accuracy in shotgun proteomics. Mol. Cell. Proteomics 2006, 7(7), 1326–1337.
Fourier Transform Ion Cyclotron Resonance Mass Spectrometry
151
108. Mann, K. The chicken egg white proteome. Proteomics 2007, 7, 3558–3568. 109. Klein, C.; Aivaliotis, M.; Olsen, J.V.; Falb, M.; Besir, H.; Scheffer, B.; Bisle, B.; Tebbe, A.; Konstantinidis, K.; Sledler, F.; Pfeiffer, F.; Mann, M.; Oesterhelt, D. The low molecular weight proteome of Halobacterium salinarum. J. Proteome Res. 2007, 6, 1510–1518. 110. Andersch-Bjorkman, Y.; Thomsson, K.A.; Holmen Larsson, J.M.; Ekerhovd, E.; Hansson, G.C. Large scale identification of proteins, mucins, and their O-glycosylation in the endocervical mucus during the menstrual cycle. Mol. Cell. Proteomics 2007, 6, 708–716. 111. Denison, C.; Rudner, A.D.; Gerber, S.A.; Bakalarski, C.E.; Moazed, D.; Gygi, S.P. A proteomic strategy for gaining insights into protein sumoylation in yeast. Mol. Cell. Proteomics 2005, 4(3), 246–254. 112. de Souza, G.A.; Godoy, L.M.F.; Mann, M. Identification of 491 proteins in the tear fluid proteome reveals a large number of proteases and protease inhibitors. Genome Biol. 2006, 7(8), R72. 113. Cooper, H.J.; Akbarzadeh, S.; Heath, J.K.; Zeller, M. Data-dependent ECD FT-ICR mass spectrometry for proteomic analyses. J. Proteome Res. 2005, 4(5), 1538–1544. 114. Creese, A.J.; Cooper, H.J. Liquid chromatography electron capture dissociation tandem mass spectrometry (LC-ECD-MS/MS) versus liquid chromatography collision-induced dissociation tandem mass spectrometry (LC-CID-MS/MS) for the identification of proteins. J. Am. Soc. Mass Spectrom. 2007, 18(5), 891–897. 115. Nielsen, M.L.; Savitski, M.M.; Zubarev, R.A. Improving protein identification using complementary fragmentation techniques in FT-MS. Mol. Cell. Proteomics 2005, 4, 835–845. 116. Savitski, M.M.; Nielsen, M.L.; Kjeldsen, F.; Zubarev, R.A. Proteomics-grade de novo sequencing approach. J. Proteome Res. 2005, 4, 2348–2354. 117. Savitski, M.M.; Nielsen, M.L.; Zubarev, R.A. New database-independent, sequence tagbased scoring of peptide MS/MS data validates Mowse scores, recovers below threshold data, singles out modified peptides and assesses the quality of MS/MS techniques. Mol. Cell. Proteomics 2005, 4, 1180–1188. 118. Sweet, S.M.M.; Creese, A.J.; Cooper, H.J. Strategy for the identification of sites of phosphorylation in proteins: Neutral loss triggered electron capture dissociation. Anal. Chem. 2006, 78, 7563–7569. 119. Crosas, B.; Hanna, J.; Kirkpatrick, D.S.; Zhang, D.P.; Tone, Y.; Hathaway, N.A.; Buecker, C.; Leggett, D.S.; Schmidt, M.; King, R.W.; Gygi, S.P.; Finley, D. Ubiquitin chains are remodeled at the proteasome by opposing ubiquitin ligase and deubiquitinating activities. Cell 2006, 127(7), 1401–1413. 120. Denison, C.; Rudner, A.D.; Gerber, S.A.; Bakalarski, C.E.; Moazed, D.; Gygi, S.P. A proteomic strategy for gaining insights into protein sumoylation in yeast. Mol. Cell Proteomics 2005, 4(3), 246–254. 121. Hoeller, D.; Crosetto, N.; Blagoev, B.; Raiborg, C.; Tikkanen, R.; Wagner, S.; Kowanetz, K.; Breitling, R.; Mann, M.; Stenmark, H.; Dikic, I. Regulation of ubiquitin-binding proteins by monoubiquitination. Nat. Cell Biol. 2006, 8(2), 163–169. 122. Villen, J.; Beausoleil, S.A.; Gerber, S.A.; Gygi, S.P. Large scale phosphorylation analysis of mouse liver. Proc. Natl. Acad. Sci. 2007, 104(5), 1488–1493. 123. Beausoleil, S.A.; Villen, J.; Gerber, S.A.; Rush, J.; Gygi, S.P. A probability-based approach for high-throughout protein phosphorylation analysis and site localisatopm. Nat. Biotechnol. 2006, 24(10), 1285–1292. 124. Chi, A.; Huttenhower, C.; Geer, L.Y.; Coon, J.J.; Syka, J.E.P.; Bai, D.L.; Shabanowitz, J.; Burke, M.J.; Trovanskaya, O.G.; Hunt, D.F. Analysis of phosphorylation sites on proteins from Saccharomyces cerevisiae by ETD mass spectrometry. Proc. Natl. Acad. Sci. 2007, 104, 2193–2198.
152
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
125. Molina, H.; Horn, D.M.; Tang, N.; Mathivanan, S.; Pandey, A. Global proteomic profiling of phosphopeptides using ETD tandem mass spectrometry. Proc. Natl. Acad. Sci. 2007, 104, 2199–2204. 126. Sweet, S.M.M.; Mardakheh, F.K.; Ryan, K.J.P.; Langton, A.J.; Heath, J.K.; Cooper, H.J. Targeted on-line liquid chromatography electron capture dissociation mass spectrometry for the localization of sites of in vivo phosphorylation in human Sprouty2. Anal. Chem. 2008, 19, 1312–1319. 127. Kjeldsen, F.; Savitski, M.M.; Nielsen, M.L.; Shi, L.; Zubarev, R.A. On studying protein phosphorylation patterns using bottom-up LC-MS/MS: The case pf human α-casein. Analyst 2007, 132, 768–776. 128. Bogdanov, B.; Smith, R.D. Proteomics by FT-ICR mass spectrometry: Top down and bottom up. Mass Spectrom. Rev. 2004, 24(2), 168–200. 129. Patrie, S.M.; Ferguson, J.T.; Robinson, D.E.; Whipple, D.; Rother, M.; Metcalf, W.W.; Kelleher, N.L. Top-down mass spectrometry of < 60 kDa proteins from Methanosarcina acetivorans using quadrupole FTMS with automated octopole CAD. Mol. Cell. Proteomics 2006, 5, 14–25. 130. Ge, Y.; ElNaggar, M.; Sze, S.K.; Oh, H.B.; Begley, T.; McLafferty, F.W.; Boshoff, H.; Barry, C.E. Top down characterization of secreted proteins from Mycobacterium tuberculosis by ECD mass spectrometry. J. Am. Soc. Mass Spectrom. 2003, 14, 253–261. 131. McLafferty, F.W. High resolution tandem FT-MS above 10 kDa. Acc. Chem. Res. 1994, 27, 379–386. 132. Winger, B.E.; Hofstadler, S.A.; Bruce, J.E.; Udseth, H.R.; Smith, R.D. High resolution accurate mass measurements of biomolecules using a new electrospray ionization ion cyclotron resonance mass spectrometer. J. Am. Soc. Mass Spectrom. 1993, 4, 566–577. 133. Marshall, A.G.; Senko, M.W.; Li, W.Q.; Li, M.; Dillon, S.; Guan, S.; Logan, T.M. Protein molecular mass to 1 Da by 13C, 15N double-depletion and FT-ICR mass spectrometry. J. Am. Chem. Soc. 1997, 119, 433–434. 134. Forbes, A.J.; Patrie, S.M.; Taylor, G.K.; Kim, Y.B.; Jiang, L.; Kelleher, N.L. Targeted analysis and discovery of posttranslational modifications in proteins from methanogenic archaea by top-down MS. Proc. Natl. Acad. Sci. 2004, 101(9), 2678–2683. 135. Sharma, S.; Simpson, D.C.; Tolic, N.; Jaitly, N.; Mayampurath, A.N.; Smith, R.D.; PasaTolic, L. Proteomic profiling of intact proteins using WAX-RPLC 2-D separations and FT-ICR mass spectrometry. J. Proteome Res. 2007, 6, 602–610. 136. Roth, M.J.; Forbes, A.J.; Boyne, M.T.; Kim, Y.B.; Robinson, D.E.; Kelleher, N.L. Precise and parallel characterisation of coding polymorphisms, alternative splicing, and modifications in human proteins by mass spectrometry. Mol. Cell. Proteomics 2005, 4, 1002–1008. 137. Pesavento, J.J.; Mizzen, C.A.; Kelleher, N.L. Quantitative analysis of modified proteins and their positional isomers by tandem mass spectrometry: Human histone H4. Anal. Chem. 2006, 78(13), 4271–4280. 138. Medzihradszky, K.; Zhang, X.; Chalkley, R.J.; Guan, S.; McFarland, M.A.; Chalmers, M.J.; Marshall, A.G.; Diaz, R.L.; Allis, C.D.; Burlingame, A.L. Characterisation of tetrahymena histone H2B variants and posttranslational populations by ECD FT-ICR MS. Mol. Cell. Proteomics 2004, 3(9), 872–886. 139. Romanova, E.V.; Roth, M.J.; Rubakhin, S.S.; Jakubowski, J.A.; Kelley, W.P.; Kirk, M.D.; Kelleher, N.L.; Sweedler, J.V. Identification and characterisation of homologues of vertebrate β-thymosin in the marine mollusk Aplysia californica. J. Mass Spectrom. 2006, 41, 1030–1040. 140. Zabrouskov, V.; Ge, Y.; Schwartz, J.; Walker, J.W. Unravelling molecular complexity of phosphorylated human cardiac troponin I by top down ECD/ETD mass spectrometry. Mol. Cell. Proteomics, 2008, 7, 1838–1849. Available on-line. DOI: 10.1074/mcp. M700524-MCP200
Analysis of 6 MS/MS Peptide–Polyphenols Supramolecular Assemblies: Wine Astringency Approached by ESI-IT-MS Benoît Plet and Jean-Marie Schmitter Contents 6.1 Introduction................................................................................................... 153 6.2 Materials........................................................................................................ 156 6.2.1 Reference Compounds....................................................................... 156 6.2.2 Peptide Synthesis............................................................................... 156 6.3 Methods......................................................................................................... 156 6.3.1 Sample Preparation............................................................................ 156 6.3.2 Energy-Resolved Mass Spectrometry (ERMS)................................. 157 6.4 Internal Energy of Ions.................................................................................. 158 6.5 Supramolecular Species Observed by Electrospray Ionization Mass Spectrometry (ESI-MS)...................................................... 160 6.6 Relative Affinity Scale in the Gas Phase....................................................... 161 6.7 Relative Affinity Scale and Astringency....................................................... 164 6.8 Conclusion..................................................................................................... 164 Acknowledgments................................................................................................... 165 References............................................................................................................... 165
6.1 INTRODUCTION Among organoleptic properties used for quality assessments of wines, astringency deserves particular attention. This complex sensation in the mouth occurs after tasting red wines, which are known for their characteristic high concentration of polyphenolic compounds called tannins. Neither the perception of astringency nor its molecular basis is understood fully, but it is accepted 153
154
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
generally that astringency arises when lubricating salivary proteins are precipitated as protein–polyphenols complexes. Astringency of red wines is estimated usually by expert wine tasters, and a fully objective analytical methodology is lacking in this domain. Wine tannins are polyphenolic compounds extracted from the skins and seeds of grapes during the process of wine-making. These phenolic compounds exhibit a wide diversity of structures having a flavonoid skeleton (Figure 6.1). Salivary proteins belonging to the proline rich proteins (PRPs) family that are major components of parotid and submandibular saliva in humans, are suspected of having an important role in the astringency phenomenon, and especially the basic PRPs [1–4]. High proline content, various repeated motifs, and common peptidic sequences are characteristic features of PRPs [5]. It has been suggested that PRPs act as a first defense line against dietary tannins by forming complexes with them and, thereby, avoiding their interaction with other body components [6,7]. Protein–polyphenols interaction has been probed by various analytical methods that include high performance liquid chromatography (HPLC) [4,8], sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) [9], binding assay [2,10–11], nephelometry [12–15], and nuclear magnetic resonance (NMR) spectrometry [1,3,16,17], but with protein models that were often far from giving a realistic representation of salivary proteins: bovine serum albumin (BSA) [18,19], gelatin [20,21], casein [1,22], polyvinylpolypyrrolidone (PVPP) [23,24], bradykinin [25,26], or diverse synthetic proteins [27,28]. Electrospray ionization mass spectrometry (ESI-MS) has a well-known ability to handle non-covalent assemblies [29]. The determination of stoichiometry, identification of binding partners and site specificity, analysis of protein folding, and determination of affinity constants are all examples of characterization of supramolecular assemblies by mass spectrometry [30–32]. The objective of this work was to gain further insight at a molecular level into the interaction of polyphenols with PRPs, and to progress in the design of a future analytical methodology for the evaluation of astringency. Following this view, we present here a study of the non-covalent interaction of polyphenols with synthetic proline rich protein (PRP) segments by means of ESI-MS. Complexes formed in solution were ionized in an electrospray source, transferred to an ion trap and disrupted by
3´ 2´ 8 7 6
9 A 5
10
O C
1´ 2
4´ B 6´
3
4
FIGURE 6.1 Structure and numbering scheme for flavonoids.
5´
155
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies OH OH OH O O O OH O HO OH CH3 O OH OH O OH O O OH OHOH Naringenin (N) OH OH Catechin (C) OH Naringenin-7-O-neohesperidoside (NN) OH O HO OH OH O HO O O OH OH O HO OH OH OH O Epicatechin (E) 7,4'-Dihydroxyflavone (F) OH OH O OH Luteolin-4'-O-glucoside (LG4) OH O HO OH OH OH OH O HO OH OH O HO OH OH O HO CH3 O OH O O OH OH O OH OHOH Quercetin (Q) OH OH OH C2 O HO OH Quercetin-3-O-rhamnoside (QR) OH
HO
HO
O
O
OH
HO OH OH
OH O Apigenin (A) HO
O
O
HO OH OH
O
OH
Daidzein (D)
HO OH
O
O
OH
O Daidzein-7-O-glucoside (DG)
O
OH OH
O
O
O
OH OH
OH OH HO OH
OH O Luteolin (L)
OH HO
O O
OH
B1
OH
HO
O
OH OH
OH O Luteolin-7-O-glucoside (LG7)
OH
O OH
B2
OH OH OH O OH
OH OH HO
OH
O OH
B3
OH OH OH O
HO
CH3 O OHOH O OH
O OH OH OH Quercetin-3-O-rutinoside (QR2)
OH
OH OH HO
OH OH HO B4
OH OH
O
OH O
OH
OH
O
OH
OH O OHOH OH Apigenin-7-O-neohesperidoside (AN)
OH
OH OH HO
O
O
OH O
OH O
HO
HO
OH
OH CH3 O
O
OH HO
OH
O
OH O Apigenin-7-O-glucoside (AG)
O
OH
HO
OH
OH
FIGURE 6.2 Structures of the 21 polyphenols studied for their affinity toward PRP peptides.
collision-induced dissociation (CID). An analysis of these supramolecular assemblies by means of energy-resolved mass spectrometry (ERMS) [33–35], has led to the creation of a relative affinity scale of these peptide–polyphenol complexes. A panel of 21 polyphenolic compounds were selected, featuring various structural properties of natural polyphenols (from monomers to trimers, bearing hydroxyl groups at several locations, glycosylated or non-glycosylated, Figure 6.2). The affinity of these compounds toward PRPs was probed with three peptides obtained by solid-phase synthesis; the three peptides encompassed repeated sequences of basic PRPs. The use of peptides of variable length gave access to different levels of investigation, from the minimal model probe (IB714, 14 amino acids) up to the size of an entire PRP (IB8c, 61 amino acids) (see Figure 6.3). An affinity scale in the gas phase was produced with these reference compounds, providing the basis for an astringency scale built in an objective way, to be compared with results obtained from alternative methods, including wine tasting.
156
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
IB714 SPP GKPQGPP PQGG IB934 (Ac)GKPQGPP PQGG NQPQGPP PPP GKPQGPP PQGGNR IB8c SPP GKPQGPP PQGG NQPQGPP PPP GKPQGPP PQGGNKPQGPP PP GKPQGPP PQGGSKS RSA
FIGURE 6.3 Primary structures of basic PRP segments IB714, IB934, and full length IB8c protein prepared by solid-phase synthesis; the repeated sequences are underlined.
6.2 MATERIALS 6.2.1 Reference Compounds Quercetin, quercetin-3-O-rhamnoside, luteolin, luteolin-4′-O-glucoside, luteolin-7O-glucoside, apigenin, apigenin-7-O-glucoside, apigenin-7-O-neohesperidoside, daidzein, and daidzein-7-O-glucoside were obtained from Extrasynthèse (Genay, France). Naringenin, naringenin-7-O-neohesperidoside, ( + )-catechin, (–)-epicatechin, and quercetin-3-O-rutinoside were supplied by Sigma-Aldrich (Saint Quentin Fallavier, France). Dimer B1 (epicatechin-4α,8-catechin), B2 (epicatechin-4α,8-epicatechin), B3 (catechin-4α,8-catechin), B4 (catechin-4α,8-epicatechin), and trimer C2 (catechin-4α,8-catechin-4α,8-catechin) were synthesized following the procedure explained by Simon et al. [17]. Acetic acid and ethanol were supplied by Fluka (SaintQuentin Fallavier, France), acetonitrile (ACN) by Fisher Scientific (Geel, Belgium). Ultrahigh-quality water was obtained through an Elga (Le Plessis Robinson, France) Ultra Pure Lab device and used for all solutions. All the above compounds were used without further purification.
6.2.2 Peptide Synthesis Solid-phase syntheses of IB714, IB934, and IB8c peptides were performed on an Applied Biosystems peptide synthesizer 433 A (PE Biosystems, Courtaboeuf, France) using the Fmoc strategy (primary structures are given in Figure 6.3). The peptides were purified by reversed-phase liquid chromatography and their purity was checked by matrix-assisted laser desorption/ionization (MALDI)-MS and liquid chromatography (LC)-ESI-MS.
6.3 METHODS 6.3.1 Sample Preparation Flavonoid and peptide solutions were mixed immediately before infusion into the mass spectrometer ion source. The peptide/polyphenol molar ratio was 1:3 for IB714. So as to obtain a ‘wine like’ final solution, 222.5 µL of the pure peptide solution (22.5 µmol L –1 in water) were mixed with 27.5 µL of the polyphenol solution
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies
157
(5.45 mmol L −1 in ethanol/acetic acid v/v 91:9). Thus, 20 µmol L −1 of peptide and 60 µmol L −1 of flavonoid were interacting in a ‘wine like’ solution (water/ethanol/ acetic acid, 89:10:1). For the longer IB934 and IB8c probes, the peptide/flavonoid molar ratio was 1:7. The final mixtures were infused into the mass spectrometer at a constant rate of 5 µL min−1 for all experiments using a syringe pump (Kd Scientific, Holliston, MA).
6.3.2 Energy-Resolved Mass Spectrometry (ERMS) Experiments were performed with an LCQ Advantage ion trap mass spectrometer (Thermo Fisher, San Jose, CA) fitted with a standard orthogonal ESI source. Room temperature was kept constant at 295±1 K and we assumed that the ambient temperature inside the instrument was 298.6±0.4 K, as reported for the same instrument set-up [36]. Xcalibur, version 1.5 (Thermo Fisher), was used for data acquisition. Mass spectra were obtained in positive ion mode by using a spray voltage of 4.5 kV, a capillary temperature of 130˚C, a capillary voltage of 10 V, and a tube lens offset of 0 V. The qz -value was maintained at 0.25. Two microscans of 200 ms per scan were used. A 10 Th isolation width was used for precursor ion selection in all experiments and the ions were monitored at their exact masses. MS/MS experiments in the ion trap used CID with helium gas at 0.1 Pa. For ERMS experiments, precursor ions corresponding to peptide/polyphenol complexes with a 1:1 stoichiometry were selected and activated at different percentages of normalized collision energy [%NCE, Equation 6.1] in steps of 0.1% NCE for 30 ms.
% NCE = Vpp ×
30
[ 0.4 + 0.002( m/z )]
(6.1),
where Vpp (in Volts) is the peak-to-peak amplitude of the effective voltage applied to the end-cap electrode during the resonance excitation. As CID is an ergodic process, the internal energy acquired during collisions with buffer gas atoms is distributed over the vibrational and rotational degrees of the isolated ion. Thus, CID efficiency is related to the size of the system, that is, to the total number of vibrational degrees of freedom of the target ion (N = 3n–6, where n is the number of atoms in the complex). To balance this dependency, %NCE has been introduced for the LCQ ion trap mass spectrometers [37]. Here, to obviate this experimental correction, N was taken theoretically into account. First, considering mass-to-charge ratio of the complex, the effective voltage applied to the end-cap electrode (Vpp) was extracted from Equation 6.1. Then, Vpp was corrected (V′pp) for the degrees of freedom in the complex (N) relative to the IB714 peptide/catechin complex chosen as a reference (Nref ) as shown in Equation 6.2,
Vpp' = Vpp ×
N ref N
(6.2).
158
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
The raw ion current (ICRaw) was normalized with respect to a pseudo-total ion current comprising the sum of the ion current (IC) of the six major ions monitored that are identified products from the dissociation reactions (Equation 6.3), [38]
ICFinal =
ICRaw
∑
6
(6.3).
ICi
i =1
Each ERMS experiment for a particular complex was run in triplicate with very small average variation for the three data sets (ca 0.01 %NCE). Each data set was fitted using a Boltzmann sigmoidal function [38] available in Microcal Origin (Microcal Software, Inc. Northampton, MA) to give plots of the relative abundance of the precursor complex vs %NCE. The fits of the three data sets were averaged and used to obtain the %NCE, and then the V′pp, at which half of the complex had fragmented, to yield the dissociation energy, DE50, corresponding to dissociation of 50% of the complex. Variations of DE50 over long period of times were always below 1%; thus, no correction for such variations was applied.
6.4 INTERNAL ENERGY OF IONS Peptide/flavonoids supramolecular complexes are sufficiently stable to accept internal energy shifts without early disruption, and sufficiently fragile to undergo dissociation by a controlled ERMS procedure. Trapped ionized complexes were submitted to stepwise activation by CID, and the intensities of the precursor and product ions were monitored while increasing %NCE. DE50 values obtained in this manner, and used as an assumed measure of the affinity between PRP and polyphenols, reflect the energy required for disruption in the ion trap of ions formed by these supramolecular complexes. This dissociation energy is related to the internal energy of ions upon entry to the ion trap; it is anticipated that tuning of each of the ion source and the interface may affect the magnitudes of DE50 values. Consequently, the influence of instrumental parameters on the internal energy of ions was examined carefully so as to be able to tune the mass spectrometer for each peptide–tannin couple without perturbing the DE50 scale. Parameters identified previously to be responsible for internal energy shifts and/ or broadening of the energy distribution of ions were kept strictly constant. These parameters are: activation time, qz, space charge, type and pressure of buffer gas. Furthermore, infusion rate of samples was kept constant, as well as the flow rate of desolvation gas. The influence of the following parameters was investigated: temperature of the transfer capillary, capillary voltage, tube lens offset, isolation width of the ions, and charge state of the selected ion. Peptide–polyphenol complexes having a wide range of affinities were used to probe the contributions of these parameters to the internal energy of the ions. The temperature of the transfer capillary was varied over the range of 40–240˚C (Figure 6.4). The influence of capillary voltage and tube lens offset variation was examined within the range of –130 to + 140 V.
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies
159
Relative intensity
1.0 0.8 0.6 0.4 0.2 0.0 4
6
8
10 12 % N.C.E
14
16
18
FIGURE 6.4 Superimposition of breakdown curves determined for the 1:1 IB714:catechin complex when the heated capillary temperature was varied throughout the range 40–240˚C in increments of 10˚C.
A clear conclusion was drawn from these experiments: capillary temperature, capillary voltage, and tube lens offset had no influence on DE50 values within the experimental error. We noticed only that the DE50 standard deviation (normally around 0.001, that is, 0.01 %NCE) was increased slightly (0.013, that is, 0.1 %NCE) when capillary temperature, capillary voltage, and tube lens offset were changed from one experiment to another. This situation is very convenient, because variations of these parameters may induce huge improvements in ionization efficiencies without compromising the affinity scale. However, the isolation width of precursor ions (∆) had a strong influence on the internal energy of these ions. This parameter is used to set the limits of the resonance excitation voltage amplitudes for the ejection of undesired ions from the ion trap during the isolation step: only ions having (m/z±0.5∆) values are retained in the ion trap. As expected, narrow isolation widths resulted in the transfer of more internal energy to the isolated ion than did wide isolation widths. The amplitude of the %NCE shifts as a function of the isolation width produced DE50 shifts corresponding to the differences found between DE50 values of several peptide/polyphenol complexes (Figure 6.5). Consequently, the isolation width was set at ∆ = 10 Th and was held constant for all experiments. For the longer peptide probes, a shift in the breakdown curves of precursor ions was observed also when moving from 3 + to 4 + charged ions, indicating that the number of charges of the ion has an influence on the magnitudes of the DE50 values (Figure 6.6). However, even when the magnitudes of DE50 differed and the shapes of the curves differed, the relative order in the affinity scale remained unchanged whatever the charge state of the peptide. Therefore, ERMS studies can be performed for any charge state, but the same global charge state must be kept constant for a given set of experiments. Consequently, there is no perturbation of the scale if larger proteins that produce ions with a higher number of charges are used in this method.
160
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Relative intensity
1.0 0.8
∆(Th)
0.6
3 5 10 20
0.4 0.2 0.0
10
12
14 % N.C.E
16
18
FIGURE 6.5 Breakdown curves determined for a 1:1 IB934: naringenine-7-O-neohesperidoside complex with isolation widths for the precursor ion set at 3, 5, 10, and 20 Th.
1.0
0.8
0.8
Relative intensity
Relative intensity
1.0
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0 8
10
12
14 16 % N.C.E
18
20
8
10
12
14 16 % N.C.E
18
20
FIGURE 6.6 Breakdown curves for the 3 + (left) and 4 + (right) charge states of the IB934: naringenin complex. For each charge state, the stoichiometry was varied from 1:1–1:4; (n = 1:1; o = 1:2; l = 1: 3; = 1: 4).
6.5 SUPRAMOLECULAR SPECIES OBSERVED BY Electrospray Ionization Mass Spectrometry (ESI-MS) Optimization of ionization, desolvation, transmission, and detection parameters modifies greatly the relative abundances of supramolecular species in ESI-MS spectra. In the present case, three types of ions detected were: (i) those from the bare peptide; (ii) aggregates of polyphenols; and (iii) peptide–polyphenol complexes. A given range of parameters is appropriate for the detection of each ion type. The tailoring of experimental conditions so as to maximize the formation of a given ion species is a typical example of the ion suppression effect that arises from the management of the electrospray cloud. In all cases studied during this work, the instrumental parameters were tuned in such a way that favored the detection of peptide–tannin complexes.
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies 1.40
161
Aggregates of two flavonoids
1.20
Complexes 1:1 with IB714
DE50
1.00 0.80 0.60 0.40
E
LG4
DG
LG7
AG
QR
AN
B3
NN
B1
QR2
B2
0.00
C2
0.20
FIGURE 6.7 Comparison of DE50 values determined for flavonoid aggregates (black bars) and IB714: flavonoid complexes (white).
Further, peptide–polyphenols complexes can be considered in different ways. For example, an aggregate of four stacked tannins may interact with the peptide on a single site, producing a ‘vertical’ species. A different situation pertains when four individual polyphenols interact in four distinct places with the peptide, producing a ‘horizontal’ species. The two types of complex have the same mass-to-charge ratio, but the conclusions that could be drawn about the interaction are different. Whatever the stoichiometry, all peptide polyphenols complexes stored in the ion trap have undergone dissociation from a 1:n peptide:polyphenol ratio to a 1:(n–1) ratio. Within the range of experimental parameters used here, a transition from a 1:n peptide– polyphenol complex direct to the bare peptide that could reveal a ‘vertical’ aggregate of n tannin molecules linked to the IB714 peptide was never observed. This behavior can be explained by the fact that polyphenol aggregates are dissociated more easily during CID experiments than are complexes with PRP peptides (Figure 6.7). Moreover, molecular dynamics runs (5 ns duration) with a system initially constituted by one IB714 peptide and four molecules of a catechin dimer (B3) ended up in a 1:3 stoichiometry with a ‘horizontal’ distribution, without observation of stacking between flavonoid molecules (the fourth tannin was ejected after 2 ns) [17]. However, observation of complexes with stoichiometries as high as 1:16 for IB714:catechin suggests that both types of behavior must be considered. Such complexes could not have been obtained with a 14 amino acid long peptide without aggregation of polyphenols upon polyphenols linked previously on the peptide. Therefore, so as to avoid the uncertainties associated with the possible spatial arrangements of polyphenols within the peptide–polyphenol complex, CID experiments were realized solely with peptide–polyphenol complexes having a 1:1 stoichiometry.
6.6 RELATIVE AFFINITY SCALE IN THE GAS PHASE Complexes of 21 flavonoids with IB714, formed in a ‘true-to-life’ solvent that mimicked the real conditions of interaction in wine, have been characterized by ERMS.
162
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
The experimental DE50 values observed are listed in Table 6.1. The lowest value for DE50 is attributed to naringenin and the highest value for DE50 to C2. Precise DE50 values were obtained for each IB714:flavonoid complex, reflecting even small structure differences between flavonoids (Table 6.1) as highlighted by the difference between the DE50 values determined for epicatechin and catechin, these isomers being differentiated only by their stereochemistry at C3. Quantitative discussion is not yet possible as we are dealing with relative values: however, current molecular modeling and NMR experiments should provide information about the apparent high degree of precision in the DE50 values. Three groups of compounds may be distinguished: (i) simple flavonoids; (ii) osidic flavonoids (having a bond between two glucose units); and (iii) oligomeric flavonols. Although DE50 values have been corrected by taking into account the size
TABLE 6.1 DE50 Values were Determined as Described Under Section 6.3.2. Methods for 1:1 Peptide: Polyphenol Complexes. Relative Standard Deviations Never Exceeded 0.001 and are Therefore, not Reported in the Table
Polyphenol
Abbreviation
DE50 IB714: Polyphenol Complex
Naringenin Daidzein
N D
0.694 0.713
Apigenin
A
0.726
7,4′-dihydroxyflavone
F
0.734
Quercetin
Q
0.750
Luteolin
L
0.797
Catechin Epicatechin
C E
0.794 0.817
Daidzein-7-O-glucoside
DG
0.827
Luteolin-4′-O-glucoside
LG4
0.830
Apigenin-7-O-glucoside Luteolin-7-O-glucoside
AG LG7
0.840 0.841
Quercetin-3-O-rhamnoside
QR
0.872
Apigenin-7-O-neohesperidoside
AN
0.840
B3 Naringenin-7-Oneohesperidoside B1 B4
B3 NN
0.936 0.995
1.356 1.346
B1 B4
0.988 0.985
1.363
Quercetin-3-O-rutinoside B2
QR2 B2
0.991 1.014
1.303
C2
1.082
1.593
C2
DE50 IB934: Polyphenol Complex 0.848
1.037
1.104
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies
163
of polyphenols, larger molecules display a higher DE50 value and, thus, a higher affinity for the peptide. This clear trend is illustrated by the affinity order: C2 (catechin trimer) > B3 (dimer) > catechin, as well as B2 > epicatechin. Furthermore, the difference found between two given monomers is transmitted to the dimers, that is, the dimer of epicatechin is bound more tightly to the peptide than is the dimer of catechin. Moreover, the DE50 value for the B1 heterodimer is exactly equidistant from the two homodimers, just like B4 (catechin–epicatechin dimer). Thus, C2 and C3 stereochemistry is definitely a very important feature in the recognition process. Among those compounds modified with glycans, a larger osidic group (such as rutinose or neohesperidose as shown in Figure 6.2) increases the affinity. It is noticeable for the luteolin structure that a switch of the osidic ring from position 7 through 4′ does not induce a higher DE50 value, but the set of compounds available for examination was not sufficient to draw any conclusions in a pragmatic way concerning the contributions of such structure variations to the relative affinity toward the peptide target. In the series of simple flavonoids, DE50 values provide information on the influence of hydroxyl group position, the presence of a ketone, and the rigidity of the skeleton with respect to the C2–C3 double bond (in the C-ring, see Figure 6.1). Epicatechin, catechin, luteolin, and quercetin bear a hydroxyl group in position 3′ (B-ring), and each of their DE50 values (above 0.739) are higher than those of flavonoids that are not hydroxylated at this position of the B-ring. Whatever the number and position of hydroxyl groups, this feature of 3′-hydroxylation seems to induce an enhanced affinity for IB714. Rigidity of the polyphenol skeleton does not appear to be an important factor for the affinity toward IB714. While a keto-group at C4 contributes to skeleton rigidity, the presence of a keto-group in luteolin affects little the DE50 value of 0.797 for luteolin when compared with that of 0.794 for catechin for which a keto-group is absent. The same conclusion applies for the rigidity of the flavonoid skeleton arising from the presence of a C2–C3 double bond: the difference between DE50 values of naringenin (no C2–C3 double bond) and apigenin (with a C2–C3 double bond) is merely 0.032, yet these compounds have a lower affinity toward IB714 than do catechin and epicatechin. Further, an affinity inversion is observed when an osidic group is present in the structure (apigenin > naringenin versus apigenin-7-Oneohesperidoside < naringenin-7-O-neohesperidoside). Also, we may observe the effect of different positions of the B-ring, as is the case for daidzein (7,4′-dihydroxyisoflavone) and 7,4′-dihydroxyflavone; the linkage of the B-ring to C2 leads to a higher DE50 value than when the B-ring is linked to C3. This conclusion is supported by the DE50 values determined for daidzein-7-O-glucoside and apigenin-7-O-glucoside. Changing the short IB714 probe for the 34 amino acids peptide IB934 does not affect the relative order of DE50 values (as is shown in Table 6.1). Data acquired with the full length PRP IB8c probe confirmed this result (data not shown). The short synthetic IB714 segment, featuring the shortest ubiquitous motif in basic PRPs population, appears to mimic the behavior of basic PRP; thus, the short IB714 segment can be used with confidence for probing the interactions between these proteins and polyphenols.
164
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
6.7 RELATIVE AFFINITY SCALE AND ASTRINGENCY When dealing with supramolecular assemblies, the consistency of conclusions drawn by mass spectrometric analysis in the gas phase and applied to the situation in solution is always a major concern. However, sensory descriptive analyses [39–42] permit astringency to be related to our affinity scale that has been constructed on the basis of interactions between salivary proteins and selected polyphenols. First, it is known that the degree of oligomerization of the catechin and epicatechin monomers is related to the intensity of astringency assessed by means of descriptive sensory analysis [39–41]. Second, in the astringency phenomenon, epicatechin is described as being more effective than is catechin in its ability to bind salivary proteins [43,44]. Third, Simon et al. [17]. have shown that B3 dimer binds to the hydrophilic side of the saliva peptide IB714 suggesting that major forces are governed by hydrogen bonds. This third conclusion is supported by our observation that flavonoids of high degree of hydroxylation are linked more tightly to the peptide than are those of lower degree of hydroxylation (Table 6.1 and Figure 6.2). The position, the stereochemistry, and the flexibility of hydroxyl groups seem to be of prime importance in the interaction with peptides, whatever the size. It is of interest to note that epigallocatechin, an important component of red wines, has a third hydroxyl group on the B-ring and is correlated negatively with astringency [42]. In the same way, the positive influence of the hydroxyl group at C3 of the flavonoid skeleton can be refined: in catechin and epicatechin, flexibility and availability of this hydroxyl group favors the formation of hydrogen bonds. This conclusion is confirmed by the weaker affinity observed for quercetin, which is constrained by the presence of the C2–C3 double bond and the keto-group in the C-ring.
6.8 CONCLUSION Provided that the internal energy of peptide:polyphenol complexes having a 1:1 stoichiometry is controlled carefully, ERMS of these ions provides a reliable method for assessing the affinity of flavonoids toward peptides belonging to the PRP family of human salivary proteins. By comparing peptide probes of variable lengths, the experimental reliance on the use of the shortest PRP probe (14 amino acids long) appears to be sound in that the shortest PRP probe mimics the interaction of salivary proteins with polyphenols. In particular, relative positions of model polyphenols on the affinity scale appear to be in agreement with available results of descriptive sensory analysis. More compounds with dedicated structural features must be screened in order to improve further our knowledge of peptide:polyphenol complexes. A better understanding of the interaction involved will allow us, on the one hand, to select from among the huge number of wine components, many of which have not yet been identified, the narrow range of compounds that are involved potentially in the astringency phenomenon. On the other hand, further research could lead to the discovery of new proteins of the human salivary proteome, which is currently under investigation, that are involved in this tactile sensation.
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies
165
ACKNOWLEDGMENTS The authors gratefully acknowledge the ‘Comité’ Interprofessionel des Vins de Bordeaux’ for a research grant allocated to Benoît Plet and the ‘Région Aquitaine’ for financial support. They also thank Yann André and Karine Barathieu for providing synthetic dimer and trimer polyphenols, Stéphane Chaignepain and Katell Bathany for their daily expertise and Emmanuel Geneste for his help in data processing.
REFERENCES
1. Baxter, N.J.; Lilley, T.H.; Haslam, E.; Williamson, M.P. Multiple interactions between polyphenols and a salivary proline-rich protein repeat result in complexation and precipitation. Biochemistry 1997, 36, 5566–5577. 2. Hagerman, A.E.; Butler, L.G. The specificity of proanthocyanidin-protein interactions. J. Biol. Chem. 1981, 256, 4494–4497. 3. Charlton, A.J.; Baxter, N.J.; Khan, M.L.; Moir, A.J.G.; Haslam, E.; Davies, A.P.; Williamson, M.P. Polyphenol/peptide binding and precipitation. J. Agric. Food Chem. 2002, 50, 593–601. 4. Kallithraka, S.; Barker, J.; Clifford, M.N. Evidence that salivary proteins are involved in astringency. J. Sens. Stud. 1998, 13, 29–43. 5. Bennick, A. Interaction of plant polyphenols with salivary proteins. Crit. Rev. Oral Biol. Med. 2002, 13, 184–196. 6. Mehansho, H.; Butler, L.G.; Carlson, D.M. Dietary tannins and salivary proline-rich proteins: interactions, induction, and defense mechanisms. Annu. Rev. Nutr. 1987, 7, 423–440. 7. Shimada, T. Salivary proteins as a defense against dietary tannins. J. Chem. Ecol. 2006, 32, 1149–1163. 8. Kallithraka, S.; Barker, J.; Clifford, M.N. Interaction of ( + )-catechin and (-)-epicatechin, procyanidin B2 and procyanidin C1 with pooled human saliva in vitro. J. Sci. Food Agric. 2000, 81, 261–268. 9. Gambuti, A.; Rinaldi, A.; Pessina, R.; Moio, L. Evaluation of aglianico grape skin and seed polyphenol astringency by SDS-PAGE electrophoresis of salivary proteins after the binding reaction. Food Chem. 2006, 97, 614–620. 10. Naurato, N.; Wong, P.; Lu, Y.; Wroblewski, K.; Bennick, A. Interaction of tannin with human salivary histatins. J. Agric. Food Chem. 1999, 47, 2229–2234. 11. Yan, Q.; Bennick, A. Identification of histatins as tannin-binding proteins in human saliva. Biochem. J. 1995, 311, 341–347. 12. Monteleone, E.; Condelli, N.; Dinnella, C.; Bertuccioli, M. Prediction of perceived astringency induced by phenolic compounds. Food Qual. Prefer. 2004, 15, 761–769. 13. De Freitas, V.; Mateus, N. Structural features of procyanidin interactions with salivary proteins. J. Agric. Food Chem. 2001, 49, 940–945. 14. De Freitas, V.; Carvalho, E.; Mateus, N. Study of carbohydrate influence on proteintannin aggregation by nephelometry. Food Chem. 2003, 81, 503–509. 15. Carvalho, E.; Mateus, N.; Plet, B.; Pianet, I.; Dufourc, E.; De Freitas, V. Influence of wine pectic polysaccharides on the interactions between condensed tannins and salivary proteins. J. Agric. Food Chem. 2006, 54, 8936–8944. 16. Wroblewski, K.; Muhandiram, R.; Chakrabarthy, A.; Bennick, A. The molecular interaction of human salivary histatins with polyphenolic compounds. Eur. J. Biochem. 2001, 268, 4384–4397.
166
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
17. Simon, C.; Barathieu, K.; Laguerre, M.; Schmitter, J.M.; Fouquet, E.; Pianet, I.; Dufourc, E.J. Three-dimensional structure and dynamics of wine tannin-saliva protein complexes. A multitechnique approach. Biochemistry 2003, 42, 10385–10395. 18. Chen, Y.; Hagerman, A.E. Characterization of soluble non-covalent complexes between bovine serum albumin and beta-1,2,3,4,6- penta-O-galloyl-D-glucopyranose by MALDIToF MS. J. Agric. Food Chem. 2004, 52, 4008–4011. 19. Papadopoulou, A.; Green, R.J.; Frazier, R.A. Interaction of flavonoids with bovine serum albumin: a fluorescence quenching study. J. Agric. Food Chem. 2005, 53, 158–163. 20. Edelmann, A.; Lendl, B. Toward the optical tongue: flow-through sensing of tanninprotein interactions based on FTIR spectroscopy. J. Am. Chem. Soc. 2002, 124, 14741–14747. 21. Llaudy, M.C.; Canals, R.; Canals, J.M.; Rozès, N.; Arola, L.; Zamora, F. New method for evaluating astringency in red wine. J. Agric. Food Chem. 2004, 52, 742–746. 22. Jobstl, E.; Howse, J.R.; Fairclough, J.P.A.; Williamson, M.P. Noncovalent cross-linking of casein by epigallocatechin gallate characterized by single molecule force microscopy. J. Agric. Food Chem. 2006, 54, 4077–4081. 23. Laborde, B.; Moine-Ledoux, V.; Richard, T.; Saucier, C.; Dubourdieu, D.; Monti, J.P. PVPP-polyphenol complexes: a molecular approach. J. Agric. Food Chem. 2006, 54, 4383–4389. 24. Siebert, K.J.; Lynn, P.Y. Comparison of polyphenol interactions with polyvinylpolypyrrolidone and haze-active protein. J. Amer. Soc. Brewing Chem. 1998, 56, 24–31. 25. Richard, T.; Lefeuvre, D.; Descendit, A.; Quideau, S.; Monti, J.P. Recognition characters in peptide-polyphenol complex formation. Biochim. Biophys. Acta. 2006, 1760, 951–958. 26. Verge, S.; Richard, T.; Moreau, S.; Richelme-David, S.; Vercauteren, J.; Prome, J.C.; Monti, J.P. First observation of non-covalent complexes for a tannin-protein interaction model investigated by electrospray ionisation mass spectroscopy. Tetrahedron Lett. 2002, 43, 2363–2366. 27. Sarni-Manchado, P.; Cheynier, V.R. Study of non-covalent complexation between catechin derivatives and peptides by electrospray ionisation mass spectrometry. J. Mass Spectrom. 2002, 37, 609–616. 28. Siebert, K.J.; Troukhanova, N.V.; Lynn, P.Y. Nature of polyphenol-protein interactions. J. Agric. Food Chem. 1996, 44, 80–85. 29. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F. Electrospray ionization-principles and practice. Mass Spectrom. Rev. 1990, 9, 37–70. 30. Loo, J.A. Studying noncovalent protein complexes by electrospray ionization mass spectrometry. Mass Spectrom. Rev. 1997, 16, 1–23. 31. Loo, J.A. Electrospray ionization mass spectrometry: a technology for studying noncovalent macromolecular complexes. Int. J. Mass Spectrom. 2000, 200, 175–186. 32. Veenstra, T.D. Electrospray ionization mass spectrometry in the study of biomolecular non-covalent interactions. Biophys. Chem. 1999, 79, 63–79. 33. McLuckey, S.A.; Glish, G.L.; Cooks, R.G. Kinetic energy effects in mass spectrometry/ mass spectrometry using a sector/quadrupole tandem instrument. Int. J. Mass Spectrom. Ion Phys. 1981, 39, 219–230. 34. Brodbelt, J.S.; Kenttamaa, H.I.; Cooks, R.G. Energy-resolved collisional activation of dimethyl phosphonate and dimethyl phosphite ions in a quadrupole ion trap and a triple quadrupole mass spectrometer. Org. Mass Spectrom. 1988, 23, 6–9. 35. Colorado, A.; Brodbelt, J. An empirical approach to estimation of critical energies by using a quadrupole ion trap. J. Am. Soc. Mass Spectrom. 1996, 7, 1116–1125. 36. Gabelica, V.; Karas, M.; De Pauw, E. Calibration of ion effective temperatures achieved by resonant activation in a quadrupole ion trap. Anal. Chem. 2003, 75, 5152–5159.
MS/MS Analysis of Peptide–Polyphenols Supramolecular Assemblies
167
37. Lopez, L.L.; Tyler, P.R.; Senko, M.W.; Schwartz, J.C. Automated strategies for obtaining standardized collisionally induced dissociation spectra on a benchtop ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 1999, 13, 663–668. 38. Kurimoto, A.; Daikoku, S.; Mutsuga, S.; Kanie, O. Analysis of energy-resolved mass spectra at MSn in a pursuit to characterize structural isomers of oligosaccharides. Anal. Chem. 2006, 78, 3461–3466. 39. Peleg, H.; Gacon, K.; Schlich, P.; Noble, A.C. Bitterness and astringency of flavan-3-ol monomers, dimers and trimers. J. Sci. Food Agric. 1999, 79, 1123–1128. 40. Vidal, S.; Francis, L.; Kwiatkowski, M.; Gawel, R.; Waters, E.J.; Guyot, S.; Marnet, N.; Cheynier, V. Taste and mouth-feel properties of different types of tannin-like polyphenolic compounds and anthocyanins in wine. J. Sci. Food Agric. 2003, 83, 564–573. 41. Vidal, S.; Francis, L.; Noble, A.; Kwiatkowski, M.; Cheynier, V.; Waters, E. The mouthfeel properties of grape and apple proanthocyanidins in a wine-like medium. Analyt. Chim. Acta. 2004, 513, 57–65. 42. Preys, S.; Mazerolles, G.; Courcoux, P.; Samson, A.; Fischer, U.; Hanafi, M.; Bertrand, D.; Cheynier, V. Relationship between polyphenolic composition and some sensory properties in red wines using multiway analyses. Analyt. Chim. Acta. 2006, 563, 126–136. 43. Thorngate, J.H.; Noble, A.C. Sensory evaluation of bitterness and astringency of 3R(-) -epicatechin and 3S( + )-catechin. J. Sci. Food Agric. 1995, 67, 531–535. 44. Kallithraka, S.; Bakker, J.; Clifford, M.N. Evaluation of bitterness and astringency of ( + )-catechin and (-)-epicatechin in red wine and in model solution. J. Sens. Stud. 1997, 12, 25–38.
and Dynamics 7 Structure of Trapped Ions Joel H. Parks Contents 7.1 Introduction................................................................................................... 169 7.2 Trapped-Ion Electron Diffraction (TIED)..................................................... 170 7.2.1 Technology......................................................................................... 171 7.2.1.1 Experimental Arrangement ............................................... 171 7.2.1.2 Electron Gun....................................................................... 172 7.2.1.3 Ion Trap Sequence............................................................... 173 7.2.1.4 Electron Interaction with Trapped Ions.............................. 175 7.2.1.5 Diffraction Detection and Analysis.................................... 176 7.2.2 Silver Clusters: Short Range to Global Order.................................... 178 7.2.3 Gold Clusters: Evolving Structural Forms........................................ 182 7.3 Trapped-Ion Fluorescence............................................................................. 186 7.3.1 Technology......................................................................................... 186 7.3.1.1 Trapped-Ion Fluorescence Spectroscopy............................ 187 7.3.1.2 Nanospray Ionization-Ion Trap Mass Spectrometry........... 190 7.3.2 Radiative Lifetimes and Conformational Fluctuations..................... 191 7.3.2.1 Polyproline Lifetime Measurements................................... 191 7.4 Conclusion..................................................................................................... 198 Acknowledgements................................................................................................. 199 References...............................................................................................................200
7.1 INTRODUCTION Our research group has developed a wide range of new experimental methods that are designed to be performed on ions stored within radio-frequency (RF) ion traps. In this chapter, we will emphasize the integration of trap technology within different experimental arrangements in order to perform unique scientific measurements. We will describe measurements of both trapped-ion electron diffraction (TIED) of metal cluster ions and radiative lifetimes of trapped biomolecular ions. Experiments will be discussed in sufficient detail to permit the advantages afforded by ion trap capabilities to be appreciated. This chapter is not intended as an extensive review of trap-related experimental measurements that are referred to and discussed elsewhere in the volumes of this series. Ion traps have enabled a remarkable range of experiments for studying the interaction of stored ions with both charged and neutral molecules, and with photons. 169
170
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Experiments performed by Dehmelt and co-workers include RF spectroscopy of stored ions [1] and single electrons [2,3], studies of electron and ion/neutral collisions [1], and fluorescence measurements of individual atomic ions [4]. Laser cooling of atomic ions [5,6], and high resolution laser spectroscopy of single trapped ions [7] by Wineland and co-workers motivated a broad range of scientific investigations in ion trap environments. Keeping pace with these applications, advances in ion trap technology for higher mass ions, such as those made by Cooks and co-workers [8], have been essential to realize flexible and reliable ion-cloud manipulation required for the electron diffraction and fluorescence measurements discussed in this chapter. The introduction of new ion sources, in particular metal-cluster aggregation sources [9] and electrospray ionization sources [10], provided unique opportunities to produce ion beams composed of metal atom clusters and biomolecules, respectively. The possibility to study the structures of these more complex species was of particular interest because it would permit us to investigate many-body properties and conformational dynamics on a size scale which is tractable theoretically. However, the sources mentioned above do not produce ions with both a specific mass and charge. Metal clusters are formed with a single charge but in a broad array of masses corresponding to the number of atoms, whereas biomolecular ions are generated for a single species in an ensemble of charge states. Because metal cluster structures are not only highly dependent on the number of atoms but differ for anions and cations, and because biomolecular conformations vary with charge state, a method for the isolation of a single ion species is essential for structural studies. The ion trap mass spectrometer presented an excellent opportunity both to isolate a specific ion species and to store a sufficient number of mass-selected ions for a sufficiently-long duration to accomplish physical measurements which could not be performed otherwise. Furthermore, collisions with the background helium gas equilibriated kinetic and internal degrees of freedom of the ion with the gas temperature to provide measurements of ion properties in thermal equilibrium. This issue is exceptionally important when characterizing ion structures and dynamics. In the following sections in this chapter, we discuss two general areas of study that have utilized an RF ion trap for the determination of the evolution of metal cluster structures with size (Section 7.2), and for probing local fluctuations of biomolecular conformations as a function of temperature (Section 7.3). In each of these areas, tradeoffs in trap design were introduced which optimized high signal-tonoise ratio (S/N) measurements of ion properties while maintaining acceptable mass spectrometric precision. This balance of ion trap capabilities was one of the most challenging issues in these experiments and the inclusion of such considerations in these applications of ion trap mass spectrometry is justified.
7.2 TRAPPED-ION ELECTRON DIFFRACTION (TIED) The size-dependent evolutions of the structures and physical and chemical properties of nanoscale clusters have been subjects of continuing basic and applied research interests. Structure determination is one of the outstanding challenges of cluster
Structure and Dynamics of Trapped Ions
171
science, and it has been approached in several ways. Structural determination via analysis of diffraction measurements is most desirable, because the scattering data are related to the spatial arrangement of the scattering atoms. This relationship allows for a direct comparison between the experimentally-measured data and theoretical structure calculations, because the measured diffraction, or interference patterns, are the Fourier transform of the pair correlation function and thus are related directly to atomic positions. However, as will be shown in discussion below, it is necessary to compare the measured diffraction patterns with patterns calculated for theoretically-optimized structures in order to identify the most probable atomic arrangement. Diffraction measurements of the structure of mass-selected atomic clusters have been particularly challenging, in part, because cluster sources produce beams with broad cluster-size distributions of atomic number with uncertain internal energy distributions. Electron diffraction studies performed previously on clusters in supersonic beams [11–14] required the full source output of the beam in order to detect signals of sufficiently high S/N. The resulting uncertainties in cluster size and internal energy prevented an unambiguous interpretation of the diffraction patterns. The development of a technique that utilizes the capabilities of an RF ion trap [15] takes advantage of current cluster source technologies yet avoids the shortcomings of beam measurements. The RF ion trap enables one to isolate a large number of size-selected clusters, to relax collisionally the vibrational energy distribution, and to store them for a time long enough to accumulate electron diffraction data which yields high S/N. A study of CsI clusters [16] has demonstrated that TIED of mass-selected clusters is particularly sensitive to size-dependent changes in structural symmetries. The technique of TIED has led to controlled investigations of quantum size effects providing the opportunity to probe the evolution of structure in metal clusters [17,18]. In Section 7.2.1, we discuss the technology assembled to develop a reliable and reproducible technique for the measurement of diffractive scattering from trapped clusters. In Sections 7.2.2 and 7.2.3, we summarize measurements on Agn+, n = 36–55 and Aunˉ, n = 11–24 clusters emphasizing details in the diffraction data and structure calculations which enable positive identification of the cluster structure.
7.2.1 Technology 7.2.1.1 Experimental Arrangement A detailed description of the diffraction technology has been presented elsewhere [19] and will be summarized here, concentrating on those aspects pertaining to the ion trap. The individual components of the current experimental apparatus are shown schematically in Figure 7.1. The metal cluster source comprises a liquid nitrogencooled aggregation tube enclosing a magnetron sputter discharge similar to that designed by the Haberland group [9]. Background helium gas within the aggregation tube extracts the heat of cluster formation and lowers the vibrational temperature. The cluster ion beam emitted by the source enters an electrostatic quadrupole bender that directs the beam either to a 1-m time-of-flight mass spectrometer (TOFMS) or
172
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Diffraction beamline Electron gun
CCD Camera Sputter/aggregation cluster source
Quadrupole bender
TOF Mass spectrometer 600 L/s Tmp
1000 L/s Tmp
FIGURE 7.1 The electron diffraction instrument consists of three main components: the UHV chamber housing the diffraction beamline, the time-of-flight mass spectrometer, and the sputter-aggregation cluster source. The ion beam propagation and various components are discussed in the text.
to the diffraction beamline. The TOFMS has sufficient resolution to optimize the cluster source parameters so that ion beam intensity can be peaked within a clustersize range of interest. The optimized beam is then deflected toward the diffraction beamline to load ions into the trap. The individual components of the diffraction beamline apparatus are shown schematically in Figure 7.2. All components are mounted on a structure composed of three 12.5-mm diameter titanium alignment rods which are fitted to the electron gun housing. This structure maintains parallel alignment of the various beamline apertures with the electron beam (e–-beam) direction by attaching the rods to a chamber flange as shown in Figure 7.2. The beamline components include a channeltron electron-multiplier detector, electrostatic bender, ion trap, Faraday cup, and microchannel plate (MCP)/phosphor screen detector which are mounted individually on the three-rod structure maintaining cylindrical symmetry around the e– -beam axis. All beamline components are fabricated from non-magnetic materials where possible and a μ-metal™ shield eliminates stray components of the earth’s magnetic field over the entire diffraction beamline to eliminate misalignment of the e–- beam. 7.2.1.2 Electron Gun The electron gun provides a collimated e–-beam of current ieb ca 0.6–1 μA and electron energy of 40.0 keV. The e–-beam traverses the trap through 2-mm diameter apertures in the end-cap electrodes and is focused gently into the Faraday cup. These
173
Structure and Dynamics of Trapped Ions Electron gun beamtube
Quadrupole bender
Alignment frame
Multi channel plate detector
Faraday cup UHV chamber flange
Electron multiplier detector
RF trap
UHV flange rail mounting
FIGURE 7.2 Components of the diffraction beamline mounted on rods to maintain a cylindrical symmetry around the electron beam axis. The e– -beam passes through the trapped ion cloud producing scattered electrons indicated by off-axis lines. The primary beam enters the Faraday cup and scattered electrons strike a multi-channel plate detector producing the diffraction pattern on a phosphor screen. This screen is imaged by a charge-coupled device camera mounted external to the UHV chamber. The electron multiplier detects ions resonantly ejected from the ion trap.
apertures have been designed to minimize scattering from both the ca 0.5-mm diameter e–-beam and forward-directed X-rays emitted by the gun. The Faraday cup is mounted on an X–Y translator to facilitate alignment. A primary effort during development of the TIED technique was directed toward minimizing the background electron scattering. The small number of trapped cluster ions (Nion ca 104) and their low density (ca 107 cm–3) results in a rate of elastic scattering relative to the incident e–beam current of ca 10 –9 which makes the design and positioning of each component critically important. The ratio of low background electron scattering current ie-back : ieb ca 2 × 10 –8 was achieved by fabricating a carbon Faraday cup to maximize the conversion of high energy electrons to X-rays which were then absorbed in the walls of the cup. The background helium gas is evacuated before the diffraction e–-beam exposure to a chamber pressure of ≤ 10 –8 Torr. The dominant detector background contributions came from high energy electrons scattered both from apertures and from residual gases in the ultra high vacuum (UHV) chamber. 7.2.1.3 Ion Trap Sequence The ion beam passes through the small bender, indicated in Figure 7.2, which redirects the beam to inject ions into the trap through an aperture of 2 mm diameter in the grounded end-cap electrode. The ion trap operates at an RF frequency of 300 kHz with an end-cap electrode spacing of 2z0 = 1 cm and ring diameter of r02 = 2z02 cm2. Cluster ions are loaded at qz = 0.3–0.5 depending on cluster mass with a range of RF amplitudes V0−p = 1–8 kV for gold and silver clusters. Ion loading in this trap
174
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
Agn+
2000
3000
4000
5000
6000
7000
3000
4000
5000
6000
7000
6000
7000
(b)
2000 (c)
Ag38+
2000
3000
4000 5000 Mass/charge
FIGURE 7.3 Mass spectra displaying (a) the detected TOF spectrum of cluster masses emitted by the source; (b) trapped cluster spectrum obtained by broadband SWIFT excitation; and (c) subsequent isolation of a single cluster mass using narrowband SWIFT excitation. In this example, silver cations of size n = 38 (dashed line) have been isolated.
configuration, with a background of ca 10 –3 Torr helium gas, yields ion clouds of ca 104 ions which represent an optimum tradeoff between ion number and space-charge limited density within the e–-beam-cloud overlap. The large number of different cluster sizes produced by the metal cluster source introduces sufficiently strong ion/ ion interactions to make the isolation of a single cluster size exceedingly difficult. For example, silver clusters, Agn+, are generated routinely over a range of cluster sizes 9 ≤ n ≤ 150. Figure 7.3 displays three mass spectra obtained from trapped silver cluster ions, Agn+. The mass spectrum in Figure 7.3a shows a range of silver cluster sizes n = 22–64 which are trapped at qz = 0.3 from the broad range of sizes generated by the source and detected by single frequency resonant ejection into the channeltron electron multiplier shown in Figure 7.2. The mass spectrum in Figure 7.3b shows a much reduced range of trapped sizes, n = 36–40, resulting from the presence of stored wave-form inverse Fourier transform (SWIFT) excitation [20,21] during the loading period, with a frequency width corresponding to a broadband mass window of 850 Th. The mass spectrum in Figure 7.3c shows the isolation of a single silver cluster ion, Ag38+, after two successive SWIFT excitations consisting of the broadband excitation followed by a second SWIFT excitation with a narrow
Structure and Dynamics of Trapped Ions
175
frequency bandwidth corresponding to 200 Th. This selection sequence resulted in an efficient, reproducible method to select a single cluster size from the large range generated by the metal cluster source. An attempt to isolate a single cluster size from the broad range of sizes trapped from the source beam as shown in Figure 9.3a is completely ineffective. It was found that loading the source beam in the presence of the broadband SWIFT excitation reduced the ion/ion interactions to a manageable level so that a single cluster size could be selected by applying the second narrowband SWIFT excitation. Upon storing the clusters for ca 5 s in helium gas at a pressure of ca 10 –3 Torr, the cluster vibrational and translational energies at the gas temperature become thermalized [22]. The gas temperature is set by channeling the helium gas through the trap superstructure which is heated and held at a constant temperature by ceramic heaters or cooled directly by flowing liquid nitrogen. For diffraction measurements of metal clusters, the clusters are relaxed collisionally to a temperature of ca 120 K. The cluster structures generated by the source have been formed in helium gas at ca 77 K to extract the heat of formation. The possibility that cluster structures are trapped within potential minima above the ground state structure is one of the more important uncertainties of structure measurements. Although such metastable clusters are heated somewhat by helium collisions as they achieve stable trajectories, these clusters can be annealed to minimize the effects of the cluster formation process. Such annealing has been found to be important for gold clusters but not for silver in our diffraction measurements. An RF potential of V0−p ca 0.2 V for ca 100 ms applied between the end-cap electrodes excites the translational motion, which heats the clusters by translational-to-vibrational (T-V) energy transfer in collisions with background neon gas at ca 10−3 Torr. After annealing, the cluster ensemble is cooled again to ca 120 K by helium gas collisions. After evacuating the helium background, the relaxed ion cloud is held at the trap parameter qz = 0.5 during the diffractive exposure. At this qz value, the ion cloud dimension is comparable to the e–-beam diameter so that the overlapped ion cloud volume includes a large fraction of the ions. 7.2.1.4 Electron Interaction with Trapped Ions The e–-beam passes through a trapped ion cloud producing scattered electrons indicated in Figure 7.2 by lines radiating from the axis. The total scattering of electrons from atoms in a cluster ion is, to a good approximation, composed of elastic scattering from each atom independently, modulated by the interference between scattering from each atomic pair, and an inelastic component arising from electron energy deposition in the cluster atoms. The inelastic processes are important to consider because they can lead to cluster heating and cluster fragmentation which introduce uncertainties in the analysis of the diffraction patterns. To identify these inelastic scattering processes, mass spectra were obtained before and after cluster exposure to the e–-beam. Figure 7.4 shows mass spectra of trapped Au55+ clusters before and after e–-beam exposure. The primary series of inelastic processes is observed to comprise multiple ionization steps accompanied by sequential losses of single atoms. The rate of this series of inelastic processes was found to increase with increasing cluster size. The fast electron interacts primarily with the electronic states of the cluster and fragmentation arises primarily from a
176
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Au+ 55 Au++ 55
Au+++ 55
4
6
8 Mass/charge
10
12×103
FIGURE 7.4 Mass spectra displaying the effect of inelastic electron scattering from gold cluster cations of size n = 55. The upper mass spectrum (dashed line) shows the trapped ion mass, Au55+, before e– -beam exposure. The lower mass spectrum was obtained after exposure and shows Au552+ and Au553+.
branching of this electronic excitation into the vibrational manifold. Note that direct vibrational heating is not observed to be a direct inelastic channel at these electron energies as evidenced by the single mass peak of the parent ion. This result for the parent ion has been observed for cluster sizes ranging from 10 to 75 atoms and materials including insulators and metals; consequently, electrons at these high energies couple predominantly to the dynamics of atomic electrons. In addition, no evidence for excessive heating is observed either in the diffraction patterns or in the comparison with calculated structures. A low-level SWIFT excitation is maintained during the diffraction exposure to eliminate possible fragments or multiply charged species produced by e–-beam inelastic processes. This low-level excitation ensures that the diffraction pattern is the result of the selected parent cluster ion only. However, as a result of the loss of parent ions during the diffraction exposure by these inelastic processes, each diffraction exposure is limited to ca 10–20 s after which the trap is reloaded with ions. To summarize, each diffraction cycle is composed of ion loading, mass selection, vibrational relaxation, annealing, e–-beam exposure, ion ejection, and a second identical sequence to record a detector background. In a typical experimental run over ca 4–5 h, the diffraction pattern is collected for ca 300–400 cycles representing an average over ca 3–4 × 106 mass-selected clusters. 7.2.1.5 Diffraction Detection and Analysis The scattered electrons strike the MCP as indicated in Figure 7.2 producing a diffraction pattern on the phosphor screen. This pattern is in the form of Debye– Scherrer rings similar to powder diffraction as a result of the orientational and spatial disorder of the trapped cluster ions. This screen is imaged by a charge-coupled device (CCD) camera mounted externally to the UHV chamber. The distance from the trap center to the MCP surface can be varied from ca 8 to 12 cm. The e–-beam crosssection and trap aperture limit the detection of scattered electrons to a maximum
177
Structure and Dynamics of Trapped Ions (a)
Diffraction
(b)
Background
(c)
Difference
FIGURE 7.5 CCD images produced by (a) cluster diffraction; (b) background in the absence of clusters; and (c) the average difference image of (a) and (b) images obtained from ca 400 trap cycles and corresponding to the diffraction from ca 2 × 106 Ag55+ clusters.
scattering angle of ± 8.1°. Figure 7.5 shows an example of the detection process for diffraction by Ag55+. Figure 7.5a shows a CCD image obtained with clusters present and Figure 7.5b shows an identical exposure without clusters. A difference image is obtained for each exposure cycle by subtracting the background image from the cluster image. The average of these difference images after ca 300 cycles, shown in Figure 7.5c, displays the diffraction pattern used for cluster structure analysis. Although the difference image subtracts the electron background, the statistical fluctuations of the background remain and determine the S/N ratio. The diffraction pattern analysis is covered thoroughly in Ref. [19] and will be only summarized here. Diffraction data are fitted to a model of the total scattering intensity, Itotal, approximated by the sum of independent scattering terms, Iindep, from each cluster atom, and terms resulting from the interference of elastic scattering from atomic pairs. Multiple scattering processes do not contribute significantly for clusters in the size range explored in metal cluster measurements. The data analysis extracts the interference, or molecular scattering, contribution to the total scattering intensity by approximating the independent scattering contribution by a polynomial in s, the momentum transfer [17,18,23]. The momentum transfer is related to the scattering angle, θ, by vs s = (4π/λ)sin(θ/2) Å–1 where λ = 0.06 Å is the de Broglie wavelength for 40 keV electrons. The resulting molecular scattering intensity, sM(s) is defined by sM(s) = s(Itotal−Iindep)/Iindep where Iindep is proportional to the atomic scattering factor. Figure 7.6 displays the cluster-size dependence of the molecular scattering intensity, sM(s) vs s calculated for (CsI)nCs+ clusters having body-centered cubic (bcc) symmetry. For smaller sizes, the weaker interference resulting from fewer pairs of CsI molecules yields relatively unstructured curves having fewer and broader peaks. As the number of interfering pairs grows, the pattern of peaks becomes increasingly denser and narrower and, ultimately, reproduces the sharp peaks of the powder pattern associated with the Bragg diffraction peaks of bulk CsI. Consequently, for cluster sizes in the range of n = 10–100 atoms, the diffraction patterns will display relatively few peaks. Nevertheless, as the data below will demonstrate, the s-values for peak positions and zero-crossings of sM(s), and the relative peak amplitudes provide sufficient information to choose between calculated isomer structures.
178
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Dodecahedral (Csl)nCs+ cluster closed shell
(211) (220) (310) (222) (321) (400) (330) (411) (420) (332) (422) (431) (510) (521) (440) (433) (530) (424) (532) (541) (631) (444) (543)
sM(s)
(200)
(110)
Csl bulk structure Bragg peaks
Fourth shell 184 Third shell 87 Second shell 32 First shell n=7 0
1
2
3
4
5
6
7
8
9
10
S (Å–1)
FIGURE 7.6 The molecular scattering curves associated with diffraction from closed shells of (CsI)nCs+ clusters. Bragg scattering peaks for bulk CsI are shown at the top of the figure.
As an example of the data analysis, Figure 7.7 shows diffraction data taken for neutral C60 clusters [15]. In this measurement, a beam of neutral C60 clusters emitted by an effusive Knudsen oven traverses the trap through 2 mm-diameter apertures in the ring electrode at right angles to the e–-beam. The total scattering intensity vs CCD pixel value is extracted from the data by averaging the intensity pattern over successive rings of CCD pixels centered on the pattern and concentric with the e–-beam axis. Pixel values are related to s by a scaling factor, which is determined by fitting to the known C60 structure having icosahedral symmetry. The total scattering vs s is shown in the upper part of Figure 7.7 together with the initial approximation for the independent scattering contribution. In the lower part of Figure 7.7, the experimental data are compared with the calculated molecular diffraction, sM(s), for the C60 structure producing the best fit of the scaling factor and the independent scattering term. The small residuals indicate the close fit of the data with the C60 structure; the experimental uncertainty shown includes only the standard deviation of the raw data. This example served to identify the experimental procedures and analysis methods that could be applied reliably to the measurements of metal cluster structures.
7.2.2 Silver Clusters: Short Range to Global Order Diffraction data were obtained from trapped Agn+ clusters [17] for a range of cluster sizes from n = 36–55 at a cluster temperature of 120 K. This range was chosen to gain a better understanding how cluster structures evolve through intermediate
179
Structure and Dynamics of Trapped Ions C60
Total scattering
8
Data Independent Scattering
6 4 2 0
Residuals molecular diffraction
1.0
4
6
8 10 s (Å–1)
12
14
Data Theory
0
0.8
∆sM(s) ±σ
0 2
4
6
s (Å–1)
8
10
12
FIGURE 7.7 Diffraction data for a neutral C60 beam. At the top, the total scattering intensity vs s is shown and the independent scattering contribution is indicated by the dashed curve. At the bottom, the experimental (solid curve) and theoretical (dashed curve) molecular scattering intensity, sM(s), are compared. The difference ∆sM(s) is shown in with the uncertainty of ±1σ (grey band) displayed at each data point.
sizes to achieve ‘magic number’ structures composed of closed electronic or atomic shells. This size range includes two such closings, an octahedral atomic shell closing at Ag38+ and an icosahedral atomic shell closing at Ag55+. The total scattering amplitudes for all the Agn+ clusters are shown in Figure 7.8. Note the change in shape of the curves in the region s = 3.5–4 Å–1 with increasing cluster size. At cluster size n = 42, this region starts to exhibit a double-peaked behavior which becomes more pronounced as the size increases. This doublet is characteristic of icosahedral symmetry and is observed also in the C60 molecular scattering in the region s = 7.0– 8.0 Å–1; a structure known to have icosahedral symmetry. In addition, these changes in the doublet structure can be shown analytically to arise from increased interference as the fraction of Ag atoms having five-fold coordination increases. The molecular scattering patterns, sM(s), for cluster sizes n = 37 to 39, 43 and 55 were fitted to structures calculated [17] by I.L. Garzón and K. Michaelian, Instituto de Física, Universidad Nacional Autónoma de México. Structures for 6 to 10 lowest energy isomers were calculated for each cluster size by density functional theory. The best fit of the diffraction pattern data to patterns calculated for the isomers of each cluster size was found to occur for only a single isomer structure although the
180
Total scattering intensity (arb. units)
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
55 46 45 44 43 42 41 40 39 38 37 36 2
3
4
5 s (1/A)
6
7
8
FIGURE 7.8 Total scattering intensity vs s Å−1 for Agn+ cluster sizes n = 36–46, 55. The inset is the CCD image of the n = 55 diffraction pattern. (From Xing, X.; Danell, R.M.; Garzón, I.L.; Michaelian, K.; Blom, M.N.; Burns, M.M.; Parks, J.H. Phys. Rev. 2005, B 72, 081405 (1–4). With permission.)
fitted structure was not always that of the lowest energy isomer. Figure 7.9 exhibits the molecular scattering data accompanied by the sM(s) curves calculated from the best-fitted isomer structures. The calculated diffraction patterns in Figure 7.9 display a shoulder on the second positive peak near s ca 5 Å–1 that is identified with the presence of five-fold symmetry [24,25]. Each calculated best-fitted structure shown on the left-hand side of Figure 7.9 exhibits evidence of five-fold order, sometimes local as in sizes 37, 38, and 39, becoming more global in size 43 and, finally, a fully global icosahedral symmetry in size 55. The calculated structures exhibiting closed atomic shells at n = 38 and 55 are shown on the right side in Figure 7.9 to emphasize the global symmetry associated with these cluster sizes. These measurements indicate an evolution from short-range order among nearest neighbors having fivefold symmetry to a global order having icosahedral symmetry at n = 55. A unique structure is observed at n = 38 which exhibits a distortion of face-centered cubic (fcc) symmetry characterized by local order having five-fold symmetry. The Agn+ diffraction measurements revealed that five-fold order dominates the structures over the size range studied. The analysis presented in Ref. [19] indicates that the number of local five-fold symmetry sites within these structures increases
181
Structure and Dynamics of Trapped Ions Ag37+
2
3
4
5
6
7
2
3
4
5
6
7
2
3
4
5
6
7
8 Ag43+
2
3
4
5
6
7
8 Ag55+
2
3
4
5 6 s (1/A)
7
8 Ag38+
8
Ag39+
8
FIGURE 7.9 Experimental molecular intensity sM(s) vs s Å−1 is displayed for silver cluster sizes n = 37–39, 43, and 55 and the sM(s) calculated from best fit isomer structure The experimental data (black line) and uncertainty (±1 σ, gray band) are shown for each cluster size. The best-fit isomer space filling structure is shown to the left of each curve indicating atoms contributing to local order having five-fold symmetry by black dots and/or lines. Detailed structures for closed shells at n = 38 and 55 are shown on the right. (From Xing, X.; Yoon, B.; Landman, U.; Parks, J.H. Phys. Rev. 2006, B 74, 165423 (1–6). With permission.)
with size leading up to the closed atomic structure at n = 55 having global order. Furthermore, local order with five-fold symmetry evolves to global order at sizes forming structures with closed atomic shells, namely n = 38 and 55. It is not known at this point whether this behavior is characteristic of metal cluster structures in general, but it is an intriguing model of structural evolution to investigate further.
182
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
7.2.3 Gold Clusters: Evolving Structural Forms Diffraction data were obtained [18] also from trapped Aun+ clusters for a range of cluster sizes from n = 11 to 24 at a cluster temperature of 120 K. This range of cluster sizes has shown a remarkable size-dependent structural evolution among different symmetries and forms. These data and their analysis show clear evidence of a planar (two-dimensional (2D)) to three-dimensional (3D) transition over the range n = 12–14, caged structures for n = 16 and 17, the development of a tetrahedral structure at n = 20, and the emergence of a highly symmetric tubular structure at n = 24. These results are in general agreement with previous mobility [26] and photoelectron spectroscopy studies [27], and with those obtained through recent photoelectron spectroscopy measurements and density functional calculations [28,29]. The molecular scattering intensities, sM(s), extracted by the diffraction analysis are shown in Figures 7.10 through 7.12 for each cluster size. These data were obtained for annealed clusters; annealing was found to be necessary to obtain reproducible diffraction data in this size range. The diffraction patterns indicate that the second positive peak near s ca 5 Å–1 changes significantly within this cluster size range and presents a unique feature related to cluster symmetry. This peak shape and the relative peak amplitudes will be shown to be the dominant characteristics identifying the transitions between the various structural motifs within the size range n = 11–24. For each cluster size, sM(s) patterns calculated for each isomer structure were fitted to structures calculated by B. Yoon and U. Landman, School of Physics, Georgia Institute of Technology. At the bottom of Figures 7.10 through 7.12, the experimental data are compared with the calculated sM(s) for the isomer structure producing the best fit. The experimental uncertainty, shown as gray shading, includes only the standard deviations of the raw data. It is important to point out that only a single isomer made a significant contribution to the fits for all sizes except n = 12 and 13. A better fit was obtained for these two cluster sizes by including a second isomer. This sum of isomer contributions is assumed to represent an ensemble of trapped clusters composed of these different structures. In this case, the fitting procedure varied the fractional contribution of each isomer structure in the sM(s) model to optimize the fit to the experimental patterns. The best-fit isomer structure(s) is shown at the lower right for each cluster size. Figure 7.10 displays molecular scattering intensity data and associated calculated low-energy structures with best fits for clusters in the size range n = 11–14. Anionic gold clusters in this range have been the subject of theoretical [29,30] and experimental [26] investigations for the identification of the presence of large planar clusters. Experimental evidence for a transition from planar to 3D structures has been observed in measurements of cluster ion mobility [26] at n = 12 and photoelectron spectra [27] at n = 13. The best-fit structures to diffraction data in Figure 7.10 indicate that n = 11 is a pure planar structure and the transition to 3D starts with n = 12 which is a mixture of 54% planar and 46% 3D structures. At size n = 13, the mixture of structures is observed to be dominated by 3D (80%) with a smaller planar contribution (20%). Finally, the size n = 14 exhibits a pure 3D structure. These diffraction measurements are consistent with previous mobility and photoelectron results but show more clearly the mixture of isomer structures present in the cluster ensemble
183
Structure and Dynamics of Trapped Ions Au11–
Au12– 0.24
1.33
sM (s)
0.21
0.54
0.09
0.32 GS
GS
0.54 46%
GS 54%
GS Au13–
Au14– 0.20
0.27
sM (s)
0.17
0.23
0.10
0.10
GS
GS
0.10 80%
2
3
4
5 6 s(1/A)
7
8
GS 20%
2
3
4
5 6 s(1/A)
7
8
0.10
FIGURE 7.10 Calculated isomer diffraction patterns of sM(s) vs s Å−1 for structures of Aun− clusters with 11 ≤ n ≤ 14 are shown (dotted curves). The energy (eV) of the isomer above the ground state (GS) is shown to the right of each isomer diffraction pattern. The experimental diffraction sM(s) (black curve) along with the best fit isomer (dotted curve) are shown in the bottom panel. The grey shading shows the data uncertainty ±σ. The best-fit isomer structure is shown to the right of each fit.
throughout the planar to 3D transition region. The smooth nature of the transition is reflected by the occurrence of sets of closely spaced energy isomers displayed in Figure 7.10 for cluster sizes in the transition region Diffraction data shown in Figure 7.11 for the size range n = 15–17 identify cage structures as best fits to calculated isomers at each size. Evidence for the cage structures at n = 16 and 17 has been observed recently in photoelectron measurements [31]. The cage structure associated with the best fit for n = 15 is not consistent with structures suggested by these photoelectron measurements. However, in this case, the best fit to a unique calculated pattern lends support to the diffraction result. The diffraction patterns observed in Figure 7.12 for clusters in the size range n = 18–20 all exhibit a shoulder on the left side of the second peak, which is evidence
184
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Au16–
sM(s)
Au15–
2
3
4
5 6 s(1/A)
7
8
0.26
0.12
0.26
0.11
0.24
0.08
GS
GS
0.26
2
3
4
5 6 s(1/A)
7
8
0.08
Au17–
0.32
sM(s)
0.08 0.08 0.06 GS
GS 2
3
4
5
6
7
8
s(1/A)
FIGURE 7.11 Similar to Figure 7.10 for diffraction data and isomer structures for Aun− clusters with 15 ≤ n ≤ 17.
for the presence of fcc symmetry corresponding to the bulk gold structure. This symmetry has been identified [32] for the tetrahedral structure of the Au20− ground state, which gives the best fit to the diffraction pattern. The diffraction patterns for Au19− and Au18− are best fitted by similar tetrahedral structures in which a single vertex atom is missing or two vertex atoms are missing, respectively. However, with the addition of a single atom that is, for Au21−, the observed diffraction pattern changes in a radical fashion as shown in Figure 7.12. Although there are several lowenergy structures that display the presence of fcc symmetry, the best-fit structure is found to be an elongated cage. Figure 7.12 displays the diffraction patterns and calculated structures for Au24−, that present evidence for the emergence of an empty single-walled tube-like
185
Structure and Dynamics of Trapped Ions (a)
Au18–
Au19–
Au20–
0.10 sM(s)
0.07 0.03 0.02 GS
2
3
4
5 6 7 s (1/A) (b)
8
0.02
0.58
0.60 0.47 0.45 0.30 GS
2
3
4
5 6 7 s (1/A)
8
Au21–
GS
0.49 0.33 0.31 GS
2
3
4
5 6 7 s (1/A)
0.55 0.44 0.40
0.10 0.09 GS
3
4 5 6 7 s (1/A)
GS
Au24– 0.20 0.16
2
8
8
0.09
0.32 GS
2
3
4 5 6 7 s (1/A)
8
GS
FIGURE 7.12 Similar to Figure 7.10 for diffraction data and isomer structures for Aun− clusters with (a) 18 ≤ n ≤ 20 and (b) n = 21 and 24.
structure. Indeed, the best-fit structure is found to be the highly-symmetric ground state isomer, 24G, whose structure is shown in more detail in Figure 7.13a. The commonality between the Au24− and Au16− structures is shown in Figure 7.13b. The tubular structure of Au24− may be viewed as obtained from the structure of Au16− by replacing the bottom-capping atom in Au16 − by the six-member ring and capping triangle of Au24− . The diffraction patterns for the best fits of n = 16 and 24 are essentially identical, both exhibiting a triangular second peak, which supports the structural assignments. Note that diffraction patterns for calculated structures similar to those for n = 16 and 24 are found for n = 14 (0.27 eV above the ground state, GS), 15 (0.26 eV), and 17 (GS), suggesting the inherent stability of the cage/ tubular structural motif. The rich array of size-dependent structural motifs described in this study is specific to gold clusters and likely to originate from the relativistically-enhanced s-d hybridization of gold bonding orbitals [29]. Most importantly, the determination of these cluster structures is essential to an understanding of the size-dependent evolution of cluster properties; for example, assisting in gaining a more complete understanding of the mechanisms that control the chemical reactivity and catalytic activity of gold clusters [33].
186
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
24 G
(b)
n = 24
n = 16
FIGURE 7.13 (a) The structure for n = 24 is shown in end-on and perspective views to emphasize the symmetry and planar aspects of the structure; (b) A comparison of the structures of gold cluster anions with n = 24 and n = 16 atoms.
7.3 TRAPPED-ION FLUORESCENCE Biomolecule conformational change in solution has been investigated widely; however, much less experimental data about structural changes are available for completely isolated gas-phase biomolecules. Our laboratory has been developing fluorescence-based methods to probe local conformational changes or fluctuations in trapped gas-phase biomolecule ions, which have been derivatized with a strongly fluorescing dye. The capability to isolate local changes offers the possibility to measure the dynamics of specific regions of the protein structure. These methods also have the advantage that measurements can be related to molecular dynamics (MD) simulations of intramolecular distances such as dye–residue and residue– residue separations. Previously, we have studied fluorescence quenching in unsolvated dye-derivatized biomolecules by performing measurements of fluorescence lifetime, intensity and emission spectrum as a function of temperature. These studies have identified that fluorescence quenching is dependent on conformational dynamics which lead to interactions between the dye and a Trp (Tryptophan) side chain in peptides [34,35] and small protein [36,37] structures. Section 7.3.1 describes the experimental apparatus for collecting and analyzing fluorescence lifetime data vs temperature. Section 7.3.2 presents the results of lifetime measurements of polyproline peptides and MD simulations to (a) relate fluorescence quenching rates to specific conformational fluctuations, and (b) calculate the implications of intramolecular electrostatic interactions for the quenching mechanism.
7.3.1 Technology Details of the instrumentation have been published elsewhere [34,35]. In brief, fluorescence experiments are performed on biomolecules stored in a home-built quadrupole ion trap by exposing the ions to laser excitation as shown in the apparatus schematic of Figure 7.14. Combined measurements of fluorescence intensity,
187
Structure and Dynamics of Trapped Ions (a)
(b) Pulse counting computer
BW
PMT
A
OG
LC
lon trap
lon trap
CD
BT BW IR P
M
LD
BT n ESI
F
W
hν
Pulse electronics
SH
IR
PD
Trig
Trig PD
Pulse picker X’tal LT M
M OF
OF IR Telescope IR
Mode locked Nd: YVO4 laser 532 nm CW Nd:YAG laser 532 nm
FIGURE 7.14 Diagram of (a) optical beam path for trapped ion laser excitation; and (b) detection optics for fluorescence emitted by trapped ions (shown enclosed in box). The nanoelectrospray source (nESI), octopole guide (OG), charge detector (CD), and UHV chamber are shown on the left. On the right at bottom, the pulsed laser (Nd:YVO4) beam passes through, optical flats (OF), an iris (IR), trap focusing lens (LT), electronic shutter (SH), periscope (P), and a second iris prior to entering the UHV chamber through a Brewster angle window (BW). The beam passes through a baffle tube section (BT), through the ion trap and exits the chamber through a second baffle tube and Brewster angle window. The path of a continuous laser (Nd:YAG) is shown at the bottom by the dashed beam path through irises, a beam-expanding telescope and is turned into the excitation beam path by the two mirrors (M). The mirror and laser path indicated by dashed lines indicates that the laser is used independently for measurements requiring continuous excitation. In the upper right, the fluorescence detection apparatus includes a fluorescencecollecting lens (LC) above the trap, a chamber window (W), mirror (M), filter (F), and focusing lens (LD) which images the cloud on the aperture (A) placed before photomultiplier (PMT) detector (PMT). Various electronic instruments are shown which prepare the laser pulse train and store the detected photon counts. Two photodiodes (PD), which are shown in the pulsed laser beam path, monitor the pulse train to trigger electronic components.
t ime-resolved lifetime, and ion trap mass spectrometry of the dye-derivatized biomolecule are performed to investigate the biomolecule dynamics leading to the observed temperature dependence of these data. The following sections discuss the ion trap operation and fluorescence measurement capabilities. 7.3.1.1 Trapped-Ion Fluorescence Spectroscopy The apparatus shown in Figure 7.14 was developed to achieve high sensitivity measurements of both the fluorescence intensity and time-resolved lifetime of the
188
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
d ye-derivatized biomolecule through careful design and placement of optical components in the laser beam path. Figure 7.14a shows the excitation optics and electronics for both intensity measurements using the continuous (CW) laser, and lifetime measurements with the mode-locked laser. Figure 7.14a indicates the optical beam path through the RF ion trap within the UHV chamber. The fluorescence detection optics and their coupling to the RF ion trap are shown in the inset Figure 7.14b. 7.3.1.1.1 Fluorescence Detection Sensitivity The apparatus in Figure 7.14 was developed to achieve high sensitivity fluorescence measurements by eliminating background detection of laser radiation scattered into the solid angle of the collecting lens (LC). The CW laser excitation beam path shown by the dashed line in Figure 7.14a passes through a beam-forming telescope to optimize overlap of the laser beam width with the trapped ion cloud and is directed to the focusing lens (LF) by mirrors (M). The focal length (50 cm) of this lens provides a focused beam at trap center of ca 200 µm diameter with a Rayleigh length (ca 12 cm) much greater than the trap ring diameter (2r 0 = 0.85 cm). This focus minimizes laser scattering at the trap ring apertures (1.2 mm diameter) and does not vary in cross-section as it passes through the ion cloud. After passing through a shutter (SH), the beam direction and angle are controlled by a periscope (P) to maintain a parallel path through the UHV chamber. The placement of irises (IR) after optical elements and light baffles (BT) within the UHV chamber are required to minimize forward-scattered laser radiation by lenses, mirrors, optical flats, and windows that, subsequently, would enter the trap and be scattered by the electrodes into the detection solid angle defined by the collecting lens (LC). The laser beam enters and leaves the UHV chamber through optical windows (BW) mounted at Brewster’s angle to reduce reflective scattering of the linearly-polarized laser beam. As indicated in Figure 7.14b, the fluorescence is emitted through a ring aperture of 1.5 mm diameter perpendicular to the laser axis. Fluorescence exits the UHV chamber through an optically-flat chamber window (6 mm thick) and is collimated by the 25 mm-collecting lens (LC) placed ca 27 mm from trap center and ca 6 mm above the window. As indicated in Figure 7.14b, the parallel fluorescence passes through filters (F) and is focused (lens LD, focal length 50 mm) at the 1 mm diameter aperture (A) placed ca 1 cm from the Ga–As PMT detector surface area of diameter 5 mm. A band-pass filter is used to isolate the fluorescence band of the dye (BODIPY-TMR, ca 550–595 nm). A second sharp-cutoff, long-wave filter was added to minimize the detection of scattered laser radiation. However, radiation at large scattering angles is transmitted in part and dominates the detected background signal; hence, the severe requirement to avoid scattering from the trap electrodes because such scattering tends to produce large angle scattering. Fluorescence intensity measurements are performed by irradiating the trapped ions with 532 nm light (frequency doubled) from a continuous wave Nd:YAG laser for 100 ms at an intensity of 130 W cm–2. The fluorescence data exhibit routinely a S/N ratio in the range of 100–400 for ca 200 ions in the laser interaction volume. An example of this sensitivity is indicated in Figure 7.15 for measurements of Rhodamine 640 fluorescence vs the number of interacting ions excited by 532 nm with an intensity of ca 260 W cm–2. The use of such small numbers of ions minimizes
189
Structure and Dynamics of Trapped Ions 1200
[CPS - (CPS)Back]
1000 800 600
200 150 100 50 0
400 200
5
10
15
20
0 0
20 40 60 80 100 Number of interacting ions
120
FIGURE 7.15 Background-subtracted fluorescence intensity (counts per second) vs number of ions of the fluorescent dye, Rhodamine 640, in the laser interaction volume. The laser beam diameter is 220 µm (fwhm). The inset shows details for ≤ 20 ions (dashed box). (From Iavarone, A.T.; Duft, D.; Parks, J.H. J. Phys. Chem. 2006, 110, 12714–12727. With permission.)
space-charge interactions resulting in higher resolution mass spectra of the trapped ions, increased density ion clouds, and more reliable ion trajectory simulations. As discussed in Section 7.3.2.1, this sensitivity has been extended recently to obtain fluorescence measurements from a single trapped polypeptide molecule. 7.3.1.1.2 Time-Resolved Lifetime Measurements The pulsed laser excitation beam path is also shown in Figure 7.14. This laser is a mode-locked, diode-pumped, solid-state Nd: YVO4 laser providing pulsed excitation at 532 nm (frequency doubled). The repetition rate is reduced from 80 to 20 MHz through pulse picking by a transverse field modulator and the pulse width is 12 ps (full width at half-maximum). The excitation and detection beam-path optics are identical to those described above with the exception of the telescope which is unnecessary for the beam diameter delivered by the laser device. A histogram-accumulating real-time processor is used for time-correlated single photon counting. The average detected count rate is maintained below 1% of the excitation rate to ensure single photon counting statistics. The fluorescence lifetime decay curves are deconvoluted and fitted by a stretched exponential model. An example of the measured lifetime decay signals from dye-derivatized polyproline peptides [M + H] + is shown in Figure 7.16a for two temperatures exhibiting the effect of fluorescence quenching. The instrument response function characterizing the time resolution of the PMT and detection electronics is shown in Figure 7.16a to have a width of ca 0.5 ns. The stretched exponential fits are shown more clearly in Figure 7.16b to deviate from a single exponential for stronger quenching observed at higher temperatures which is consistent with a quenching process occurring in an inhomogeneous ensemble. In this case, the inhomogeneity is believed to derive from an ensemble of different quenching conformers leading to a distribution of quenching rates.
190
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
200 150
(b)
Decay signal (arb.units)
100
303 K
50 438 K 0 60
70
80
90
2 100 6 4
τ = 9.6(4) ns
2 10 6 4
τ = 3.0(5) ns
2 1 60
70
Time (ns)
80
90
FIGURE 7.16 (a) Example of fluorescence decays of Pro4[M + H] + ions at 303 and 438 K and the instrument response function (light solid curve); (b) fits to the measured decays by a stretched exponential model (solid black curves) showing the fit decay constant. Linear and logarithmic intensity scales are used in (a) and (b), respectively. (From Iavarone, A.T.; Duft, D.; Parks, J.H. J. Phys. Chem. 2006, 110, 12714–12727. With permission.)
7.3.1.2 Nanospray Ionization-Ion Trap Mass Spectrometry Droplets and biomolecular ions formed by a home-built nanoelectrospray ionization (nESI) source [34] are sampled from atmospheric pressure through a stainless steel capillary maintained at ca 350 K. As shown in Figure 7.14a, ions from the nESI source pass through an octopole guide of 23 cm length into the quadrupole ion trap through an aperture in the end-cap of 1.5 mm diameter. The ion trap operates at an RF frequency of 600 kHz with an end-cap electrode spacing of 2z0 = 0.6 cm and ring diameter of r02 = 2z02 cm2. Cluster ions are loaded at qz = 0.5 with a range of RF amplitudes, V0–p = 0.5–2 kV, for the singly-protonated polyproline peptides discussed in this paper. Upon storing the clusters for ca 5 s in helium gas at a pressure of ca 10 –3 Torr, the cluster vibrational and translational energies at the gas temperature become thermalized. Ion loading in this trap configuration yields ion clouds of ca 2000 ions that are adequate for these fluorescent measurements.
Structure and Dynamics of Trapped Ions
191
The background helium gas temperature is set by channeling gas through the trap superstructure which is heated and held at a constant temperature by ceramic heaters or cooled directly by flowing liquid nitrogen. Fluorescence measurements can be performed over a temperature range of 150–550 K with a precision of ±1 K. The ions of interest are isolated by ejecting ions occurring at higher or lower m/z-ratios using RF ramping and SWIFT excitation [20,21]. The trap background helium gas pressure is maintained at ca 3 × 10 –5 Torr and is pulsed to ca 2 × 10 –3 Torr for ion loading and thermalization. Ions are thermalized in the trap for ca 1–2 s at qz = 0.50 before fluorescence lifetime measurements are performed. The trapped ions undergo > 105 collisions which equilibrate them with the helium background gas which is maintained at the temperature of the trap during lifetime measurements. After exposure to laser excitation, a mass spectrum is obtained by ejecting ions resonantly into the channeltron electron multiplier indicated in Figure 7.14a. The ion trap design with apertures in the ring and end-cap electrodes limit the mass resolution to ≤ 300, which is sufficient for the experiments described here.
7.3.2 Radiative Lifetimes and Conformational Fluctuations Temperature-dependent quenching of the fluorescence emitted by dye-derivatized biomolecules has been observed to depend on biomolecule composition, sequence, and charge state. Quenching measurements have been performed for peptides [34,35], proteins [36,37], and non-covalent complexes [38] and each species exhibits a similar functional dependence on temperature as shown in Figure 7.17. Figure 7.17 shows the lifetime vs temperature for (a) Trp-cage protein in the 1 + , 2 + and 3 + charge states, (b) polyproline peptides of different sequences and chain lengths, and (c) Vancomycin non-covalent complexes with tripeptides having different chirality. The following section, which summarizes lifetime experiments and calculations [39] for unsolvated dye-derivatized polyproline peptides over the temperature range 150– 460 K, helps in the interpretation of the physical basis for the lifetime temperature dependence and its relationship to conformational fluctuations. A conclusion that can be drawn from these studies is that the observed temperature dependence of the fluorescent lifetime is related to the increasing rate of conformational fluctuations. In particular, those specific conformations which enhance the probability of quenching via photoinduced electron transfer. 7.3.2.1 Polyproline Lifetime Measurements 7.3.2.1.1 Quenching Rate: Dye–Trp Proximity The following peptide sequences were chosen to investigate variation in the quenching rate as the separation of the fluorescing dye and the Trp residue is increased by a relatively rigid polyproline chain. These studies will help to characterize the quenching mechanism and indicate qualitative trends which can be compared with MD simulations. Measurements were performed on the [M + H] + charge state of the derivatized peptides BoTMR-(Pro)n-Arg-Trp (n = 4 or 10; ‘Pro4’ and ‘Pro10’, respectively). The BODIPY® analog of tetramethylrhodamine (BoTMR) is neutral in solution and consequently Arg is the most probable protonation site. The
192
Lifetime (ns)
(a)
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
12
(b) 12
10
10 8
8 6 4 2
6
Trp-cage 3+ Trp-cage 2+ Trp-cage 1+ 200
250
300
4 2 350
400
450
0
Pro4sansTrp Pro4sansLinker Pro10 Pro4 150 200 250 300 350 400 450 500
Temperature (K)
Temperature (K)
(c)
Lifetime (ns)
12 10 8
V_R6G WKAA_LDLL WKAA_DLDD WKAA_LLDD
6 4 150
200
250 300 350 Temperature (K)
400
FIGURE 7.17 Fluorescence lifetime vs temperature measurements for gas-phase biomolecular ions of (a) Trp-cage protein charge states; (b) Dye-(Pro)n-Arg-Trp peptide sequences, [M + H] + ; (c) Vancomycin-peptide non-covalent complexes, [M + H] + . Best-fits for the quenching rate model are shown by lines through each set of data points.
sequences BoTMR-(Pro)4-Arg without Trp and BoTMR-(Pro)4-Arg–Trp without an aminohexanoate linker (‘Pro4 sans Trp’ and ‘Pro4 sans X’, respectively) were also studied. The derivatized peptides were synthesized commercially and purified by reversed-phase HPLC to a stated purity of > 70% prior to shipment. BoTMR is obtained with or without an aminohexanoate linker and the C-termini of the peptides were amidated. Electrospray solutions contain the dye-derivatized peptides at 10 μM in 50% acetonitrile/50% water. The BoTMR excitation and emission spectra have been published elsewhere [40,41]. Attachment of BoTMR to peptides does not alter significantly the emission or excitation spectra (data not shown). The lifetime vs temperature data for Pro4 sans Trp, Pro4, Pro4 sans X, and Pro10 are shown in Figures 7.18a through d, respectively, over the temperature range 150–460 K. Assuming the fluorescence is emitted by a population decaying from a single excited electronic state, the lifetime, τ, can be shown from a rate equation analysis for the excited state population to be given by 1/τ = 1/τ0 + kq where τ0 is the unperturbed lifetime and the non-radiative rate kq defines the quenching rate constant. The temperature dependence of the quenching rates, kq vs T, is also shown in Figure 7.18 for each species. It is important to point out that fragmentation was not observed in any of the fluorescence lifetime measurements as a result of the
0
Lifetime (ns)
(b)
0.0
6 4 2 0
4.0 Pro4
Lifetime Quenching rate Model fit
3.0 2.0 1.0 0.0
150 200 250 300 350 400 450 500 Temperature (K)
6 4 2 0
150 200 250 300 350 400 450 500 Temperature (K)
10 8
1.0
8
(d)
1.0
6 4 2 0
0.6 0.4 0.2 0.0
150 200 250 300 350 400 450 500 Temperature (K) 2.0
10 8
0.8
Pro4 sans x Lifetime Quenching rate Model fit
Pro10 Lifetime Quenching rate Model fit
1.6 1.2 0.8 0.4 0.0
Quenching rate (×108s–1)
2
2.0
10
Lifetime (ns)
4
3.0
(c)
Lifetime (ns)
6
Pro4 sans trp Lifetime Quenching rate Model fit
Quenching rate (×108s–1)
8
4.0
Quenching rate (×108s–1)
10
Lifetime (ns)
(a)
Quenching rate (×108s–1)
193
Structure and Dynamics of Trapped Ions
150 200 250 300 350 400 450 500 Temperature (K)
FIGURE 7.18 Fluorescence lifetime and quenching rate vs temperature for the [M + H] + ions of (a) Pro4 sans Trp showing the decay without Trp; (b) Pro4 with Trp and the flexible linker; (c) Pro4 with Trp but without the flexible linker; (d) Pro10 with Trp. Lifetimes are indicated on the left hand axis, and quenching rates are on the right hand axis. The legends identify the data for each sequence. Best-fits to the quenching rate model are shown by solid lines through each set of data points. (From Shi, X.; Duft, D.; Parks, J.H. J. Phys. Chem. 2008, 112, 12801–12815. With permission.)
low average laser power (15 mW) and short exposure time (50–200 ms) used to measure decay times. When temperature measurements are extended to higher temperatures, the lifetime data for Pro4 are observed to approach slowly a minimum value. However, the exposure times must be reduced to 30 ms to avoid fragmentation during the lifetime measurement at 550 K. In the absence of Trp (Pro4 sans Trp), the lifetime is unchanged essentially as shown in Figure 7.18a, exhibiting a decrease of ≤ 10% even at the highest temperatures. This slight decrease in lifetime is probably related to weaker interactions of the dye with the remaining residues resulting in quenching rates less than 2 × 107 s–1. The measurements of Pro4 sans Trp also emphasize that the charge interactions per se do not contribute to fluorescence quenching. Pro4 exhibits the strongest Dye–Trp quenching interaction resulting in a lifetime varying by a factor of five over the full temperature range as shown in Figure 7.18b. The quenching rates shown in Figure 7.18b exhibit values for which τ0 kq ≥ 1, indicating that quenching is a dynamic process occurring during the unperturbed lifetime of τ0 ca 10 ns. The quenching rates for Pro4 sans X in Figure 7.18c are reduced probably as a result of more constrained dye motion which limits the proximity to Trp without the flexible linker. Pro10 exhibits quenching rates intermediate between Pro4 and Pro4 sans X. As shown in Figure 9.18d, the lifetime
194
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
varies by a factor of ca 3 over the temperature range and the maximum kq is reduced by a factor of ca 2 from Pro4. The dye linker and the presence of backbone flexibility partially compensates for the longer proline chain length. 7.3.2.1.2 Quenching Rate: Electrostatic Fields A second peptide sequence was chosen to investigate variation in the quenching rate resulting from the intramolecular electrostatic fields. In these sequences, the dye without a linker was positioned adjacent to the Trp residue and the Arg charge site proximity was varied by the length of the rigid polyproline chain. Measurements were performed on the [M + H] + charge state of the derivatized peptides BoTMRTrp-(Pro)n-Arg (n = 0, 4 and 10; ‘DW’, ‘DW-Pro4’ and ‘DW-Pro10’, respectively). Each of these peptides was synthesized without the linker. The lifetime measurements for these peptides demonstrate that electrostatic interactions introduced by the Arg protonation site influence strongly the quenching rates. The absence of the dye linker ensures Dye–Trp proximity for a larger fraction of the time than expected for the Pro4 sans linker peptide in which the dye and Trp are separated by the polyproline chain. However, as shown in Figure 7.19a, the quenching rates for DW Pro4 are comparable to Pro4 sans linker suggesting that although Dye–Trp proximity is probably necessary for strong quenching, other factors play a significant role. The quenching rates shown in Figure 7.19b for different polyproline chain lengths indicate a strong dependence on the average separation of the charged Arg side chain from the Dye–Trp pair. These data provide clear evidence that the electrostatic fields of the protonation site are involved intimately in determining the quenching rate. 7.3.2.1.3 Rate Model The quenching interactions in these peptide structures are occurring in the presence of conformational fluctuations which lead not only to large changes in the intramolecular separations of Dye–Trp, but also in the electrostatic interactions of Arg + with the dye and side chain polarizations. In the presence of spatial fluctuations,
Quenching rate (×108 s–1)
1.0 0.8 0.6
Arg+
Dye trp Pro4 Dye Pro4 Arg+ Trp
0.4 0.2 0.0
150 200 250 300 350 400 450 Temperature (K)
1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0
Quenching rate (×108 s–1)
(b)
(a)
Dye Trp Arg+ Dye Trp Pro10 Arg+ Dye Trp Pro4 Arg+
150 200 250 300 350 400 450 Temperature (K)
FIGURE 7.19 Quenching rates vs temperature for the [M + H] + ions (a) comparing DW Pro4 and Pro4 sans X; and (b) comparing DW, DW Pro4 and DW Pro10. The legends identify the data for each sequence. Best-fits to the quenching rate model are shown by solid lines through each set of data points. (From Shi, X.; Duft, D.; Parks, J.H. J. Phys. Chem. 2008, 112, 12801–12815. With permission.)
Structure and Dynamics of Trapped Ions
195
an unshielded charge can produce electrostatic field strengths the order of 107–108 V cm–1 at the position of nearby side chains. As a result, the dye or side chains can become trapped in fluctuating potential well depths of ca 0.1–0.6 eV. Consequently, the temperature dependence of the fluctuations is expected to influence strongly the quenching rates because these rates will depend on the Dye–Trp separation and also the Dye–Arg + and Trp–Arg + separations which determine the strength of electrostatic fields in the presence of the quenching pair. A model of the lifetime temperature dependence is based on the assumption that the quenching rate, kq, is limited by the rate of conformational fluctuations, kf, and not by the rate of a specific quenching interaction. The assumption that fluctuations are the rate-limiting kinetic process is reasonable and consistent with prior research [42–46]. Almost exclusively, these previous experiments have shown that interconversion among a relatively small number of dominant conformers was the rate-limiting step in the quenching process. The following phenomenological model relates the quenching rate to the peptide conformational fluctuations
kq = kf = Af exp( − Ebf /kT )
(7.1),
where kf is the rate of fluctuations which result in quenching conformers. Af represents an ensemble-averaged rate characterizing those fluctuations which lead to quenching conformations; for example, Af could represent collective side chain fluctuations having rates ca 1011–1012 s–1 or slower backbone motions ca 1010 s–1 but probably not faster single atom fluctuations ca 1013 s–1. The spatial fluctuations leading to quenching conformers are constrained by energy barriers on the peptide potential energy surface having an ensemble average Ebf. The temperature T is the thermal equilibrium temperature established by collisions of trapped peptide ions with background helium gas. The quenching rate model fits the lifetime and data very closely for each peptide species studied as shown in Figures 7.17 through 7.19. For example, the fit parameters for Pro4 are Af = 4.1±0.6 × 1011 s–1 and Ebf = 0.28±0.01 eV, and for DW-Pro4 are Af = 1.1±0.02 × 1010 s–1 and Ebf = 0.20±0.01 eV. 7.3.2.1.4 Field-Induced Electron Transfer The MD simulations presented in Ref. [39] help to characterize conformers which contribute to the quenching rate as suggested by the data for the polyproline sequences. Consider the simulation trajectories and histograms of the Dye–Trp, Dye–Arg+, and Trp–Arg+ separations at 350 K shown in Figure 7.20. Structures of five overlaid conformers are shown below the trajectories in Figure 7.20; these structures were chosen at arbitrary times during the trajectory durations A and B denoted by dashed rectangles. The group of conformer structures associated with region A (ca 3.5 ns) displays a Dye–Trp separation of ca 9.5 Å given by the peak in the Dye–Trp histogram. Note that the Trp–Arg+ separation is ca 9 Å and the Dye–Arg+ separation shows steps of ca 7 Å and 5 Å during trajectory region A. The conformers associated with Region B (ca 2.5 ns) exhibit a group of nearly identical structures, which are expected to provide a greater probability for quenching the dye fluorescence. In all the conformers occurring in region B, the Dye–Trp separation
196
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V A
20
B
Dye–Trp
15 10 5
Separation (Å)
0 20
Dye–Arg+
15 10 5 0 20
Trp–Arg+
15 10 5 0
0
1
2
3 Time (ns)
4
5
6
(b)
(a) Arg Dye
+
Dye
Arg
+
Trp Trp
FIGURE 7.20 Trajectories and histograms of the Dye–Trp, Dye–Arg + and Trp– Arg + separations are shown at 350 K. Below the trajectories are five superimposed structures obtained from the trajectory time durations indicated by the rectangles labeled A and B. (From Shi, X.; Duft, D.; Parks, J.H. J. Phys. Chem. 2008, 112, 12801–12815. With permission.)
is ca 6 Å, the Trp–Arg + separation ca 8 Å and the Dye–Arg+ separation ca 5 Å. These conformers satisfy the conditions for proximity of the dye and Trp side chain and, simultaneously, the presence of strong electrostatic fields required for faster quenching rates in the experimental data shown in Figure 7.19. The electrostatic field at the position of the dye for the trajectories shown in region B is calculated in Ref. [39] to fluctuate between ca 3 and 9 × 107 Vcm−1. The exothermicity, ∆E, of a photoinduced charge transfer reaction is defined as
197
Structure and Dynamics of Trapped Ions
the difference in energy of the charge transfer state and the dye excited state. The exothermicity for charge transfer is calculated for conformations described by the trajectories shown in region B by including contributions of the intramolecular interaction energies of the charge and side chain polarizabilities in the energy of the charge transfer state. The fluctuating fields induce exothermicities up to ca ∆E = −2 eV, that is, the charge transfer state lies 2 eV below the excited state of the dye. In the absence of these field interactions, the energetics are endothermic by ca ∆E = + 2 eV. What is of interest here are not the absolute values of the exothermicity, which are only rough estimates, but the large variations induced by the fluctuating fields. The presence of such quenching conformations is interesting; however, a more critical question is the rate of occurrence of such conformers as a function of temperature. Figure 7.21 presents a first estimate of the trend with temperature derived from the MD simulations. Figure 7.21 displays the rate that fluctuations lead to a Dye–Trp separation < 8 Å and simultaneously an exothermicity ΔE < −0.2 eV. The rate of occurrence of these quenching conformers at each temperature is calculated from trajectories with different starting structures at each temperature and is fitted to an Arrhenius rate model. The separation limit of 8 Å is based on single molecule studies [47] of electron transfer quenching in solution. The exothermicity limit of −0.2 eV is chosen somewhat arbitrarily based on the Rhem–Weller observations [48]. A more complete analysis is currently in progress which considers rates for a range of these limits. Figure 7.21 indicates that the MD rates are fitted reasonably well by the model parameters A = 1±0.5 × 1013 s−1 and Ea = 0.18 ± 0.02 eV. The estimate of a rate of occurrence for quenching conformers is perhaps the most appropriate comparison between the quenching rate model (Equation 7.1) used to characterize the experimental data and the MD simulations. The simulation prefactor rate is A >> Af found for Pro4 which is to be expected because A (i) does not correspond to an ensemble average and (ii) requires a quantum chemistry analysis to provide realistic proximity and energy constraints and account for the dependence on the relative
Fluctuation rate (1011 s–1)
2.0 Dye–Trp < 8Å ∆E < – 0.2 eV
1.5 1.0 0.5 0.0 200
250
300
350
400
450
500
Temperature (K)
FIGURE 7.21 The rate of fluctuations that result in Dye–Trp separation < 8 Å and, simultaneously, an electron transfer exothermicity ΔE < –0.2 eV vs temperature (black squares). The best-fit to an Arrhenius model is shown also (black curve). (From Shi, X.; Duft, D.; Parks, J.H. J. Phys. Chem. 2008, 112, 12801–12815. With permission.)
198
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
orientations of the dye and Trp rings [49,50]. It is interesting that the simulation activation barrier Ea ca 0.2 eV is consistent with the data fit parameter Ebf ca 0.3 eV of the quenching rate model. The temperature dependence displayed in Figure 7.21 is a more significant result of the calculations. Although fluctuations are expected to increase with temperature, it is not obvious that the rate of quenching conformers will increase, nor that the rate would follow an Arrhenius model. In fact, this was not the case for all starting structures. A more systematic comparison with experimental data is in progress in which an ensemble average of trajectories, over many starting structures, will be calculated to model reliably the distribution of conformations related to quenching. In addition, the trends of the calculated occurrence rates for all polyproline peptides will be compared with the experimental trends. Energy transfer cannot contribute significantly to the observed quenching process because there are no singlet or triplet energy levels of Trp [51] near the first singlet excited state of the dye at ca 2.3 eV. The intersystem crossing rate for S1 → T0 of the dye core molecule BODIPY has been measured in solution [52] to be k ISC ca 106 s−1 and, as a result, does not contribute to quenching on the timescale of the dye lifetime. On the basis of these considerations, Ref. [39] concludes that the measured lifetime dependence on temperature results from specific conformational fluctuations that increase the probability of electron transfer between the dye and Trp. The possibility to identify directly such conformation fluctuations in the gas phase has motivated the development of methods to perform fluorescence lifetime measurements on a single trapped polypeptide molecule. The recent successful implementation [53] of this method for gas-phase biomolecules opens the possibility to study conformational fluctuations using methods developed for single molecules in solution [54,55].
7.4 CONCLUSION This chapter has reviewed several new experimental methods to investigate the structures and the dynamics of structures. These methods have contributed to our understanding of how the atoms are organized in small metal clusters and how to observe collective fluctuations of the molecular substructures within biomolecules. In both these research areas, ion traps have been essential; (i) to select the ions of interest from a much larger array generated by the source; (ii) to concentrate the ions into a volume providing significant overlap with the probe, either an e− - beam or laser photons; (iii) to control the internal degrees of freedom; and (iv) to allow observation of atomic and molecular arrangements over extended timescales. The ion traps used in these investigations were integrated into the larger experimental apparatus in which the interface between the trap and its surrounding components were the most important barriers to overcome for successful measurements. In each of these research areas, the flexibility of trap design and application of unique electronics to manipulate the ions were central to optimizing experimental performance. In many respects, these experiments are allowing us to explore new and exciting science. The most intriguing and unexpected result of the metal cluster studies is that most of the available structural isomers do not appear to contribute to the measured diffraction patterns. Perhaps even more remarkable is that the comparison of diffraction data with calculated structures is capable of identifying a single best
Structure and Dynamics of Trapped Ions
199
fit, and in most cases a much better fit than other isomers with close-lying energies. The possibility to probe biomolecule dynamics with fluorescence measurements is a fascinating opportunity provided we can understand explicitly how the dye interacts with the local environment within the biomolecule. I think we have made significant progress toward this goal by demonstrating that fluctuation-induced electron transfer can be an important contributor to the dye–Trp quenching process, consistent with the results of experiments and calculations. These initial studies will have to be extended to test this interpretation by probing biomolecules, such as a beta hairpin, for which there are well-defined spatial dynamics. So where to next? There are many rich opportunities to consider, but I believe two very important areas stand out. Small metal clusters of less that 50 atoms have been shown to be efficient catalysts even at low temperatures. However, the relationships between catalytic properties and cluster structures are only beginning to be examined. Comprehension of these relationships requires demanding calculations and equally arduous experiments. Diffraction of trapped clusters should be able to contribute here by correlating catalytic activity with structural motifs. The study of biomolecules in gas phase provides new insights into complex systems and applications to health science that will drive continued expansion of this field. Although it is generally appreciated that studies of hydrated biomolecules can forge links to the behavior of conformational dynamics in solution, probing the conformations of hydrated species in a way that does not significantly perturb the water molecule binding or distribution is not a trivial undertaking. Yet, there are interesting possibilities under consideration to apply fluorescence methods to this opportunity.*
Acknowledgments It is a pleasure to identify the group of postdoctoral fellows who have contributed to the development of the TIED science and technology. In order of their tenure in my laboratory, they are Douglas Cameron, Mathias Maier-Borst, Stefan Kruekeberg, Detlef Schooss, Rongbin Huang, Xiaopeng Xing, and Xi Li. It is difficult to imagine the success we have experienced without the broad accomplishment and creative efforts of these individuals. I acknowledge fruitful discussions with Mordechai Rokni and Abraham Szöke who helped to originate the ideas leading to TIED, and for conversations leading to a clearer view of the issues involved. I thank my Rowland Institute colleagues Michael Burns and Frans Spaepen for many helpful conversations and advice regarding analysis and interpretation of diffraction measurements. Although the theoretical contributions of my collaborators in this work have been mentioned above, Uzi Landman’s physical insight and clarity are greatly appreciated, and the suggestions and encouragement of Ignacio Garzón and Karo Michaelian were particularly helpful. The biofluorescence research was performed by a group of dedicated postdoctoral fellows including Sandra Rodriguez-Cruz, Joseph Khoury, Allison Danell, Ryan Danell, Anthony Iavarone, Denis Duft, and Xiangguo Shi. These exceptionally hard-working * See Volume 4, Chapter 21: Chemical and Photochemical Studies of Metal Dication Complexes in an Ion Trap by Guohua Wu and Anthony J. Stace.
200
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
individuals have brought these difficult experimental methods to fruition and have applied them to exciting new research areas of gas-phase biomolecules. I acknowledge Evan Williams, Graham Cooks, and Martin Karplus who were particularly encouraging at critical times and opened my thinking to a wider range of biomolecule studies. The help and insight gained from collaboration with David van der Spoel has been necessary and welcomed. The cluster diffraction research was supported by the Department of Energy (DOE) under Grant No.DE-FG02-01ER45921 and the biomolecule fluorescence research was supported by the Rowland Institute at Harvard.
References
1. Dehmelt, H.G. Radiofrequency spectroscopy of stored Ions I: storage. Adv. At. Mol. Phys. 1967, 3, 53–72; Dehmelt, H.G. Radiofrequency spectroscopy of stored ions II: spectroscopy. Adv. At. Mol. Phys. 1969, 5, 109–153. 2. Wineland, D.; Ekstrom, P.; Dehmelt, H.G. Monoelectron oscillator. Phys. Rev. Lett. 1973, 31, 1279–1282. 3. Gabrielse, G.; Dehmelt, H.G. Observation of inhibited spontaneous emission. Phys. Rev. Lett. 1985, 55, 67–70. 4. Neuhauser, W.; Hohenstatt, M.; Toschek, P.E.; Dehmelt, H.G. Localized visible Ba + mono-ion oscillator. Phys. Rev. 1980, A22, 1137–1140. 5. Wineland, D.J.; Itano, W.M. Laser cooling of atoms. Phys. Rev. 1979, A 20, 1521–1540. 6. Itano, W.M.; Bergquist, J.C.; Bollinger, J.J.; Wineland, D.J., ‘Laser cooling of trapped ions’ in Laser Manipulation of Atoms and Ions, Proc. Enrico Fermi Summer School, Course CXVIII, Varenna, Italy, July, 1991, edited by E. Arimondo, W.D. Phillips, and F. Strumia (North-Holland, Amsterdam, 1992) pp. 519–537S. 7. Itano, W.M.; Bergquist, J.C.; Wineland, D.J. Laser spectroscopy of trapped atomic ions. Science 1987, 237, 612–617. 8. Williams, J.D.; Cox, K.A.; Schwartz, J.C.; Cooks, R.G. ch 1, High Mass, High Resolution Ion Trap Mass Spectrometry, Vol. 2, Ion Trap Instrumentation, in: Practical Aspects of Ion Trap Mass Spectrometry, R.E. March and J.F.J. Todd, (Eds.), CRC Press, Boca Raton, 1995, pp. 3–50. 9. Haberland, H.; Karrais, M.; Mall, M.; Thurner, Y. Thin films from energetic cluster impact. J. Vac. Sci. Technol. 1992, A10, 3266–3271. 10. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; Whitehouse, C.M. Electrospray ionization for mass spectrometry of large biomolecules. Science 1989, 246, 64–71. 11. De Boer, B.G.; Stein, G.D. A metal cluster generator for gas-phase electron diffraction and its application to bismuth, lead, and indium: variation in microcrystal structure with size. Surf. Sci. 1981, 106, 84–94; Yokozeki, A.; Stein, G.D. Production and electron diffraction studies of silver metal clusters in the gas phase. Appl.Phys. 1978, 49, 2224–2230. 12. Farges, J.; Feraudy, M.F.; Raoult, B.; Torchet, G. Structure and temperature of rare gas clusters in a supersonic expansion. Surf. Sci. 1981, 106, 95–100; Farges, J.; Raoult, B.; Torchet, G. Crystalline and noncrystalline effects in electron diffraction patterns from small clusters in an argon cluster beam. J. Chem. Phys. 1973, 59, 3454–3458. 13. Hovick, J.W.; Bartell, L.S. Structural aspects of nanocrystals of transition-metal hexafluorides. J. Phys. Chem. 1998, B 102, 534–539; Bartell, L.S. Diffraction studies of clusters generated in supersonic flow. Chem. Rev. 1986, 86, 491–504. 14. Reinhard, D.; Hall, B.D.; Berthoud, P.; Valkealahti, S.; Monot, R. Unsupported nanometer-sized copper clusters studied by electron diffraction and molecular dynamics. Phys. Rev. 1998, B 58, 4917–4926; Reinhard, D.; Hall, B.D.; Ugarte, D.; Monot, R. Size-
Structure and Dynamics of Trapped Ions
201
independent fcc-to-icosahedral structural transition in unsupported silver clusters: an electron diffraction study of clusters produced by inert-gas aggregation. Phys. Rev. 1997, B 55, 7868–7881. 15. Maier-Borst, M; Cameron, D.B.; Rokni, M.; Parks, J.H. Electron diffraction of trapped cluster ions. Phys. Rev. 1999, A 59, R3162–R3165. 16. Krückeberg, S.; Schooss, D.; Maier-Borst, M.; Parks, J.H. Diffraction of trapped (CsI)nCs + : the appearance of bulk structure. Phys. Rev. Lett. 2000, 85, 4494–4497. 17. Xing, X.; Danell, R.M.; Garzón, I.L.; Michaelian, K.; Blom, M.N.; Burns, M.M.; Parks, J.H. Size-dependent fivefold and icosahedral symmetry in silver clusters. Phys. Rev. 2005, B 72, 081405 (1–4). 18. Xing, X.; Yoon, B.; Landman, U.; Parks, J.H. Structural evolution of Au nanoclusters: from planar to cage to tubular motifs. Phys. Rev. 2006, B 74, 165423 (1–6). 19. Parks, J.H.; Xing, X. Trapped ion electron diffraction: structural evolution of silver and gold clusters, Vol. 12, Atomic Clusters: From Gas Phase to Deposited, in: The Chemical Physics of Solid Surfaces D. Woodruff (Ed.), Elsevier, New York, 2007, p. 377–407. 20. Guan, S.; Marshall, A.G. Stored waveform inverse Fourier transform (SWIFT) ion excitation in trapped-ion mass spectrometry: theory and applications. Int. J. Mass Spectrom. Ion Processes 1996, 157/158, 5–37; Marshall, A.G.; Wang, T-C.L.; Ricca, T.L. Tailored excitation for Fourier transform ion cyclotron resonance mass spectrometry. J. Am. Chem. Soc. 1985, 107, 7893–7897. 21. Julian, Jr, R.K.; Cooks, R.G. Broad-band excitation in the quadrupole ion trap mass spectrometer using shaped pulses created with the inverse Fourier transform. Anal. Chem. 1993, 58, 1827–1833; Soni, M.H.; Cooks, R.G. Selective injection and isolation of ions in quadrupole ion trap mass spectrometry using notched waveforms created using the inverse Fourier transform. Anal. Chem. 1994, 66, 2488–2496. 22. Asano, K.G.; Goeringer, D.E.; McLuckey, S.A. Thermal dissociation in the quadrupole ion trap: ions derived from leucine enkephalin. Int. J. Mass Spectrom. 1999, 185/186/187, 207–219. 23. Schooss, D.; Blom, M.N.; Parks, J.H.; v. Issendorf, B.; Haberland, H.; Kappes, M.M. The structures of Ag55+ and Ag55−: trapped ion electron diffraction and density functional theory. Nano Lett. 2005, 5, 1972–1978. 24. Kelton, K.F.; Lee, G.W.; Gangopadhyay, A.K.; Hyers, R.W.; Rathz, T.J.; Rogers, J.R.; Robinson, M.B.; Robinson, D.S. First X-ray scattering studies on electrostatically levitated metallic liquids: demonstrated influence of local icosahedral order on the nucleation barrier. Phys. Rev. Lett. 2003, 90, 195504 (1–4). 25. Sachdev, S.; Nelson, D.R. Theory of the structure factor of metallic glasses. Phys. Rev. Lett. 1984, 53, 1947–1950. 26. Furche, F.; Ahlrichs, R.; Weis, P.; Jacob, C.; Gilb, S.; Bierweiler, T.; Kappes, M.M. The structures of small gold cluster anions as determined by a combination of ion mobility measurements and density functional calculations. J. Chem. Phys. 2002, 117, 6982–6990. 27. Häkkinen, H.; Yoon, B.; Landman, U.; Li, X.; Zhai, H-J.; Wang, L-S.. On the electronic and atomic structures of small AuN− (N = 4-14) clusters: a photoelectron spectroscopy and density-functional study. Phys. Chem. 2003, A 107, 6168–6175. 28. Yoon, B.; Koshkinen, P.; Huber, B.; Kostki, O.; von Issendorff, B.; Hakkinen, H.; Moseler, M.; Landman, U. Size-dependent structural evolution and chemical reactivity of gold clusters. Chem. Phys. Chem. 2007, 8, 157–161. 29. Häkkinen, H.; Moseler, M.; Landman, U. Bonding in Cu, Ag, and Au clusters: relativistic effects, trends, and surprises. Phys. Rev. Lett. 2002, 89, 33401(1–4). 30. Gilb, S. Ph.D. thesis, University of Karlsruhe, 2001. 31. Bulusu, S.; Li, X.; Wang, L-S.; Zeng, X.C. Evidence of hollow golden cages. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 8326–8330.
202
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
32. Li, J.; Li, X.; Zhai, H-J.; Wang, L-S. Au20: a tetrahedral cluster. Science 2003, 299, 864–867. 33. Heiz, U.; Landman, U. Nanocatalysis, Springer, Berlin, 2006. 34. Iavarone, A.T.; Meinen, J.; Schulze, S.; Parks, J.H. Fluorescence probe of polypeptide conformational dynamics in gas phase and in solution. Int. J. Mass Spectrom. 2006, 253, 172–180. 35. Iavarone, A.T.; Duft, D.; Parks, J.H. Shedding light on biomolecule conformational dynamics using fluorescence measurements of trapped ions. J. Phys. Chem. 2006, 110, 12714–12727. 36. Iavarone, A.T.; Patriksson, A.; van der Spoel, D.; Parks, J.H. Fluorescence probe of Trpcage protein conformation in solution and in gas phase. J. Am. Chem. Soc. 2007, 129, 6726–6735. 37. Iavarone, A.T.; Parks, J.H. Conformational change in unsolvated Trp-cage protein probed by fluorescence. J. Am. Chem. Soc. 2005, 127, 8606–8607. 38. Shi, X.; Duft, D.; Parks, J.H. A fluorescence probe of conformational dynamics of noncovalent complexes in gas phase. Proc. 55th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, June 3–7, 2007. Indianapolis, IN 39. Shi, X.; Duft, D.; Parks, J.H. Fluorescence quenching induced by conformational fluctuations in unsolvated polypeptides. J. Phys. Chem. 2008, 112, 12801–12815. 40. Danell, A.S.; Parks, J.H. FRET measurements of trapped oligonucleotide duplexes. Int. J. Mass Spectrom. 2003, 229, 35–45. 41. http://probes.invitrogen.com, accessed 3/25/08; http://www.invitrogen.com/site/us/en/ home/References/Molecular-Probes-The-Handbook/Fluorophores-and-Their-AmineReactive-Derivatives/BODIPY-Dye-Series.html, accessed 5/24/09. 42. Wegewijs, B.; Hermant, R.M.; Verhoeven, J.W.; Kunst, A.G.M.; Rettschnick, R.P.H. Determination of the barrier to intramolecular exciplex formation in a jet-cooled, bichromophoric molecule. Chem. Phys. Letters 1987, 140, 587–590. 43. Oevering, H.; Paddon-Row, M.N.; Heppener, M.; Oliver, A.M.; Cotsaris, E.; Verhoeven, J.W.; Hush, N.S. Long-range photoinduced through-bond electron transfer and radiative recombination via rigid nonconjugated bridges: distance and solvent dependence. J. Am. Chem. Soc. 1987, 109, 3258–3269. 44. Jortner, J.; Bixon, M.; Wegewijs, B.; Verhoeven, J.W.; Rettschnick, R.P.H. Long-range, photoinduced charge separation in solvent-free donor-bridge-acceptor molecules. Chem. Phys. Letters 1993, 205, 451–455. 45. Brunschwig, B.S.; Sutin, N. Directional electron transfer: conformational interconversions and their effects on observed electron-transfer rate constants. J. Am. Chem. Soc. 1989, 111, 7454–7465. 46. Hoffman, B.M.; Ratner, M.A. Gated electron transfer: when are observed rates controlled by conformational interconversion? J. Am. Chem. Soc. 1987, 109, 6237–6243. 47. Yang, H.; Luo, G.; Karnchanaphanurach, P.; Louie, T-M.; Rech, I.; Cova, S.; Xun,L.; Xie, X.S. Protein conformational dynamics probed by single-molecule electron transfer. Science 2003, 302, 262–266. 48. Rhem, D.; Weller, A. Kinetics of fluorescence quenching by electron and H-atom transfer. Z. Phys. Chem. 1970, 69, 183–200. 49. Vaiana, A.C.; Neuweiler, H.; Schulz, A.; Wolfrum, J.; Sauer, M.; Smith, J.C. Fluorescence quenching of dyes by tryptophan: interactions at atomic detail from combination of experiment and computer simulation. J. Am. Chem. Soc. 2003, 125, 14564–14572. 50. Siders, P.; Cave, R.J.; Marcus, R.A. A model for orientation effects in electron-transfer reactions. J. Chem. Phys. 1984, 81, 5613–5624. 51. Borisevich, N.A.; Ivanov, A.L.; Kazakov, S.M.; Kukhto, A.V.; Mit’kovets, A.I.; Murtazaliev, D.V.; Povedailo, V.A.; Khristoforov, O.V. Interaction of electrons with indole, tryptophan, and their derivatives in the gas phase. J. Appl. Spect. 2005, 72, 503–508.
Structure and Dynamics of Trapped Ions
203
52. Ulrich, G.; Ziessel, R.; Harriman, A. The chemistry of fluorescent bodipy dyes: versatility unsurpassed. Angew. Chem. Int. Ed. 2008, 47, 1184–1201. 53. Duft,D.; Shi,X.; Parks, J.H. Fluorescence measurements of a single unsolvated biomolecule: conformation fluctuations. Proc. 55th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, June 3–7, 2007. Indianapolis, IN. 54. Xie, X.S.; Trautman, J.K. Optical studies of single molecules at room temperature. Annu. Rev. Phys. Chem. 1998. 49, 441–480. 55. Bai,C.; Wang, C.; Xie, X.S.; Wolynes, P.G. Single molecule physics and chemistry. Proc. Natl. Acad. Sci. USA 1999, 96, 11075–11076.
of Traveling 8 Applications Wave Ion Mobility-Mass Spectrometry Konstantinos Thalassinos and James H. Scrivens Contents 8.1 Introduction...................................................................................................205 8.2 Drift Cell Ion Mobility Spectrometry (Dcims)...........................................206 8.2.1 Drift Cell Ion Mobility-Mass Spectrometry (Dcim-Ms).................207 8.3 Field Asymmetric Waveform Ion Mobility Spectrometry (Faims).............209 8.4 Traveling Wave Ion Mobility-Mass Spectrometry (Twim-Ms)..................209 8.4.1 Instrumentation..................................................................................209 8.4.1.1 Traveling Wave Ion Guides (TWIGs).................................209 8.4.1.2 Synapt High Definition Mass Spectrometer (Hdms)........ 210 8.4.1.3 Cross-Sectional Calibration................................................ 215 8.5 Applications................................................................................................... 219 8.5.1 Protein Structure................................................................................ 219 8.5.2 Proteomics......................................................................................... 223 8.5.3 Ambient Ionization............................................................................224 8.5.4 Imaging.............................................................................................. 225 8.5.5 Other Application Areas.................................................................... 226 8.6 Future Development....................................................................................... 229 References............................................................................................................... 230
8.1 INTRODUCTION Ion mobility spectrometry (IMS) is now a well-established analytical technique that is employed throughout the world for the detection of explosives, drugs, and chemical warfare agents. The predominant approach is based on the use of a drift cell in which ions migrate through a counter flowing buffer gas in the presence of a low electric field. Separation of ions takes place as a result of interactions between these ions and the buffer gas, and depends on the mass, charge, and shape of the ion. Because the drift cell was employed in the first ion mobility (IM) approach used (and is still the most common), the use of the drift cell in this manner is often referred to as IMS. In discussions of the development of the field of IM, it would be prudent to differentiate this experimental technique in which the drift cell is used from other
205
206
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
IM approaches; thus, this experimental technique in which the drift cell is used will be referred to as drift cell IMS (DCIMS). Experiments that employ DCIMS can be characterized further as being carried out at either ambient or reduced pressure. Ambient pressure experiments have the advantage of a larger number of ion/ molecule interactions per second, which can lead to higher separation efficiency, whilst the reduced pressure approach provides significant advantages when interfacing IMS with mass spectrometry. The field of IMS* has been described recently in a book by Eiceman and Karpas [1]. Recently, two other approaches to IMS have been introduced commercially; these are high field asymmetric waveform IMS (FAIMS) [2] and traveling wave IMS (TWIMS) [3].
8.2 Drift Cell Ion Mobility Spectrometry (DCIMS) IMS, also known as plasma chromatography and (somewhat confusingly!) as ion chromatography, is an analytical technique that separates ions based on their mobility though a counter flowing neutral target gas [13]. Ions are separated on the basis of their different velocities attained when accelerated through a drift tube, filled with a neutral gas, by a constant electric field (E). The drift tube has a series of electrodes in order to provide the constant electric field. Ions are accelerated by the electric field while collisions with the gas decelerate them and this leads to a quasi-constant velocity vd. The IM K is the ratio of vd to electric field strength E
K=
vd E
(8.1).
The IM is expressed usually as reduced mobility, K0, which is the mobility of the ion at standard temperature and pressure (P0 of 760 Torr and T0 of 273.15 K) so that
K0 = K
P × 273.15 PT0 =K T × 760 TP0
(8.2).
Using kinetic theory [14], the reduced mobility K0 can be related to the collision cross-section of the ion species by using the following equation
K0 =
3ze 1 2π × 16 N 0 σ µ k B T
(8.3),
* See Volume 5, Chapter 15: The Study of Ion/Molecule Reactions at Ambient Pressure with Ion Mobility Spectrometry and Ion Mobility/Mass Spectrometry by Gary A. Eiceman and John A. Stone.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
207
where z is the number of charges, e is the electronic charge, N0 is the buffer gas number density at standard temperature and pressure, µ is the reduced mass of the buffer gas and the ion, k B is Boltzmann’s constant, T is the effective temperature, and σ is the average collision integral or collision cross-section. The average collision cross-section can be obtained by averaging over all possible collision geometries [15]. Exact calculation of an average collision cross-section is not trivial and many models have been proposed. Using the hard-sphere model, the polyatomic ion is treated as a collection of hard spheres [16,17]. An extension of this model incorporates long-range interactions so that mobilities can be calculated as a function of temperature [18]. Several other approaches have been employed including the use of neural networks [19]. In the most widely-used form of DCIMS, applied in explosives, drugs, and chemical warfare agent screening, the most common way to produce ions in atmospheric pressure is using a 63Ni source, which effects ionization of the analyte through a series of ion/molecule reactions. These ions are then injected (via an electric potential) through a drift region, where separation takes place, and they are detected subsequently by an ion collector (usually a Faraday plate) [20].
8.2.1 Drift Cell Ion Mobility-Mass Spectrometry (DCIM-MS) IM has been coupled with a range of mass spectrometers. Both electrospray ionization (ESI) [21,22], and matrix-assisted laser desorption/ionization (MALDI) [18] ionization sources have been used to generate ions prior to IM analysis. Different configurations of mass analyzers have been used to perform mass analysis including quadrupole mass analyzers [23], time-of-flight (TOF) analyzers [24] and Fourier transform ion cyclotron resonance spectrometry* (FT-ICR) [25]. Recently, an instrument utilizing two and three drift cells in tandem has been described [26]. IM measures the time it takes for ions to traverse a drift tube. For this reason, pulsed sources, such as MALDI, are well suited to IM studies. ESI sources generate ions in a continuous manner so that a large proportion of the ions (ca 99%) can be lost leading to a loss in sensitivity. An ESI-ion trap interface has been developed in an attempt to overcome this problem. The ion trap was used effectively to store the ions and to pulse them to the drift region at certain intervals thus increasing the sensitivity [27]. DCIM-MS approaches have been used to study a wide variety of molecules from various types of clusters [17,28], polymers [29,30], peptides [31], proteins [23,32], and nucleic acids [33]. Bowers and co-workers have pioneered the approach of relating theoretical cross-sections to those obtained from experiment [18]. This approach involves the use of molecular mechanics/molecular dynamics calculations to generate three-dimensional structures of peptides (and small proteins), minimizing these structures to determine the lowest-energy conformation of these molecules, and using programs [29] to calculate theoretical cross-sections. These values are then correlated to the cross-sections obtained from the IM experiment [34,35]. These calculations were shown to be in excellent agreement with experiment [32]. * See Volume 5, Chapter 5: Fourier Transform Ion Cyclotron Resonance Mass Spectrometry in the Analysis of Peptides and Proteins by Helen J. Cooper.
208
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
DCIM-MS has been coupled to liquid chromatography (LC) for analyzing complex peptide samples [36]. Peptides eluting from the LC column were analyzed on an IMSQ-TOF mass spectrometer. The ions generated from the source were accumulated in an ion trap and injected periodically into the drift tube. After mass analysis by the quadrupole, ions were subjected to collision-induced dissociation (CID) within an octopole collision cell and the product ions were analyzed by a TOF analyzer. Using this instrumental configuration, the urinary proteome [37], the Drosophila melanogaster head proteome [38], and the human plasma proteome [39] have been analyzed. While many additional measurements, compared to standard mass spectrometrybased proteomics experiments, were obtained (for example, collision cross-sections), these were not used to improve upon protein identification results. The majority of DCIM-MS instruments described to date have been built in-house in a small number of laboratories of leading physical chemists. Because the construction of these instruments requires very significant mechanical, electrical, engineering, and software support, the wider application of the technique has been limited markedly. The laboratories concerned must not only manufacture and integrate the mobility device but must assemble mass spectrometric instrumentation that competes with commercial offerings in terms of sensitivity, ease of use, reliability, capability of interfacing with separation science approaches, and acquisition and processing software. A drift cell-based IM device has been incorporated recently into a commercial instrument [40]. A drift cell and ancillary ion optics were placed before the first analyzer in a Q-TOF hybrid instrument by the inclusion of an additional vacuum chamber. The mobility cell, which was based on a design first reported by the Bowers group, is capable of being operated under temperature control. The results were shown to be in good agreement with those obtained previously using in-house built instruments, and the combination offers potential advantages in increased mass range and use of commercial software. A different design of mobility cell has been added also to a commercial Q-TOF hybrid instrument [41]. The modifications were made in the two transfer regions between the source and the initial (quadrupole) mass analyzer. In the first of these, ion storage and IM regions replaced the radio frequency (RF)-only ion guide and, in the second, the RF-only guide was modified to allow the generation of an axial voltage gradient. All ion guides were of stacked ring-electrode geometry; an RF voltage was applied to each electrode such that the phase of the RF voltage on each electrode was 180° out-of-phase with each of its neighboring electrodes. No loss of instrument sensitivity was observed. A range of peptides and proteins were studied using the device and good agreement was obtained with other published drift cell-based approaches. Calibration was carried out by comparing arrival times of ‘standard’ compounds, whose rotationally-averaged cross-sections had been obtained previously, with those of compounds of interest and, after allowing for the time the ions spend outside the mobility device, applying a linear correction. This instrument has been employed subsequently in the analysis of drug formulations using ambient ionization approaches, in which mobility separation was employed to simplify the mass spectra obtained [42], an ambient ionization study of tryptic peptides [43], and in experiments to characterize urinary metabolites [44]. Both these commerciallymodified experiments benefited significantly from the ability of the TOF analyzer to acquire mass spectra on a time scale significantly faster than the time required
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
209
for mobility separation (that is, microseconds rather than milliseconds). This ability of the TOF analyzer enables a large number (typically 200) of mass spectra to be obtained during a single mobility separation. A number of excellent reviews covering the IM-mass spectrometry field have been published [45–47].
8.3 FIELD ASYMMETRIC WAVEFORM ION MOBILITY SPECTROMETRY (FAIMS) Unlike DCIMS and TWIMS, in which ions have to be pulsed into a mobility cell containing buffer gas, FAIMS* can be operated with continuous introduction of ions. It is operated conventionally at atmospheric pressure and ambient temperature, although these may be varied when required. A periodic electrical waveform is applied to conductive surfaces that are typically 2 mm apart. These plates may be either parallel electrodes or, more commonly, concentric cylinder electrodes. The electric waveform is asymmetric with a significant difference between the peak positive and negative voltages. The applied field drags the ions through the background gas and they reach a terminal velocity that is roughly proportional to the electric field. At high electric fields this proportionality changes, making FAIMS possible. The fact that the proportionality is compound-dependent enables the separation of ions to be carried out. During separation, ions will move toward the plates. IM varies between applied high and low electric field, and the motion of the ions can be stabilized by the application of a small direct current (DC) voltage to one of the plates. When the DC voltage is of an appropriate magnitude and polarity, ions of certain mobility can be prevented from hitting the metal plates. This applied voltage is called the compensation voltage (CV). A mixture of ions in a gas flow can be separated, therefore, by scanning the CV to allow ions with different mobilities to achieve successively stable trajectories through the FAIMS device before being detected subsequently. Some ions have mobility that increases with electric field whilst others have mobility that decreases. Applications of FAIMS have included charge selection for the characterization of peptides in proteomics experiments [4–6] and the study of protein folding [7–11]. Because the theory of FAIMS separation has not been developed fully as yet, it is difficult to relate mobility to applied CV [12] and, therefore, to predict from theory the elution order of compounds studied using this approach.
8.4 TRAVELING WAVE ION MOBILITY-MASS SPECTROMETRY (TWIM-MS) 8.4.1 Instrumentation 8.4.1.1 Traveling Wave Ion Guides (TWIGs) Unlike drift cell IM experiments, where a constant low electric field is applied to the mobility cell, TWIMS uses a traveling wave (T-Wave) comprising a series of * See Volume 4, Chapter 5: High-Field Asymmetric Waveform Ion Mobility Spectrometry (FAIMS) by Randall W. Purves.
210
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
transient DC voltages to propel ions through a stacked-ring ion guide (SRIG) to which RF voltages have been applied to consecutive electrodes [48]. The SRIG consists of a series of ring electrodes that are arranged orthogonally to the ion transmission axis, and opposite phases of RF voltage are applied to adjacent rings which creates a radially-confining effective potential barrier preventing ion loss via diffusion. The ring geometry creates axial traps that can slow or even stop ion motion. Ions are propelled through the SRIG by superimposing a DC potential on the RF potential applied to one pair of adjacent ring electrodes. This potential moves along ring electrode pairs across the length of the SRIG at regular time intervals generating a sequence of T-Waves. The voltage applied is superimposed on top of the RF voltage and the RF provides a potential well, which keeps the ions confined radially within the device, significantly improving sensitivity. This particular configuration of SRIG will from now on be referred to as a traveling wave ion guide (TWIG). Each TWIG has a 5 mm diameter ion transmission aperture, with each electrode having a thickness of 0.5 mm and adjacent electrodes being separated by 1.5 mm (Figure 8.1). 8.4.1.2 Synapt High Definition Mass Spectrometer (HDMS) A concatenation of three TWIGs has been incorporated within a Q-TOF geometry to create the Synapt High Definition Mass Spectrometer (HDMS) system (Waters Corp., Milford, USA) [3], a commercial instrument incorporating IM separation. A schematic representation of this instrument is shown in Figure 8.2. Samples are ionized by means of ESI, and more recently MALDI, and enter the instrument via a ‘Z-spray’ source. Three TWIGs, which form the ‘TriWave’ device, are placed between a quadrupole and an orthogonal acceleration (oa)TOF with the first TWIG referred to as the trap, the second as the mobility cell, and the third as the transfer TWIG. The trap and transfer TWIGs are 100 mm long and each composed of 31 electrode pairs. One difference between these two TWIGs is that the final electrode of the trap TWIG is DC-only and its voltage R.F. (+)
lon exit
lon entry
Electrode thickness 0.5 mm
Nominal spacing 1.5 mm
R.F. (–)
Nominal aperture diameter 5 mm
FIGURE 8.1 Schematic diagram of the stacked-ring ion guide. (From Figure 1(a) in Giles, K., Bateman, R.H. Proc. 49th ASMS Conf. on Mass Spectrometry and Allied Topics, Chicago, IL, May 27–31, 2001. Waters Corporation.)
Oil-free scroll pump
DRE lens
Trap
Non mobility separation Transfer
Air-cooled turbomolecular pumps
Quadrupole
FIGURE 8.2 Schematic diagram of the Synapt instrument.
Analyte spray
T-wave lons guide
Lockmass reference spray Pusher
Detector
Applications of Traveling Wave Ion Mobility-Mass Spectrometry 211
212
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
is modulated so that, periodically, it gates ions into the mobility TWIG for separation to occur. The trap and transfer TWIGs share a common gas supply from a 0–10 mL min –1 mass flow controller such that the pressure in each TWIG is in the region of 10 –2 mbar. While the trap TWIG has no T-Wave applied, the transfer TWIG has a T-Wave applied (1–2 V, 300 m s –1) in order to maintain the mobility separation of the ions, carried out in the mobility cell TWIG, and to transfer the ions to the oa-TOF. The mobility TWIG, also referred to as the traveling wave IM separator (TWIMS), is 185 mm long and consists of 61 electrode pairs. The entrance and exit apertures of the TWIMS are 2 mm in diameter, and it can be operated at pressures of up to 1 mbar with the pressure in the cell being controlled by a 0–200 mL min–1 mass flow controller. The repeat pattern of the T-Wave in this cell is six pairs which means that the DC pulse is applied to the first and seventh pairs of electrodes then the pulse is applied to the second and eighth pairs of electrodes, and so on. The T-Wave velocity is given in meters per second, and is derived from the distance between pairs of electrodes divided by the time the pulse remains in each pair. Given that the spacing between pairs of electrodes is ca 3 mm, a pulse duration of 10 μs will result in an average velocity of 300 m s–1. Velocities applied to the cell can be in the range of 200–600 m s–1. Mobility separation of ions is achieved because ions with high mobility are carried with the wave and exit the device faster than ions of lower mobility that roll over the wave top and, as a consequence, exit the device later. In order to record the arrival time distribution (ATD) of mobility-separated ions, the oaTOF acquisition is synchronized with the gated release of a packet of ions from the trap TWIG into the TWIMS cell. Packets of ions are released typically every 100 µs and for each of these packets, 200 oa pushes (mass spectra) of the TOF analyzer are recorded. The overall mobility separation time is 200 × tp where tp is the pusher period. For a pusher period of 64 µs, the overall mobility separation is, therefore, 12.8 ms. A further 200 mass spectra, which are acquired for the next gated release of ions, are added to the data from the previous gated release. The process is repeated until a desired signal-to-noise ratio, for the mass spectrum, is obtained. This mode of mobility acquisition is one of the major advantages of the Synapt instrument. In one single experiment, both mass spectra and ATD profiles, for each ion in the mass spectrum, are recorded. Another advantage of this instrumentation is that CID experiments can be performed in either, or both, of the trap and transfer TWIGs. This facility allows a number of different experiments to be carried out, such as:
1. Mobility separation of ions in the TWIMS and subsequent fragmentation of these ions in the transfer TWIG 2. Fragmentation of ions in the trap TWIG and subsequent mobility separation of the product ions in the TWIMS 3. Fragmentation of ions in the trap TWIG, mobility separation of the fragments in the TWIMS cell, and further fragmentation in the transfer TWIG
This procedure is similar to that of an MS3 experiment, with the difference that all product ions (MS2 and MS3) are recorded at the same time. The different modes of operation of the Synapt instrument are summarized in Table 8.1.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
213
TABLE 8.1 Modes of Operation of the TWIG Components of the Triple-Stage TWIG Synapt Instrument Trap TWIG
IMS TWIG
Transfer TWIG
Fragmentation Fragmentation Fragmentation Fragmentation
Mobility-mode Mobility-mode Mobility-mode Mobility-mode
Fragmentation Fragmentation
Experiment MS/MS MS/MS Mobility MS/MS-Mobility Mobility-MS/MS MS/MS-Mobility-MS/MS (MS)3
To add to this flexibility, the quadrupole present before the TriWave can be o perated either in the RF-only mode to allow passage of all ions or in the resolving mode to isolate mass-selected ions of a particular m/z-value. Experimental parameters available for mobility separation include mobility gas, mobility gas pressure, T-Wave height (voltage), T-Wave velocity, and injection voltage. In addition the T-Wave voltage and the T-Wave velocity can be programed, either linearly or using an operator-defined function. Processing of the experimental data is carried out using specially-written commercial software. There are two ways in which mobility data can be extracted from the raw data files. First, the mobility information can be embedded in the raw files; however, this can be done only during acquisition and the mobility data are saved in a separate channel of the raw data. For lengthy acquisitions, however, writing of the mobility data can take a significant time (in comparison to the very fast timescale of the experiment) preventing the acquisition of further data during this time. The second way is to reconstruct the mobility data afterward using a program called DriftScope (Waters Corporation, Milford, USA). The raw data files are viewed in DriftScope using a mass/charge ratio vs arrival time plot. The entire data, or regions of this plot, can be selected to recreate the mobility data and save it to a separate raw file. Once the mobility data have been saved, extraction of ATDs for ions of interest is carried out as follows. The entire ATD profile for all ions, which can be referred to as a ‘mobilogram,’ is viewed as a chromatogram using the instrumentation software. This mobilogram is spread over 200 scans, each scan representing one push of the oa-TOF. Combining data from all 200 scans will recreate the entire mass spectrum (Figure 8.3). Selecting certain mass ranges in the mass spectrum will recreate the ATD for that mass range which can then be displayed. 8.4.1.2.1 Synapt Advantages The initial paper describing the TWIMS instrument [3] outlines clearly the advantages and limitations of the new approach. The advantages can be summarized as follows.
1. High sensitivity. It was demonstrated that introduction of the TWIMS device does not compromise the intrinsic sensitivity of the hybrid Q-TOF instrument.
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
1000 1500
2000 2500 3000
(b) Mass spectrum
3500
m/z
214
(a) Output of the driftscope program (abscissa represents scan number ordinate represents m/z)
20
(c) ATD profile for all ions in mass spectrum
40
60
80
100
120
140
160
Scan 180 200
FIGURE 8.3 (a), Output of the Driftscope program for a truncated Syrian Hamster protein (fragment 90–231); (b), corresponding mass spectrum; and (c), overall arrival time distribution for all ions in the mass spectrum, often referred to as a ‘mobilogram.’
2. Excellent reproducibility. The reproducibility shown was better than one scan (in the 200 collected for each mobility injection) with standards having drift times in very good agreement with that observed for the same compound in complex mixtures. 3. Versatility. The ability of the instrument to collect MS (and later MS/MS) data and also to perform MS/MS experiments before and/or after the mobility cell provides an analytical technique of significant flexibility. 4. Ease of use. Because the instrument is essentially a modified version of a successful hybrid Q-TOF, then interfaces to separation science components (under full computer control), data acquisition, data processing, and other conventional mass spectrometry experiments are already developed fully.
8.3.1.2.2 Synapt Limitations The limitations can be summarized as follows:
1. Determination of cross-sections. The ion motion within the TWIMS device is complex and not yet understood fully. The direct calculation of cross-sectional values from the measured drift times is not yet possible. As
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
215
we describe below, however, it is possible to obtain estimated cross-sections by acquiring data on calibration standards of established cross-sections and calibrating via reference. 2. Flexibility of drift cell design. As launched, the TWIMS device is not capable of temperature control and has an upper pressure limit of 1 mbar. 3. Resolution. The resolving power of the device, although comparable with that of some published drift cell results, falls short of that obtained with specialized in-house developed DCIMS instruments and, in particular, atmospheric pressure devices.
The instrument as launched had only ESI capability although a MALDI source has been introduced recently. Three types of quadrupole mass filter have been supplied having a 4, 8, or 32 kTh upper transmission range which are appropriate for studies of materials with a variety of molecular weights. For the first time, therefore, it was possible to purchase an IM-enabled commercial mass spectrometer to one’s own specifications. This opportunity for client-designed instruments has led to a significant increase in the IM user base. 8.4.1.3 Cross-Sectional Calibration The use of a T-wave to separate ions using IM results in complex ion trajectories through the mobility cell; thus, to date, direct calculation of an absolute crosssection, for ions of interest, is not possible. Despite this difficulty, calibration of the TWIMS device to obtain estimates of cross-sections has been demonstrated [49–55]. The most common calibration approach relies on using the Synapt to measure the mobility of compounds of known cross-section values, and to derive estimates of cross-section of unknown compounds from a calibration curve. Slight variations to this framework exist, with some practitioners plotting cross-section vs arrival time, corrected cross-sections vs corrected arrival times (see protocol below), or reciprocal cross-sections vs arrival time. Both linear and power series fits to the data have been used. While a linear fit is appropriate when studying small peptides of the same charge state, calibration using differently-charged protein ions requires a power fit [56]. A recent study has highlighted the need of deriving TWIMS calibrations in terms of mobilities and not cross-sections [56]. What this means is that in order for a calibration to be valid, the mobilities (in this case, the corrected arrival times), and not the cross-sections, of the calibrant need to bracket the corrected arrival times of the analytes to be calibrated. It has been shown that the mobilities of ions produced from ESI depend weakly on their mass [57] and that ions with a small spread in their mobilities can span a broad range of cross-sections [58–60]. This is why myoglobin (with crosssections from 2352 to 3313 Å 2 for charge states [M + 8H]8+ to [M + 16H]16+) can be used successfully to create a calibration used for the study of intact hemoglobin (Hb) tetramers [61] (cross-sections from 3900 to 4500 Å 2). Unfortunately, calibration standards do not exist in order to calibrate the Synapt for the study of large protein complexes. In such cases, a linear extrapolation of the calibration line has been used [54].
216
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
One of the challenges in the field is to develop a theoretical understanding of the separation process in the TWIMS device in order to understand, to model, and to predict the mobility process of the TWIMS device to enable cross-sections to be calculated from measured drift times. An initial theory describing the fundamentals of the TWIMS device using derivations and ion dynamics simulations has been published [56]. A hard-sphere collision model to predict drift times depending on varying wave height voltages, which was in very good agreement with experiment, has been also shown [62]. A method to calculate directly the crosssection from drift time has been presented but requires measurement of the ion at the precise pulse height at which it starts to travel at the same speed as the T-Wave [55]. This approach, while potentially useful, is not applicable practically for cross-section measurement of a large number of ions analyzed in the same experiment. A protocol used to calibrate the TWIMS is presented here and an Excel spreadsheet simplifying its use is available online (http://www2.warwick.ac.uk/fac/sci/ bio/research/jscrivens/synapt_calibration/). To create a calibration, values of absolute cross-sections obtained on DCIM-MS instrumentation are required. Such values have been published [23,59,63] and are available free from the Web site of Professor D.E. Clemmer (http://www.indiana.edu/~clemmer/). The procedure is as follows.
1. Measure arrival time in scans (scan number (n)) 2. Convert to time (multiply by pusher time) t d = n × pusher time (ms)
(8.4).
3. Correct for m/z-independent TOF. (a) This time is related to the T-Wave velocity and is the time an ion spends in the mobility region (tm) and the transfer region (tt). The time is calculated for each pair of plates. The mobility cell has 61 pairs of plates so 61 × tm needs to be subtracted from td. The transfer region has 31 pairs of plates so an additional 31 × tt needs to be subtracted. At 300 m s–1, tm, and tt are equal to 10 µs, subtraction would therefore be [(61 × 10) + (31 × 10)] µs = 920 µs. (This time is not dependent upon the mass/charge ratio.) Correcting for these values one gets
td′ = td − 920
(8.5).
( b) Correct for m/z-dependent TOF. For m/z 1000, the TOF flight time is 44 µs and the transit time from the exit of the TriWave to the TOF is 41 µs, giving a total time of 85 µs
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
217
to be subtracted from the drift time. The correction time to the drift time is proportional to the square root of the m/z value.
m z td′′ = td′ − × 0.085 1000
(8.6),
td″ is termed the corrected effective drift time. 4. Obtain calibration coefficients from published cross-section data. (σ) 5. Correct published cross-sections by taking into account reduced mass and charge state σ′ =
σ
1 1 e + mi mn
(8.7),
e = charge, mi = mass of ion, mn = mass of mobility gas.
6. Plot σ′ against td″ 7. Fit appropriate curve to data points ( a) Power fit y = Ax B, or (b) Linear fit y = Ax + B.
8. Calculate A and B from fit. 9. Convert experimental measurements to estimated cross-sections by calculating (a) For a power fit 1 1 σ = td′′ B × A × e × + mi mn
(8.8);
1 1 σ = [( td′′× A) + B ] × e × + mi mn
(8.9).
(b) For a linear fit
Figure 8.4 shows a plot of corrected effective drift time vs corrected published cross-sections used to create a calibration. Because a number of parameters such as the TWIMS velocity, height, and pressure in the cell all affect the resulting ATD profile, the
218
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Corrected cross-sections (Å2)
1700 1600 1500 1400
y = 573.9×0.433 R2 = 0.985
1300 1200 1100 1000
3
4
5
6
7
8
9
10
11
Corrected arrival times (ms)
FIGURE 8.4 Normalized cross-sections vs corrected arrival times for myoglobin. Plot used to create a cross-section calibration for the Synapt instrument. The sample analyzed was equine myoglobin and charge states [M + 8H]8+ –[M + 16H]16+ were used.
Published cross-section (Å2)
2650 2600 2550 2500 2450
R2 = 0.9856
2400 2350 2300 2250 2200 2150 2200
2250
2300
2350
2400
2450
2500
Estimated cross-section (Å2)
2550
2600
FIGURE 8.5 A Comparison of published and estimated cross-sections for bovine cytochrome c charge states [M + 9H]9+ to [M + 15H]15+. The calibration used for estimating crosssections is presented in Figure 8.4.
optimized parameters for the sample of interest need to be established before analyzing the calibrant compound under the exact same conditions. The actual TWIMS parameters do not affect the outcome of the calibration. Figure 8.5 shows the excellent agreement obtained between estimated and absolute cross-sections for bovine cytochrome c (a calibration was created using values for equine myoglobin), while Figure 8.6 shows an equivalent agreement for a set of peptides. Because proteins with the same function, for example, myoglobin, yet obtained from different organisms have been shown to have different cross-sections [51], it is important to use the correct protein when creating a calibration. The fact that absolute cross-sections were obtained on DCIMS instrumentation using helium as the mobility gas whereas the TWIMS cell is filled usually with nitrogen, does not appear to affect the validity of the calibration [49].
219
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
Estimated cross-section (Å2)
450 400 350 300
y = 0.998× + 0.248 R2 = 0.990
250 200 200
y = 0.991× + 2.609 R2 = 0.990
250
300
350
400
450
Published cross-section (Å2)
FIGURE 8.6 A comparison of estimated and published cross-sections for a set of 32 peptides obtained from four different tryptic protein digests (alcohol dehydrogenase (yeast), aldolase (rabbit), creatine phosphokinase (bovine), and hemoglobin (rabbit)). For each peptide the 2 + charge state was used. The same peptides were analyzed twice with an interval of eight months between experiments. (Reproduced from Thalassinos, K.; Grabenauer, M.; Slade, S.E.; Hilton, G.R.; Bowers, M.T.; Scrivens, J.H.; Anal. Chem. 2009, 81, 248–254. With permission from the American Chemical Society.)
8.5 APPLICATIONS 8.5.1 Protein Structure The majority of applications using the Synapt have been focused thus far on studying the conformation of proteins and protein complexes. A prototype of the commercial Synapt instrument was used to study the architecture of trp RNA binding protein (TRAP) [52]. Molecular dynamics methods, in which each TRAP subunit was modeled as a sphere, were used to calculate theoretical cross-sections for ring and collapsed structures of the complex. Estimated cross-sections obtained from the Synapt were correlated to these theoretical structures that allowed a number of important biological conclusions to be drawn. It was shown that the 11-membered ring topology of the complex was maintained in the gas-phase in the absence of bulk water, and that the complex was stabilized with the addition of tryptophan or binding of RNA. The ability of the Synapt to record mass spectra in addition to mobility, showed that the RNA-bound complex co-existed in solution with the unbound form. The ability to distinguish between these two co-existing forms and to study selectively the conformation of the biologically-important holo-form is of tremendous value in biology. Other low-resolution characterization approaches used in structural biology, such as X-ray scattering and circular dichroism spectroscopy, are unable to provide this distinction and, therefore, results from these latter approaches provide information on the overall structures only. This limitation of X-ray scattering and circular dichroism spectroscopy can lead to incomplete biological conclusions. The capability of TWIM-MS to monitor and to separate distinct conformational families co-existing in solution has been used also to study the amyloidogenic protein
220
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
β2-microglobulin (β2m) [64]. Unfolding of the protein is followed by self-aggregation that, in turn, leads to formation of amyloid fibrils. Deposition of these amyloids in the musculo-skeletal system is the cause of dialysis-related amyloidosis. Acid-unfolding of β2m revealed the presence of three discrete conformational states corresponding to folded, partially-folded, and acid-unfolded populations. The emergence of the partially-folded population occurred at pH ca 4.8, while that of the acid-unfolded population at pH ca 3.0. The authors then studied three different forms of the β2m, the wild-type, the single mutant I7A, and the double mutant I7A/P32G, under the same physiological conditions (pH 6.8). For the single mutant, profiles of m/z-value vs drift time were the same as those of the wild-type and revealed the presence of only folded populations. Profiles for the double mutant, on the contrary, revealed the presence of both folded and partially-folded populations indicating the unstable nature of this mutant. These data were in agreement with other observations from which it had been concluded that the double mutant forms fibrils de novo under physiological conditions while the single mutant forms fibrils only in the presence of seeds. This study exemplifies that TWIM-MS is able to provide protein stability information, which is an application that could be of great value for the screening of proteins produced using recombinant synthesis. TWIM-MS has been used to study the prion protein [65], a fibril-forming protein involved in prion diseases. Prions are a class of fatal, infectious, neurodegenerative diseases that affect both humans and animals. The infectious agent PrPSc is thought to be a conformational iso-form of the host-encoded prion protein (PrPC). It is the conversion of the PrPC, which is predominantly α-helical, to the β-sheet rich PrPSc and its subsequent aggregation that is considered to be the cause of neuronal death. Recombinant Syrian hamster PrP protein was expressed in Escherichia coli and purified. The conversion of α-helical PrP to β-sheet-rich structures was carried out by thermal shock and the two iso-forms were distinguished successfully, at pH 5.5, using TWIM-MS. Separation of different synthetic Aβ oligomers (Aβ40, Aβ42, and Aβmet) has been achieved using the TWIM-MS approach [66]. Aβ oligomers are believed to be the neurotoxic agents in Alzheimer’s disease, a fatal brain disease affecting millions of people worldwide. The greatest mobility separation of the different Aβ constructs was observed for the lowest charge states of the Aβ dimers. Addition of proton transfer reagents (triethylene, tributylamine) enhanced Aβ separation because these reagents scavenge protons from multiply-protonated proteins, enhancing the intensity of the lower charge states. TWIM-MS analysis shows great promise in studying protein mis-folding diseases because the aberrant forms of these proteins are often unstable and tend to aggregate in very short time scales. Further research could lead to the development of ante-mortem diagnostic tests capable of screening mis-folded proteins in biological samples such as body fluids. The structures of virus capsids [67,68] and virus tails [69] have been studied using TWIM-MS. The Hepatitis B virus capsid protein can form two different icosahedral capsids of different size, one having a triangulation number of T = 3 and the other having T = 4. In virology, the triangulation number refers to the number of smaller identical equilateral triangles into which each triangular face of an icosahedron can be subdivided. A good image can be found at http://www.virology.wisc.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
221
edu/virusworld/tri_number.php. In a study by Uetrecht et al., it was observed that the T = 3 and T = 4 capsids were present as different conformers for all charge states observed [67]. The authors suggested that the two conformers represented alternative, approximately isoenergetic, states that were populated in the ionization/ desolvation process. Estimated cross-sections were in good agreement with radii obtained from cryo electron microscopy (EM) suggesting that the capsids retained their hollow-sphere structure in the gas-phase. A number of collisional activation experiments followed by IM measurements also supported this finding. During the study of the P22 phage tail assembly and stoichiometry, it was shown, using TWIM-MS, that the binding of protein gp4 (a structural protein, which along with gp10 and gp26, plays a role in sealing the phage capsid) to the P22 tail induces a major conformational change in the phage portal ring [70]. The conformational change stabilizes a closed conformation of the portal protein that aids in sealing the DNA in the phage capsid and facilitates the assembly of the other tail proteins. A large number of biological processes are initiated within a cell, after key proteins, such as protein receptors and enzymes, are activated following the binding of ligands such as metals or small molecules. This ligand binding leads usually to a conformational change of the protein that results in the protein becoming active.TWIM-MS is ideally suited, and has been used, for the study of such events [70–72]. Type I cyclic guanosine 3′,5′-monophosphate (cGMP)-dependent protein kinase (PKG) is part of the family of cGMP-dependent kinases which are implicated in the regulation of a number of important biological functions, such as smooth-muscle relaxation and platelet function. A TWIM-MS study of the binding of cGMP to PKG showed that the conformation of PKG was increased upon cGMP binding [72]. Hydrogen/deuterium (H/D) exchange monitored by ESI-MS was used also, in the same study, to monitor the process and the results from both techniques were in excellent agreement. An advantage of TWIM-MS compared to H/D exchange is that less sample preparation is required and the experiment is much faster to perform. TWIM-MS and EM have been used to probe the conformational changes and stoichiometry of the Drosophila Toll receptor bound to its ligand [71]. Binding of a cytokine ligand, Spätzle (Spz), to the Toll receptor induces a conformational change that, in turn, initiates downstream signal transduction events. The plots of mass/ charge ratio vs drift time of Toll and Spz revealed a number of overlapping ion series corresponding to Toll1, Toll1:Spz1, Toll2:Spz1, and Toll2:Spz2. Whereas the Toll1:Spz1 complexes had a bimodal ATD, the Toll2:Spz2 had unimodal distributions leading the authors to propose that the Toll2:Spz2 complex was the predominant stoichiometry in solution. This observation was in agreement with the results obtained from EM and taken together were used to propose a new mode of dimerization for the Toll receptor which was induced indirectly, by an endogenous protein ligand, and not directly as thought previously. Another study [51] has also provided strong evidence that protein structural studies carried out using TWIMS-based mobility experiments are valid. Data that were in good agreement with X-ray and nuclear magnetic resonance (NMR) experiments were provided. A number of proteins were analyzed under near-native physiological and under denaturing conditions. Cross-sections for ions observed in the mass spectrum were estimated based on the calibration method presented here (see Section
222
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
8.4.1.2. Cross-sectional Calibration). MobCal, a program to calculate mobilities [73,74], was used to calculate cross-sections for structures obtained from the PDB database (http://www.rcsb.org) via three methods: the projection approximation (PA), the exact hard-sphere scattering (EHSS), and the trajectory method (TM). PA is the simplest method that replaces the cross-section of an ion with its projection (shadow) and averages the projections created by every orientation of the ion. The PA is an adequate approximation for small molecules but underestimates the crosssection of proteins with highly convex structures where interactions with the buffer gas become important. The TM gives the most reliable estimate, incorporating all interactions, but is computationally intense. A compromise is to use the EHSS method, which ignores electrostatic interactions and so requires substantially less computational time. The values obtained from the TM and EHSS methods were found to be within a few percent of each other. The cross-section value of the lowest charge state using TWIMS, for all the proteins studied, was between the PA and TM values indicating that this conformation corresponds to that which is biologically most relevant. The study also found that the flexibility and unfolding of proteins in the gas-phase could be monitored using the Synapt; the change in cross-section, as more charges were added, was smaller for proteins containing intact disulfide bonds compared to proteins with no disulfide bonds present. Using a similar approach, the assembly process of intact normal and aberrant Hb tetramers has been examined [61]. Hb is a tetramer consisting of four globin chains, two α and two β-chains, each associated with a heme group. Because Hb is the major oxygen-transport protein found in the red blood cells of all vertebrates, it is of significant biological importance. The most common of all inherited disorders are those of Hb. One of the most incapacitating Hb variants is the one that causes sickle-cell anemia. The sickle-cell mutation results in the production of a β-chain with a single amino-acid substitution (β6 Glu→Val) and changes the conformation of the assembled tetramer to allow molecular stacking which, in turn, causes red blood cells to have a sickle shape (compared to the normal disk-shape found in normal individuals). The cross-sections measured in the experiment for normal (Hb A) and sickle hemoglobin (Hb S) were different so that the two forms could be differentiated easily. The Hb S variant was found to have a larger cross-section than Hb A in agreement with theoretical cross-section values. The study showed that assembly of the Hb tetramer proceeds with two heterodimers coming together to form the tetramer. Monomers of α and β-chains both with and without heme bound were also observed, but binding of heme did not result in conformational changes of these molecules. A study of the tetrameric transthyretin (TTR) complex by means of TWIM-MS, computational modeling, and activation of selected ions by acceleration in the source region of the instrument, revealed that during the dissociation of protein complexes a number of partially-folded intermediate states were stable during the time-scale of the experiment (5–25 ms) [75]. The results indicated that upon activation, unfolding of one or more subunits occurred while the remainder of the subunits remained folded. This occasion was the first time that partially-folded intermediates had been studied; such types of measurements will allow future studies into the understanding of the dissociation mechanism of other large protein complexes to be carried out.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
223
The advantage of coupling high performance liquid chromatography (HPLC) to TWIM-MS for the mass spectrometric analysis of H/D exchange data have been exemplified during the analysis of Hck SH3 domain [76]. Analysis of the intact protein by means of TWIM-MS revealed that deuteration of the protein did not affect the mobility of the sample. Similar results were observed for peptides with varying degrees of deuteration. Mass analysis of the peptides using standard Q-TOF instrumentation, without any mobility separation, gave similar results. It was concluded, therefore, that incorporation of TWIMS did not affect the peptide deuteration results. Using the additional separation capability of the TWIM-MS, however, the authors were able to separate and to study peptides that were co-eluting during the chromatographic separation step. A study of the gas-phase conformation of small oligonucleotides (dTG, dC6, and C6) by H/D exchange and, independently, by TWIM-MS showed that the two techniques are complementary [77]. The authors make a distinction between hydrogen accessibility and compactness of a structure, which are connected to different physical parameters. In some cases, H/D exchange can resolve different families of conformers that IM is not able to and vice-versa, and so a combination of the two techniques needs to be used for complete characterization of ion conformation.
8.5.2 Proteomics High-throughput proteomic studies using the Synapt have not yet been undertaken. A report by Thalassinos et al., however, has evaluated the use of the Synapt for the screening of phosphorylated peptides in complex peptide mixtures [49]. A plot of cross-sections vs the mass/charge ratio of the non-phosphorylated peptides for doubly-charged ions was found to be linear while phosphorylated peptide ions, doubly-charged also, were shown to exhibit a negative deviation from linearity. This negative deviation from linearity is believed to be due to the adoption of a more compact conformation by the peptides upon phosphorylation. The deviation was greater for peptides having more than one phosphorylated residue. All measurements were performed twice, with an interval of eight months between measurements, and were shown to be highly reproducible. While some non-phosphorylated peptides also showed a negative deviation from the line, no positive deviation was observed for any of the phosphorylated peptides; all phosphorylated peptides always showed a negative deviation. This observed behavior meant that 40% of the peptides analyzed in the study could be discarded as not being phosphorylated. Incorporation of this approach into a proteomics workflow would allow for the rapid screening of phosphorylated peptides. A subset of the peptides was analyzed by means of DCIMS; the cross-sectional values obtained by TWIM-MS and DCIMS were in excellent agreement. The study showed that post-mobility fragmentation of peptides was of high quality allowing sequence information to be obtained. In an extension of the above study [51], the authors compared estimated cross-section measurements from the Synapt with cross-sections of structures calculated theoretically and obtained using a simulated annealing protocol. A software program was developed to automate the extraction of ATD profiles from the raw data.
224
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
8.5.3 Ambient Ionization Desorption ESI (DESI), introduced by Cooks and co-workers [78], was one of the first ambient ionization approaches which allowed the analysis of condensed phase analytes by mass spectrometry under ambient conditions, without the need for a matrix or sample pre-treatment. This novel technique appears to have been the catalyst for the development of a considerable variety of surface sampling and ionization approaches coupled with mass spectrometry. The field has been reviewed recently [79]. These techniques are well suited to being coupled to mobility separation. The rapid sampling enabled by these methods is an ideal introduction method for mobility MS and MS/MS studies with significant potential as a rapid screening approach. DESI has been coupled to a drift cell-based IM mass spectrometer [80]. The coupling was described as straightforward and was used to study protein conformations. It was discovered that DESI was a softer ionization method than ESI and that proteins could be desorbed in their native state (under appropriate solvent conditions). The results supported the droplet pickup mechanism proposed for the DESI experiment. DESI has been employed also as an ionization method for a series of experiments carried out with a modified commercial Q-TOF instrument [42]. Pharmaceutical drug formulations were studied and the active components characterized by DCIMS. The approach showed great promise, both for rapid screening and as a method of separating the background excipients from the molecules of interest. The same instrument was used to study the use of DESI as a rapid screen for the identification of a mixture of peptides from a digested protein [43]. The mobility experiment provided excellent separation of charge states thus simplifying the peptide mass fingerprinting approach to protein identification. The Synapt instrument has been coupled with DESI in order to characterize complex pharmaceutical formulations [50]. Accurate mass MS and MS/MS data were obtained on a number of samples with separation of both background ions and isobaric ions of particular advantage. In the same publication, the instrument was coupled with neutral desorption extractive electrospray ionization (EESI) [81,82] to study both pharmaceutical formulations and synthetic polymer mixtures. For both ionization approaches, interfacing the ionization sources with the Synapt instrument was facile with high information content data being obtained. A wider study on the combination of DESI ionization and mobility separation for the characterization of synthetic polymers have been undertaken using the Synapt [50]. Mass spectra of the complex mixtures were obtained in seconds, an excellent match to the timescale for TWIM-MS and TWIM-MS/MS experiments. The low requirements for sample preparation coupled with the high information content of the experiments indicate significant potential for this approach. In a complementary series of experiments not involving ambient ionization, the Synapt has been used to provide detailed structural information on a series of synthetic (Poly(ethylene glycol)s) polymer mixtures [83]. The mobility separation was employed, coupled with MS/MS experiments to provide end-group and sequence information on the polymer mixture. Effective precursor ion resolution of more than 6000 was obtained (significantly better than could be obtained by using the quadrupole ion analyzer on the instrument for ion isolation) on a series of isobaric ions from the complex mixture. The MS/MS data obtained on these
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
225
obility-separated ions were found to be in excellent agreement with conventional m product ion mass spectra of standard materials. McEwen et al. have utilized another ambient technique called atmospheric pressure solids analysis probe (ASAP) [84,85] and interfaced it with TWIM-MS [86]. This source is now available commercially on the Synapt instrument. They have used m/z vs drift time plots to define compound classes in crude oil analysis and to resolve isobaric compounds. Excellent agreement was demonstrated for drift times obtained from standards run individually and doped into the crude oil mixture. This work demonstrates the utility of mobility experiments to provide a degree of separation and classification in the study of very complex mixtures where chromatographic separation is either not possible or too time consuming.
8.5.4 Imaging Mass spectrometry is rapidly becoming an analytical method* of choice to probe the distribution of molecular species within tissue sections [87]. Both ambient ionization [88] and MALDI [89] approaches have been shown to be successful. Among the challenges presented by this approach are the complex mass spectra that are obtained, the high degree of chemical noise present, and the need for specificity in the profiling. Coupling mobility experiments with imaging offers potential alleviation of a number of these problems. McLean et al. have developed an approach based on a drift cell of length 13.9 cm coupled with an oa-TOF [90]. They used MALDI as an ionization source with pulse repetition rates of 300 Hz for profiling and 100 Hz for imaging. A spatial resolution of 200 µm was obtained. The resolving power of their drift cell device was demonstrated to be 30–50 (t/Δt FWHM). Because the mobility experiment was over in less than 600 µs (less than the 1 ms between laser pulses) no experimental time was lost. Sub-attomole levels of detection for doped materials were demonstrated. Among the advantages claimed for this approach were: (i) qualitative identification of molecular classes; (ii) exclusion of chemical noise; (iii) potential for accurate mass measurement; and (iv) possibility (not shown) for tandem mass spectrometry (MS/MS). The Synapt instrument can be fitted with a MALDI source in order to carry out imaging experiments. Both MS (with < 5 ppm mass accuracy) and MS/MS data (coupled with mobility separation) can be obtained. The laser is operated normally with a repetition rate of 200 Hz and a resolution of 250 µm can be obtained. Using a Synapt instrument, Ridenour et al. demonstrated the capability to acquire high quality small molecule MS/MS data from samples of liver tissue separate from the high background observed [91]. In addition, Clench et al. demonstrated the ease of obtaining high quality MS/MS data from small molecules adsorbed on skin [92] while Trim et al. (also from the Clench group) visualized the distribution of the drug vinblastine and its metabolites in rat tissue, and demonstrated the capability of the * See Volume 5, Chapter 14: The Role of Trapped Ion Mass Spectrometry for Imaging by Timothy J. Garrett and Richard A. Yost.
226
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
approach for removing isobaric interferences [93]. In another study from the Clench group, Djdja et al. visualized (and characterized) proteins in adenocarcinoma tissue sections using in situ enzymatic digestion [94]. Mobility data provided facile separation of singly-charged ions from doubly-charged ions. Stauber et al. demonstrated the acquisition of good quality MS/MS data from a study of brain tissue using the mobility separation to select the ion of interest [95]. It is clear that the ability of mass spectrometry-based imaging to produce images of high information content offers significant potential for further research and application. The inclusion of mobility measurements coupled with MS and MS/MS experiments provides additional information. Very high dimensionality of data are obtained with x-position, y-position, mobility, mass/charge ratio and, potentially, MS/MS information being obtained for each laser spot. The accumulation of data of high dimensionality generate very large data files, which require sophisticated data classification and characterization software.
8.5.5 Other Application Areas Tandem mass spectrometric experiments on peptides (often formed by tryptic digestion of proteins) play a key role in the field of proteomics. An understanding of the fragmentation processes involved is important for data interpretation and database searching. Riba-Garcia et al. have used the Synapt instrument to probe the structural diversity of peptide fragments [96]. Studying a model YAGFL-NH2 peptide, they demonstrated from mobility experiments that the b5 ion formed from CID experiments on the protonated molecule had a stable macromolecular structure and that the a5 ion was present as two structures one of which, based on subsequent fragmentation studies, was attributable also to a macromolecular structure. These initial experiments on model systems indicate that significant additional structural complexity than expected may be present in many peptide MS/MS experiments and demonstrates the power of the Synapt instrument, with its unique combination of collisional and mobility regions, to probe these features. Characterization of glycosylation profiles is a difficult task. Mass spectrometry has made a significant contribution both to the measurement of glycoprotein heterogeneity and, often combined with separation science approaches, to the structural characterization of individual glycans released using enzymatic or chemical means. Mobility studies offer the potential for another dimension of separation, which may simplify the complex mixtures found in glycan studies. Olivova et al. have utilized mobility separation in a detailed study of the glycosylation of monoclonal antibodies [97]. They demonstrated that mobility separation of light and heavy chains of partially-reduced IgG1 was as effective as that carried out using a liquid chromatography-based approach with significant advantages in speed of analysis. In one of the first published experiments to make full use of the versatility of the trap, mobility cell, and transfer regions of the Synapt, the group carried out glycopeptide analysis in which fragmentation of the glycopeptide of interest was carried out in the trap, the fragments separated using mobility with the resulting mobilityseparated ions undergoing a second stage of fragmentation in the transfer region. This experiment is closely related to MS3 experiments carried out traditionally using
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
227
ion-trapping devices; however, in this case, the method has the advantage that all first and second-generation product ions are recorded in parallel. The application of low collision energy in the first fragmentation stage is used to create a peptide with a single N-acetylglucosamine (often the most abundant fragment) while the peptide backbone remains intact. The second fragmentation stage provides information on the peptide sequence and the site of glycan attachment. This use of two separate fragmentation steps (with different collision energy regimes) affords an opportunity to tailor an analysis protocol and has significant potential in both glycan and other complex analyses. Harvey et al. have used the Synapt instrument to characterize glycan structures which have been released, in this case, from glycopeptides [98]. They show the ability of the technique to simplify complex mass spectra by using a plot of mass/ charge ratio vs drift time to classify the data into separate charge state regions. They also show that residual glycoproteins can be separated using this approach. Some isomeric glycan structures could be resolved using mobility separation with comprehensive sequence data being obtained from the MS/MS spectra of the mobility-resolved isomers. Both positive and negative ions were studied with the negative product ion mass spectra providing the most detailed information. No prior chromatographic separation was undertaken and this approach has significant potential in the rapid screening of released glycan structures in, for example, the recombinant synthesis of post-translationally-modified pharmaceuticals. Holland et al. have investigated the gas-phase structure of glucose polymers [99]. Both dextran (a 1–6 linked glucose polymer) and maltodextran (a 1–4 linked glucose polymer) were studied with oligomer lengths from 1 to 40 observed. Estimated cross-sections were obtained by calibration against a peptide mixture covering the range of mobilities studied. It was observed that variation of cation attachment makes little difference to estimated cross-section but that the two systems, although isomeric, exhibit significant differences in estimated cross-sections as the number of oligomer units increases. Differences in estimated cross-section can be observed for singly-sodiated oligomers of seven and above. These relative differences increase as the number of cations (and, therefore, charges) on the molecule increases. There is considerable interest in the biopharmaceutical industry in reacting active molecules with polyethylene glycol (PEG). These PEGylated materials often offer increased circulatory half-life and reduced immunological response. The characterization of these materials offers significant analytical challenges. The large numbers of molecules present, over a wide dynamic range, coupled with the wide molecular weight distributions present, make conventional mass spectrometry experiments and chromatographic approaches difficult. Bagal et al. utilized gas-phase super bases to manipulate the charge states of the ions of interest in order to measure molecular weight distributions [100]. In their complex mixtures, they showed that adding the mobility dimension to the data offers improved dynamic range, better sensitivity, improved selectivity, and the ability to visualize compound classes. They demonstrated that the transmission of a wide molecular weight range of ions through the mobility cell was uniform and that the mobility cell did not introduce bias into the mass spectrometric ion signal intensities. This finding was in contrast to that of
228
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
FAIMS experiments in which the intensity was found to depend on CV [101]. They developed their own in-house software, written using MATLAB®, to process the mobility data and utilized NMR-based approaches to quantify the compound classes visualized using a plot of mass/charge ratio vs drift time. These values were found to be in good agreement with those predicted. The approach enabled them also to see small components, which were not visible in the conventional mass spectrometry experiment such as 2% free PEG in a PEGylated peptide. In the first published example of the integration of liquid chromatography and mobility separation using the Synapt instrument, Eckers et al. used the fact that mobility separation and mass spectrometric detection operate on a much faster timescale than does chromatographic separation to provide additional selectivity in the detection of drug actives [102]. PEG background ions, which co-elute with the compound of interest (in this case Zidovudine a major constituent in a number of anti-viral drugs), can be excluded effectively using mobility separation, because the mobility of PEG ions differs from that of the active ingredient. In another example of using the Synapt instrument to simplify the complexity of mass spectral data, Williams et al. have used mobility separation to analyze tryptically-digested peptides obtained from Hb samples [103]. The work involves the identification of human Hb variants, the relevant peptides of which are often minor components hidden in the complex mass spectra obtained. Mobility experiments allow easy charge state separation enabling the relevant peptides to be observed more clearly and sequenced using MS/MS experiments. Isobaric compounds, found in the complex samples, can be separated and also characterized. The use of direct infusion coupled with mobility separation, a mass scan, and tandem mass spectrometric analysis constitutes a rapid experimental approach yielding high information content that has great potential in the development of diagnostic techniques. The need for additional software to process and to classify the significant data provided by TWIM-MS and TWIM-MS/MS experiments are shown in the work by Jung et al. who are using this approach to characterize N-terminal peptides obtained from the Glu-C enzymatic digestion of histone H3 [104]. The complex mobilityseparated MS and MS/MS spectra were interpreted, classified, and compared using an in-house developed bioinformatics platform. Site-specific post-translational modifications induced by different cell treatment regimes were identified. The Synapt has been used to investigate the conformational structure of the serine octomer [105]; this is an interesting experiment because this octomer is related to the development of homochirality. Serine clusters composed of eight amino-acids have been found to have a remarkable preference for homochirality. Mobility MS and MS/ MS experiments were carried out and demonstrated that two major isomeric forms of the molecule are present; the more compact form is able to dissociate readily to the protonated dimer and serine monomer, whilst the less compact form is much more resistant to dissociation. These results are to be compared with theoretical calculations to throw new light on the structure of these potentially very important molecules. In an application well-suited to mobility studies, organophosphate chemical warfare simulants were studied using the traveling wave approach [106]. The ability of
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
229
the instrument to separate components at high speed, together with the provision of MS and MS/MS information on the separated components, yields excellent selectivity that bestows great confidence in the identification of the components of interest. The use of this approach to identify a wide range of the compounds scheduled under the Chemicals Warfare Convention is anticipated.
8.6 FUTURE DEVELOPMENT Clearly, the Synapt instrument has enabled a wide range of research groups to carry out some highly informative IM experiments. In the short time since the introduction of the instrument, a wide range of applications has been studied and it is abundantly clear that a substantial growth in the mobility literature can be expected. It is likely that, because the technology is still in the early stages of development, experimental improvements will be made in the short to medium term. Calibration of the instrument with standards of known cross-section has enabled estimated cross-sections to be obtained for peptides, proteins and, with extrapolation, higher molecular weight molecular complexes. These methods are now well established. Exciting data on conformational changes in biological systems (at biologically-relevant concentrations) have been obtained already and it is certain that these experiments will become even more important as more groups are able to make the measurements. Further development could involve a deeper understanding of the traveling wave mobility experiment so that cross-sections could be calculated or predicted from experimental drift time data. There is a need for well-characterized calibration standards that cover a wider mobility space extending the calibration range that can be used without extrapolation. The technique has significant potential in providing an additional dimension of information in proteomics experiments. In order to fulfill this potential, a number of issues need to be addressed. The ability to acquire mobility-separated MS and MS/ MS data within 20 ms provide one of the major advantages of the technique. There is an excellent match in timescales between the seconds of the chromatographic separation, the milliseconds of the mobility experiment, and the microseconds of the MS data acquisition. In order for these experiments to become routinely part of the classic liquid chromatography-ESI-MS/MS proteomics approach, dynamic range considerations imposed by the instrument need to be addressed. There must be no loss of information in carrying out the mobility stage of analysis. This is an experimental challenge but one that can be met. Another feature of mobility-enabled proteomics experiments is the very large volume of data that will be produced. For high throughput experiments, new software approaches will be required in order to collect, process, classify, and to extract useful information from the multi-dimensional data sets that will be acquired. This need for additional application-specific software is common across the current and potential application areas of the Synapt instrument. These programs will need to be provided by a combination of efforts of the vendor and the significant and expanding user base. The resolution of the device is defined commonly as resolving power = t/Δt FWHM where t is the drift time and Δt FWHM is the width of peak at full width half maximum. Currently, the resolving power of the Synapt is limited to ca 10. This value
230
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
is comparable with some published drift cell experiments but is inferior to those incorporating longer drift cell designs or other novel design features, and is certainly less than can be obtained with atmospheric mobility instruments. Although the ability to acquire MS and MS/MS data improves significantly the selectivity of the device, improved resolution would be desirable. From observation of the drift cell literature, there are a number of potential ways by which this objective could be achieved, such as longer cell geometry, higher pressures, lower temperature, and multi-pass designs, as well as a number of others. It is essential, however, that improved resolution does not come at the expense of sensitivity, ease of use, and information content. Because the implementation of the traveling wave design for mobility studies is relatively recent, it is possible that design improvements particular to the traveling wave approach may be made. What is already clear is that the number of installed Synapt instruments has increased considerably and rapidly among the research community of IM practitioners. The enthusiastic acceptance of Synapt instruments will ensure that, as well as building on the excellent work of the pioneers of the technique, new and exciting application areas will be explored and developed.
REFERENCES
1. Eiceman, G.A.; Karpas, Z. Ion Mobility Spectrometry. 2nd Edn. CRC Press, Boca Raton, FL, 2005. 2. Guevremont, R.; Purves, R.W. Atmospheric pressure ion focusing in a high-field asymmetric waveform ion mobility spectrometer. Rev. Sci. Instr. 1999, 70, 1370–1383. 3. Pringle, S.D.; Giles, K.; Wildgoose, J.L.; Williams, J.P.; Slade, S.E.; Thalassinos, K.; Bateman, R.H.; Bowers, M.T.; Scrivens, J.H. An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/traveling wave IMS/ oa-ToF instrument. Int. J. Mass Spectrom. 2007, 261, 1–12. 4. Guevremont, R.; Barnett, D.A.; Purves, R.W.; Vandermey, J. Analysis of a tryptic digest of pig hemoglobin using ESI-FAIMS-MS. Anal. Chem. 2000, 72, 4577–4584. 5. Venne, K.; Bonneil, E.; Eng, K.; Thibault, P. Improvement in peptide detection for proteomics analyzes using NanoLC-MS and high-field asymmetry waveform ion mobility mass spectrometry. Anal. Chem. 2005, 77, 2176–2186. 6. Tang, K.; Li, F.; Shvartsburg, A.A.; Strittmatter, E.F.; Smith, R.D. Two-dimensional gasphase separations coupled to mass spectrometry for analysis of complex mixtures. Anal. Chem. 2005, 77, 6381–6388. 7. Shvartsburg, A.A.; Bryskiewicz, T.; Purves, R.W.; Tang, K.; Guevremont, R.; Smith, R.D. Field asymmetric waveform ion mobility spectrometry studies of proteins: Dipole alignment in ion mobility spectrometry? J. Phys. Chem. B 2006, 110, 21966–21980. 8. Shvartsburg, A.A.; Li, F.; Tang, K.; Smith, R.D. Characterizing the structures and folding of free proteins using 2-D gas-phase separations: Observation of multiple unfolded conformers. Anal. Chem. 2006, 78, 3304–3315. 9. Purves, R.W.; Barnett, D.A.; Ells, B.; Guevremont, R. Gas-phase conformers of the [M + 2H](2+) ion of bradykinin investigated by combining high-field asymmetric waveform ion mobility spectrometry, hydrogen/deuterium exchange, and energy-loss measurements. Rapid Commun. Mass Spectrom. 2001, 15, 1453–1456. 10. Purves, R.W.; Barnett, D.A.; Ells, B.; Guevremont, R. Elongated conformers of charge states + 11 to + 15 of bovine ubiquitin studied using ESI-FAIMS-MS. J. Am. Soc. Mass Spectrom. 2001, 12, 894–901.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
231
11. Borysik, A.J.; Read, P.; Little, D.R.; Bateman, R.H.; Radford, S.E.; Ashcroft, A.E. Separation of beta2-microglobulin conformers by high-field asymmetric waveform ion mobility spectrometry (FAIMS) coupled to electrospray ionization mass spectrometry. Rapid Commun. Mass Spectrom. 2004, 18, 2229–2234. 12. Shvartsburg, A.A.; Smith, R.D. Optimum waveforms for differential ion mobility spectrometry (FAIMS). J. Am. Soc. Mass Spectrom. 2008, 19, 1286–1295. 13. Cohen, M.J.; Karasek, F.W. Plasma chromatography TM-new dimension for gas chromatography and mass spectrometry. J. Chromatogr. Sci. 1970, 8, 330–337. 14. Mason, E.A.; McDaniel, E.W. Transport Properties of Ions in Gases. Wiley, New York, 1988. 15. Clemmer, D.E.; Jarrold, M.F. Ion mobility measurements and their applications to clusters and biomolecules. J. Mass Spectrom. 1997, 32, 577–592. 16. Lee, S.; Wyttenbach, T.; Von Helden, G.; Bowers, M.T. Gas phase conformations of Li+, Na+, K+, and Cs+ complexes with 18-Crown-6. J. Am. Chem. Soc. 1995, 117, 10159–10160. 17. Von Helden, G.; Hsu, M.; Gotts, N.; Bowers, M. Carbon cluster cations with up to 84 atoms: Structures, formation mechanism, and reactivity. J. Phys. Chem. 1993, 97, 8182–8192. 18. Von Helden, G.; Wyttenbach, T.; Bowers, M.T. Inclusion of a MALDI ion source in the ion chromatography technique: Conformational information on polymer and biomolecular ions. Int. J. Mass Spectrom. Ion Processes 1995, 146/147, 349–364. 19. Wessel, M.D.; Jurs, P.C. Prediction of reduced ion mobility constants from structural information using multiple linear regression analysis and computational neural networks. Anal. Chem. 1994, 66, 2480–2487. 20. Collins, D.C.; Lee, M.L. Developments in ion mobility spectrometry-mass spectrometry. Anal. Bioanal. Chem. 2002, 372, 66–73. 21. Wittmer, D.; Yong Hong, C.; Luckenbill, B.K.; Hill, H.H. Electrospray ionization ion mobility spectrometry. Anal. Chem. 1994, 66, 2348–2355. 22. Wu, C.; Siems, W.F.; Asbury, G.R.; Hill, H.H. Electrospray ionization high-resolution ion mobility spectrometry-mass spectrometry. Anal. Chem. 1998, 70, 4929–4938. 23. Shelimov, K.B.; Clemmer, D.E.; Hudgins, R.R.; Jarrold, M.F. Protein structure in vacuo: Gas-phase conformations of BPTI and cytochrome c. J. Am. Chem. Soc. 1997, 119, 2240–2248. 24. Srebalus, C.A.; Li, J.; Marshall, W.S.; Clemmer, D.E. Gas-phase separations of electrosprayed peptide libraries. Anal. Chem. 1999, 71, 3918–3927. 25. Bluhm, B.K.; Gillig, K.J.; Russell, D.H. Development of a Fourier-transform ion cyclotron resonance mass spectrometer-ion mobility spectrometer. Rev. Sci. Instr. 2000, 71, 4078–4086. 26. Merenbloom, S.I.; Bohrer, B.C.; Koeniger, S.L.; Clemmer, D.E. Assessing the peak capacity of IMS-IMS separations of tryptic peptide ions in He at 300 K. Anal. Chem. 2007, 79, 515–522. 27. Hoaglund, C.S.; Valentine, S.J.; Clemmer, D.E. An ion trap interface for ESI-ion mobility experiments. Anal. Chem. 1997, 69, 4156–4161. 28. Lerme, J.; Dugourd, P.; Hudgins, R.R.; Jarrold, M.F. High-resolution ion mobility measurements of indium clusters: Electron spill-out in metal cluster anions and cations. Chem. Phys. Lett. 1999, 304, 19–22. 29. Wyttenbach, T.; Von Helden, G.; Batka, J.J.; Carlat, D.; Bowers, M.T. Effect of the long-range potential on ion mobility measurements. J. Am. Soc. Mass Spectrom. 1997, 8, 275–282. 30. Gidden, J.; Bowers, M.T.; Jackson, A.T.; Scrivens, J.H. Gas-phase conformations of cationized poly(styrene) oligomers. J. Am. Soc. Mass Spectrom. 2002, 13, 499–505.
232
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
31. Kaleta, D.T.; Jarrold, M.F. Noncovalent interactions between unsolvated peptides: Dissociation of helical and globular peptide complexes. J. Phys. Chem. B 2003, 107, 14529–14536. 32. Wyttenbach, T.; Von Helden, G.; Bowers, M.T. Gas-phase conformation of biological molecules: Bradykinin. J. Am. Chem. Soc. 1996, 118, 8355–8364. 33. Gidden, J.; Bowers, M.T. Gas-phase conformations of deprotonated trinucleotides (dGTT-, dTGT-, and dTTG-): The question of zwitterion formation. J. Am. Soc. Mass Spectrom. 2003, 14, 161–170. 34. Barran, P.E.; Polfer, N.C.; Campopiano, D.J.; Clarke, D.J.; Langridge-Smith, P.R.; Langley, R.J.; Govan, J.R.; Maxwell, A.; Dorin, J.R.; Millar, R.P. Is it biologically relevant to measure the structures of small peptides in the gas-phase? Int. J. Mass Spectrom. 2005, 240, 273–284. 35. Bernstein, S.L.; Wyttenbach, T.; Baumketner, A.; Shea, J.E.; Bitan, G.; Teplow, D.B.; Bowers, M.T. Amyloid beta-protein: Monomer structure and early aggregation states of Abeta42 and its Pro19 alloform. J. Am. Chem. Soc. 2005, 127, 2075–2084. 36. Lee, Y.J.; Hoaglund-Hyzera, C.S.; Srebalus Barnes, C.A.; Hilderbrand, A.E.; Valentine, S.J.; Clemmer, D.E. Development of high-throughput liquid chromatography injected ion mobility quadrupole time-of-flight techniques for analysis of complex peptide mixtures. J. Chromatogr. B 2002, 782, 343–351. 37. Moon, M.H.; Myung, S.; Plasencia, M.; Hilderbrand, A.E.; Clemmer, D.E. Nanoflow LC/ion mobility/CID/TOF for proteomics: Analysis of a human urinary proteome. J. Proteome Res. 2003, 2, 589–597. 38. Taraszka, J.A.; Gao, X.; Valentine, S.J.; Sowell, R.A.; Koeniger, S.L.; Miller, D.F.; Kaufman, T.C.; Clemmer, D.E. Proteome profiling for assessing diversity: Analysis of individual heads of Drosophila melanogaster using LC-ion mobility-MS. J. Proteome Res. 2005, 4, 1238–1247. 39. Liu, X.; Valentine, S.J.; Plasencia, M.D.; Trimpin, S.; Naylor, S.; Clemmer, D.E. Mapping the human plasma proteome by SCX-LC-IMS-MS. J. Am. Soc. Mass Spectrom. 2007, 18, 1249–1264. 40. McCullough, B.J.; Kalapothakis, J.; Eastwood, H.; Kemper, P.; MacMillan, D.; Taylor, K.; Dorin, J.; Barran, P.E. Development of an ion mobility quadrupole time of flight mass spectrometer. Anal. Chem. 2008, 80, 6336–6344. 41. Thalassinos, K.; Slade, S.E.; Jennings, K.R.; Scrivens, J.H.; Giles, K.; Wildgoose, J.; Hoyes, J.; Bateman, R.H.; Bowers, M.T. Ion mobility mass spectrometry of proteins in a modified commercial mass spectrometer. Int. J. Mass Spectrom. 2004, 236, 55–63. 42. Weston, D.J.; Bateman, R.; Wilson, I.D.; Wood, T.R.; Creaser, C.S. Direct analysis of pharmaceutical drug formulations using ion mobility spectrometry/quadrupole-timeof-flight mass spectrometry combined with desorption electrospray ionization. Anal. Chem. 2005, 77, 7572–7580. 43. Kaur-Atwal, G.; Weston, D.J.; Green, P.S.; Crosland, S.; Bonner, P.L.; Creaser, C.S. Analysis of tryptic peptides using desorption electrospray ionization combined with ion mobility spectrometry/mass spectrometry. Rapid Commun. Mass Spectrom. 2007, 21, 1131–1138. 44. Harry, E.L.; Weston, D.J.; Bristow, A.W.; Wilson, I.D.; Creaser, C.S. An approach to enhancing coverage of the urinary metabonome using liquid chromatography-ion mobility-mass spectrometry. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2008, 871, 357–361. 45. Kanu, A.B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill, H.H. Ion mobility-mass spectrometry. J. Mass Spectrom. 2008, 43, 1–22. 46. Creaser, C.S.; Griffiths, J.R.; Bramwell, C.J.; Noreen, S.; Hill, C.A.; Thomas, C.L.P. Ion mobility spectrometry: A review. Part 1. Structural analysis by mobility measurement. Analyst 2004, 129, 984–994.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
233
47. Wyttenbach, T.; Bowers, M.T. Gas-Phase Conformations: The Ion Mobility/Ion Chromatography Method. Modern Mass Spectrometry (Topics in Current Chemistry), Springer, Berlin, 2003. 48. Giles, K.; Pringle, S.D.; Worthington, K.R.; Little, D.; Wildgoose, J.L.; Bateman, R.H. Applications of a travelling wave-based radio-frequency-only stacked ring ion guide. Rapid Commun. Mass Spectrom. 2004, 18, 2401–2414. 49. Thalassinos, K.; Grabenauer, M.; Slade, S.E.; Hilton, G.R.; Bowers, M.T.; Scrivens, J.H. Characterization of phosphorylated peptides using travelling wave-based and drift cell ion mobility mass spectrometry. Anal. Chem. 2009, 81, 248–254. 50. Williams, J.P.; Scrivens, J.H. Coupling desorption electrospray ionization and neutral desorption/extractive electrospray ionization with a travelling-wave based ion mobility mass spectrometer for the analysis of drugs. Rapid Commun. Mass Spectrom. 2008, 22, 187–196. 51. Scarff, C.A.; Thalassinos, K.; Hilton, G.R.; Scrivens, J.H. Travelling wave ion mobility mass spectrometry studies of protein structure: Biological significance and comparison with X-ray crystallography and nuclear magnetic resonance spectroscopy measurements. Rapid Commun. Mass Spectrom. 2008, 22, 3297–3304. 52. Ruotolo, B.T.; Giles, K.; Campuzano, I.; Sandercock, A.M.; Bateman, R.H.; Robinson, C.V. Evidence for macromolecular protein rings in the absence of bulk water. Science 2005, 310, 1658–1661. 53. Wildgoose, J.L.; Giles, K.; Pringle, S.D.; Koeniger, S.L.; Valentine, S.J.; Bateman, R.H.; Clemmer, D.E. A comparison of travelling wave and drift tube ion mobility separations. Proc. 54th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, May 28–June 1, 2006, ThP 64.AQ 54. Ruotolo, B.T.; Benesch, J.L.P.; Sandercock, A.M.; Hyung, S.J.; Robinson, C.V. Ion mobility-mass spectrometry analysis of large protein complexes. Nat. Protocols 2008, 3, 1139–1152. 55. Giles, K.; Wildgoose, J.L.; Langridge, D. Determining ion mobility values using a travelling wave separator. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, WP 043. 56. Shvartsburg, A.A.; Smith, R.D. Fundamentals of travelling wave ion mobility spectrometry. Anal. Chem. 2008, 80, 9689–9699. 57. Robinson, E.W.; Shvartsburg, A.A.; Tang, K.; Smith, R.D. Control of ion distortion in field asymmetric waveform ion mobility spectrometry via variation of dispersion field and gas temperature. Anal. Chem. 2008, 80, 7508–7515. 58. Valentine, S.J.; Counterman, A.E.; Clemmer, D.E. Conformer-dependent proton-transfer reactions of ubiquitin ions, J. Am. Soc. Mass Spectrom. 1997, 8, 954–961. 59. Valentine, S.J.; Anderson, J.G.; Ellington, A.D.; Clemmer, D.E. Disulfide-intact and reduced lysozyme in the gas phase: Conformations and pathways of folding and unfolding. J. Phys. Chem. B 1997, 101, 3891–3900. 60. Shelimov, K.B.; Jarrold, M.F. Conformations, unfolding, and refolding of apomyoglobin in vacuum: An activation barrier for gas phase protein folding. J. Am. Chem. Soc. 1997, 119, 2987–2994. 61. Scarff, C.A.; Patel, V.J.; Thalassinos, K.; Scrivens, J.H. Probing hemoglobin structure by means of travelling wave ion mobility mass spectrometry. J. Am. Soc. Mass Spectrom. 2009, 20, 625–631. 62. Langridge, D.; Giles, K.; Hoyes, J.B. Simulation of ion motion in a travelling wave mobility separator using a hard-sphere collision model. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, WP 064. 63. Valentine, S.J.; Counterman, A.E.; Clemmer, D.E. A database of 660 peptide ion cross sections: Use of intrinsic size parameters for bona fide predictions of cross sections. J. Am. Soc. Mass Spectrom. 1999, 10, 1188–1211.
234
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
64. Smith, D.P.; Giles, K.; Bateman, R.H.; Radford, S.E.; Ashcroft, A.E. Monitoring copopulated conformational states during protein folding events using electrospray ionizationion mobility spectrometry-mass spectrometry. J. Am. Soc. Mass Spectrom. 2007, 18, 2180–2190. 65. Slade, S.E.; Thalassinos, K.; Hilton, G.R.; Pinheiro, T.; Blindauer, C.A.; Bowers, M.T.; Scrivens, J. Travelling wave ion mobility mass spectrometry-based conformational studies of prion protein – effects of metal cation binding and buffer gas. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, MP 457. 66. Pang, E.S.; Loo, R.O.; Yin, S.; Boontheung, P.; Teplow, D.B.; Loo, J.A. Ion mobility mass spectrometry and proton transfer reactions of non-covalent amyloid β-protein oligomers. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, TP 231. 67. Uetrecht, C.; Versluis, C.; Watts, N.R.; Wingfield, P.T.; Steven, A.C.; Heck, A.J.R. Stability and shape of hepatitis B virus capsids in vacuo. Angew. Chem. Int. Ed. 2008, 47, 6247–6251. 68. Morton, V.L.; Stockley, P.G.; Stonehouse, N.J.; Ashcroft, A.E. Insights into virus capsid assembly from non-covalent mass spectrometry. Mass Spectrom. Rev. 2008, 27, 575–595. 69. Lorenzen, K.; Olia, A.S.; Uetrecht, C.; Cingolani, G.; Heck, A.J.R. Determination of stoichiometry and conformational changes in the first step of the P22 tail assembly. J. Mol. Biol. 2008, 379, 385–396. 70. Sobott, F.; Watt, S.J.; Campuzano, I. The use of ion mobility/time-of-flight mass spectrometry for the study of protein conformations. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, ThP 067. 71. Gangloff, M.; Murali, A.; Xiong, J.; Arnot, C.J.; Weber, A.N.; Sandercock, A.M.; Robinson, C.V.; Sarisky, R.; Holzenburg, A.; Kao, C.; Gay, N.J. Structural insight into the mechanism of activation of the Toll receptor by the dimeric ligand Spatzle. J. Biol. Chem. 2008, 283, 14629–14635. 72. Alverdi, V.; Mazon, H.; Versluis, C.; Hemrika, W.; Esposito, G.; van den Heuvel, R.; Scholten, A.; Heck, A.J.R. cGMP-binding prepares PKG for substrate binding by disclosing the C-terminal domain. J. Mol. Biol. 2008, 375, 1380–1393. 73. Shvartsburg, A.A.; Jarrold, M.F. An exact hard-spheres scattering model for the mobilities of polyatomic ions. Chem. Phys. Lett. 1996, 261, 86–91. 74. Mesleh, M.F.; Hunter, J.M.; Shvartsburg, A.A.; Schatz, G.C.; Jarrold, M.F. Structural information from ion mobility measurements: Effects of the long-range potential. J. Phys. Chem. 1996, 100, 16082–16086. 75. Ruotolo, B.T.; Hyung, S.J.; Robinson, P.M.; Giles, K.; Bateman, R.H.; Robinson, C.V. Ion mobility-mass spectrometry reveals long-lived, unfolded intermediates in the dissociation of protein complexes. Angew. Chem. Int. Ed. 2007, 46, 8001–8004. 76. Iacob, R.E.; Murphy, J.P.; Engen, J.R. Ion mobility adds an additional dimension to mass spectrometric analysis of solution-phase hydrogen/deuterium exchange. Rapid Commun. Mass Spectrom. 2008, 22, 2898–2904. 77. Balbeur, D.; Widart, J.; Leyh, B.; Cravello, L.; De Pauw, E. Detection of oligonucleotide gas-phase conformers: H/D exchange and ion mobility as complementary techniques. J. Am. Soc. Mass Spectrom. 2008, 19, 938–946. 78. Takats, Z.; Wiseman, J.M.; Gologan, B.; Cooks, R.G. Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science 2004, 306, 471–473. 79. Van Berkel, G.J.; Pasilis, S.P.; Ovchinnikova, O. Established and emerging atmospheric pressure surface sampling/ionization techniques for mass spectrometry. J. Mass Spectrom. 2008, 43, 1161–1180.
Applications of Traveling Wave Ion Mobility-Mass Spectrometry
235
80. Myung, S.; Wiseman, J.M.; Valentine, S.J.; Takats, Z.; Cooks, R.G.; Clemmer, D.E. Coupling desorption electrospray ionization with ion mobility/mass spectrometry for analysis of protein structure: Evidence for desorption of folded and denatured States. J. Phys. Chem. B 2006, 110, 5045–5051. 81. Chen, H.; Yang, S.; Wortmann, A.; Zenobi, R. Neutral desorption sampling of living objects for rapid analysis by extractive electrospray ionization mass spectrometry. Angew. Chem. Int. Ed. 2007, 46, 7591–7594. 82. Chen, H.; Wortmann, A.; Zenobi, R. Neutral desorption sampling coupled to extractive electrospray ionization mass spectrometry for rapid differentiation of biosamples by metabolomic fingerprinting. J. Mass Spectrom. 2007, 42, 1123–1135. 83. Hilton, G.R.; Jackson, A.T.; Thalassinos, K.; Scrivens, J.H. Structural analysis of synthetic polymer mixtures using ion mobility and tandem mass spectrometry. Anal. Chem. 2008, 80, 9720–9725. 84. McEwen, C.N.; McKay, R.G.; Larsen, B.S. Analysis of solids, liquids, and biological tissues using solids probe introduction at atmospheric pressure on commercial LC/MS instruments. Anal. Chem. 2005, 77, 7826–7831. 85. McEwen, C.; Gutteridge, S. Analysis of the inhibition of the ergosterol pathway in fungi using the atmospheric solids analysis probe (ASAP) method. J. Am. Soc. Mass Spectrom. 2007, 18, 1274–1278. 86. McEwen, C.; Major, H.; Green, M.; Petkovska, V. Analysis of polymers, polymer additives, and petroleum using the ASAP method on Synapt and Orbitrap mass spectrometers. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, ThOB 3:30. 87. Walch, A.; Rauser, S.; Deininger, S.O.; Hofler, H. MALDI imaging mass spectrometry for direct tissue analysis: A new frontier for molecular histology. Histochem. Cell Biol. 2008, 130, 421–434. 88. Wiseman, J.M.; Ifa, D.R.; Venter, A.; Cooks, R.G. Ambient molecular imaging by desorption electrospray ionization mass spectrometry. Nat. Protocols 2008, 3, 517–524. 89. Cornett, D.S.; Frappier, S.L.; Caprioli, R.M. MALDI-FTICR imaging mass spectrometry of drugs and metabolites in tissue. Anal. Chem. 2008, 80, 5648–5653. 90. McLean, J.A.; Ridenour, W.B.; Caprioli, R.M. Profiling and imaging of tissues by imaging ion mobility-mass spectrometry. J. Mass Spectrom. 2007, 42, 1099–1105. 91. Ridenour, W.B.; Snel, M.F.; Claude, E.; Mclean, J.A.; Frappier, S.L.; Caprioli, R.M. Simultaneous imaging of small molecules from thin tissue sections using MALDI travelling wave ion mobility-mass spectrometry. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, MP 139. 92. Clench, M.; Burrell, M.; Trim, P.J.; Earnshaw, C.; Francese, S.; Muharib, T.; Anderson, D.M.G.; Atkinson, S. Strategies for small molecule imaging by MALDI-MSI. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 2008, WOG pm 2:50. 93. Trim, P.J.; Avery, J.L.; McEwen, A.; Snel, M.F.; Claude, E.; Marshall, P.S.; West, A.; Princivalle, A.P.; Clench, M. MALDI-ion mobility separation-MS imaging of Vinblastine and its metabolites in rat tissue. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, MP 130. 94. Djidja, M-C.; Sutton, C.W.; Loadman, P.M.; Scriven, P.; Snel, M.F.; Claude, E.; Clench, M.R. Visualisation and in situ characterisation of proteins in adenocarcinoma tissue sections by direct MALDI-mass spectrometry imaging. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, WP 169. 95. Stauber, J.; Kaletas, B.K.; Wiel, I.M.v.d.; Snel, M.F.; Claude, E.; Heeren, R.M.A. Ion mobility imaging mass spectrometry: A new tool for in situ identification. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, ThOC pm 2:30.
236
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
96. Riba-Garcia, I.; Giles, K.; Bateman, R.H.; Gaskell, S.J. Evidence for structural variants of a- and b-type peptide fragment ions using combined ion mobility/mass spectrometry. J. Am. Soc. Mass Spectrom. 2008, 19, 609–613. 97. Olivova, P.; Chen, W.B.; Chakraborty, A.B.; Gebler, J.C. Determination of N-glycosylation sites and site heterogeneity in a monoclonal antibody by electrospray quadrupole ionmobility time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 2008, 22, 29–40. 98. Harvey, D.J.; Scrivens, J.; Holland, R.; Williams, J.; Wormald1, M.R. Ion-mobility separation coupled with negative ion fragmentation of N-linked carbohydrates. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, MOG 9:10 a.m. 99. Holland, R.; Thalassinos, K.; Scrivens, J. Shape selective studies of cationised gas phase structures of glucose polymers using travelling wave-based ion mobility mass spectrometry. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver CO, June 1–5, 2008, ThP 049. 100. Bagal, D.; Zhang, H.; Schnier, P.D. Gas-phase proton-transfer chemistry coupled with TOF mass spectrometry and ion mobility-MS for the facile analysis of poly(ethylene glycols) and PEGylated polypeptide conjugates. Anal. Chem. 2008, 80, 2408–2418. 101. Robinson, E.W.; Garcia, D.E.; Leib, R.D.; Williams, E.R. Enhanced mixture analysis of poly(ethylene glycol) using high-field asymmetric waveform ion mobility spectrometry combined with Fourier transform ion cyclotron resonance mass spectrometry. Anal. Chem. 2006, 78, 2190–2198. 102. Eckers, C.; Laures, A.M.; Giles, K.; Major, H.; Pringle, S. Evaluating the utility of ion mobility separation in combination with high-pressure liquid chromatography/mass spectrometry to facilitate detection of trace impurities in formulated drug products. Rapid Commun. Mass Spectrom. 2007, 21, 1255–1263. 103. Williams, J.P.; Giles, K.; Green, B.N.; Scrivens, J.H.; Bateman, R.H. Ion mobility augments the utility of mass spectrometry in the identification of human hemoglobin variants. Rapid Commun. Mass Spectrom. 2008, 22, 3179–3186. 104. Jung, H.R.; Liu, W.; Jensen, O.N. Determination of post-translational modifications of Histone H3 using chromatography, ion mobility tandem mass spectrometry and customized bioinformatics tools. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, MP 394. 105. Souza, G.H.M.F.; Uria, D.; Eberlin, M.N. Ion mobility of serine octamer clusters using tri-wave high-definition mass spectrometry. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, ThP 042. 106. D’Agostino, P.; Chenier, C.; Baker, A. Ion mobility separation of organophosphates using a quadrupole time-of-flight mass spectrometer. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008, MP 175.
Part III Ion Spectroscopy
Spectroscopy of 9 The Ions Stored in Trapping Mass Spectrometers Matthew W. Forbes, Francis O. Talbot, and Rebecca A. Jockusch Contents 9.1 9.2 9.3 9.4
Introduction...................................................................................................240 Some Practical Considerations...................................................................... 242 Infrared Multiphoton Dissociation (Irmpd) Using Co2 Lasers.................. 245 Infrared (IR) Action Spectroscopy of Biological Molecules.........................246 9.4.1 Infrared Multiphoton Dissociation (IRMPD) Mechanism and Interpretation of IRMPD Action Spectra.......................................... 249 9.4.2 A Final Caution................................................................................. 251 9.5 Electronic Action Spectroscopy.................................................................... 252 9.5.1 Photodissociation (PD)...................................................................... 252 9.5.2 Electron Photodetachment................................................................. 254 9.6 Fluorescence Spectroscopy............................................................................ 254 9.7 Case Study: Modification of A Commercial QIT for Optical Spectroscopic Experiments........................................................................... 257 9.7.1 Apparatus Overview and Design Considerations.............................. 257 9.7.2 Tuning the RF Circuit Incorporating the Ring Electrode................. 259 9.7.3 What Size Holes to Drill?..................................................................260 9.7.4 Ion Behavior in Real Traps................................................................ 261 9.7.5 Ion Trajectory Calculations...............................................................264 9.7.5.1 SIMION Models.................................................................264 9.7.5.2 Single Ion Trajectories in Vacuum......................................266 9.7.5.3 Mass-Selective Instability Scan with Collisions and Axial Modulation................................................................ 272 9.7.5.4 Spatial Distribution of the Ion Cloud.................................. 278 9.7.6 Characterization of a QIT Modified for Optical Spectroscopic Experiments............................................................... 281 9.8 Summary and Outlook.................................................................................. 282 Acknowledgments...................................................................................................284 References...............................................................................................................284
239
240
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
9.1 INTRODUCTION Optical spectroscopy of mass-selected ions stored within a trapping mass spectrometer provides access to new dimensions of information. The activation of ions using light, or photo-excitation, has been used in conjunction with trapping mass spectrometers to effect photodissociation (PD, also referred to as photo-induced dissociation or PID) of simple molecular ions since the 1960s [1]. More recently, photo-excitation of protonated biomolecules using infrared (IR) or ultraviolet (UV) fixed-wavelength lasers has been used for peptide and protein sequencing [2–4], determination of dissociation energetics [5–7], and to access distinctive high-energy dissociation pathways [3]. An optical spectrum can be constructed by stepping or scanning the excitation source through a range of wavelengths while monitoring some signal, such as ion intensities, as a function of excitation wavelength. When the extent of PD is plotted as a function of excitation wavelength, an optical spectrum is generated indirectly; this process is termed ‘action’ or ‘consequence’ spectroscopy. Action spectroscopy has been used to generate vibrational (IR) and vibronic (UV-visible) spectra of mass spectrometric precursor and product ions. A more direct form of optical spectroscopy is fluorescence spectroscopy, in which radiative emission from activated ions is monitored using a photon detector. Fluorescence is highly sensitive to a chromophore’s local environment, making it an excellent probe of conformation. The detection of fluorescence from organic molecular ions in trapping mass spectrometers has been demonstrated recently in several laboratories [8–16]. In volume II of this series, published in 1995, Chapter 5 was devoted to PD studies in ion traps which focused on small (by today’s mass spectrometry (MS) standards) ions [17]. In this chapter, we discuss practical aspects of experimental design and highlight recent applications from our research using PD, electron photodetachment, and fluorescence of trapped, mass-selected ions to characterize gas-phase properties of larger organic ions. Absorption spectroscopy is a primary tool for molecular characterization. Both electronic (UV/visible) and rovibrational (IR) spectroscopy provide information complementary to that obtained from mass spectrometric experiments. Unfortunately, the densities of ions in mass spectrometers are much lower than those needed for standard, or direct, absorption experiments. For example, 1000 ions confined in a sphere of 1 mm radius corresponds to an ion density of 2.4 × 105 ions cm–3, which is equivalent to the density of ions in a 0.4 fM liquid solution. The space charge limit in a typical Paul-type 3-D Quadrupole ion trap (QIT) is ca 1500 times higher than this; however, when expressed in terms of the equivalent density of ions in solution, the corresponding concentration is less than 1 pM, still far below the 10 –4 –10 –6 M typical limit of detection for conventional UV/visible absorption spectrophotometers. Direct absorption experiments have a high limit of detection because the signal monitored is a decrease in photon fluence, which is generally a small change in signal on a large background. In principle, a multi-pass arrangement such as cavity ring-down, may be used in conjunction with trapping mass spectrometers to provide a long path length for absorption and increased sensitivity [18]. Cavity ring-down spectroscopy (CRDS) is a technique that employs an optical cavity for measuring the absorbance of light through a substance that is normally in the gas phase [19,20]. A brief pulse
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
241
of laser light is directed into the cavity, which consists of a pair of mirrors facing each other, and is reflected back and forth between the mirrors. The small amount (ca 0.1%) of laser light that leaks out of the cavity on each reflection is measured by a detector. When an absorbing substance is placed in the cavity, the light intensity impinging on each mirror in turn decreases more rapidly, hence the light intensity that leaks out of the cavity on each reflection decreases also. A CRDS arrangement measures the time taken for the leaked light intensity to fall to a fixed percentage of the original leaked light intensity; this ‘ring-down’ time can be used to calculate the concentration of the absorbing substance in the cavity. Unfortunately, such cavities are difficult to implement in practice and we are not aware of any successful implementation of this technique with a trapping mass spectrometer. One way to generate an absorption-based spectrum whilst circumventing the problem of low ion density is to use action (or consequence) spectroscopy. In this approach, the change in ion intensity as a result of absorption of photons (photodissociation, PD), Equation 9.1, is monitored as a function of the excitation irradiation wavelength. In this way, the mass spectrometer is used to monitor the result of photon absorption, without the need to detect light directly. Several methods exist for calculating a dissociation yield; either the appearance of product ions, disappearance of precursor ions, or both may be monitored. In our work, unless otherwise noted, we take the dissociation yield to be the sum of fragment ion intensities divided by the total ion intensity. The normalization to total ion signal should compensate for small fluctuations in the isolated ion population, but is not appropriate for cases in which the products are neutral and/or below the low mass cut-off of a radiofrequency (RF) trapping device. Another form of action spectroscopy is electron photodetachment, Equation 9.2. Here, the result of photon absorption is the loss of an electron and consequent increase in charge state of the molecular ion. Although electrons can be detached, in principle, from both positive and negative multiply-charged ions, the barrier for detachment is lower in negative ions due to Coulomb repulsion and it is with anions that this technique has been used primarily. Note that the large ion species, such as protonated or deprotonated peptides and DNA, which are the focus of many current investigations, are usually even-electron ions formed by electrospray ionization (ESI) or matrix-assisted laser desorption ionization (MALDI). The PD products are generally, though not exclusively, also even-electron ions. However, photodetachment of an electron converts an even-electron ion to an odd-electron (radical) species that may undergo subsequent fragmentation processes with relative ease. Action spectroscopy can be a highly-sensitive technique because it is based on measuring the change in ion intensity, which is straightforward to measure in a mass spectrometer. However, it is important to note that the use of action spectroscopy, rather than direct spectroscopy (in which light is detected directly) is subject to several potential difficulties, primarily associated with intensity variations, of which the user must be aware. These will be discussed in greater detail later in Section 9.4.
ν AB + h → A+ + B
(9.1)
242
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V ν M n− h → M ( n−1)− + e −
(9.2)
Fluorescence spectroscopy is an alternative approach to spectroscopic characterization of trapped ions in conjunction with trapping MS. An electronically-excited ion, Mn + *, is created by absorption of a UV or visible photon. Fluorescence emission, a radiative transition between the excited electronic state and ground state of the same spin state, is one pathway for de-excitation back to the ground electronic state (Equation 9.3). Other de-excitation pathways, which compete with fluorescence, are available, including internal conversion and fragmentation (PD).
ν1 M n+ h → M n+ * → M n+ + hν2
(9.3)
A significant advantage of fluorescence is the sensitivity of the technique, which makes it compatible with the low ion densities found in trapping mass spectrometers. Also, in contrast to PD, which requires an increasing amount of energy with molecular size (and thus an increasingly powerful laser), the energy needed to excite fluorescence does not scale with the size of the molecules. Disadvantages of combining fluorescence with trapping MS include: (i) the need to implement methods for light collection and detection; and (ii) limited applicability as not all ions fluoresce significantly. These factors will be discussed in greater detail later in this chapter.
9.2 SOME PRACTICAL CONSIDERATIONS Modification of trapping mass spectrometers to interface with lasers is, in principle, straightforward. The laser beam must pass into the vacuum chamber of the mass spectrometer and intersect with the ion cloud. Some possible arrangements for the laser beam in a Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer, and in a 3-D QIT are shown in Figure 9.1. In linear ion traps and the ion storage cells of FT-ICR instruments, there is often a pre-existing line-of-sight available for the laser beam path (Figure 9.1a). A multi-pass arrangement has been used also in ICR cells (Figure 9.1b). In Paul-type QITs, the most usual approach is to make additional holes in the ring electrode (Figure 9.1c) or to use, or to enlarge, pre-existing holes in the end-cap electrodes (not shown). In Paul-type traps in which the electrodes share an asymptote, an alternative is to send the excitation beam along the asymptote, avoiding modification of the hyperbolic surfaces of the electrodes (Figure 9.1d). Modification of the electrodes changes the electric trapping potential and, therefore, may effect adversely operation of the instrument as a mass spectrometer. However, in practice, the operation of QITs is relatively insensitive to modifications to the ring electrode as mass resolution depends primarily on axial rather than radial displacement of the ions. Another key consideration in the design of trapping MS combined with optical spectroscopy experiments is the choice of light source. Here, we limit the discussion to that convenient modern light-source, the laser, which emits typically light that is highly directional and monochromatic (except for ultrashort pulses that can exhibit spectral widths of several dozens of nanometers). Laser selection criteria include
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
243
(b)
(a)
(c)
(d)
Ring electrode Ring electrode End-cap electrode
FIGURE 9.1 Possible geometries for trapped ion photo-excitation: (a,b) open-cylindrical cell of an ion cyclotron resonance mass spectrometer; (c,d) 3-D quadrupole ion traps shown in cross-section. (a) The laser is directed on the central axis of the cell (or quadrupole rod set of a linear trap) in line with the ion cloud. (b) A multi-pass arrangement where the laser intersects the ion cloud several times as it is reflected between the inner surfaces of the detection plates. (c) Two holes bored in the ring electrode of a quadrupole ion trap allow the beam to pass through the center of the trap assembly intersecting the ion cloud along a radial path. (d) The ion cloud is accessed along the common asymptote of the ring and end-cap electrodes without any modification to their hyperbolic surfaces.
desired wavelength (or wavelength range), power, continuous wave (cw) vs pulsed sources, and pulse repetition rate. For PD-based experiments, sufficient laser irradiation power is necessary to fragment the ion species. Dissociation energies range from less than 0.1 eV (equivalent to a single IR photon of 800 cm–1) for some van der Waals complexes up to 5 eV (equivalent to a single 250 nm UV photon) or more for a covalent bond. However, the ion internal energy must be raised significantly over threshold for dissociation to occur on the experimental time-scale for large ions with efficient energy flow among vibrational modes; this energy excess is described as the kinetic shift [21]. A wide variety of lasers have been used in combination with trapping mass spectrometers. Two lasers in particular are incorporated already into many commercial mass spectrometers: the CO2 (λ = 10.6 µm or 0.12 eV photon−1) and nitrogen (λ = 337 nm or 3.68 eV photon–1) lasers. The former is used for performing infrared multiple photon dissociation (IRMPD) experiments (mostly in ICR instruments, see Section 9.3) while the latter is used commonly in MALDI ionization sources. Other commonly-used lasers are Q-switched Nd:YAG lasers (fundamental: λ = 1064 nm
244
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
or 1.17 eV photon–1; 2nd harmonic λ = 532 nm or 2.33 eV photon–1), and vacuum UV fluorine lasers (λ = 157 nm or 7.87 eV photon–1). The wavelengths of these lasers differ greatly; consequently, the optical components used in the experiments must be chosen appropriately. The CO2 laser, in particular, stands out as its wavelength is very far from the more conventional UV and visible wavelengths of the other lasers. The optical materials that offer the best transmission for the CO2 wavelength (ZnSe, ZnS) do not transmit visible and UV wavelengths. Conversely, most materials used to transmit UV-VIS wavelengths are opaque at 10.6 µm. To measure excitation spectra, a tunable light source is needed. Much recent work has used tunable IR lasers to access vibrational transitions and tunable UV/visible lasers to access electronic transitions. In such experiments, where the laser wavelength is scanned over a certain range, it is especially important to make certain that the whole optical pathway up to the ions has a flat transmission curve over the scanned wavelengths. Tunable lasers will be discussed in more detail in Sections 9.4 and 9.5. The photon flux required, and hence the choice of excitation source, depends on whether the experiment entails the absorption of a single photon or of multiple photons. Many dissociation experiments rely on the absorption of tens or hundreds of photons (or even more for larger molecules) for fragmentation to occur. For these experiments, it is essential to use a sufficiently-high photon flux to populate adequately the dissociative mode in order to effect dissociation, and to overwhelm the competing radiative and collisional de-excitation processes. For example, the power of cw CO2 lasers used for IRMPD in FT-ICR and QIT mass spectrometers is generally in the range 25–50 W while the pulsed free-electron lasers (FELs) have average powers of only ca 500 mW. However in FELs, dissociation is activated by a single 5 µs or 8 µs-long ‘macropulse’ (see Section 9.4) during which the average power is effectively kilowatts and the peak power in the megawatt range. The interpretation of information from band intensities is complicated by the dependence of fragmentation, and hence of band intensity, on the absorption of multiple photons (see Section 9.4). On the other hand, low power lasers are suitable for spectroscopic experiments that rely on the absorption of a single photon. For example, laser powers of less than 10 mW are often sufficient for fluorescence studies in which the absorption of a single UV or visible photon populates a fluorescent state. Absorption of a single UV photon is believed also to effect electron photodetachment [22,23] and PD [3,24] in some cases. The effect of a single photon may be monitored by measuring the increase in PD rate, due to absorption of an additional resonant photon, of a population that is already undergoing dissociation. For example, Bush and coworkers monitored the wavelength-dependent increase in dissociation of arginine-metal ion complexes which were stored in a heated FT-ICR mass spectrometer and irradiated with a low-power tunable IR laser [25]. Another single-photon scheme, which is applicable to cold molecules and clusters, is the ‘messenger’ technique in which the analyte is tagged with a weakly-bound messenger atom such as argon that will readily dissociate from the analyte and that (ideally) does not affect the structure of the analyte [26,27]. Importantly, interpretation of band intensities is usually more straightforward in single-photon activated processes.
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
245
Gas-phase spectroscopy of neutral molecules is usually carried out with a cold beam, such as that produced in a supersonic jet [28], because low temperatures increase dramatically spectroscopic resolution. In a supersonic jet, translational and rotational temperatures of less than 10 K can be achieved and vibrational temperatures are generally less than 100 K. Useful cooling techniques to lower the internal energy of trapped molecular ions include cooling by collisions with cold buffer gas and by removal of energy from large clusters via evaporation [26,27,29,30]. In order to avoid collisional heating in instruments with trapping gas and to avoid radiative heating of ions stored for long times in Penning traps, a cooled enclosure is necessary. While several laboratories have reported using cooled FT-ICRs [31–33], the cooling of ions in Paul-type traps is more difficult due to RF heating. Although beyond the scope of this chapter, the work of the Gerlich group in developing a cooled 22-pole trap is especially notable [34]. The use of a higher-order multi-pole for ion trapping results in reduced RF heating; however, the ion cloud in such traps is more diffuse than in a quadrupolar trap, which is a disadvantage for some spectroscopic applications. The Rizzo group has used such a cooled 22-pole trap to measure well-resolved vibronic and IR spectra of cold ions as large as a 12-residue peptide [35]. Very recently, a cryogenically-cooled 3-D Paul trap that can be cooled to 10 K and is coupled to an electrospray ion source has been described [36].
9.3 Infrared Multiphoton Dissociation (IRMPD) USING CO2 LASERS IR multiple photon dissociation wrought by CO2 lasers has been used primarily as a convenient activation method for ion structural characterization in FT-ICR mass spectrometers. CO2 lasers have the advantage of being compact, rugged, reliable, relatively inexpensive, and capable of producing the high powers needed for IRMPD. The 25 W or 50 W lasers are used typically in conjunction with trapping mass spectrometers. CO2 lasers produce cw IR light of 10.6 µm (943 cm−1), which is in the region of the C–C stretch. Most biological polymers appear to absorb this frequency, even when computed (harmonic linear) absorption spectra do not show modes closely corresponding to this energy. The experimental fact that absorption occurs has been explained in terms of a quasi-continuum of vibrational states, in which vibrational energy levels are broadened due to a high density of states facilitating intramolecular vibrational redistribution (IVR), of energy, resulting in ‘semi-resonant’ absorption processes [37,38]. It is also worth keeping in mind that non-resonant (or semi-resonant) absorptions are facilitated by the extremely high photon flux from CO2 lasers (1021 photons s−1 at 25 W). The absorption of tens to hundreds (or even more for large biopolymers) of IR photons is necessary to produce dissociation due to the low energy of IR photons (0.12 eV photon−1 for CO2 lasers). Furthermore, ions need additional or excess internal energy significantly above threshold in order to dissociate at a reasonable rate (the ‘kinetic shift’), even in the relatively long time-scales accessible in trapping mass spectrometer experiments. For example, given a typical peptide bond dissociation energy of ca 1.5 eV, 13 CO2-laser photons would be sufficient for dissociation if all the quanta of energy remained in the dissociating bond. However, assuming complete IVR for a 1 kDa peptide that has a dissociation energy of 1.5 eV and a
246
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
loose transition state, the average internal energy of the ion population must be about 4.1 eV (or 34 photons) to achieve a dissociation rate constant of 0.07 s−1, while for a peptide which is four-times longer, ca 16.5 eV of internal energy (138 photons) are needed [39]. Of course, for longer time-scale experiments in an FT-ICR, de- excitation via radiative emission must be considered, while in an ion trap de-excitation through collisions must be overcome also. Thus, energy significantly in excess of the threshold is necessary to photo-fragment large biomolecules. Laser activation with a cw CO2 laser is a (very) slow heating method that produces fragments from low-energy dissociation pathways similar to those obtained by blackbody infrared radiative dissociation (BIRD) or sustained off-resonance irradiation collisionally-induced dissociation (SORI-CID) in an FT-ICR or CID in an ion trap [40]. CO2 laser activation is the slow heating method of choice in FT-ICR instruments, many of which are now sold with CO2-laser based IRMPD modules. Such a module is more convenient to use (though also more expensive to implement) than BIRD and SORI-CID, which require long times to heat the vacuum system (BIRD) and the addition of collision gas to the ICR cell, resulting in pump-down delay periods and/or decreased resolution (SORI-CID). The use of IRMPD is less popular in ion traps, which are well-suited for CID due to the presence of neutral trapping gas and ease of applying appropriate excitation waveforms, because collisional cooling in the ion trap can out-compete heating rates. Nevertheless, several laboratories have used CO2 laser-activation to photodissociate ions stored in radio frequency traps [41–44]. It should be noted that IRMPD can result in significant second (or higher) generation product ions as both parent ion and fragments are irradiated in contrast to CID, which is usually restricted to activation of a single mass-to-charge ratio value through resonance at the unique ion secular frequency.
9.4 Infrared (IR) ACTION SPECTROSCOPY OF BIOLOGICAL MOLECULES An exciting application of IR activation that has been exploited over the past decade is the action spectroscopy of ions derived from large molecules. A vibrational action spectrum is constructed by determining experimentally the PD yield as a function of IR activation wavelength. This procedure has been used in conjunction with extensive computational work to determine structure and conformation of many biomolecules including cationized amino acids [25,45–53], derivatized amino acids [54–57], proton-bound amino acid complexes [58,59], small peptides [60–64], DNA bases [65], and the dissociation products from CID [61,66] and electron capture dissociation (ECD) [67] of biomolecules. It has been applied also to larger systems including the protein cytochrome c [68], though the conformational conclusions for these larger systems are less clear. The majority of this work has been done using powerful, scannable free electron lasers (FELs) at facilities in both Nieuwegein (The Netherlands) [38,69] and Orsay (France) [70]. The FELs are coupled with both FT-ICR mass spectrometers and 3-D Paul-type ion traps. Because FELs are large instruments, occupying a full room, and expensive to build and to maintain, they are located at dedicated facilities. FELIX, the FEL in The Netherlands, is tunable over a broad range (40–4000 cm–1), and produces typically 5 µs-long macropulses
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
247
of ca 60 mJ at a repetition rate of 5 Hz. CLIO, the FEL in France, has similar c haracteristics: a tuning range of 110–3300 cm–1, with 8 µs-long macropulses of ca 25 mJ at a repetition rate of 25 Hz. FELs naturally have a fairly broad bandwidth at higher powers, which is set generally to ca 0.8% of the central frequency; this fraction translates to a bandwidth of 12 cm−1 at 1500 cm−1, thus limiting the spectral resolution. Alternatively, several laboratories have interfaced recently bench-top tunable IR lasers based on optical parametric oscillator (OPO) technology with trapping mass spectrometers [25,59,71,72]. These laser systems have reasonable pulse energies (5–20 mJ pulse−1) in the hydride stretching region (2800–4000 cm−1), narrower bandwidths than FELs (ca 3 cm−1), and are run usually at 10 Hz with ca 5 ns-long pulses. Very recently, non-linear crystals to access the mid-IR (fingerprint region between 1000–2000 cm−1) have been developed. However, pulse energy in this range remains very low, <1 mJ, in commercially-available laser systems, making them unsuitable for use in the majority of IRMPD studies. Several studies have used IR action spectroscopy to examine the stabilization of zwitterions in the gas phase [25,45–52,59,71]. Amino acids, peptides, and proteins exist as zwitterions in aqueous solution. While the charge-separated zwitterionic form is favored in solution, the non-zwitterionic (also called canonical) form of amino acids is favored in the gas phase. However, the zwitterionic forms of gasphase amino acids can be stabilized preferentially by interactions with appropriate metal ions. Our research group has performed IRMPD experiments to determine the conformation of the most basic amino acid, arginine, when it is complexed with several cations [47]. The vibrational signatures of zwitterions and non-zwitterions are readily distinguishable in the mid IR because the non-zwitterions possess a carbonyl stretch above 1700 cm−1 (an otherwise bare area of the vibrational spectrum), while in the zwitterionic form, this mode is replaced by symmetric and antisymmetric stretching vibration of the negatively-charged carboxylate group, which appear at lower frequencies. In addition, the N-H bending frequencies of the zwitterionic and canonical forms of arginine are distinct, as the zwitterionic form features a hydrogen bond, between the ε nitrogen of the side chain and the N-terminal nitrogen, which is not present in the canonical form. The hydrogen-bound N-H bend appears at slightly higher frequencies (ca 1690 cm−1) than other N-H bends (ca 1660 cm−1). Thus, the IR spectrum of arginine in its canonical form features a characteristic band corresponding to the carbonyl stretch while the zwitterionic form shows a characteristic band corresponding to an H-bonded N-H bend. Figure 9.2 shows IRMPD action spectra of arginine complexed with a series of cations. The presence of an intense band at ca 1730–1760 cm−1 in the spectra for lithium-bound (Figure 9.2a) and silver-bound (Figure 9.2d) arginine, corresponds to the carbonyl stretch of a carboxylic acid functional group; that is, this band indicates that the arginine is predominantly non-zwitterionic in these complexes. The presence of an intense band at ca 1690 cm−1 corresponds to the presence of an H-bonded N–H bend in the sodium-bound and potassium-bound complexes, in Figure 9.2b and c, respectively, indicating the presence of the zwitterionic form of arginine in these complexes. The existence of a weak band at 1745 cm−1 in the [Arg + Na]+ spectrum (Figure 9.2b) suggests that a small amount of arginine may be present in its canonical form. The results from the IRMPD action spectra agree with minimum energy
248
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V H-bonded imine Nε–H bend
Carbonyl stretch of carboxylic acid
(a) [Arg+Li]+ Charge solvated
IRMPD yield
(b) [Arg+Na]+
Salt bridge (c) [Arg+K]+ Salt bridge (d) [Arg+Ag]+
0 dB 3 dB Charge solvated
800
900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 Wavenumber (cm–1)
FIGURE 9.2 Infrared multiple photon action spectra of cationized arginine complexes easured using the free electron laser FELIX shown together with minimum energy geometries m optimized at the B3LYP/6-31 + G(d,p) level of theory. (a) [Arg + Li] + ; (b) [Arg + Na] + ; (c) [Arg + K] + ; and (d) [Arg + Ag] + . For [Arg + Ag] + , the action spectrum was measured both at full power and with 3 dB attenuation for the N-H bend. The absorptions due to the H-bonded imine Nε – H bend (at ca 1690 cm–1) and the carbonyl stretch of the carboxylic acid functional group (ca 1730–1760 cm–1) are indicated in the figure. (Adapted from Forbes, M.W.; Bush, M.F.; Polfer, N.C.; Oomens, J.; Dunbar, R.C.; Williams, E.R.; Jockusch, R.A., J. Phys. Chem. A. 2007, 111, 11759–11770.)
conformations computed at the MP2/6-311 + + G(2d,2p)//B3LYP/6-31 + G(d,p) level of theory. Figure 9.3 compares the measured IRMPD action spectrum of [Arg + Li]+ with linear absorption spectra computed using density functional calculations at the B3LYP/6-31 + G(d,p) level of theory and scaled by 0.98. The global minimum structure identified computationally features arginine in its canonical (non-zwitterionic) form and is a reasonable match with the measured spectrum (compare Figure 9.3a and b). Notably, the position of the band at 1730 cm−1, corresponding to the carbonyl stretch of the carboxylic acid, is reproduced well by the calculations. The lowest energy structure of [Arg + Li]+ that contains zwitterionic arginine is computed to be 15 kJ mol−1 higher in energy than the global minimum, and its computed spectrum (Figure 9.3c) does not
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
249
(a) [Arg+Li]+ IRMPD action spectrum
(b) CS_A ∆E0 = 0 kJ mol–1
(c) SB_D ∆E0 = +15 kJ mol–1
800
900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 Wavenumber (cm–1)
FIGURE 9.3 Comparison of an observed infrared multiple photon action spectrum of lithiated arginine with calculated linear absorption spectra for two structures. (a) Infrared multiple photon action spectrum of lithiated arginine measured using the free electron laser FELIX; (b) the calculated linear absorption spectrum for the charge-solvated (non-zwitterionic) conformer that, computationally, has the most stable structure; and (c) the calculated linear absorption spectrum for the salt bridge (zwitterionic) conformer. Structures and frequencies are calculated at the B3LYP/6-31G(d,p) level of theory. Frequencies are scaled uniformly by 0.98 and are shown as both sticks and with a 20 cm–1 Gaussian convolution to approximate the bandwidth of FELIX. Relative energies listed are computed at the MP2/6-311 + + G(2d,2p)//B3LYP/6-31 + G(d,p) level of theory. The three gray vertical bands identify the wavenumber regions of particular interest. (Reproduced from Forbes, M.W; Bush, M.F.; Polfer, N.C.; Oomens, J.; Dunbar, R.C.; Williams, E.R.; Jockusch, R.A.; J. Phys. Chem. A. 2007, 111, 11759–11770. With permission.)
match the measured IRMPD spectrum (Figure 9.3a). The computed spectrum for the zwitterionic complex is missing a band at 1730 cm−1 and has a band at ca 1690 cm−1 arising from a hydrogen-bonded Nε –H bend that is not present in the experimental spectrum. Thus, by comparing the IRMPD action spectrum and computed absorption spectra, assignments to specific structures or to structural families can be made. It is also worth noting that computed intensities do not match measured intensities as well as any researcher would like. In particular, the measured intensities for lower energy modes, those below 1300 cm−1, are not well reproduced by the calculations. A number of reasons for this discrepancy are discussed in Sections 9.4.1 and 9.4.2.
9.4.1 Infrared Multiphoton Dissociation (IRMPD) Mechanism and Interpretation of IRMPD Action Spectra The current picture of the IRMPD mechanism invokes sequential semi-resonant absorption processes made possible by IVR of energy [38]. The first essential
250
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
step is the absorption of an IR photon. For absorption to occur, the energy of the photon must coincide with or be close to that of a vibrational mode of the molecule (resonant or ‘semi-resonant’ absorption). When this is not the case, no absorption occurs and there is no path to dissociation. ‘Semi-resonant’ absorption processes occur (although with a lower probability than true resonant absorption) more easily for large molecules than for small molecules due to the higher density of states in larger molecules. The internal energy of the molecule is increased by photon absorption. However, the energy does not stay within the resonant mode into which it was deposited; rather, IVR (typically < 1 ns timescale) ensures that the energy is distributed into other modes. In other words, the molecule acts as an energy sink, allowing energy to flow out of the resonant vibrational mode into other vibrational motions. The resonant vibration is then free to absorb another photon of the same energy, which is dissipated through IVR to the molecular energy sink, and so on. After the (semi) resonant absorption of several photons, the molecule’s internal energy is increased to the point that the high density of states facilitates the further absorption of photons. Eventually, the internal energy of the molecule is raised enough over the threshold dissociation energy that dissociation occurs. Note that a distinction has been made between this incoherent multiple photon process, which relies on sequential photon absorption followed by dissipation through IVR, and coherent IR multiphoton dissociation in which the vibrational ladder of a single mode is ascended without significant coupling between modes and energy flow via IVR. True multiphoton dissociation (ladder climbing) is impeded by anharmonicity of vibrational modes, resulting in decreased energy level spacing as the vibrational ladder is ascended, thus, causing the available transition energies to move out of resonance with the excitation irradiation. As mentioned in Section 9.2, there is an important distinction between action spectroscopies based on the absorption of multiple photons, as is possible with FEL excitation, and those based on single photon absorption, such as messenger spectroscopy [26,27], laser-induced reaction spectroscopy [73] or indeed direct absorption experiments. When the signal (in this case, the depletion of the parent ion and creation of product ions) depends on the absorption of multiple photons, the signal intensity may not be compared directly to computed linear (single-photon) absorption spectra, as is the normal fashion. There are several reasons why intensities may differ. Thus, band position (frequency) is considered usually to be more reliable than band intensity when comparing computed and experimentally-measured spectra. Band intensities and, to a lesser extent, positions measured using IRMPD action spectroscopy may differ from computed intensities and positions due to one or more of the following factors: anharmonicity, restricted energy flow, dynamics of the vibrationally-excited state, and inadequacies in the computation methods used. For example, one consequence of anharmonicity combined with semi-resonant absorption is that IRMPD bands may appear somewhat broadened to the red, or lower energy side of the band [45]. Anharmonicity combined with restricted IVR, that is, less mode coupling than usual, can result in an ‘anharmonic bottleneck’ in which further photon absorption (and hence dissociation) is hindered. This ‘anharmonic bottleneck’ arises because the lack of efficient IVR traps the absorbed energy in the initial mode and, due to anharmonicity, the molecule cannot be excited further, leading to low
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
251
dissociation despite the fact that the light is resonant with a vibrational mode. This effect is believed to be responsible for the lower than expected intensity for some carbonyl stretch modes [55]. Using a decreased laser power may alleviate minor saturation problems by giving the molecule time to undergo IVR but, in extreme cases with very little coupling between modes, bands are believed to disappear entirely from the action spectrum. This process has been referred to evocatively as an ‘IVR killer mode’. It should be noted also that for many modes there is a non-linear power dependence, such that below a threshold laser intensity, no dissociation is observed even though absorption takes place. Another way in which multiple photon dissociation experiments differ from single photon experiments is that changes to the system that occur during excitation may affect the results. For example, in the FEL-IRMPD experiments, PD is produced within a macropulse of duration 5 or 8 µs. During this time, as the ions’ internal energies increase, dynamics may occur which affect the subsequent absorption of photons. For example, several conformations may be sampled in a 5 µs-time-frame when the barrier for interconversion is low, resulting in absorption cross-sections which do not match that of the ground state configuration. Computational approaches incorporating dynamics can account for these effects [74,75], but they are more expensive computationally and less straightforward to implement than the frozen ensemble of low-energy structures which are modeled usually. Other deficiencies also exist in current computational methods. The most common computational approach is to use conformation searching or molecular dynamics in conjunction with molecular mechanics force fields to identify an ensemble of candidate minimum energy structures. Each candidate conformation is geometry optimized subsequently and harmonic frequencies are calculated using electronic structure theory methods such as the popular hybrid density functional theory B3LYP. All relevant structures may not be identified in the initial computational steps either due to an inadequate search of the potential energy surface, which becomes more difficult as the molecular size increases, or due to deficiencies in the molecular mechanics methods used. An additional concern is that computed harmonic frequencies are overestimates of their true values. The extent of overestimation depends on the method and basis set used as well as on the particular vibration being examined. The usual approach is to use an empirical scaling factor to compensate for some of these effects. Recommended scaling factors for different method/basis set combinations have been tabulated by National Institute of Standards and Technology (NIST) [76]. As a general rule, lower energy modes, coupled motions, and stronglyinteracting vibrations (for example, those involved in a hydrogen bond) are more anharmonic and, therefore, less well reproduced in the harmonic normal mode analysis. This enhanced anharmonicity of lower energy modes may explain why many FEL-IRMPD spectra measured below 1000 cm−1 do not match computed harmonic absorption spectra, while for those above 1300 cm−1 the match is often excellent.
9.4.2 A Final Caution Fragments formed using IRMPD may reflect conformation and dynamics of excited vibrational states rather than ground state characteristics. Computational simulations
252
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
of the IRMPD process have shown unusual internal energy distributions [38]. One consequence of this phenomenon is that branching ratios measured using IRMPD may reflect more strongly characteristics of the excited state rather than of ground state populations, and caution should be used when interpreting IRMPD fragmentation pathways.
9.5 ELECTRONIC ACTION SPECTROSCOPY Visible and UV light sources, which excite electronic transitions, can be used also for PD spectroscopy. By scanning the frequencies of the radiation emitted from the UV/vis light source and measuring PD or electron photodetachment as a function of excitation wavelength, an electronic action spectrum can be constructed in the same way as a vibrational action spectrum is constructed using an IR source. There is a variety of tunable UV and visible lasers with sufficient power for action spectroscopy. These include dye laser systems, OPO systems, Titanium:Sapphire (Ti:Saph) lasers, and supercontinuum lasers. Dye lasers pumped by Nd:YAG or excimer lasers provide narrow linewidths (< 0.5 cm−1) and are tunable over a wide range of wavelengths (200 nm−1 µm) by using a variety of laser dyes. The repetition rate is defined by the repetition rate of the pump laser, and this may be a limiting factor in the experiment as the most popular nanosecond pump lasers have a repetition rate of 20 Hz or lower. Also, use of laser dyes is somewhat inconvenient as these must be replaced frequently because of their limited lifetimes. Tunable lasers based on OPOs (pumped usually by Nd:YAGs) are becoming more popular. Modern OPOs are can feature relatively narrow linewidths (1.5 cm−1), broad tunability (200 nm-IR) and some do not have the tuning range gap, known as the ‘degeneracy gap’ which was present in older models. Tunability can be achieved also by performing wavelength selection on the output of a supercontinuum laser source that generates broadband light between 400–2500 nm. A fourth option is to use a mode-locked Ti:Saph laser in combination with second and/or thirdharmonic generation. Mode-locked Ti:Saph systems are ‘ultrafast’ lasers, with pulse lengths of picoseconds or 100’s of femtoseconds and high repetition rates (80 MHz). Commercial supercontinuum lasers have picoseconds pulse widths with 20 MHz repetition rates, making Ti:Saph and supercontinuum-based sources wellsuited for fluorescence lifetime measurements.
9.5.1 Photodissociation The energy of UV and visible photons is comparable with covalent bond dissociation energies; thus, in principle absorption of a single UV/vis photon can cause PD if the excitation is to a short-lived dissociative state. Indeed, single-photon dissociation appears to occur in some cases when the photon energy is sufficiently high. Reilly and co-workers have observed that excitation with 157 nm (Vacuum UV light, 7.87 eV) produces peptide fragmentation patterns that differ from those observed with lower energy excitation; they interpreted this observation as evidence for a singlephoton process [3,24]. Note that if energy redistribution is efficient compared to dissociation processes occurs, significant energy over the dissociation threshold and
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
253
thus, multiple photons would be necessary to effect photo-fragmentation of large molecules due to the large magnitude of their kinetic shift [21,40]. The absorption of multiple photons followed by IVR should produce fragments similar to those accessed using slow heating methods such as multiple collisional activation, BIRD, and IRMPD; this is what is observed frequently with excitation by visible photons. Figure 9.4 shows the visible action spectra measured in our laboratory of three charge states of gas-phase fluorescein formed by ESI and isolated and photo-activated inside a QIT [77]. Details of our experimental set-up can be found in Section 9.7.1. The squares in panels (a) and (b) show the measured gas-phase dissociation yield (calculated as the sum of fragment ion intensities divided by total ion intensity) produced using a range of excitation wavelengths at constant laser power for (a) protonated fluorescein [Fl + H]+ and (b) singly-deprotonated fluorescein [Fl−H]−. For comparison, the solid lines shown in the figures are: (a) the absorption spectrum of dilute (200 µM) aqueous fluorescein adjusted to pH 1.95, which corresponds to predominantly protonated fluorescein, and (b) the absorption spectrum of dilute aqueous (b) [Fl–H]–
(a) [Fl+H]+
1.00
1.00
O OH
0.40 0.20
0.80
Relative absorbance
+
0.60
OH
Dissociation yield
O
HO
Relative absorbance
Dissociation yield
0.80
0.60 0.40 0.20 0 400 420 440 460 480 500 520
0 400 420 440 460 480 500 520
0.50 0.40 0.30
O–
O
O O
O–
0.20 0.10
Relative absorbance
Electron photodetachment yield
(c) [Fl–2H]2–
0 400 420 440 460 480 500 520 Wavelength (nm)
FIGURE 9.4 Visible action spectra of different charge states of gaseous protonated/deprotonated fluorescein (Fl) compared with solution absorption spectra. (a) [Fl + H] + ; (b) [Fl−H]−; and (c) [Fl−2H]2−. Data points (n) indicate the fraction of photodissociation (for a and b) and the fraction of electron detachment (for c); the error bars indicate one standard deviation. Solid lines are absorption spectra of aqueous fluorescein: (a) pH 1.95; (b) pH 5.68; and (c) pH 11.65 (where the pH was adjusted either with acetic acid or with sodium hydroxide). The dashed lines represent the solution phase spectrum shifted in each case: (a) –10 nm; (b) + 110 nm; and (c) + 10 nm.
254
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
fluorescein adjusted to pH 5.68, at which ca 85% of fluorescein is thought to exist in its singly-deprotonated form. The dashed lines correspond to the aqueous spectra shifted such that their maxima coincide with the measured gas-phase maxima. For [Fl + H]+ (panel (a)), the shift is 10 nm to higher energy while for [Fl−H]−, (panel (b)) the shift is 110 nm to lower energy. The similarity of the measured gas-phase visible action spectra and aqueous absorption spectra suggest that the gas-phase structures of protonated fluorescein and doubly-deprotonated (dianionic) fluorescein (see below) resemble their solution phase counterparts. In contrast, the electronic properties of the singly-deprotonated (monoanionic) fluorescein differ markedly from those of dilute aqueous fluorescein adjusted to pH 5.68, which may indicate the presence of a different prototropic form of [Fl−H]−. Electronic structure theory calculations support this hypothesis: different sites of deprotonation are favored depending on whether the fluorescein is in the gas phase or in solution.
9.5.2 Electron Photodetachment In some cases, the consequence of photon absorption is electron detachment rather than dissociation, as shown in Equation 9.2. Electron detachment is straightforward to monitor in a mass spectrometer because the charge state number increases by one. Note that for monoanions, the products of electron detachment are neutral molecules and electrons; because the neutral species escape the trap undetected, the dissociation yield must be calculated from the disappearance of the parent ion. The photodetachment spectrum of doubly-deprotonated fluorescein [Fl−2H]2− is depicted as solid squares in Figure 9.4c; [Fl−2H]2− (m/z 165) converts readily to [Fl−2H]−• (m/z 330) upon excitation by visible radiation [77]. Also shown in the figure is the absorption spectrum of 200 µM fluorescein in aqueous solution adjusted to pH 11.65 with NaOH (solid line) and the solution spectrum shifted 10 nm to lower energy (dashed line) to facilitate comparison with the gas-phase data. The similarity of the gas-phase and solution-phase spectra, including the vibronic feature at + 1600 cm−1 (lying at ca 470 nm in the gas-phase data), suggest that the solvent does not have a strong influence on the structure of doubly-deprotonated fluorescein.
9.6 FLUORESCENCE SPECTROSCOPY Fluorescence spectroscopy may be used also to characterize ions in trapping mass spectrometers. Fluorescence is a highly-sensitive technique, which is applied increasingly to single molecule/ion studies.* Because fluorescence may be excited using a single UV or visible photon, difficulties associated with the interpretation of multiple-photon induced dissociation spectra are avoided. However, only selected molecules/ions fluoresce significantly, limiting the applicability of the technique or requiring labeling of analytes with fluorophores. Quantities of interest that may be measured in fluorescence experiments include fluorescence excitation maxima, * See Volume 5, Chapter 10: Sympathetically-Cooled Single Ion Mass Spectrometry by Peter Frøhlich Staanum, Klaus Højbjerre and Michael Drewsen, and Volume 5, Chapter 11: Ion Trap: A Versatile Tool for the Atomic Clocks of the Future by Fernande Vedel.
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
255
emission maxima, emission intensity (also known as ‘brightness’), and fluorescence lifetimes. All of these quantities are sensitive to the molecular environment of a fluorophore, and thus provide a useful probe of ion conformation and dynamics. At this point it is relevant to note the terminology employed in this chapter: the expression laser-induced fluorescence (LIF) is used as a general term describing any fluorescence that is excited using a laser. A fluorescence excitation spectrum shows fluorescence emission yield as a function of excitation wavelength; that is, it is similar to an absorption spectrum when that absorption results in radiative emission. It is noted that some authors reserve LIF as a synonym for fluorescence excitation spectroscopy. Dispersed fluorescence refers to dispersion of the emitted fluorescence light into its component wavelengths, that is, production of an emission spectrum. An enticing prospect for gas-phase structure determination is the use of fluorescence resonance energy transfer (FRET) measurements [13,78]. In FRET studies, a donor fluorophore is excited and, when an acceptor fluorophore is nearby, energy can be transferred from the donor to the acceptor via a dipole–dipole interaction. The efficiency of the energy transfer is dependent on the inverse sixth power of distance between the donor and acceptor [79]. The use of FRET has undergone a resurgence in recent years as biologists have employed it as a molecular ruler; it can be used to probe biologically-relevant length scales (10–100 Å) and has been used extensively in condensed phase studies of biomolecule folding and interactions [80]. In fluorescence spectroscopy, the signal that is measured is light (hν2 in Equation 9.3), as distinct from the action spectroscopic methods discussed in previous sections, in which the signal measured is ion intensity as a function of m/z-value, that is, a mass spectrum. Thus, in addition to interfacing an excitation laser with the mass spectrometer as is discussed in Section 9.2, an appropriate photon detector must be interfaced with the mass spectrometer. Once photons are absorbed, excited ions in the ion cloud become point sources of divergent light. This situation poses a significant challenge to the development of appropriate instrumentation. Single molecule studies employ fluorescence collection optics that collect light over essentially a 4π solid angle (that is, light is emitted in all directions and the emission is described as being isotropic). The detection of fluorescence is generally not straightforward in most mass spectrometers because either the ion cloud is somewhat diffuse (which is the case for FT-ICR or linear ion traps, see Figure 9.1) or (in the case for Paul-type QITs) the electrodes used to confine the ions to a small volume limit optical access. Figure 9.5 shows a dispersed emission spectrum of gaseous ions of the commonly-used laser dye rhodamine 590 (C27H29N2O3+, m/z 429.22) (solid line) stored inside a 3-D QIT and measured in our laboratory [15,81]. Also shown on the figure is a fluorescence excitation spectrum for gaseous rhodamine 590 (circles), measured by integrating the dispersed fluorescence signal as a function of excitation wavelength. Although rhodamine 590 was marketed as the popular laser dye rhodamine 6G by Exciton Corporation, mass spectrometric analysis showed that it was the methyl ester variant (m/z 429.22) of rhodamine 6G, which is an ethyl ester (m/z 443.22). Exciton updated their data sheets after we contacted them with this information. A subsequent gas-phase emission spectrum of rhodamine 6G purchased from Sigma Aldrich Corporation (m/z 443.22) showed that the gas-phase dispersed emission spectra of the two dyes are indistinguishable. The measured
256
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 1.0 0.9
Normalized intensity
0.8 0.7
H N
H+ N
O
O
0.6
O
0.5 0.4 0.3 0.2 0.1 400 420 440 460 480 500 520 540 560 580 600 Wavelength (nm)
FIGURE 9.5 Fluorescence emission (solid line) and excitation (circles) spectra of the ionic dye rhodamine 590 (C27H29N2O3 + , m/z 429.22). The dotted line joining the excitation data (circles) is a guide to the eye. It should be noted that fluorescence excitation and emission spectra are most often mirror images because the excitation spectrum (mostly) originates from the vibrational ground state of the ground electronic state while the emission spectrum originates from the vibrational ground state of the excited electronic state; the spectra ought to be mirror images when the ground state and excited electronic states exhibit basically the same geometry and have the same vibrational spacings.
gas-phase absorption and emission maxima of rhodamine 590 (λabmax and λem max, respectively) are at λabmax = 493±3 nm and λem max = 503±2 nm, respectively. For comparison, solution absorption and emission maxima of rhodamine 6G are lower in energy, ranging from λabmax = 526 nm and λem max = 563 in water to λabmax = 534 nm and λem max = 568 in octanol [82,83]. The fact that it takes more energy to excite the electronic transition in the gas phase than it does in solution indicates that the excited state is stabilized preferentially by interaction with the solvent. This indication is consistent with a simple picture in which excited states are somewhat more diffuse and hence more polarizable than is the ground state, leading to more favorable interactions with solvent molecules. However, the observed electronic transition for several negatively-charged ions is lower in energy in the gas phase than in some solvents, suggesting that this explanation may be too simplistic [77]. The small Stokes shift (the energy difference between λabmax and λem max) of rhodamine 590 measured in the gas phase (400 cm−1) as well as the ‘mirror image’ appearance of the excitation and emission spectra are indicative of little geometric change between the gaseous ground and excited states. The electronic spectra for fluorescein and rhodamine, shown in Figures 9.4 and 9.5, respectively, illustrate the utility of having a tunable laser source. Not only does this facility enable the construction of absorption type spectra (PD action spectra or excitation spectra) but, for fluorescence studies, it is useful to be able to tune the excitation wavelength so as to excite the system of interest in the most
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
257
efficient manner. While compendia of solution absorption maxima exist, little data are available for gas-phase ions. Thus, the flexibility of a tunable excitation source is advantageous.
9.7 case STUDY: MODIFICATION OF A COMMERCIAL QIT FOR OPTICAL SPECTROSCOPIC EXPERIMENTS 9.7.1 Apparatus Overview and Design Considerations Our laboratory has modified a 3-D Paul trap for use in fluorescence spectroscopic experiments. An overview of our experimental set-up is shown in Figure 9.6. Frequency-doubled light from a pulsed tunable Titanium:Sapphire laser system (80 MHz repetition rate, ca 150 fs-pulse width) (Tsunami pumped with a 10W Millenia Pro Nd:YAG, Spectra Physics Laser Division of Newport, Inc., Richmond, CA) is directed into a modified QIT (Esquire 3000 + , Bruker Daltonics, Bremen, Germany), which is equipped with an ESI source. Fluorescence radiation is collected orthogonally to the excitation path and dispersed by a spectrograph (Shamrock 303i, Andor Technologies, Ireland) onto an electron multiplying charge-coupled device (EM-CCD) (Newton, Andor Technologies, Ireland) to measure a fluorescence emission spectrum. A fluorescence excitation spectrum can be constructed by stepping through a range of wavelengths of the excitation light and measuring the integrated emission signal as a function of excitation wavelength. Beam stop
Ir2
Sh
Ir1 PH2
M5
ND filter
PH1 L2 BBO L1
M4
M8 BW1
M2
M3
BW2
LP filter
L4
Spectrograph/ CCD camera
Ti:Sapph M7
M6
Nd:YAG
M1
Vacuum chamber containing QIT
FIGURE 9.6 Overview of the experimental apparatus for measuring laser-induced fluorescence and photodissociation action spectra of gas phase ions stored in a 3-D quadrupole ion trap. Optical component abbreviations: Lenses (L1–L4); β-BaB2O4 crystal (BBO); Mirrors (M1-M8); Shutter (Sh); Irises (Ir1-Ir2); Pinholes (PH1–PH2); Brewster-angle windows (BW1– BW2); Neutral density filter (ND); Long-pass filter (LP). L3 is inside the ring electrode (see Figure 9.7).
258
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Primary considerations for instrument design were: (i) maximization of uorescence emission collection efficiency with minimized background signal from fl scattered laser light; (ii) maintenance of the commercial QIT performance characteristics including sensitivity, mass range, and mass resolution; and (iii) ease of alignment for the optical excitation and detection. Based on these considerations, we chose to create an optical excitation path through the ring electrode such that after intersection of the ion cloud, the excitation beam is directed back out of the instrument in order to minimize background from scattered laser light. Fluorescence light is collected orthogonally to the path of the laser using an antireflection coated BK7 plano-convex lens inserted into a third hole bored in the ring electrode. The lens is positioned as close to the ion cloud as possible, minimally inset from the electrode surface, in order to maximize the solid angle of collection. A cross-sectional view of the apparatus is shown in Figure 9.7. At the heart of our experimental apparatus is an optical assembly, placed inside the existing vacuum chamber and mounted directly to the ion trap assembly, straddling the ring electrode (Figure 9.7). The optical assembly consists of a series of optical baffles and two 45˚ mirrors that provide a defined optical path for the excitation beam, which passes through the two holes in the ring electrode and through the center of the ion trap where the ion cloud is located. The optical assembly is grounded, but it is located only a few millimeters from the ring electrode to which an RF potential up to 15 kV (V0-p) is applied. As a result, the capacitance of the RF circuit was perturbed sufficiently to require significant re-tuning. The re-tuning procedure is described in more detail below. (a)
Fluorescence
Laser
Brewster window
(b)
Vacuum chamber Baffles
RF feedthrough Adjustable mirror
Ring electrode RF Lead
3 mm lens 2 mm hole
FIGURE 9.7 (a) Diagram of the ion trap housing showing the optical assembly for photoexcitation and fluorescence detection as it sits straddling the ring electrode of the QIT; and (b), close-up view of the fluorescence collection hole and lens inserted into the ring electrode.
259
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
9.7.2 Tuning the RF Circuit Incorporating the Ring Electrode The Esquire 3000 + is so named because it is capable of scanning a mass range from 50–3000 Th under ‘normal’ operational parameters that deliver acceptable resolution with modest scan speeds (Standard: 10,000 Th s−1, Enhanced: 5550 Th s−1). This mass range is achieved using a mass-selective instability scan with axial modulation at qz = 0.78 and requires a zero-to-peak RF voltage (V) of ca 15 kV. The RF potential is generated by driving a circuit, made up of a three-coil inductor and QIT assembly, with a fixed-frequency quartz crystal oscillating at 781.250 kHz. A schematic diagram of the circuit is shown in Figure 9.8a. The primary coil has two turns, the secondary coil has ca 400 turns and the tertiary coil has five turns. The circuit is characterized by a sharp frequency-gain curve (Figure 9.8b) with a resonant frequency described by fresonant =
1
2π LC
(9.4)
where fresonant is the resonant frequency in Hertz, L is the inductance of the circuit, and C is the capacitance of the circuit. The tertiary coil is coupled to a bank of capacitors for fine-tuning the resonance. As a result of the narrow resonant band-width, the RF circuit is sensitive to relatively small changes in the capacitance of the ion trap assembly. One of the initial modifications that we made was to rotate the ring electrode and to re-route the RF lead to allow installation of the optic assembly (see Figure 9.7). A new lead was fabricated from 18 gauge copper magnet wire* and connected to the ring electrode in its new orientation, resulting in an increase in capacitance of ca 0.25 pF. Subsequently, (a)
(b)
Pulse generator
3o Coil
781 kHz
1 0.8
1o coil 2o coil
Antenna Oscilloscope
Relative gain
Tuning capacitors
0.6 0.4 773
777
781 f (kHz)
785
789
FIGURE 9.8 (a) Diagram of the RF circuit incorporating a three-coil inductor and the ring electrode. A method for measuring the RF gain curve using a frequency variable pulse generator and antenna is shown. (b) RF gain curve measured as a function of drive frequency. * The commonly-used term for wire that is used to make electromagnets, but is itself non-magnetic, is ‘magnet wire’; it is an excellent conductor with a thin coating of insulation.
260
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
the RF current was found to overload at m/z < 200, indicating that the RF circuit was no longer in resonance with the drive frequency. Upon installation of the grounded optical assembly ca 1 mm away from the ring electrode and re-routing the RF lead, ca 1.7 pF of capacitance was added to the circuit, reducing the resonant frequency by nearly 24 kHz and far beyond the range of tuning that could be achieved by adjusting the capacitors attached to the tertiary coil. In order to compensate for the added capacitance, the inductance of the circuit was reduced to regain resonance at 781.25 kHz. This change was accomplished by trimming turns from the secondary coil. The effect of removing turns from the coil was monitored by assessing the change in the resonance frequency of the RF circuit in the following manner. The relative gain of the RF circuit was measured using an antenna, fixed near the inductor, while driving the circuit using a square waveform of variable frequency (Figure 9.8). In order to achieve resonance with the drive frequency, 17 and 2/3 turns were removed from the secondary coil, reducing its inductance by 85 µH. After tuning, the RF current (monitored by the instrument control software) was found to be well within the expected range and, after several days of conditioning, the full mass range up to m/z 3000 was re-established. No loss of sensitivity, isolation, fragmentation efficiency or mass resolution was observed.
9.7.3 What Size Holes to Drill? An important design decision was choosing the size and number of holes to be bored in the ring electrode. If the fluorescence signal is assumed to be a point source originating at the center of the QIT, the fraction, F, of fluorescence light passing through an aperture may be estimated from the radius of the collection hole r1 and the distance between the collection hole and the point source r 2 as
F=
r2 area of collection hole = 12 surface area of sphhere radius r2 4r2
(9.5)
From Equation 9.5, it is apparent that maximizing the radius of the collection aperture(s) while minimizing their distance from the ion cloud will increase the fraction of light that may be collected. In the QIT, the closest a collection lens can be placed to the center of the ion trap (that is, the minimum possible value of r 2) is equal to the radius of the ring electrode, r0, which is ca 10 mm in our trap. Note that placement of a lens inside the trapping volume is not feasible as it will perturb the trapping field and become charged up as ions hit the lens. Neither is it possible to place a lens in the axial direction because there must be an unobstructed path for the ions to enter and exit the trap. Thus, for a collection hole of r1 = 1 mm bored in the ring electrode, only 0.25% of all the light emitted from the ion cloud will follow the collection path. To alter the hyperboloidal surfaces of the electrodes would be to introduce additional non-linear field components that can affect adversely the performance of the
261
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
mass spectrometer. In order to characterize the effects of drilling different numbers and different sizes of holes in the ring electrode, the behavior of ions stored within modified versions of the Esquire 3000 + QIT was studied using SIMION (SIMION v. 8, Scientific Instrument Services Inc., New Jersey, USA). Before discussing the results from the SIMION models of our trap, the expected behavior of ions in QITs is reviewed briefly.
9.7.4 Ion Behavior in Real Traps The dynamics of ion motion in QITs have been investigated in detail using both theoretical and experimental methods, and have been summarized concisely in Volume I of this series [84]. Here, we review briefly some of the key concepts before discussing ion behavior in real (non-ideal) traps. The basic theoretical treatment is restricted to ions stored in a pure quadrupolar electric field (as it is referred to colloquially) that is generated by applying a potential ϕ0 to perfectly hyperbolic electrodes having ideal geometry (r 02 = 2z02). The operating conditions (RF potential V = V0-p, DC potential U and drive radial frequency Ω) for which an ion of mass m and charge e will adopt a stable trajectory are defined by the trapping parameters au and qu (where u is r or z) given by
ar =
8eU m ( r + 2 z02 ) Ω 2
qr =
− 4eV m ( r + 2z02 ) Ω 2
(9.6)
az =
−16eU m ( r + 2 z02 ) Ω 2
qz =
8eV m ( r + 2 z02 ) Ω 2
(9.7)
2 0
2 0
2 0
2 0
The trapping parameters, expressed generally as au and qu, are the axes of the well-known stability diagram. The boundaries of the stability regions in the axial and radial directions are defined by an additional trapping parameter βu calculated from a continued fraction of the form
βu2 = au +
qu2
(βu + 2)2 − au −
qu2
(βu + 4 )2 − au −
+
(β u + 6 )
qu2 2
− au −
qu2
(βu − 2)2 − au −
qu2
(βu − 4 )2 − au −
(β u − 6 )
qu2 2
− au −
(9.8)
262
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Ions confined by a pure quadrupolar field and within the bounds 0 ≤ βu ≤ 1 will adopt regular, stable trajectories dominated by fundamental frequencies of oscillation in the r and z directions, ωr,n and ωz,n, given by Equations 9.9 and 9.10, where n is an integer. The characteristic fundamental secular frequencies (ωr,0 and ωz,0) are unique for a given value of βu. Equations 9.9 and 9.10 describe the principal frequencies of ion motion in a quadrupolar field.
β ω u ,n = n + u Ω 2 β ω u ,n = − n + u Ω 2
0≤n<∞
−∞ < n ≤0
(9.9)
(9.10).
In a pure quadrupolar field, the radial and axial components of an ion’s motion are independent of each other. Also, the magnitude of the electric field varies linearly with distance from the center of the trap in both the axial and radial coordinate. Hence, a quadrupolar field is sometimes referred to as being ‘linear’. The theoretical treatment of ion motion in a pure quadrupolar field summarized above is a simplification of the true behavior of ions stored in a real instrument. In reality, electrodes are never perfectly hyperbolic and must be truncated to fit inside a vacuum chamber and to avoid electrical discharge between the ring and end-cap electrodes. Almost invariably, electrodes will have surface imperfections and they may be imperfectly aligned or spaced. In mass spectrometers, holes must also be bored in the end-cap electrodes to permit ion injection/detection. These changes from the ideal electrode geometry introduce higher-order fields (hexapole, octopole, etc.) that perturb the regular ion motion predicted by the pure quadrupolar model. The presence of higher-order fields was described first by Paul [85] and, for a time, they were considered detrimental to the performance of QIT mass spectrometers and were to be avoided. Unlike the pure quadrupolar field, higher-order fields vary nonlinearly with displacement in either r or z. This non-linearity translates into a spatial dependence of the secular frequency; that is, ions of a given m/z-value confined near the center of the trap will have different values of ωu than ions of the same m/z-value possessing trajectories having larger displacements from the trap center. This variation of ωu, particularly ωz, could result in decreased mass resolution. Higher-order terms in the electric field have also cross-terms coupling r and z so that motion in the radial and axial directions can no longer be described independently [84]. The contribution of higher-order terms in the electric field (so called ‘non-linear fields’) is minor near the center of the trap (small values of r and z), where quadrupolar terms dominate, but become increasingly important as the displacement from the center of the trap increases [86]. Thus, collisional cooling of ions to the center of the trap is advantageous as it limits the effects of non-linear field components. A direct consequence of the presence of non-linear fields is that under certain operating conditions, ions will have their regular motions altered significantly and may develop unstable trajectories. In particular, these effects are expected to be most
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
263
severe during injection, detection, and resonant activation when the ion ensemble has large spatial displacements. The term ‘non-linear resonance’ describes the resonant absorption of potential energy from higher-order trapping fields. Such absorption occurs when combinations of an ion’s secular frequencies match harmonic sidebands of the RF drive frequency. Conditions for non-linear resonance exist in all QITs but the effect on ion motion is dependent upon the order, sign (+ or −) and strength of the superimposed non-linear fields. Due in large part to the work of Franzen and co-workers [87–92], the phenomenon of non-linear fields and their contribution to non-linear resonance effects in ion traps is now well understood. In addition to the quadrupolar resonance, there are a series of resonance conditions described by Equations 9.11 and 9.12 where ν is an integer ≥0 and n is the order of the multipole. The overall order of the resonance (N) is described by Equation 9.13.
nr βr + nz β z = 2 ν
(9.11)
nr ω r + nz ω z = ν Ω
(9.12)
nr + nz = N
(9.13).
As can be seen from Equation 9.12, non-linear resonances arise when an integral multiple of the drive frequency is equal to a linear combination of integral multiples of the secular frequencies, that is, when the combination bands of secular frequencies and side bands of the drive frequency overlap [93]. The effects of non-linear resonances were observed originally in experiments using commercial instruments; the effects were attributed to ‘black holes’ [94] or ‘black canyons’ [95] in the stability diagram, and are known to be responsible (in conjunction with collisions) for the chemical mass shift [96]. However, as has been discussed by Franzen [86,89], the contribution of multipole fields of any order can be controlled by adjusting the geometry of the electrode assembly. In particular, Franzen has described how the appropriate selection of the hyperbolic angle of the electrodes can be used to take advantage of non-linear fields, so as to improve the operating characteristics of the QIT such as resonant ejection of ions at the non-linear resonance βz = 2/3 (nr = 0, nz = 3, ν = 1) due to the hexapolar component of the trapping field [92]. In order to evaluate the changes to the performance of the ion trap, caused by the perturbation to the trapping field due to the addition of holes in the ring electrode to enable our fluorescence measurements, we have carried out ion trajectory calculations in several models of the modified Esquire 3000 + QIT. Franzen [87,91,92] demonstrated previously the utility of ion trajectory calculations in investigation of the modified hyperbolic angle trap, but considered only ion motion in the axial direction. Here, we discuss results from several SIMION models, which have been constructed with different numbers and sizes of holes in the ring electrode. Fourier analysis of ion trajectories is used to determine secular frequencies, frequency shifts,
264
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
and the strength of combination bands. The spatial and temporal distributions of the ion cloud have been investigated also with the inclusion of a hard-sphere collision model to estimate the overlap of the ion cloud with the excitation laser. Finally, a mass-selective instability scan with resonant excitation has been simulated for a large ensemble of ions to explore the effects of added higher-order fields on ion ejection and detection.
9.7.5 Ion Trajectory Calculations Trajectory calculations of ions stored in a Bruker Esquire 3000 + QIT were performed for several reasons: (i) to examine the effects of drilling larger and different numbers of holes in the ring electrode; (ii) to gain insight into the behavior of the ion cloud under conditions best suited to obtaining fluorescence signal; and (iii) to learn more about the modified hyperboloidal angle QIT geometry. Ion trajectories were calculated using the SIMION package. The methods used by the SIMION package to perform ion trajectory calculations have been described in detail elsewhere. [97–101] Briefly, a SIMION model consists of a region of space divided into an array of points, called the potential array (PA). The array consists of electrode points (to which a potential may be applied) and non-electrode points for which the potential and electric field are determined from solutions to the Laplace equation calculated by a method of finite differences [97,98]. This method for calculation of electric fields is known also as the field interpolation method that has been incorporated recently into version 6.0 of the Ion Trajectory Simulation software package (ITSIM) [102]. Once a PA has been constructed, ions may be created with user-specified parameters of mass, charge, initial position (xi, yi, zi), and velocity ( x, y, z ). Ion trajectories are computed by applying Newton’s equations of motion at software-controlled time steps. At each integration step, the ion’s mass, charge, position, and velocity vector are known and the force acting on the ion is determined from the magnitude of the electric field at its position in the PA. Finally, an acceleration vector due to the electric field is calculated, the ion’s velocity vector is updated, and the trajectory continues to the next step. 9.7.5.1 SIMION Models The SIMION models created for this investigation are an attempt to characterize the fundamental features of the modified hyperboloidal angle electrodes unique to the Bruker design. Estimates of the relevant dimensions and operating parameters for the Bruker Esquire 3000 + QIT are given in Table 9.1 (see also Figure 9.9). We are aware of no reports in the literature describing a complete simulation of the Esquire 3000 + geometry. However, a series of papers by Franzen [88–92] in the early 1990s described in detail the theoretical aspects of the non-linear ion traps and used computations, of the axial component of ion motion only, to characterize operational applications such as the mass-selective instability scan and axial modulation. It has since been accepted that non-linear fields are present to some extent in all commercial QITs and need not necessarily be detrimental to instrument performance, provided the effects of these fields are known and the geometry
265
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
Table 9.1 Dimensions and Operating Parameters for the Quadrupole Ion Trap Used in the Simulations SIMION Model Parameters r0 z0 θ
10 mm 7 mm 1.8
ρ0 fRF fAC Number and diameter of entrance end-cap holes Number and diameter of exit end-cap holes
2.3r0
(a)
781.25 kHz 1/3fRF 1×2 mm 7×1.5 mm
Fluorescence
(b)
x y
ρ0
z
z0 r0
La
s er
FIGURE 9.9 (a) Cross-section of the quadrupole ion trap electrodes. (b) A picture of the truncated electrode assembly modeled using SIMION v. 8.0.
is controlled carefully. For this reason, we sought to establish whether our SIMION models could aid our understanding of the operation of the Esquire 3000 + and to predict subsequently whether drilling holes in its ring electrode perturbs the ‘desirable’ non-linear effects. The resolution of a SIMION model is limited by memory constraints. The accuracy of ion trajectory simulations is highly dependent on the spatial resolution of the PA; naturally, higher resolution models provide better approximations of smooth electrode surfaces and hence a better description of the electric field. Unfortunately, highresolution models are memory-intensive; each point in the PA requires 8–10 bytes of dynamic memory. SIMION v. 8.0 has an upper limit of 2 × 108 points, which corresponds to ca 1.8 GB of RAM. Thus, the maximum cubic PA allowed is approximately 580 × 580 × 580 points. However, in order to run efficiently simulations of dynamic
266
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
devices such as the QIT, which have time-dependent voltages on the electrodes, multiple Potential arrays (PAs) must be included in the simulation. SIMION creates a ‘fast adjustable’ PA for each dynamic electrode as well as a master PA that contains the sum of the potentials due to each electrode. All of these potentials must be dimensioned to sum to less than the maximum memory limit of 1.8 GB. The use of symmetry elements reduces the memory requirements of PAs. The cylindrical symmetry of an ideal trap is lost in the Esquire 3000 + due to the presence of seven holes in the exit end-cap electrode and a single hole in the entrance end-cap electrode. The symmetry is further reduced, to just two orthogonal planes, upon inclusion of two holes in the ring electrode for the laser. With the addition of a third hole for fluorescence collection, this symmetry is reduced further to a single plane, the x–z plane (see Figure 9.9). Given the memory limitations and the low symmetry, the spatial extent of the model was truncated in order to maintain sufficient resolution. To strike a balance between resolution and electrode truncation, the model was truncated along the asymptote at ρ0 = 2.3r0 (Figure 9.9a) yielding a final resolution of ca 0.076 mm per grid unit. A total of nine SIMION models (Table 9.2, see also Figure 9.9) were constructed with varying numbers and sizes of holes in the ring electrode. Model #0 is a highresolution model of an ideal geometry trap, which has no holes in the ring or end-cap electrodes. Model #1 is a model of the modified-angle electrode assembly before any further modification; this model, which includes holes in the end-cap electrodes, was used to establish a baseline for describing the characteristics of the modified hyperboloidal angle geometry. Model #2 was as for Model #1 plus two 2 mm-diameter holes for delivering the laser light (‘laser holes’) located on opposite sides of the ring electrode and along the y-axis. Models #3, 4, and 5 were as for Model #2 plus a single fluorescence hole of diameter 2, 3, or 4 mm, respectively; the fluorescence hole was located along the x-axis and passed through one side only of the ring electrode. This arrangement of three holes introduces an asymmetry in the ring electrode. In order to establish whether this asymmetry affected adversely the shape and location of the ion cloud, three more models (#6–8) were constructed. Models #6, 7, and 8 were as for Model #2 plus two fluorescence holes of diameter 2, 3, or 4 mm, respectively; the two fluorescence holes were located on opposite sides of the ring electrode and along the x-axis. 9.7.5.2 Single Ion Trajectories in Vacuum To compare the models, single ion trajectories under collision-free trapping conditions were calculated using SIMION. Particular attention was paid to ions at βz = 2/3 where there exists a strong non-linear resonance, used for ion ejection in the Esquire 3000 +. Ion trajectories were calculated for a period of ca 13 ms at a sampling rate of 10 MHz and the spatial coordinates at each time step were recorded. The data were apodized using a sine function and the frequency power spectra were computed using the fast Fourier transform algorithm implemented by Matlab• v. 7.0 (The Mathworks, Inc. Nattick, MA). The Fourier transform of an ion trajectory reveals the component frequencies present in the ion’s motion including secular frequencies and sideband frequencies (Equations 9.9 and 9.10 for a quadrupolar field).
End-cap holes+Laser holes+1 Fluorescence hole ″ ″
#3
2×3 mm 2×4 mm
2×2 mm
1×3 mm 1×4 mm
1×2 mm
No holes Entrance 1×2 mm Exit 7×1.5 mm 2×2 mm
ωx
34.33 34.10
34.41
34.41 34.32
34.41
34.48
34.49 34.48 34.41
34.48 34.56
34.41
34.41 34.48
34.41
34.41
69.43 69.35
69.50
69.50 69.43
69.50
69.50
ωx
111.3 110.7
111.7
111.5 111.2
111.8
111.8
111.94 111.9 111.8
111.8 112.2
111.7
111.7 111.8
111.6
111.6
111.94 111.9 111.8
ωy
ωz 69.64 69.66 69.58
ωy 34.49 34.48 34.41
ωz
259.3 258.8
259.6
259.6 259.2
259.7
259.8
260.42 260.5 259.9
Note: The theoretical frequencies were calculated using ITSIM v. 5.0 and all ion trajectories were calculated with SIMION, using identical simulation parameters under collision-free conditions. According to Equations 9.9 and 9.10, secular frequencies, ωu, should be expressed in Rad s–1; regrettably, it is common practice to express secular frequencies in kilohertz.
#7 #8
#6
End-cap holes+Laser holes+2 Fluorescence holes ″ ″
End-cap holes+Laser holes
#2
#4 #5
Theoretical Ideal geometry End-cap holes
Model #0 #1
qz = 0.785
qz = 0.249
Table 9.2 A Summary of the Characteristic Secular Frequencies (ωu Expressed in kHz) from Single Ion Trajectories in QIT Models with Progressively More Holes Bored in the Electrodes
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers 267
268
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
The presence of higher-order multipole components of the electric field acting on the ions has several effects that should be apparent on analysis of ion trajectories [93]. First, in accordance with Equation 9.12 an ion’s motion will have component frequencies in addition to the fundamental quadrupolar resonance frequencies (ωr and ωz). Second, the ion’s secular frequency may deviate slightly from the quadrupolar secular frequency. Furthermore, due to cross-terms (terms dependent upon both r and z) in the expansions for higher-order fields, ion motion in the axial and radial directions no longer can be treated independently; there will be coupling between motion in r and z, observable as additional frequency components in the Fourier analyses. Third, the secular frequencies will depend on the axial and/or radial displacement of the ion in the trap. These conditions are often manifested as a beating motion of an ion’s trajectory, in which case a broadened peak in the Fourier analysis for frequency determination should also be observed. When the ion’s secular frequency matches that of a non-linear resonance, the kinetic energy of the ion can increase rapidly as the amplitude of its excursions increases. However, the increase in ion axial or radial displacement can cause the secular frequency to change such that the ion falls out of resonance, subsequently losing kinetic energy by destructive interference. The rapid oscillation in and out of resonance results in the ‘breathing’ behavior of the ion’s trajectory. Finally, for the strongest resonance conditions, ions that gain sufficient kinetic energy can be ejected from the trap by the same mechanism as ejection at the βz = 1 boundary of the stability diagram. An example of the effect of a non-linear resonance is shown in Figure 9.10 for a single ion stored at qz = 0.785 (near βz = 2/3) in model #1 with two different starting conditions. In Figure 9.10a through d, the ion is started near the center of the trap and its excursions are limited to approximately ± 0.5 mm in r and z. In Figure 9.10e through h, the ion’s initial position was displaced to xi = yi = zi = 1 mm, whereupon it was observed that the ion made wide excursions from the center of the trap (zmax = ±5 mm and rmax = ±2.5 mm); the change in its frequency spectrum was striking. The ion that started near the center of the trap showed a fairly clean Fourier spectrum (Figure 9.10b and d) in both the x and z-directions and the beat motion did not alter significantly the secular frequencies, as indicated by the sharpness of the peaks (ωz = 259.9 kHz, ωr = 111.8 kHz). Several component frequencies predicted by Equation 9.12 are identified and are indicated by [•] in Figure 9.10b and d. These frequencies correspond to coupling bands between the axial and radial frequencies. Furthermore, several frequencies appear in both the axial and radial projections, providing additional evidence of coupling between these dimensions. In contrast, the power spectra of the ion with large oscillations (Figure 9.10f and h) no longer exhibit such sharp peaks for the secular frequencies; note, however, that the vertical scale in the Fourier spectra is logarithmic. There is significant beating in both the r and z-directions and the secular frequencies are split into multiple component frequencies, although the principal fundamental secular frequencies are still present. The axial secular frequency (ωz = 262.5 kHz) is shifted higher by ca 2.6 kHz from that measured for the ion with low amplitude oscillations, while the radial secular frequency is shifted lower by 0.8 kHz (ωr = 111.0 kHz). These single-ion trajectories illustrate several aspects of the modified hyperboloidal angle geometry. The observation of non-linear resonance lines in the power
269
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers (a) z (mm)
0.5
(b) 1010
0
−0.5 0 (c)
{xi, yi, zi } = 0 mm
100 0.002 0.004 0.006 0.008 t (s)
0.01
x (mm)
−0.5 0 (e) z (mm)
5
x (mm)
0.002 0.004 0.006 0.008 t (s)
0.01
200
2ωz = f – ωz
•
•
••
400
f
• •• • • •• •
600 f (kHz)
ωr
• •
10
800
200
••
• •
400
•
• ••
1000
f + ωr
f – ωr
•
–10
f + ωz
•
f
• •
600 800 f (kHz)
••
1000
(f )
{xi, yi, zi } = 1 mm
ωz
1010
f – ωz
f + ωz
100 0.002 0.004 0.006 0.008 t (s)
0.01
10–10
(h)
3
1010
0
−3 0
•
••
100
0
−5 0 (g)
1010
0
•
10–10
(d)
0.5
ωz
200
400
ωr
600 f (kHz)
800
f – ωr
1000
f + ωr
100 0.002 0.004 0.006 0.008 t (s)
0.01
10–10
200
400
600 f (kHz)
800
1000
FIGURE 9.10 Plots analyzing single ion trajectories from SIMION calculations of an ion stored at qz = 0.785 in trap model #1 starting at (a through d) the center of the trap and (e through h) displaced from the center of the trap. Panels (a, e) show plots of axial (z) position as a function of time, and panels (c, g) show plots of radial (x) position as a function of time. Panels (b), (d), (f), and (h) show frequency power spectra, plotted on a log scale, of the radial (x) and axial (z) motion. The solid circles (•) in panels (b) and (d) mark frequencies predicted by Equation 9.12.
spectra confirm that higher-order fields, introduced by the holes in the end-cap electrodes and the modified geometry, are of sufficient strength to affect noticeably the ion trajectories. There is also significant coupling observed between the axial and radial components of the ion’s motion; the secular frequencies for both r and z-motions appear in the frequency analysis of the axial and radial projections, respectively. Finally, the amplitude dependence of the non-linear fields is illustrated clearly by the differences in the frequency spectra in Figure 9.10. The motions of the ion that remains near the center of the trap, where the quadrupolar field dominates, are regular and are perturbed only moderately, whereas beating becomes predominant for the second ion with larger oscillation amplitudes.
270
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
To investigate systematically the effects on the ion motions introduced by addition of holes in the ring electrode, single ion trajectories in vacuum were calculated for each model geometry (#0–#8). In Table 9.2 are summarized the results from the simulations for a single ion of m/z 100 at two working points on the az = 0 axis of the stability diagram: qz = 0.249 (standard conditions for ion storage, ion isolation, and collision induced dissociation in the Esquire 3000 +) and qz = 0.785 (the working point for ion ejection during the mass-selective instability scan with axial modulation at βz = 2/3). The initial position of the ion was defined at the center of the trap (as in Figure 9.10a through d) such that the secular frequencies could be observed clearly in the Fourier power spectra. One consequence of adding more and larger holes to the ring electrode is an overall decrease of the principal secular frequencies (Table 9.2). Model #1, which corresponds to the unmodified trap such that it has no holes in the ring electrode, is used as a baseline for comparison of the secular frequencies. The largest differences in secular frequencies from those of model #1 are observed for model #8 which has the largest fluorescence holes (2 × 4 mm); at qz = 0.249 (qz = 0.785), the axial frequency is decreased by 0.2 (1.1) kHz while the radial (x) frequency is decreased by 0.3 (1.1) kHz and the radial (y) frequency is increased by 0.2 (0.4) kHz. The difference in radial x and y-frequencies is a consequence of the broken symmetry; the fluorescence holes lie along the x-axis while the smaller laser holes are on the y-axis. The largest splitting between the radial frequencies in the x and y-directions (0.5 kHz for ions at qz = 0.249, 1.5 kHz for ions at qz = 0.785) are observed for model #8, which has the largest holes. It is interesting to note that the splitting between ωx and ωy for model #8 at qz = 0.785 (1.5 kHz) is more than twice as large as that for model #5 (0.6 kHz), which has a single 4 mm fluorescence hole. The fact that some arrangements of holes introduce differences in the radial frequencies (and even an increase in ωy for model #8) may affect adversely the efficiency of ion ejection, because the contribution of hexapole and octopole fields has been selected to facilitate preferential ejection toward the exit end-cap electrode [87]. The calculated shifts in secular frequency are small, < 0.5% even for the model with the largest holes, and it should be possible to compensate for the shifts by mass calibration. Thus, the frequency shifts caused by adding holes to the ring electrode were not a source of concern. Multiple coupling frequencies are present in the power spectra of the ions at qz = 0.249 due to beating of the ion motions (data not shown). However, the rather dramatic effect due to increase of the oscillation amplitude that was demonstrated very clearly for the ion at qz = 0.785 in Figure 9.10f and h was not observed to nearly the same extent at qz = 0.249. This finding is not surprising because this working point on the stability diagram does not fall near one of the stronger non-linear resonance βz -lines. Figure 9.11 illustrates the effect of adding progressively more and larger holes to the ring electrodes by comparing the frequency analysis for single ions stored at qz = 0.785 in several of the models described in Table 9.2. The frequency spectra of the high-resolution model simulations of the ideal trap show the expected frequency components at ωu, f−ωu, and f + ωu where f = Ω/2π. In the heading of Table 9.2 there is a note concerning the units of ωu; it is an idiosyncrasy of this field that one finds nomenclature such as f + ωu where the units of f are hertz while those of ωu are
271
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
1010
(a)
Model #1 End-cap holes
1010
f + ωz
(b) ωr
f – ωr f + ωr
f
(c)
(d)
(e)
(f)
100 10−10
Model #2 End-cap holes + two laser holes
f – ωz
f
Model #0 Ideal geometry 0 high resolution 10
10−10
ωz
1010 100
10−10 (g)
(h) 1010 Model #5 End-cap holes, two laser holes 100 + one fluorescence hole 10−10 (i) (j) 1010 Model #8 End-cap holes, two laser holes 100 + two fluorescence holes 10−10 0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200 Axial frequency (kHz) Radial frequency (x) (kHz)
FIGURE 9.11 Fast Fourier analyses of the axial (z) and radial (x) components of the trajectory for a single ion stored at qz = 0.785 (βz = 2/3) using five electrode configurations. (a,b), Model #0, ideal geometry trap; (c,d), Model #1, End-cap holes; (e,f), Model #2, 2 × 2 mm Laser holes; (g,h), Model #5, 1 × 4 mm Fluorescence hole; and (i,j), Model #8, 2 × 4 mm Fluorescence holes.
strictly radians s−1, yet ωu is expressed often in hertz. Some other very small features (note the logarithmic vertical scale) are observed also. The frequency spectra for the unmodified trap (Model #1, Figure 9.11c and d) and the trap with 2 × 2 mm laser holes in the ring electrode (Model #2, Figure 9.11e and f) are virtually indistinguishable, suggesting that adding two holes of this size to the ring electrode will not affect unduly the operation of the QIT. However, the addition of larger holes, that is, holes of diameter greater than 3 mm, introduces additional frequency components to the ions’ trajectories. These additional lines are most prominent for model #5, which has a single fluorescence hole of diameter 4 mm in the ring electrode (Figure 9.11g and h). Coupling between the axial and radial components of an ion’s motion is observed to some extent for all of the models, indicating that higher-order fields
272
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(having non-zero cross terms in r and z) are present, in addition to the axial-only resonance βz = 2/3. The results of the single ion simulations in vacuum confirm the presence of higher-order fields, as expected. While the simulations suggest that only minor changes to secular frequencies result from modifications to the ring electrode, it cannot be determined from the single-ion trajectories at fixed qz whether the modifications cause a loss of performance in either ejection efficiency or mass resolution. In order to investigate these possibilities, the ion ejection process during a mass-selective instability scan was simulated for large ensembles of ions to ascertain effects of hole number, symmetry, and hole size on mass resolution and detection efficiency. 9.7.5.3 Mass-Selective Instability Scan with Collisions and Axial Modulation The Esquire 3000 + generates a mass spectrum using a mass-selective instability scan during which the RF voltage applied to the ring electrode is ramped (thus ramping the axial secular frequency of all ions) with a concurrent application of an auxiliary dipolar AC potential to the end-cap electrodes to effect ion ejection. The frequency of the auxiliary AC potential (fa = 260.417 kHz) is fixed for all scans at exactly one third the frequency of the RF drive potential (f = 781.250 kHz). This judicious selection of the AC frequency coincides with the secular frequency of ions held at qz = 0.785, that is, at the non-linear resonance βz = 2/3. An important parameter in tuning the instrument is selection of the appropriate phase shift γa for the auxiliary AC potential in relation to the RF drive to ensure coherent motion of the ions (in resonance with fa), optimizing ejection efficiency by minimizing destructive interference with the RF potential. For the simulations presented here, the RF and AC potentials, ϕ0 and ϕa, respectively, applied to the electrodes are given by Equations 9.14 and 9.15, respectively, with γa = π/3 radians. For the ring electrode φ 0 = V sin Ω t
(9.14),
φ a = Va sin(Ω t /3 + γ a )
(9.15),
for the entrance end-cap electrode
and -ϕa is applied to the exit end-cap electrode. An ensemble of 10,000 singly-charged ions representing the rhodamine 101 cation (even electron ion: C32H31N2O3 +, m/z = 491.23) was created composed of the appropriate abundance of isotopomers (m/z 491, 492, 493, and 494). To approximate reasonable starting conditions, the ions’ initial conditions were generated as follows, where TOB is the time of birth: (xi, yi, zi) 3-D Gaussian distribution around the trap center with σ = 0.5 Ek 0.1 eV ( x, y, z ) randomized TOB uniform distribution over 1 RF cycle
273
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
To represent the true environment of the QIT, a hard-sphere collision model (HS1 coded by Manura) [97] was implemented with a user program in SIMION. Ion trajectories were computed individually from an initial qz = 0.70 for m/z 491. The RF voltage was scanned at a rate equivalent to 5500 Th s−1, allowing approximately 2 ms prior to ion ejection for the ions to equilibrate with the helium collision gas held at a pressure of 1 mTorr. Simulations of a mass-selective instability scan using axial modulation were run to compare ejection efficiency, detection efficiency, and mass resolution for three model geometries (#1, #5, and #8). The ion properties (time of flight, position, and kinetic energy) were recorded at the end of the simulation when the ion strikes (or splats, in SIMION parlance) an electrode or is ejected through a hole in one of the end-cap electrodes. A ‘detector’ stop was placed immediately behind the holes in the entrance and exit end-cap electrodes (see Figure 9.12). The time of splat (on either of the end-cap electrodes or the detector stops) for each ion was recorded and the entire set of splat times for the ensemble of 10,000 ions in each model geometry (#1, #5, and #8) was used to reconstruct a mass spectrum by assuming that each ion was ejected at qz = 0.785 (see Figure 9.13). Figure 9.12 illustrates the fate of the 10,000 ions striking the entrance end-cap electrode (left-hand side), exit end-cap electrode (right-hand side) or the ‘detect’ plate for the model geometry #8 (2 × 4 mm fluorescence holes). It is clear that the ions are, indeed, ejected preferentially in the ‘correct’ direction, toward the exit end-cap electrode. (a)
N
4000
Entrance end-cap 30%
Exit end-cap “Detect” 48% 22%
2000
Radial (x) (mm)
(b)
0 −10 5
−8
−6
−4
−2
0
2
4
6
8
10
8
10
“Detect” plate 0
−5 −10
−8
−6
−4
−2 0 2 Axial (z) (mm)
4
6
FIGURE 9.12 Axial distributions of ion splat events following a simulated mass-selective instability scan using Model #8 with γa = π/3 radians. (a) A histogram showing the percentages of ions striking the various regions of the end-cap electrodes and the ‘detect’ plate. (b) The spatial distribution of ion splat events in the axial (z) direction with respect to the radial (x) direction.
274
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 100
Model #1 End-cap holes
491
fAC=260.4 kHz
50 0 490
0.6
491‡ 491
492
493
492
494
100
Model #5 End-cap holes laser holes 50 1×4 mm fluorescence hole 0 490 100 Model #8 End-cap holes laser holes 50 2×4 mm fluorescence holes 0 490
493
495
496
497
498
499
500
498
499
500
498
499
500
491 0.6
492
491‡ 491
492
493
494
495
493
496
497
491 0.6 491‡ 491
492
493
494
495 m/z
492 496
493 497
FIGURE 9.13 Simulated mass spectra for rhodamine 101 using three model geometries: #1, #5, and #8. Note that the m/z scale is reconstructed from the recorded time of ion splat.
Table 9.3 A Summary of the Ion Ejection and ‘Detection’ Efficiency for Three Model Geometries: #1, #5, and #8. γa = π/3 Radians 10,000 Ions Rhodamine 101 Model #1 End-cap holes #5 1×4 mm Fluorescence hole #8 2×4 mm Fluorescence hole
Entrance End-cap
Exit End-cap
Detect
29% 28% 30%
50% 50% 48%
21% 22% 22%
Note: The data summarized in this table for Model #8 are depicted in Figure 9.12.
Approximately 70% of the ions are ejected toward the exit end-cap electrode. However, only 22% of the ions exit the trap and hit the detect plate. The majority of ions strike the exit end-cap electrode in the region immediately surrounding the central exit hole. The results from simulations using all three models are presented in Table 9.3. No significant differences in the simulated overall detection efficiencies were observed, suggesting that the insertion of holes in the ring electrode does not affect adversely detection efficiency. It is perhaps not surprising that the axial ejection of ions is relatively insensitive to changes in the radial geometry of the trap. The simulated mass spectra for rhodamine 101 are shown in Figure 9.13 for three model geometries, #1, #5, and #8. The comparison of the three simulated
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
275
mass spectra revealed little or no difference in mass resolution. In all cases, the monoisotopomer m/z 491 has a FWHM of 0.6 Th. Upon closer examination, ions are largely, though not exclusively, ejected in groups; these groups give rise to the jaggedness of the simulated peaks and correspond to cycles of the AC potential. For the monoisotopic peak the majority of the ions are ejected over 12 cycles of the AC or ca 46 µs. For all the models there is an ejection delay, causing the apparent m/z-value to be higher than the true m/z-value of the ion, and the magnitude of this difference increases with the number of holes in the ring electrode. Note that the m/z scale plotted in Figure 9.13 is merely a representation of the time that the ions are detected; the m/z-value is calibrated to the theoretical qz (of 0.785) at which ions ought to be ejected. Thus, the shift to higher apparent m/z-value is a natural consequence of the additional time required for ions to attain sufficient kinetic energy after achieving resonance with the auxiliary potential. This ejection delay has been observed previously and explained by Plass [96] (see below for additional discussion) and, in practice, is dealt with readily and empirically by mass calibration. Figure 9.13a shows the simulated mass spectrum using model #1, which exhibits an apparent m/z shift of + 3 Th. As holes are added to the ring electrode, the ions are ejected later, resulting in even higher apparent m/z-value for model #5 (+ 3.5 Th) and model #8 (+ 4 Th). The later ejection time for ions in traps with more holes is consistent with the shift toward lower axial secular frequency identified from single ion trajectories calculated under collision-free conditions in these traps (see Table 9.2). The results from these more sophisticated simulations suggest that this effect is still operative under collisional conditions. Overall, the shifts in secular frequency caused by modifying the ring electrode are expected to be of little consequence because the difference of only two or three thomsons is well within the calibration range of the instrument. Taken together, the results of these large-scale simulations do not suggest that there should be any significant losses of performance with the addition of multiple holes up to 4 mm in diameter. Furthermore, and somewhat to our surprise, symmetric placement of holes in the ring electrode appears to be neither necessary nor particularly advantageous based on the results of these simulations. We have since verified that there has been no noticeable loss of mass resolution or sensitivity for the Esquire 3000 + using a modified ring electrode that has 2 × 1.5 mm laser holes and a single 2 mm fluorescence hole. Mass calibration and tuning the phase between the drive frequency and the auxiliary AC were necessary to re-establish ion trap performance. Our experimental peak width (both with and without ring electrode holes) is about 0.3 Th for rhodamine 101, or about half that of the simulation. Possible reasons for this discrepancy include inaccuracies in the shapes of the model electrodes, insufficient cooling time in the model prior to ion ejection (only ca 2 ms was used to shorten the length of the simulation), or an incomplete description of the ion ejection waveform. It has not escaped our attention that there is a significant number of ions, ca 20% of the total number and identified as m/z 491, that are ‘pre-ejected’ before the main peak in the simulations. There are similar relative contributions of these pre-ejected ions for each isotopomer. For m/z 491, the pre-ejection peak is centered ca 2 Th lower
276
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
than the main peak and has a width of ca 1.5 Th. The explanation for the pre-ejection is highly suggestive of a manifestation in the simulations of the so-called chemical mass shift. The origin of the chemical mass shift and its role in directing the path for commercialization of the first QITs has been summarized previously [84]. The theory of chemical mass shifts has been addressed also thoroughly by Plass and coworkers [96], and we will discuss the theory only briefly here. The reader is directed elsewhere [96] for a detailed description that is beyond the scope of the present discussion. In general, the chemical mass shift describes the difference between the measured m/z-value and the true m/z-value for an ion in a mass spectrum, that is, ions may be ejected slightly sooner or later than they should otherwise be. According to the definition suggested by Plass et al., ions that are ejected early have a lower apparent m/z-value, such as those seen in Figure 9.13, and exhibit a positive chemical mass shift; correspondingly, ions that are ejected later have a higher apparent m/z-value, and exhibit a negative chemical mass shift. There are several mechanisms of the chemical mass shift, yet all are due principally to higher-order fields and to some extent, the nature of individual collision events. During the mass-selective instability scan, the ions’ secular frequencies are increasing ‘toward’ the applied AC frequency. For those ions that are near to the center of the trap, there is an ejection delay (mentioned above) corresponding to the time required for ions to absorb resonant energy from the auxiliary potential and their oscillation amplitude to increase. When the ion ensemble is well-cooled and near the center of the trap, the energy uptake amongst the ions will be relatively uniform resulting in sharper peaks and superior resolution in the mass spectra. The ejection process in our models is facilitated further by the exponential increase in the strength of the hexapole field (with increasing axial displacement) due to the non-linear resonance βz = 2/3 [87,91,92]. However, in the case where an ion already has a large axial displacement at the onset of resonance with the auxiliary potential, the exponential increase in kinetic energy occurs rapidly, with minimal delay, and the ions can be ejected ‘early’. Individual collision events are another possible source of chemical mass shift during a mass-selective instability scan. Based on results from simulations using ITSIM, Plass et al. [96] reported that elastic collisions causing a change in the ion’s oscillation amplitude can reduce the normal ejection delay, particularly when there are scattering events causing a sudden increase in axial displacement. The origin of the pre-ejection peaks is explored in Figure 9.14. Figure 9.14a shows the simulated mass spectrum for ions of m/z 491. A pre-ejection peak is clearly visible and appears to be composed of two overlapping groups of ions. The first group is centered near m/z 491.1 (indicating nearly zero ejection delay) while a second larger group is centered around m/z 491.8 (having a modest ejection delay). Figure 9.14b illustrates the correspondence between ions’ initial positions in the axial direction (z) and apparent m/z-value. Clearly, ions with initial axial positions within 0.5 mm of the center of the trap are not pre-ejected in the first group. This observation, which suggests that ions with larger axial displacements are more susceptible to rapid ejection at the onset of resonance, is consistent with similar observations described by
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers (a)
277
1
Relative abundance
m/z 491
m/z 491‡ Pre-ejection
490 (b)
491
492
491
492
493 m/z
494
495
493
494
495
2
zi (mm)
1 0 –1 –2 490
m/z
FIGURE 9.14 (a), Simulated mass spectrum of m/z 491 from rhodamine 101 striking the end-cap electrode using model geometry #1. A significant number of ions, ca 20% of the total number and identified as m/z 491‡, are ‘pre-ejected’ before the main peak and are observed at ca 2 Th lower than the main peak. (b), A plot of the apparent m/z-value with respect to ion initial axial position (zi) for the individual ions contributing to the simulated mass spectrum in panel (a).
Plass et al. [96] and earlier, by Franzen [91]. In contrast, we have found no similar relationship to initial conditions (xi, yi, zi, or Ek) that can distinguish the second preejection group. While none of the ions in the first pre-ejection group have small initial axial displacements, the converse is not true; many of the ions in the main peak have large initial axial displacements. At 1 mTorr, ions are subject to 8–10 collisions ms−1. Thus, those ions that are pre-ejected might have experienced fewer effective cooling collisions or a higher number of scattering collisions; in either scenario, the ion retains a larger axial oscillation and experiences a more intense hexapole field. A more realistic time-frame for the simulation would be required to account for statistical outliers of collision events, and it would be necessary to track each individual ion throughout the ejection process to determine the precise origin of the observed subpopulations. In practice, we do not observe pre-ejection peaks in the Esquire 3000 + instrument.
278
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
9.7.5.4 Spatial Distribution of the Ion Cloud Knowledge of the spatial distribution of the ion cloud is helpful when planning efficient photo-excitation and detection of fluorescence of trapped ions. A number of ion tomography studies have appeared in the literature providing estimates of the size of the ion cloud under various operational conditions [103–105]. Cooks and coworkers have investigated the size [103–105] and spatial oscillations [106,107] of the ion cloud by measuring PD efficiency as a function of laser beam position. They determined that the equilibrium and radially-averaged axial distribution of the cloud was roughly Gaussian (FWHM ca 0.8 mm) after a period of collisional cooling. Later investigations probed the radial distribution of the ion cloud (FWHM ca 1–1.2 mm) [105] using several ion trap geometries and as a function of total ion counts. More complex aspects of ion motion were modeled using ITSIM in conjunction with PD/DC pulse ion tomography experiments to study axial secular frequencies of ions [106] and to photodissociate selectively a particular ionic species in a mixture [107]. Estimating the size of the ion cloud using simulations is complex because the effects of collisional cooling must be taken into account. Indeed, it was recognized early on that collisions with buffer gas were of paramount importance for improving the performance of the first QITs [108]. For large numbers of ions (> 103) the effects of space charge become an important and possibly overriding factor determining the dimensions of the ion cloud. However, at this time it was not considered feasible to compute (simultaneously) the trajectories of such a large ensemble of ions. Instead, the trajectories of an ensemble of 50 ions under collisional conditions were recorded for a period of 10 ms for three different trap geometries: model #1 (no holes in the ring electrode), model #5 (2 × 2 mm laser holes + 1 × 4 mm fluorescence hole) and model #8 (2 × 2 mm laser holes + 2 × 4 mm fluorescence holes). Initial ion positions were generated by SIMION to create a 3-D Gaussian distribution (σ = 0.5 mm) that was centered radially and displaced axially −1 mm toward the entrance end-cap electrode. Initial kinetic energies were 0.1 eV with randomized velocity vectors and the ions’ TOB distributed evenly over one RF cycle. Trapping parameters for these simulations corresponded to the standard isolation conditions in the Esquire 3000 + at qz = 0.249 and ion trajectories were recorded at 1 µs intervals to maintain sufficient temporal resolution. Collisions were implemented using the HS1 hard-sphere collision model [98] for a background helium pressure of 1 mTorr. Ions suffer an average of 9.5 collisions ms−1 yielding an average mean free path of ca 2.5 mm. Each of the 10,000 time points for the fifty 10 ms-long 3-D ion trajectories (500,000 points in total) was projected into two dimensions and binned at a resolution of 0.01 mm to generate density plots from the perspective of the laser path (xz-plane), the fluorescence path (yz-plane), and the axial direction (xy-plane). The central portion of Figure 9.15 shows an ion density plot for Model #5 (asymmetric 1 × 4 mm fluorescence hole) projected in the xz-plane. Accompanying the central 2-D projection are almost identical single-dimension projections for both Model #5 (solid line) and Model #1 (no ring electrode holes, dashed line). The distributions on the left-hand side compare the relative density of the ion cloud in the axial (z)
279
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers 0.5
1.0
Rel abnd
Model #5 Model #1 FWHM= 1.3 mm
Rel abnd
2
0
z (mm)
FWHM= 0.6 mm
1
−1
−2
−1
0 x (mm)
1
−2 2
FIGURE 9.15 Spatial distribution of an ensemble of ions stored at qz = 0.249 under collisional conditions with 1 mTorr of helium buffer gas. The distribution represents the trajectories for 50 ions, tracked for 10 ms at 1 µs time steps, resulting in a total of 500,000 points. The central density plot illustrates the shape of the ion cloud computed for model #5 as may be viewed from the perspective of the laser beam (along the y-axis). The accompanying projections in single dimensions illustrate the Gaussian-like profiles of the axial and radial dispersions for model #5 (solid line) and model #1 (dashed line).
direction for models #5 and #1, while the curves on the top compare the relative ion densities along the radial x-direction for the two models. These simulations indicate that the spatial distribution of the ions is not altered significantly by drilling holes in the ring electrode. Of importance is that the ion cloud is still centered within the trap, despite the presence of a single hole for fluorescence collection in model #5. Also, it is clear that the randomizing effect of collisions negates any minor differences in starting conditions for even this relatively small ensemble of 50 ions. Because the ions are cooled rapidly to the center of the trap where the effects of the non-linear fields are weak and the quadrupolar potential dominates, over a longer time-scale it is not altogether surprising that there are no observed differences in the spatial distributions. The shape of the cloud is ellipsoidal (an oblate spheroid) and the singledimension projections indicate that the axial distribution has a FWHM = 0.6 mm and, in the radial x- and radial y-directions, the FWHM = 1.3 mm. Obviously, space charge, which has been neglected in our current implementation of the ion trajectory calculations, will play a significant role in limiting the ultimate number density of the ion cloud. However, the ion cloud dimensions calculated here agree reasonably
280
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
well with previous reports [105–107], and can be approximated as roughly 10–15% of the electrode dimensions. According to these simulations, an 800 µm-diameter laser beam traveling through the laser holes along the y-axis will intersect ca 36% of the ion cloud. Put another way, an ion will spend about 36% of its time within the path of an 800 µm-diameter laser beam. This interpretation and, indeed, our construction of an ion cloud of 500,000 points from 50 ion trajectories, assumes that collisions provide sufficient randomization to negate the effects of starting conditions. These results indicate the photo-excitation of ions can be quite efficient. This situation contrasts with the fluorescence collection efficiency. As discussed in Section 9.7.3, fluorescence is radiated in all directions so only a small portion of the ions (ca 0.25% in our set-up) that are excited optically will emit fluorescence in the direction of the fluorescence collection hole. Figure 9.16 shows the computed overlap between an 800 µm-diameter laser beam and the ion cloud as a function of qz. Increasing qz during ion storage yields a deeper potential well and decreases the size of the ion cloud, resulting in a better overlap between laser beam and ion cloud. At the highest qz value modeled, qz = 0.85, the overlap decreases due to expansion of the ion cloud near the boundary of the stability diagram. While storing ions at high qz maximizes photo-excitation, it is not usually required for PD because, in practice, 100% dissociation can be achieved frequently by irradiating ions trapped at a lower value of qz for an extended period of time as individual ions cycle in and out of the laser beam. Moreover, the increased low mass cut-off and potential for RF heating associated with high qz values may make it desirable to perform experiments at lower qz values. 1 0.9
Relative overlap
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
0.15
0.25
0.35
0.45
qz
0.55
0.65
0.75
0.85
FIGURE 9.16 Calculated fractional overlap between the ion cloud and an 800 µm-diameter laser beam as a function of qz of the stored ions. The solid circle (•) at qz = 0.68 corresponds to the qz value at which a dip is observed in the measured fluorescence signal (see Figure 9.17d).
281
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
9.7.6 Characterization of a QIT Modified for Optical Spectroscopic Experiments We have obtained strong fluorescence signals from ionic dyes and dye-labeled peptide ions confined in a 3-D QIT (see Figure 9.5) by optimization of appropriate experimental parameters. Figure 9.17 shows the effect of several experimental parameters on the fluorescence signal observed for rhodamine 590 ions (C27H29N2O3+, m/z 429.22), mass-selected and stored in a modified Bruker Esquire 3000 + QIT [15,81]. The set-up used for these experiments is described in Section 9.7.1 and is illustrated in Figures 9.6 and 9.7. The effect of laser power and pressure of buffer gas are illustrated in Figure 9.17a, which shows the dispersed fluorescence spectra measured as a function of laser power, and in Figure 9.17b, which shows the integrated fluorescence intensity as a function of laser power and trapping gas pressure. As the power from the frequency doubled output of the Ti:Sapphire laser is increased, the fluorescence signal first increases due to increased excitation of trapped ions, then decreases due to the PD of the ions into product ions which do not fluoresce as strongly as the parent rhodamine ion. Higher pressures result in increased fluorescence signals. In part, this behavior may be due to the smaller size of the ion cloud as pressure is
6
1.5
4
1.2 0.9
2
0.6
1
0.3 520
Integrated intensity
(c)
8
10 12 14 16
540
8.0
1 mW 2 mW 4 mW 6 mW 8 mW 10 mW 12 mW 14 mW 16 mW
560 580 600 Wavelength (nm)
620
1 2 3 ICC number
2.46×10–5 mbar 1.91×10–5 mbar 0
3
6 9 12 Laser power (mW)
×107
2
1
2.95×10–5 mbar
2.0
15
(d)
×106 PHe,ext.=3.0×10–5 mbar 6 q = 0.6 z 5 Normal MS/MS 4 operating conditions ICC~30,000 3
0
4.0
0.0
640
×108
6.0
Integrated intensity
Intensity (CPS)
1.8
(b)
×106 PHe,ext.=2.95×10–5 mbar
Integrated intensity
(a)
4
5x105
ICC 416 000
3.0 2.5
ICC 250 000
2.0 1.5 1.0 0.1
0.2
0.3
0.4
qz
0.5
0.6
0.7
0.8
FIGURE 9.17 Fluorescence intensity as a function of experimental parameters. (a) Dispersed emission as a function of laser power. (b) Integrated fluorescence intensity as a function of laser power measured at three different pressures of helium trapping gas. (c) Effect of the ion charge control (ICC) value, a relative measure of the number of detected ions, and (d) effect of qz on fluorescence intensity monitored with two different numbers of trapped ions.
282
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
increased (spatial cooling). The lower laser power threshold for the onset of PD at lower pressures suggests that the ions are cooled internally via collisions with the bath gas. Figure 9.17c illustrates the effect of ion number on the measured fluorescence signal. The number of ions stored in the trap is approximately proportional to the ion charge control (ICC) value in Bruker QITs.* Maximum fluorescence signal is obtained at very high ICC values, 20-fold or more higher than used in typical MS experiments. At these very high ICC values, mass resolution is lost due to space charge effects. In Figure 9.17d, the intensity of the fluorescence signal is plotted as a function of qz for two different ICC values. As predicted (see Figure 9.16), the measured fluorescence signal intensity increases as a function of qz. Interestingly, there is a dip in fluorescence signal intensity at qz = 0.68. We believe this is due possibly to the presence of a non-linear resonance at this qz value; however, a resonance was not apparent in our models at this value.
9.8 SUMMARY AND OUTLOOK The past decade has seen an explosion in investigations of molecular ions using a variety of optical spectroscopic techniques in conjunction with trapping mass spectrometers. The mass selection and ion storage capabilities of instruments such as 3-D QITs and FT-ICR mass spectrometers provide valuable control over the ion population under investigation. Moreover, thanks to modern ion sources, the size of molecules is no longer a limitation for gas-phase ion spectroscopic studies. A number of spectroscopic techniques have been developed to probe gas-phase molecules that will be fruitful when applied to the spectroscopy of trapped ions. The most popular technique for ion spectroscopy in trapping mass spectrometers is PD action spectroscopy, both in the IR and UV/visible ranges. The use of this technique circumvents the difficulties associated with photon collection and detection with sufficient signal to noise ratio. The coupling of trapping mass spectrometers to the powerful IR output from scannable FELs has enabled the measurement of IRMPD action spectra of a variety of molecules, providing a wealth of conformational information for smaller biomolecules. IRMPD of larger species, up to medium-sized proteins, has been accomplished also; however, structural interpretation for such large species is difficult because the large number of overlapping IR modes compromises spectral resolution and because frequency calculations at a reasonable level of theory are not currently possible for such large molecules. Thanks to advances in laser technology, a growing number of bench-top laser systems that are continuously tunable in the IR are becoming commercially available. With the adoption of this technology in MS laboratories, a better understanding of the intrinsic properties of small to medium sized biomolecules and small non-covalent clusters should emerge. UV/visible PD action spectroscopy is currently not as popular as IR action spectroscopy and, to date, has yielded relatively little structural information on * See Volume 4, Chapter 13: An Examination of the Physics of the High-Capacity Trap (HCT) by Andreas Brekenfeld, Ralf Hartmer, Desmond Kaplan, Carsten Baessmann, Jochen Franzen, and Michael Schubert.
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
283
species containing more than a few dozen atoms. One reason for this is that not all molecules absorb strongly in the UV/visible region; thus, PD in this range is not as universally applicable as in the IR. However, there is a multitude of interesting species that are well suited for this technique; proteins and peptides containing aromatic residues absorb strongly below 300 nm and many metal centers absorb strongly in the visible. Another reason underlying the relative lack of popularity is possibly the difficulty in extracting conformational information from UV/visible spectra. In principle, vibronic spectra contain within them a plethora of information including electronic excitation energies and excited state vibrational energy level spacings. However, extracting and interpreting this information in terms of structure is made difficult because electronic structure theory calculations do not yet routinely reproduce well the electronic excitation energies. Moreover, when the trapped ions are not cold or when the excited state lifetimes are too short, the spectra are broadened (as they are in solution) to the extent that few separate features are distinguishable (see for example Figure 9.4). The implementation of ion cooling in trapping mass spectrometers will resolve features blurred out due to temperature effects. This, in turn, will allow conformational selectivity via well-established spectroscopic methods such as ion dip, to enable elucidation of individual conformers. The use of ion dip spectroscopy to distinguish conformers present in a cooled 22-pole trap has been demonstrated already [33]. We anticipate many more investigations using this powerful combination of techniques. In addition to spectroscopic schemes for probing individual conformers, conformational selectivity may be achieved also by implementing a spatial separation step, such as a drift tube, a high-field asymmetric waveform, or traveling wave ion mobility separation, prior to spectroscopic interrogation. Photo-excitation of gas-phase ions may result in the photodetachment of an electron rather than photo-fragmentation. Coulombic considerations dictate that this process is more prevalent for anions than for cations. Electron photodetachment action spectroscopy of trapped anions has proved also to be a valuable source of molecular information. In some systems, electron photodetachment and PD compete. The mechanisms for these two processes in large molecules are yet to be understood fully; consequently, their branching ratios in specific experimental conditions cannot be predicted as yet. One exciting possibility is the idea of using frequency and phase-shaped pulses to promote selected photochemical pathways. Many established fluorescence techniques can be applied profitably to the study of trapped ions. These techniques include fluorescence steady-state excitation and emission spectroscopy as well as lifetime and fluorescence anisotropy measurements. Fluorescence has the advantage over most PD experiments in that it is insensitive to molecular size; while an increasing number of photons is required to dissociate larger and larger molecules (at least if they undergo IVR), such is not the case for fluorescence. An enticing prospect is the use of fluorescence energy transfer schemes to glean conformation and structural information on the 10–100 Å-length scale. For example, measurement of fluorescence energy transfer as a function of temperature provides a route to protein folding thermodynamics [10]. Likewise, monitoring FRET efficiency as a function
284
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
of solvent attachment should provide insight into how interactions with solvent affect protein conformation. Trapping MS experiments are inherently slow, taking milliseconds to seconds. The use of pulsed lasers brings the ability to probe events that occur on much shorter time scales than this, as even push-button femtosecond lasers are now available. Time-resolved spectroscopic experiments may be performed when a high enough signal to noise ratio is achieved. Such experiments could provide not only a series of snapshots showing how ion structure and conformation evolve during trapping or ion activation but also the ability to track ion populations as they move in a mass spectrometer.
ACKNOWLEDGMENTS The authors are grateful to Desmond Kaplan of Bruker Daltonics for helpful discussions and for financial support provided by the National Science and Engineering Research Council of Canada, the Canada Research Chairs Program, the American Society for Mass Spectrometry, and the Canadian Foundation for Innovation.
References
1. Dehmelt, H.G.; Jefferts, K.B. Alignment of the H2 + molecular ion by selective photodissociation. I. Phys. Rev. 1962, 125, 1318–1322. 2. Little, D.P.; Speir, J.P.; Senko, M.W.; O’Connor, P.B.; McLafferty, F.W. Infrared multiphoton dissociation of large multiply-charged ions for biomolecule sequencing. Anal. Chem. 1994, 66, 2809–2815. 3. Thompson, M.S.; Cui, W.D.; Reilly, J.P. Fragmentation of singly charged peptide ions by photodissociation at lambda = 157 nm. Angew. Chem.-Int. Ed. 2004, 43, 4791–4794. 4. Williams, E.R.; Furlong, J.J.P.; McLafferty, F.W. Efficiency of collisionally-activated dissociation and 193-nm photodissociation of peptide ions in Fourier-transform massspectrometry. J. Am. Soc. Mass Spectrom. 1990, 1, 288–294. 5. Dunbar, R.C. Kinetics of low-intensity infrared-laser photodissociation-the thermal-model and application of the Tolman theorem. J. Chem.Phys. 1991, 95, 2537–2548. 6. Freitas, M.A.; Hendrickson, C.L.; Marshall, A.G. Gas phase activation energy for unimolecular dissociation of biomolecular ions determined by Focused RAdiation for Gaseous Multiphoton ENergy Transfer (FRAGMENT). Rapid Commun. Mass Spectrom. 1999, 13, 1639–1642. 7. Jockusch, R.A.; Paech, K.; Williams, E.R. Energetics from slow infrared multiphoton dissociation of biomolecules. J. Phys. Chem. A 2000, 104, 3188–3196. 8. Wang, Y.; Hendrickson, C.L.; Marshall, A.G. Direct optical spectroscopy of gas-phase molecular ions trapped and mass-selected by ion cyclotron resonance: Laser-induced fluorescence excitation spectrum of hexafluorobenzene (C6F6 + ). J. Chem.Phys. Lett. 2001, 334, 69–75. 9. Iavarone, A.T.; Duft, D.; Parks, J.H. Shedding light on biomolecule conformational dynamics using fluorescence measurements of trapped ions. J. Phys. Chem. A 2006, 110, 12714–12727. 10. Iavarone, A.T.; Parks, J.H. Conformational change in unsolvated Trp-cage protein probed by fluorescence. J. Am. Chem. Soc. 2005, 127, 8606–8607.
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
285
11. Cage, B.; Friedrich, J.; Little, R.B.; Wang, Y.S.; McFarland, M.A.; Hendrickson, C.L.; Dalal, N.; Marshall, A.G. Wavelength resolved laser-induced fluorescence emission of C6F3H3 + trapped in an ion cyclotron resonance cell. Chem. Phys. Lett. 2004, 394, 188–193. 12. Cage, B.; McFarland, M.A.; Hendrickson, C.L.; Dalal, N.S.; Marshall, A.G. Resolution of individual component fluorescence lifetimes from a mixture of trapped ions by laser-induced fluorescence/ion cyclotron resonance. J. Phys. Chem. A 2002, 106, 10033–10036. 13. Danell, A.S.; Parks, J.H. FRET measurements of trapped oligonucleotide duplexes. Int. J. Mass Spectrom. 2003, 229, 35–45. 14. Dashtiev, M.; Azov, V.; Frankevich, V.; Scharfenberg, L.; Zenobi, R. Clear evidence of fluorescence resonance energy transfer in gas-phase ions. J. Am. Soc. Mass Spectrom. 2005, 16, 1481–1487. 15. Jockusch, R.A.; Bian, Q.; Talbot, F.O.; Forbes, M.W. Development and characterization of laser induced fluorescence spectroscopy coupled with ion trap mass spectrometry, Proc. 56th ASMS Conf. on Mass Spectrometry and Allied Topics, Denver, CO 2008. 16. Wright, K.C.; Blades, M.W. Fluorescence emission spectroscopy of trapped molecular ions, Proc. 51st ASMS Conf. on Mass Spectrometry and Allied Topics, Montreal, Canada 2003. 17. Stephenson Jr., J.L.; Yost, R.A. Photodissociation in the Ion Trap. In Practical Aspects of Ion Trap Mass Spectrometry: Volume II; March, R.E., Todd, J.F.J., Eds.; CRC Press: Boca Raton, FL, 1995, pp 163–203. 18. Huang, Y.L.; Jackson, G.; Kim, H.S.; Guan, S.H.; Marshall, A.G. Instrumental configuration for direct measurement of optical-absorption of ion-cyclotron resonance massselected trapped ions. Physica Scripta. 1995, T59, 387–391. 19. O’Keefe, A.; Deacon, D.A.G. Cavity ring-down optical spectrometer for absorptionmeasurements using pulsed laser sources. Rev. Sci. Instrum. 1988, 59, 2544–2551. 20. Berden, G.; Peeters, R.; Meijer, G. Cavity ring-down spectroscopy: Experimental schemes and applications. Int. Rev. Phys. Chem. 2000, 19, 565–607. 21. Lifshitz, C. Kinetic shifts. Eur. J. Mass Spectrom. 2002, 8, 85–98. 22. Joly, L.; Antoine, R.; Broyer, M.; Lemoine, J.; Dugourd, P. Electron photodetachment from gas phase peptide dianions. Relation with optical absorption properties. J. Phys. Chem. A 2008, 112, 898–903. 23. Antoine, R.; Joly, L.; Tabarin, T.; Broyer, M.; Dugourd, P.; Lemoine, J. Photo-induced formation of radical anion peptides. Electron photodetachment dissociation experiments. Rapid Commun. Mass Spectrom. 2007, 21, 265–268. 24. Cui, W.D.; Thompson, M.S.; Reilly, J.P. Pathways of peptide ion fragmentation induced by vacuum ultraviolet light. J. Am. Soc. Mass Spectrom. 2005, 16, 1384–1398. 25. Bush, M.F.; O’Brien, J.T.; Prell, J.S.; Saykally, R.J.; Williams, E.R. Infrared spectroscopy of cationized arginine in the gas phase: Direct evidence for the transition from nonzwitterionic to zwitterionic structure. J. Am. Chem. Soc. 2007, 129, 1612–1622. 26. Ayotte, P.; Weddle, G.H.; Kim, J.; Johnson, M.A. Mass-selected ‘matrix isolation’ infrared spectroscopy of the I-•(H2O)2 complex: Making and breaking the inter-water hydrogen-bond. Chem. Phys. 1998, 239, 485–491. 27. Okumura, M.; Yeh, L.I.; Myers, J.D.; Lee, Y.T. Infrared-spectra of the cluster ions H7O3 + .H2 and H9O4 + .H2. J. Chem. Phys. 1986, 85, 2328–2329. 28. Smalley, R.E.; Wharton, L.; Levy, D.H. Molecular optical spectroscopy with supersonic beams and jets. Acc. Chem. Res. 1977, 10, 139–145. 29. Lee, S.W.; Freivogel, P.; Schindler, T.; Beauchamp, J.L. Freeze-dried biomolecules: FT-ICR studies of the specific solvation of functional groups and clathrate formation observed by the slow evaporation of water from hydrated peptides and model compounds in the gas phase. J. Am. Chem. Soc. 1998, 120, 11758–11765.
286
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
30. Schindler, T.; Berg, C.; Niednerschatteburg, G.; Bondybey, V.E. Solvation of hydrochloric-acid in protonated water clusters. Chem. Phys. Lett. 1994, 229, 57–64. 31. Wong, R.L.; Paech, K.; Williams, E.R. Blackbody infrared radiative dissociation at low temperature: Hydration of X2 + (H2O)n for X = Mg, Ca. Int. J. Mass Spectrom. 2004, 232, 59–66. 32. Jockusch, R.A.; Lemoff, A.S.; Williams, E.R. Hydration of valine-cation complexes in the gas phase: On the number of water molecules necessary to form a zwitterion. J. Phys. Chem. A 2001, 105, 10929–10942. 33. Guo, X.H.; Duursma, M.; Al-Khalili, A.; McDonnell, L.A.; Heeren, R.M.A. Design and performance of a new FT-ICR cell operating at a temperature range of 77-438 K. Int. J. Mass Spectrom. 2004, 231, 37–45. 34. Gerlich, D.; Horning, S. Experimental investigations of radiative association processes as related to interstellar chemistry. Chem. Rev. 1992, 92, 1509–1539. 35. Stearns, J.A.; Boyarkin, O.V.; Rizzo, T.R. Spectroscopic signatures of gas-phase helices: Ac-Phe-(Ala)(5)-Lys-H + and Ac-Phe-(Ala)(10)-Lys-H + . J. Am. Chem. Soc. 2007, 129, 13820–13821. 36. Wang, X.B.; Wang, L.S. Development of a low-temperature photoelectron spectroscopy instrument using an electrospray ion source and a cryogenically controlled ion trap. Rev. Sci. Instrum. 2008, 79, 173108. 37. Beauchamp, J.L.; Thorne, L.R. Infrared Photochemistry of Gas Phase Ions. In Gas Phase Ion Chemistry Volume 3; Bowers, M.T., Ed.; Academic Press, Inc.: London, 1984, pp. 42–97. 38. Oomens, J.; Sartakov, B.G.; Meijer, G.; Von Helden, G. Gas-phase infrared multiple photon dissociation spectroscopy of mass-selected molecular ions. Int. J. Mass Spectrom. 2006, 254, 1–19. 39. Price, W.D.; Williams, E.R. Activation of peptide ions by blackbody radiation: Factors that lead to dissociation kinetics in the rapid energy exchange limit. J. Phys. Chem. A 1997, 101, 8844–8852. 40. McLuckey, S.A.; Goeringer, D.E. Slow heating methods in tandem mass spectrometry. J. Mass Spectrom. 1997, 32, 461–474. 41. Payne, A.H.; Glish, G.L. Thermally assisted infrared multiphoton photodissociation in a quadrupole ion trap. Anal. Chem. 2001, 73, 3542–3548. 42. Colorado, A.; Shen, J.X.X.; Vartanian, V.H.; Brodbelt, J. Use of infrared multiphoton photodissociation with SWIFT for electrospray ionization and laser desorption applications in a quadrupole ion trap mass spectrometer. Anal. Chem. 1996, 68, 4033–4043. 43. Stephenson, J.L.; Booth, M.M.; Shalosky, J.A.; Eyler, J.R.; Yost, R.A. Infrared multiplephoton dissociation in the quadrupole ion-trap via a multipass optical arrangement. J. Am. Soc. Mass Spectrom. 1994, 5, 886–893. 44. Hughes, R.J.; March, R.E.; Young, A.B. Multi-photon dissociation of ions derived from 2-propanol in a QUISTOR with low-power cw infrared-laser radiation. Int. J. Mass Spectrom. Ion Processes. 1982, 42, 255–263. 45. Kapota, C.; Lemaire, J.; Maitre, P.; Ohanessian, G. Vibrational signature of charge solvation vs salt bridge isomers of sodiated amino acids in the gas phase. J. Am. Chem. Soc. 2004, 126, 1836–1842. 46. Bush, M.F.; Forbes, M.W.; Jockusch, R.A.; Oomens, J.; Polfer, N.C.; Saykally, R.J.; Williams, E.R. Infrared spectroscopy of cationized lysine and epsilon-N-methyllysine in the gas phase: Effects of alkali-metal ion size and proton affinity on zwitterion stability. J. Phys. Chem. A 2007, 111, 7753–7760. 47. Forbes, M.W.; Bush, M.F.; Polfer, N.C.; Oomens, J.; Dunbar, R.C.; Williams, E.R.; Jockusch, R.A. Infrared spectroscopy of arginine cation complexes: Direct observation of gas-phase zwitterions. J. Phys. Chem. A 2007, 111, 11759–11770.
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
287
48. Bush, M.F.; Oomens, J.; Saykally, R.J.; Williams, E.R. Effects of alkaline earth metal ion complexation on amino acid zwitterion stability: Results from infrared action spectroscopy. J. Am. Chem. Soc. 2008, 130, 6463–6471. 49. Armentrout, P.B.; Rodgers, M.T.; Oomens, J.; Steill, J.D. Infrared multiphoton dissociation spectroscopy of cationized serine: Effects of alkali-metal cation size on gas-phase conformation. J. Phys. Chem. A 2008, 112, 2248–2257. 50. Rodgers, M.T.; Armentrout, P.B.; Oomens, J.; Steiw, J.D. Infrared multiphoton dissociation spectroscopy of cationized threonine: Effects of alkali-metal cation size on gasphase conformation. J. Phys. Chem. A 2008, 112, 2258–2267. 51. Dunbar, R.C.; Polfer, N.C.; Oomens, J. Gas-phase zwitterion stabilization by a metal dication. J. Am. Chem. Soc. 2007, 129, 14562–14563. 52. Polfer, N.C.; Oomens, J.; Dunbar, R.C. IRMPD spectroscopy of metal-ion/tryptophan complexes. Phys. Chem. Chem.Phys. 2006, 8, 2744–2751. 53. Polfer, N.C.; Oomens, J.; Moore, D.T.; von Helden, G.; Meijer, G.; Dunbar, R.C. Infrared spectroscopy of phenylalanine Ag(I) and Zn(II) complexes in the gas phase. J. Am. Chem. Soc. 2006, 128, 517–525. 54. Correia, C.F.; Balaj, P.O.; Scuderi, D.; Maitre, P.; Ohanessian, G. Vibrational signatures of protonated, phosphorylated amino acids in the gas phase. J. Am. Chem. Soc. 2008, 130, 3359–3370. 55. Simon, A.; MacAleese, L.; Maitre, P.; Lemaire, J.; McMahon, T.B. Fingerprint vibrational spectra of protonated methyl esters of amino acids in the gas phase. J. Am. Chem. Soc. 2007, 129, 2829–2840. 56. MacAleese, L.; Simon, A.; McMahon, T.B.; Ortega, J.M.; Scuderi, D.; Lemaire, J.; Maitre, P. Mid-IR spectroscopy of protonated leucine methyl ester performed with an FTICR or a Paul type ion-trap. Int. J. Mass Spectrom. 2006, 249, 14–20. 57. Lucas, B.; Gregoire, G.; Lemaire, J.; Maitre, P.; Glotin, F.; Schermann, J.P.; Desfrancois, C. Infrared multiphoton dissociation spectroscopy of protonated N-acetyl-alanine and alanyl-histidine. Int. J. Mass Spectrom. 2005, 243, 105–113. 58. Fridgen, T.D.; MacAleese, L.; Maitre, P.; McMahon, T.B.; Boissel, P.; Lemaire, J. Infrared spectra of homogeneous and heterogeneous proton-bound dimers in the gas phase. Phys. Chem. Chem. Phys. 2005, 7, 2747–2755. 59. Oh, H.B.; Lin, C.; Hwang, H.Y.; Zhai, H.L.; Breuker, K.; Zabrouskov, V.; Carpenter, B.K.; McLafferty, F.W. Infrared photodissociation spectroscopy of electrosprayed ions in a Fourier transform mass spectrometer. J. Am. Chem. Soc. 2005, 127, 4076–4083. 60. Polfer, N.C.; Oomens, J.; Dunbar, R.C. Alkali metal complexes of the dipeptides PheAla and AlaPhe: IRMPD spectroscopy. Chemphyschem. 2008, 9, 579–589. 61. Polfer, N.C.; Oomens, J.; Suhai, S.; Paizs, B. Infrared spectroscopy and theoretical studies on gas-phase protonated leu-enkephalin and its fragments: Direct experimental evidence for the mobile proton. J. Am. Chem. Soc. 2007, 129, 5887–5897. 62. Polfer, N.C.; Paizs, B.; Snoek, L.C.; Compagnon, I.; Suhai, S.; Meijer, G.; von Helden, G.; Oomens, J. Infrared fingerprint spectroscopy and theoretical studies of potassium ion tagged amino acids and peptides in the gas phase. J. Am. Chem. Soc. 2005, 127, 8571–8579. 63. Lucas, B.; Gregoire, G.; Lemaire, J.; Maitre, P.; Ortega, J. M.; Rupenyan, A.; Reimann, B.; Schermann, J.P.; Desfrancois, C. Investigation of the protonation site in the dialanine peptide by infrared multiphoton dissociation spectroscopy. Phys. Chem. Chem. Phys. 2004, 6, 2659–2663. 64. Fukui, K.; Takada, Y.; Sumiyoshi, T.; Imai, T.; Takahashi, K. Infrared multiphoton dissociation spectroscopic analysis of peptides and oligosaccharides by using Fourier transform ion cyclotron resonance mass spectrometry with a midinfrared free-electron laser. J. Phys. Chem. B. 2006, 110, 16111–16116.
288
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
65. Salpin, J.Y.; Guillaumont, S.; Tortajada, J.; MacAleese, L.; Lemaire, J.; Maitre, P. Infrared spectra of protonated uracil, thymine and cytosine. Chemphyschem. 2007, 8, 2235–2244. 66. Polfer, N.C.; Oomens, J.; Suhai, S.; Paizs, B. Spectroscopic and theoretical evidence for oxazolone ring formation in collision-induced dissociation of peptides. J. Am. Chem. Soc. 2005, 127, 17154–17155. 67. Frison, G.; van der Rest, G.; Turecek, F.; Besson, T.; Lemaire, J.; Maitre, P.; ChamotRooke, J. Structure of electron-capture dissociation fragments from charge-tagged peptides probed by tunable infrared multiple photon dissociation. J. Am. Chem. Soc. 2008, 130, 14916–14917. 68. Oomens, J.; Polfer, N.; Moore, D.T.; van der Meer, L.; Marshall, A.G.; Eyler, J.R.; Meijer, G.; von Helden, G. Charge-state resolved mid-infrared spectroscopy of a gasphase protein. Phys. Chem. Chem. Phys. 2005, 7, 1345–1348. 69. Valle, J.J.; Eyler, J.R.; Oomens, J.; Moore, D.T.; van der Meer, A.F.G.; von Helden, G.; Meijer, G.; Hendrickson, C.L.; Marshall, A.G.; Blakney, G.T. Free electron laser-Fourier transform ion cyclotron resonance mass spectrometry facility for obtaining infrared multiphoton dissociation spectra of gaseous ions. Rev. Sci. Instrum. 2005, 76. 023103. 70. Lemaire, J.; Boissel, P.; Heninger, M.; Mauclaire, G.; Bellec, G.; Mestdagh, H.; Simon, A.; Caer, S.L.; Ortega, J.M.; Glotin, F.; Maitre, P. Gas phase infrared spectroscopy of selectively prepared ions. Phys. Rev. Lett. 2002, 89, 273002–273004. 71. Kong, X.L.; Tsai, I.A.; Sabu, S.; Han, C.C.; Lee, Y.T.; Chang, H.C.; Tu, S.Y.; Kung, A.H.; Wu, C.C. Progressive stabilization of zwitterionic structures in [H(Ser)(2-8)]( + ) studied by infrared photodissociation spectroscopy. Angew. Chem.-Int. Ed. 2006, 45, 4130–4134. 72. Atkins, C.G.; Rajabi, K.; Gillis, E.A.L.; Fridgen, T.D. Infrared multiple photon dissociation spectra of proton- and sodium ion-bound glycine dimers in the N-H and O-H stretching region. J. Phys. Chem. A. 2008, 112, 10220–10225. 73. Asvany, O.; Kumar, P.; Redlich, B.; Hegemann, I.; Schlemmer, S.; Marx, D. Understanding the infrared spectrum of bare CH5 + . Science 2005, 309, 1219–1222. 74. Gregoire, G.; Gaigeot, M.P.; Marinica, D.C.; Lemaire, J.; Schermann, J.P.; Desfrancois, C. Resonant infrared multiphoton dissociation spectroscopy of gas-phase protonated peptides. Experiments and Car-Parrinello dynamics at 300 K. Phys. Chem. Chem.Phys. 2007, 9, 3082–3097. 75. Marinica, D.C.; Gregoire, G.; Desfrancois, C.; Schermann, J.P.; Borgis, D.; Gaigeot, M.P. Ab initio molecular dynamics of protonated dialanine and comparison to infrared multiphoton dissociation experiments. J. Phys. Chem. A 2006, 110, 8802–8810. 76. NIST Computational Chemistry Comparison and Benchmark Database, NIST Standard Reference Database Number 101, section XIII.B.4a: Precomputed vibrational scaling factors, Ed. Russell D. Johnson III. http://cccbdb.nist.gov/ 77. McQueen, P.D.; Jockusch, R.A., Visible action spectroscopy and fluorescence of gasphase fluorescein. In preparation. 78. Li, G.Z.; Vining, V.A.; Guan, S.H.; Marshall, A.G. Laser-induced fluorescence of Ba + ions trapped and mass-selected in a Fourier transform ion cyclotron resonance mass spectrometer. Rapid Commun. Mass Spectrom 1996, 10, 1850–1854. 79. Förster, T. 10th Spiers Memorial Lecture-Transfer mechanisms of electronic excitation. Disc. Farad. Soc. 1959, 27, 7–17. 80. Selvin, P.R. The renaissance of fluorescence resonance energy transfer. Nature Structural Biology 2000, 7, 730–734. 81. Bian, Q.; Forbes, M.W.; Talbot, F.O.; Jockusch, R.A. An instrument for fluorescence excitation and emission spectroscopy of trapped, mass-selected gas-phase ions. In preparation.
The Spectroscopy of Ions Stored in Trapping Mass Spectrometers
289
82. Magde, D.; Wong, R.; Seybold, P.G. Fluorescence quantum yields and their relation to lifetimes of rhodamine 6G and fluorescein in nine solvents: Improved absolute standards for quantum yields. Photochem. Photobiol. 2002, 75, 327–334. 83. Hinckley, D.A.; Seybold, P.G.; Borris, D.P. Solvatochromism and thermochromism of rhodamine solutions. Spectrochimica Acta Part A-Molecular and Biomolecular Spectroscopy 1986, 42, 747–754. 84. Todd, J.F.J. Introduction to Practical Aspects of Ion Trap Mass Spectrometry. In Practical Aspects of Ion Trap Mass Spectrometry: Volume I, Chapter 1; March, R.E., Todd, J.F.J., Eds.; CRC Press: Boca Raton, FL, 1995, pp. 3–24. 85. von Busch, F.; Paul, W. Über nichtlineare Resonanzen im elektrischen Massenfilter als Folge von Feldfehlern. Z. Phys. 1961, 164, 588–594. 86. Franzen, J.; Gabling, R-H.; Schubert, M.; Wang, Y. Nonlinear Ion Traps. In Practical Aspects of Ion Trap Mass Spectrometry: Volume I, Chapter 3; March, R.E., and Todd, J.F.J., Eds.; CRC Press: Boca Raton, FL, 1995, pp. 49–167. 87. Franzen, J. Simulation study of an ion cage with superimposed multipole fields. Int. J. Mass Spectrom. Ion Processes 1991, 106, 63–78. 88. Wang, Y.; Franzen, J. The nonlinear resonance Quistor 1. Potential distribution in hyperboloidal Quistors. Int. J. Mass Spectrom. Ion Processes 1992, 112, 167–178. 89. Wang, Y.; Franzen, J.; Wanczek, K.P. The nonlinear resonance ion trap 2. A general theoretical analysis. Int. J. Mass Spectrom. Ion Processes 1993, 124, 125–144. 90. Wang, Y.; Franzen, J. The nonlinear ion-trap 3. Multipole components in 3 types of practical ion-trap. Int. J. Mass Spectrom. Ion Processes 1994, 132, 155–172. 91. Franzen, J. The nonlinear ion-trap 4. Mass-selective instability scan with multipole superposition. Int. J. Mass Spectrom. Ion Processes 1993, 125, 165–170. 92. Franzen, J. The nonlinear ion-trap 5. Nature of nonlinear resonances and resonant ion ejection. Int. J. Mass Spectrom. Ion Processes 1994, 130, 15–40. 93. March, R.E.; Todd, J.F.J. Quadrupole Ion Trap Mass Spectrometry 2nd Edition; John Wiley & Sons: Toronto, 2005. 94. Guidugli, F.; Traldi, P. A phenomenological description of a black-hole for collisionally induced decomposition products in ion-trap mass-spectrometry. Rapid Commun. Mass Spectrom. 1991, 5, 343–348. 95. Guidugli, F.; Traldi, P.; Franklin, A.M.; Langford, M.L.; Murrell, J.; Todd, J.F.J. Further thoughts on the occurrence of black-holes in ion-trap mass-spectrometry. Rapid Commun. Mass Spectrom. 1992, 6, 229–231. 96. Plass, W.R.; Li, H.; Cooks, R.G. Theory, simulation and measurement of chemical mass shifts in RF quadrupole ion traps. Int. J. Mass Spectrom. 2003, 228, 237–267. 97. Manura, D.; Dahl, D.A.; SIMION Manual, Scientific Instruments Services, Inc.: Ringoes, NJ, 2006. 98. www.simion.com. Accessed 1 July 2007. 99. Appelhans, A.D.; Dahl, D.A. Measurement of external ion injection and trapping efficiency in the ion trap mass spectrometer and comparison with a predictive model. Int. J. Mass Spectrom. 2002, 216, 269–284. 100. Forbes, M.W.; Sharifi, M.; Croley, T.; Lausevic, Z.; March, R.E. Simulation of ion trajectories in a quadrupole ion trap: A comparison of three simulation programs. J. Mass Spectrom. 1999, 34, 1219–1239. 101. Doroshenko, V.M.; Cotter, R.J. Injection of externally generated ions into an increasing trapping field of a quadrupole ion trap mass spectrometer. J. Mass Spectrom. 1997, 32, 602–615. 102. Wu, G.; Cooks, R.G.; Ouyang, Z.; Yu, M.; Chappell, W.J.; Plass, W.R. Ion trajectory simulation for electrode configurations with arbitrary geometries. J. Am. Soc. Mass Spectrom. 2006, 17, 1216–1228.
290
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
103. Hemberger, P.H.; Nogar, N.S.; Williams, J.D.; Cooks, R.G.; Syka, J.E.P. Laser photodissociation probe for ion tomography studies in a quadrupole ion-trap mass-spectrometer. Chem. Phys. Lett. 1992, 191, 405–410. 104. Williams, J.D.; Cooks, R.G.; Syka, J.E.P.; Hemberger, P.H.; Nogar, N.S. Determination of positions, velocities, and kinetic energies of resonantly excited ions in the quadrupole ion-trap mass-spectrometer by laser photodissociation. J. Am. Soc. Mass Spectrom. 1993, 4, 792–797. 105. Cleven, C.D.; Cooks, R.G.; Garrett, A.W.; Nogar, N.S.; Hemberger, P.H. Radial distributions and ejection times of molecular ions in an ion trap mass spectrometer: A laser tomography study of effects of ion density and molecular type. J. Phys. Chem. 1996, 100, 40–46. 106. Lammert, S.A.; Cleven, C.D.; Cooks, R.G. Determination of ion frequencies in a quadrupole ion-trap by using a fast direct-current pulse as pump and a laser probe. J. Am. Soc. Mass Spectrom. 1994, 5, 29–36. 107. Cleven, C.D.; Nappi, M.; Cooks, R.G.; Garrett, A.W.; Nogar, N.S.; Hemberger, P.H. Selective photodissociation of trapped ions after ion cloud manipulation with an impulsive quadrupolar electric field. J. Phys. Chem. 1996, 100, 5205–5209. 108. Stafford, G.C.; Kelley, P.E.; Syka, J.E.P.; Reynolds, W.E.; Todd, J.F.J. Recent improvements in and analytical applications of advanced ion trap technology. Int. J. Mass Spectrom. Ion Processes 1984, 60, 85–98.
10 Sympathetically-Cooled Single Ion Mass Spectrometry Peter Frøhlich Staanum, Klaus Højbjerre, and Michael Drewsen Contents 10.1 Introduction.................................................................................................. 292 10.2 Single-Ion Mass Measurements................................................................... 293 10.2.1 Common Single-Ion Mass Spectrometry Techniques................... 293 10.2.2 The SCSI-MS Technique............................................................... 293 10.2.3 Advantages and Disadvantages of the SCSI-MS Technique......... 295 10.3 Experimental Realization............................................................................ 296 10.3.1 The Linear RF Ion Trap................................................................. 296 10.3.2 Laser-Cooling of Atomic Ions....................................................... 299 10.3.2.1 Loading of Atomic Ions................................................ 299 10.3.2.2 Cooling of the Atomic Ions........................................... 299 10.3.2.3 Fluorescence Detection.................................................300 10.3.3 Inducing Forced Motion of the Ions..............................................300 10.3.3.1 Electrical Forces........................................................... 301 10.3.3.2 Radiation Pressure Forces............................................. 301 10.3.4 Detection of the Motion of the Ions...............................................302 10.3.4.1 Amplitude Detection.....................................................304 10.3.4.2 Phase Detection............................................................ 305 10.4 Experimental Results................................................................................... 305 10.4.1 Experiments with Atomic Ions......................................................307 10.4.1.1 Two Calcium Ions.........................................................307 10.4.1.2 One Calcium and One Magnesium Ion........................307 10.4.1.3 Precision Mass Measurements of Calcium Isotopes....309 10.4.2 Experiments with Molecular Ions.................................................. 310 10.4.2.1 CaO + Ions..................................................................... 310 10.4.2.2 MgH + and MgD + Ions.................................................. 310 10.4.2.3 The Aniline Ion and its Photofragments....................... 311 10.5 Accuracy of the SCSI-MS Technique.......................................................... 312 10.5.1 Laser-Cooling Force...................................................................... 312 10.5.2 Non-Linearity in the Coulomb Interaction.................................... 313 291
292
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
10.5.3 Ion-Trap Imperfections.................................................................. 315 10.5.3.1 Anharmonicity.............................................................. 315 10.5.3.2 Trapping Field Imperfections........................................ 316 10.5.3.3 Residual Magnetic Fields.............................................. 317 10.5.4 Ion Loading.................................................................................... 317 10.5.5 Background Gas............................................................................. 319 10.5.6 Photon Detection............................................................................ 319 10.5.6.1 Spatial Resolution......................................................... 319 10.5.6.2 Scattering Rate and Collection Efficiency.................... 320 10.5.7 Reference Measurements............................................................... 321 10.5.8 Choice of Parameters..................................................................... 323 10.6 Conclusion and Outlook.............................................................................. 323 Acknowledgment.................................................................................................... 324 References............................................................................................................... 324
10.1 INTRODUCTION Many mass spectrometry methods have been demonstrated, refined and applied since the first technique was introduced nearly a century ago [1]. As is documented, for example, by the previous contributions to the present series of books Practical Aspects of Ion Trap Mass Spectrometry [2], several of these techniques involve ion trapping. An obvious question to ask at this juncture is: do we need still more methods? We believe that the answer to this question is ‘yes’, and we hope that we are able to present a good case through this contribution to the current volume. New mass spectrometric methods emerge typically when a higher precision in measurements is needed and when new experimental approaches become feasible. It is with respect to the latter circumstance that the present contribution should be viewed. Recently, we have been interested in exploring the possibilities of carrying out experiments with a single trapped molecular ion cooled sympathetically by a single laser-cooled atomic ion trapped simultaneously. The prospects for such single molecular ion experiments are manifold. First, this situation (to be described in more detail below) makes it possible to perform experiments with molecules one at the time, such that ensemble averaging is obviated. Second, due to the sympathetic cooling, the molecular ion can be translationally very cold as well as spatially localized. Third, under ultra-high vacuum (UHV) conditions, experiments can be carried out in the absence of perturbations of the internal state of the molecules. Consequently, the above features make such single molecular ions ideal targets for a wealth of investigations including coherent manipulations of single molecular ions by laser light. For the proposed investigations, the objective was to develop a simple in situ and non-destructive mass measurement method for the identification of a single trapped molecular ion. In Section 10.2 is discussed the principal idea of our novel SCSI-MS (Sympathetically-Cooled Single Ion Mass Spectrometry) technique for performing single-ion mass measurements in the light of already well-established techniques. This discussion is followed by both a description of the current experimental arrangement, used for single molecular ion experiments (Section 10.3), and an account of the
Sympathetically-Cooled Single Ion Mass Spectrometry
293
present status of experiments with single molecules (Section 10.4). Thereafter, various issues that may limit the potential of this method are examined and the effects of these issues in practical applications (Section 10.5) are evaluated. This chapter concludes with a brief survey of experiments that may be facilitated through exploitation of this technique in the near future (Section 10.6).
10.2 SINGLE-ION MASS MEASUREMENTS In this section, a brief discussion is presented of existing techniques for performing single-ion mass measurements. This discussion is followed by a description of the nondestructive SCSI-MS technique that has been developed recently in this laboratory [3], and a comparison of the SCSI-MS technique with methods established previously.
10.2.1 Common Single-Ion Mass Spectrometry Techniques Detection of a single ion is generally fairly easy, because the direct detection of a particle with a single elementary charge can be achieved with close to 100% efficiency, provided the particle is accelerated to an energy of a few electron-volts. This feature is used, for example, in more traditional mass spectrometers such as the quadrupole mass filter (QMF) [4,5] and the time-of-flight (TOF) instrument [4,6], where the ion in each case hits eventually an electron multiplier detector (for example, a channeltron detector). In the more recent technique of Fourier Transform Ion Cyclotron Resonance (FT-ICR) mass spectrometry [7], single ion sensitivity can again be obtained, but this time through a non-destructive measurement of cyclotron motion induced image charges in pick-up plates. While FT-ICR can be used to reach formidable accuracies in mass measurements through high precision frequency measurements (relative mass resolution [8] m/∆m = 1011 compared to 102–106 for QMF/ TOF measurements [5,6], where m is the mass of interest and ∆m the uncertainty with which it can be determined), it is limited in its ability to make spatial and timeresolved measurements as compared to the QMF and TOF techniques. This ability to make spatial and time-resolved measurements means that the QMF and TOF mass measurement techniques are employed usually in fast dynamic experiments where the mass of the particles of interest may change due to reactions or photofragmentation, while FT-ICR is applied often in ‘static’ precision measurements and slower dynamics situations.
10.2.2 The SCSI-MS Technique A schematic of the SCSI-MS technique is presented in Figure 10.1. The technique relies on the measurement of the resonant excitation frequency of one of the two oscillatory modes of a trapped and crystallized linear two-ion system consisting of one laser-cooled atomic ion of known mass and the a priori unknown atomic or molecular ion, whose mass is to be determined. From this measured frequency, the mass of the unknown ion can be deduced from a simple relation between the frequency and the relative masses of the two ions (see Section 10.3).
294
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
CCD x
Shutter
z y
Trap
EOC Laser beam
Laser beam
Volt. mod.
Laser beam
Light mod.
FIGURE 10.1 Sketch of the SCSI-MS technique. Shown in the figure are the linear RF ion-trap electrodes, the cooling laser beams and the CCD camera used to monitor the fluorescence from the laser-cooled atomic ion. An image-intensifier-based shutter that can be gated phase-locked to a periodic driving force is installed in front of the camera. A driving force, for excitation of the ion motion along the z-axis, is applied either in the form of a sinusoidallyvarying voltage applied to the end-cap electrodes of the ion trap or through modulation of the scattering force on the atomic ion by using an EOC for intensity modulation. (Reproduced from Drewsen, M.; Mortensen, A.; Martinussen, R.; Staanum, P.; SØrensen, J.L., Phys. Rev. Lett. 2004, 93, 243201. With permission from the American Physical Society.)
The crystallization of the two-ion system results from the sympathetic cooling of the unknown ion through the Coulomb interaction with the laser-cooled ion. This crystallization can be observed by imaging the fluorescence light emitted by the lasercooled ion onto a charge-coupled device (CCD) camera chip. Here, a well-localized spot appears with the atomic ion displaced a specific distance away from the iontrap center when it is trapped together with a non-fluorescing unknown ion, see, for example, Figure 10.4b. In the linear RF trap used in these experiments (see Section 10.3), the two-ion system is aligned along the trap’s main axis (the z-axis in Figure 10.1). The resonant excitation is promoted either by applying a sinusoidally-varying electric field along this axis (through sinusoidally-varying voltages applied to the end electrodes of the ion trap) that will exert a force on both ions, or by modulating periodically the laser intensity of at least one of the cooling laser beams propagating along the main axis; this latter action leads to a periodically-varying scattering force on the laser-cooled ion. The resonance frequencies are determined by monitoring the fluorescence light from the laser-cooled ion by the CCD camera while scanning the period of the applied driving force. When the period is equal to the period of one of the two oscillatory modes of the two-ion system (that is, the center-of-mass (COM) mode, where the ions move in phase, or the breathing (BR) mode, where the ions move with opposite phase [9,10]), the motion of the ions is most highly excited (neglecting damping exerted by the cooling lasers). For CCD camera exposure times larger than
Sympathetically-Cooled Single Ion Mass Spectrometry
295
the oscillation period of the ions, an enlarged axial extension of the fluorescence spot is observed close to these mode resonance frequencies. The simplest way to obtain an estimate for the mode frequency, and hence the mass of the unknown ion (see Section 10.3), is to look for a maximum amplitude in the motion of the laser-cooled ion while changing the period of the force. This detection method leads easily to a relative mass resolution below 102, and can lead, for optimized conditions, to a resolution at the 104 level. However, when the phase of the motion of the laser-cooled ion is monitored over a longer period, a more precise measurement is expected (see Section 10.5.6.2). The phase of the motion is monitored by gating the CCD camera such that light emitted only at a certain phase of the motion with respect to the phase of the driving force is detected (see Section 10.3.4). Such measurements have been shown [3] to lead to relative mass resolutions, m/∆m, of about 3 × 104. It is worth mentioning here that both detection methods are non-destructive, and that the ion of interest (the unknown ion) is translationally cold and extremely welllocalized spatially both before and after each a mass measurement. This feature of the measurements is highly relevant to the successful achievement of several experiments proposed for the future (see Section 10.6).
10.2.3 Advantages and Disadvantages of the SCSI-MS Technique It is interesting first to note that in contrast to standard mass measurement techniques, a strong Coulomb coupling between ions (the ion of interest and the laser-cooled atomic ion) is essential, rather than being problematical. However, as is discussed in Section 12.5, the non-linear nature of the Coulomb interaction between the two ions can lead to unwanted systematic errors in the mass measurements. While the relative mass resolution of ca 104 of the SCSI-MS technique can compete very well with standard QMF and TOF mass spectrometers, the mass accuracy of the SCSI-MS technique will probably never be able to challenge that of FT-ICR mass spectrometry; nevertheless, this shortcoming does not mean that the new technique is irrelevant. On the contrary, the SCSI-MS technique encompasses the advantages of high spatial and temporal resolution from QMF and TOF together with the ability to perform non-destructive detection from FT-ICR. Due to the simple and open ion-trap structure, laser and molecular beams can be integrated more easily into the SCSI-MS technique (see Figures 10.2 and 10.3,[11]) than into a FT-ICR mass spectrometer with its large bulky super-conducting solenoid cooled cryogenically. Furthermore, because the SCSI-MS technique is compatible with micro-traps that are under development currently by the ion-trapping community (see for example Stick et al. [12]), this technique has the potential for possible future commercialization. A limitation of the SCSI-MS technique in its present linear RF trap version is the requirement that the mass of the unknown singly-charged ion be within the range (ca 0.3–3)mat.ion, where mat.ion is the mass of the laser-cooled atomic ion, in order to achieve stable operating conditions with respect to the dynamical confining potential (see Section 10.3.1). With the availability of more atomic ion species that can be laser cooled, it will be possible to extend the range of masses that can be measured with the present complex experimental arrangement beyond the Mg+ and Ca+ ions that are
296
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a) –½Vcos (Ωt) 2 r0
½Vcos (Ωt)
2R Zendcap
2z0
Zendcap
(b)
FIGURE 10.2 (a) Sketch of the linear ion trap used in the experiments. The various parameters are defined in the text of Section 10.3.1. (b) Photograph of the actual ion trap used. As a scale, the center part of the electrode structure (2z0) is 5.4 mm long. (Reproduced from Drewsen, M.; Jensen, L.; Lindballe, J.; Nissen, N.; Martinussen, A.; Mortensen, A.; Staanum, P.; Voight, D., Int. J. Mass Spectrom. 2003, 229, 83–91. With permission from Elsevier.)
investigated presently in Aarhus. In principle, singly-charged ions from 1 to ca 600 Th should be measurable. Because the equilibrium position of the laser-cooled ion depends on the charge state of the unknown ion, the SCSI-MS technique measures both the charge and the mass of the unknown ion, and not merely the charge-to-mass ratio as do most mass spectrometers. This feature means that higher-charge states can be used, in some cases, to expand the measurable range of masses. Recently, the axial alignment of two ions in a Penning trap has been demonstrated [13]. When a sufficiently general procedure can be realized such that any two simultaneously-trapped ions can be so aligned, the above-mentioned mass range for a single laser-cooled atomic ion species may be increased also.
10.3 EXPERIMENTAL REALIZATION In this section is presented the experimental arrangement for the experiments to be discussed in the succeeding sections.
10.3.1 The Linear RF Ion Trap The type of ion trap used in the experiments discussed below is a so-called linear RF(Paul) ion trap [14,15]. A schematic diagram of this ion trap is shown in Figure 10.2a
297
Sympathetically-Cooled Single Ion Mass Spectrometry
8
6
7 5
1
2
4 3
FIGURE 10.3 Photo of the linear RF ion trap inside the cylindrical vacuum chamber used in the experiments. The chamber has an inner diameter of 30 cm. The numbers in the picture indicate the essential components of the instrument. (1) The linear RF ion trap. (2) The magnesium and calcium ovens, which are the sources for the atomic beams passing through the center of the ion trap. (3) Oven shutter. (4) Skimmer plates, used to collimate the effusive beams from the ovens in order to avoid contamination of the ion-trap electrodes. (5) Electron gun including deflection plates installed to make an electron beam cross the atomic beams in the center of the ion trap for ion production. (6) One of two sets of oppositely-positioned window openings for sending in laser beams for laser cooling of the ions along the main ion-trap axis (the z-axis) and along an axis perpendicular to the z-axis. (7) A piece of optical fiber, with known diameter, that can be translated into the ion-trap center for calibration of the magnification of the lens system used to image the fluorescence of the laser-cooled ions. (8) One of two opposed window openings that permit a laser beam to cross the atomic beams nearly at right angles in the ion-trap center for producing ions by photoionization. Not seen in this picture is a CCD camera to monitor the fluorescence from the laser-cooled ions through a window in a top-flange of the vacuum chamber (not shown). Also not seen is a leak valve controlled inlet for the admission of various gasses into the ion-trap region. (Reproduced from Drewsen, M.; Jensen, L.; Lindballe, J.; Nissen, N.; Martinussen, A.; Mortensen, A.; Staanum, P.; Voight, D., Int. J. Mass Spectrom. 2003, 229, 83–91. With permission from Elsevier.)
while a photograph of the ion trap is shown in Figure 10.2b. The ion trap, similar to a standard QMF, consists of four electrodes with an RF voltage of the same phase applied to diagonally-opposed electrodes, but with a phase difference of 180° between the two sets of opposite electrodes. This arrangement leads to a near-ideal two-dimensional quadrupole field in the plane perpendicular to the direction defined by the electrodes (that is, the z-direction). Each electrode is divided into three parts such that a positive DC voltage, U, can be applied to the eight end-cap electrodes. While this DC voltage leads to static confinement along the z-axis of the ion trap, it gives rise to a defocusing force in the radial plane perpendicular to this axis. By appropriate choices of U, and the RF peak-to-peak voltage, V, as well as the RF frequency, Ω, both axial and radial confinement can be obtained for the given ion-trap dimensions [16,17]. The confinement can be described in terms of the stability parameters
298
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
az ≡ η
8neU m Ω 2 z02
(10.1)
for the static confinement along the z-axis,
ar ≡ −
az 4 neU = −η 2 m Ω 2 z02
(10.2)
and
qr ≡
2neV m Ω 2r02
(10.3)
for the motion in the radial plane, where m is the mass of the ion, n is the number of elementary charges, e, on the ion, z0 and r0 are the ion-trap dimensions as defined in Figure 10.2a, and η is a positive geometric parameter dependent on z0 and r0. Whenever az > 0 and ar and qr lie within certain stability regions [16], stable single-ion trajectories exist. The motion is generally complex but, when qr < 0.4, a single ion not too far from the trap center will experience an effective harmonic potential, φ(z,r)
φ( z , r ) =
1 m(ω 2z z 2 + ω r2r 2 ) 2
(10.4)
where ωz and ωr, the oscillatory frequencies along the z-axis and in the radial plane, respectively, are given by
ωz =
ωr =
1 Ω az 2
1 Ω 2
qr2 2
+ ar
(10.5)
(10.6).
In the radial plane, the ions will be subject to the force from the effective potential above in addition to a fast oscillating force at the RF frequency; however, it can be shown that the amplitude of the corresponding motion (often named the micromotion) will be much smaller than the distance of the ion from the z-axis when qr < 0.4 [18]. In the experiments to be presented here, two ions are cooled such that both of them lie on the z-axis, hence we can neglect generally this rapid quiver motion. The ion trap used in the experiments and shown in Figure 10.2b has the following dimensions: r0 = 3.50 mm, z0 = 2.70 mm, zendcap = 20.00 mm, and R = 4.00 mm; these values have been chosen in order to achieve a nearly perfect radial quadrupole RF field [19] and a near-harmonic DC axial potential over a few millimeters. Numerical simulations show that these choices of ion-trap dimensions result in η = 0.248. The applied RF field is coupled resonantly to the ion-trap electrodes at a radial frequency
Sympathetically-Cooled Single Ion Mass Spectrometry
299
Ω = 2π × 3.88 MHz with a peak-to-peak amplitude V of typically 300–500 V. With a typical DC voltage of ca 1 V, Equation 10.4 gives rise to a potential-well depth of ca 1 eV. Because this energy is many orders of magnitude higher than the thermal energy of the cooled ions and the thermal energy of the background gas, the storage time of the ions can be hours or more at a typical vacuum of ca 10 –10 mbar. In Figure 10.3, a photograph of the ion trap (see also Figure 10.2b) and other important components in the experimental vacuum chamber are presented.
10.3.2 Laser-Cooling of Atomic Ions 10.3.2.1 Loading of Atomic Ions The atomic ions used in the experiments presented in Section 10.4 are created by crossing a skimmed atomic beam originating from an effusive oven with a photoionizing laser beam at the center of the ion-trapping region. This method has many advantages in comparison to the more usual electron impact ionization technique. First of all, resonance-enhanced photoionization is not only element selective but is isotope selective also [20,21]. Second, because photoionization is very efficient, these experiments may be carried out using low laser powers and low atomic beam fluxes. The adequacy of low atomic beam fluxes for these experiments means that ovens can be operated at relatively low temperatures with the accompanying benefit of a minimal pressure rise in the vacuum chamber. During ion loading, the pressure in the vacuum chamber increases by less than 10 –10 mbar above the base pressure of ca 2 × 10 –10 mbar. Lastly, but actually very importantly for the SCSI-MS technique presented in this contribution, the problem of electrical charging of the insulating parts near the ion-trapping zone is kept at a minimum because only one free electron is created for each atomic ion created. In our laboratory, both singly-charged magnesium and calcium ions can be produced currently by photoionization [20]. 10.3.2.2 Cooling of the Atomic Ions To laser-cool a trapped particle in an harmonic potential well characterized by three different oscillation frequencies, only a single laser beam propagating in a direction different from the three major axes of motion is needed. However, in our experiments to decouple the cooling along the z-axis from the cooling in the radial plane, as well as being able to establish a balanced force along the z-axis, the atomic ions produced are Doppler laser cooled [22] by two counter-propagating laser beams (of the appropriate wavelengths for the particular ion species in use) along the z-axis of the trap and by one beam propagating perpendicularly to this axis (see Figure 10.3). With this cooling configuration, the laser-cooled ion(s) can reach readily a temperature below 10 mK, a temperature at which the ions become spatially localized to ca 1 µm or less (see Figure 10.4a). Through the Coulomb interaction between the lasercooled ion and a simultaneously-trapped unknown ion, both ions can be cooled. For appropriate trapping parameters, this sympathetic cooling [23] can lead to temperatures of a few milliKelvins and can bring into alignment the two ions along the z-axis (Figure 10.4b).
300
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a)
(b)
(c)
(d)
100 µm
FIGURE 10.4 Images of fluorescence from 40Ca+ ions. (a) Two 40Ca+ ions at thermal equilibrium in the trap. (b) A laser-cooled 40Ca+ ion trapped together with a sympatheticallycooled 40Ca16O+ ion at thermal equilibrium. (c) The same two 40Ca+ ions as in (a) but with a modulation voltage applied at a frequency near the COM mode resonance frequency of 98.7 kHz. (d) The same ions as in (b) but with a modulation voltage applied at a frequency close to the COM mode resonance frequency of 89.4 kHz. In all experiments, the radial trap frequencies were 380 kHz and the exposure time of the CCD chip was 100 ms. (Reproduced from Drewsen, M.; Mortensen, A.; Martinussen, R.; Staanum, P.; SØrensen, J.L., Phys. Rev. Lett. 2004, 93, 243201. With permission from the American Physical Society.)
10.3.2.3 Fluorescence Detection The fluorescence imaging system used in these experiments consists, briefly, of an imaging lens system that, with a magnification of typically × 10, images light from the central ion-trapping region to the surface of an image intensifier (the shutter in Figure 10.1) that is connected to a CCD camera via a relay lens of magnification × 2. During the laser-cooling process, the cooled ion emits typically ca 107 fluorescence photons per second. With a total photon detection efficiency of the imaging system of about 10−4–10−3, some 103−104 photons per second are detected. Because typically a few hundred detected photons are needed for a satisfactory ion-position determination, camera integration times of the order of ca 100 ms are needed. Examples of fluorescence images are presented in Figure 10.4a for the case of two atomic ions in the ion trap and in Figure 10.4b for an atomic ion together with an unknown and non-fluorescing ion.
10.3.3 Inducing Forced Motion of the Ions The SCSI-MS technique relies on the determination of the common mode resonance frequencies for a two-ion system in which the mass is known for only one of the ions. In order to determine these resonance frequencies, the two-ion system is subjected to a periodic driving force. The forced motion of the ions can be induced in two ways: (1) by applying a sinusoidally-varying voltage to some of the end electrodes of the ion trap; and (2) by modulating the radiation pressure force from one of the laser beams propagating along the z-axis by chopping its intensity using an electro-optical chopper (EOC).
301
Sympathetically-Cooled Single Ion Mass Spectrometry
10.3.3.1 Electrical Forces A sinusoidally-varying driving force can be obtained by applying a modulation voltage to some of the ion-trap end-cap electrodes (see Figure 10.2). The COM mode can be excited effectively by applying a time-varying homogenous electric field along the z-axis. This maneuver can be achieved easily by modulating the voltage applied to two diagonally-opposed end-cap electrodes at one end of the ion trap. In order to excite effectively the BR mode, a time-varying electric field that has a gradient along the z-axis is needed. This situation can be realized when a common modulation voltage is applied simultaneously to sets of end-cap electrodes at both ends of the ion trap. Three potentials are combined in order to achieve the desired degree of control of the electrode potential; these potentials are (1) a constant DC potential for static trapping, (2) a modulation potential of ca 100 kHz for COM and BR mode excitation, and (3) an RF potential of ca 4 MHz for ion trapping in the radial plane. As indicated in the diagram shown in Figure 10.5, the modulation voltage is added to the DC voltage before it is combined with the RF voltage and led ultimately to the ion trap. The L1,C1 filter suppresses the modulation voltage on the DC end; the R1,C3 filter ensures a strong suppression of RF voltage at the DC end while allowing a fraction of the modulation voltage to pass the other way. The adjustable capacitor, C5, is used for fine-tuning the RF voltage on the electrode. A home-built programmable frequency synthesizer provides the modulation voltage for the COM and for the BR mode excitation. This synthesizer is able to step through a series of frequencies and, thus, to perform a scan across the expected resonance while delivering a trigger signal to the camera to ensure a well-defined frequency for each single image. In addition, the synthesizer outputs a short gating pulse with a variable phase delay that can be used to gate the image intensifier for phase measurements (see Section 10.3.4.2, below). The size of the frequency steps, the time between each step, and the modulation amplitude are adjustable and provide flexibility for the tailoring of scans. Furthermore, the synthesizer can hold in memory up to 10 predefined scans which, together with an option to loop when reaching the end of the last scan, permits repeated frequency scans of relevant masses. 10.3.3.2 Radiation Pressure Forces Not only can the radiation pressure force lead to ion cooling through the Dopplereffect, but it can be used further to exert a periodic force on the atomic ions. By modulating the intensity of one of the two laser-cooling beams propagating along Modulation addition UDC
R1
L1 C1
C2 Umod
RF addition
C3
Electrode C4
C5
URF
FIGURE 10.5 Electrical diagram showing how the modulation voltage is added to the iontrap voltages.
302
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
the z-axis, the atomic ions can be induced to experience a periodic driving force. In these experiments, one axial laser-cooling beam is directed through an electro-optic modulator (EOM) and a linear polarizer in tandem. By appropriate adjustment of the polarization of the laser beam entering the EOM and the orientation of the polarizer, the two optical elements can together act as an EOC when the correct voltages are applied to the EOM. The programmable frequency synthesizer described in the previous section controls the shutter. Test mass measurement experiments have shown that, at least at the level of m/∆m ca 103, the two methods for introducing a periodic driving force do not lead to any systematic errors.
10.3.4 Detection of the Motion of the Ions Before discussing the various ways of detecting the excited motion of the ions, first let us discuss briefly some basic theoretical issues of forced dynamics. For the linear RF ion trap described in Section 10.3.1, the potential energy for two singly-charged ions aligned along the z-axis is described well by the one-dimensional harmonic potential Φ(z1,z2)
Φ( z1 , z2 ) =
1 2 1 2 e2 κz1 + 2 κz2 + 2 4 π ε 0 z2 − z1
(10.7),
where z1 and z2 are the positions of the two ions, respectively, κ is a spring constant, and ε0 is the permittivity of vacuum. The equilibrium positions of the ions when cooled to low temperatures are z2 = –z1 = ∆z/2 (z2 > z1), with 1/3
e2 (10.8) ∆z = 2πε 0κ being the equilibrium ion distance. Under such conditions, the strongly-coupled motion of the ions, mediated by the Coulomb interaction and for small excursions, gives rise to two axial normal motional modes. Using the fact that the spring constant κ can also be expressed in terms of the single ion oscillation frequency of the two ions ωi (i = 1,2) through
κ = miω i2
(10.9)
the eigenfrequencies of these two modes can be found to be [10]
1 1 1 ω + ,−2 = 1 + ± 1 − + 2 ω 12 µ µ µ
(10.10),
where µ = m2/m1. The solutions ω+ and ω– correspond to the mode with eigenvectors where the ions move in phase (COM mode) and out of phase (BR mode), respectively, with mass-dependent amplitudes.
Sympathetically-Cooled Single Ion Mass Spectrometry
303
Equation 10.10 displays the mass dependence of the oscillation frequency for the two normal modes that is essential for the non-destructive SCSI-MS technique. From a measurement of either ω+ or ω− as well as ω1 (the oscillation frequency of the known laser-cooled ion species), the mass ratio µ and, hence, the mass m2 of an unknown ion species can be determined. Alternatively, by measuring the ratio between ω+ and ω−, Equation 10.10 gives rise to an equation with solutions µ and 1/µ. In such a case, µ can be determined without measurement of ω1, provided that it is known that either ω + > ω1 or ω + < ω1. In the experiment, we try to determine ω1, ω + , and ω− by resonant excitation of the ion motion through the application of a periodic driving force (see Section 10.3.3). In practice, the situation is complicated slightly by the fact that the lasercooled ion(s) is (are) subject to a damping force due to the laser-cooling process. In this case, the equations of motion read
z1 −
FDopp,1 ( z1 ) F e2 + ω12 z1 = − + drive,1 cos(ω t ) 4 π ε 0 m1 ( z2 − z1 )2 m1 m1
(10.11)
z2 −
FDopp,2 ( z 2 ) F e2 + ω 22 z2 = + drive,2 cos(ω t ) 2 4 π ε 0 m2 ( z2 − z1 ) m2 m2
(10.12),
and
where FDopp,i denotes the Doppler cooling force applied to ion i and Fdrive,i is the driving force applied to ion i. For two identical laser-cooled ions, one obtains FDopp,1 = FDopp,2, while for two different ions where only ion 1 is assumed to be laser cooled, one obtains FDopp,2 = 0. For end-cap electrode excitation, we choose Fdrive,1 = Fdrive,2 for COM mode excitation and Fdrive,1 = –Fdrive,2 for BR mode excitation. For excitation by laser modulation we have either Fdrive,1 = Fdrive,2 (identical ions) or Fdrive,2 = 0 (different ions). For small ion velocities the Doppler cooling force in Equations 10.11 and 10.12 can be approximated by a viscous damping force that is proportional to velocity, FDopp,i = − mi γ i zi , where γi is a constant that depends on the laser wavelength, intensity, detuning, and the ion mass [21]. Assuming small velocities and small excursions from the equilibrium positions, the motion for two identical ions can be described in terms of normal mode coordinates z + = (z2 + z1)/ 2 and z− = (z2 –z1–∆z)/ 2 for the COM and the BR mode, respectively, by the uncoupled equations of motion
z+ ,− + γ 1z + ,− + ω 2+ ,− z+ ,− =
F+ ,− cos(ω t ) m1
(10.13),
where F+ , –= (Fdrive,2 ± Fdrive,1)/ 2 . Each of these equations of motion is that of a classical driven, damped harmonic oscillator with the solution leading to z1 (t ) =
F+ ,− 2 m1 (ω − ω 2
) +γ ω
2 2 + ,−
2 1
2
cos(ω t + ϕ ) −
∆z ∆z ≡ z0 (ω )cos(ω t + ϕ ) − 2 2 (10.14),
304
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
where cos(ϕ ) =
ω + ,− 2 − ω 2 (ω + ,− 2 − ω 2 )2 + γ 12 ω 2
(10.15).
For ions with different masses, the solution of the equation of motion for the ‘detector’ ion (ion 1) is much more complex [24] and reads as z1 (ω , t ) =
Fdrive ,2 κ
2 2 cos(ω t + ϕ )(2ω 2 − ω )
(2ω12 − ω 2 ) +
Fdrive ,1 Fdrive , 2
∆z − cos(ω t ) − 2 + γ 12 (2ω 22 − ω 2 )2 ω 2 2
ω12 + γ 12 ω 2
[ω 4 − 2(ω12 + ω 22 )ω 2 + 3ω12 ω 22 ]2
(10.16),
where cos(ϕ ) =
[(2ω12 − ω 2 )Fdrive,2 + ω12 Fdrive,1 ] ⋅ [ω 4 − 2(ω12 + ω 22 )ω 2 + 3ω12 ω 22 ] + Fdrive,2 γ 12 (2ω 22 − ω 2 )ω 2
(2ω12 − ω 2 )Fdrive ,2 + ω12 Fdrive ,1 + iγ 1ωFdrive,,2 ⋅ ω 4 − 2(ω12 + ω 22 )ω 2 + 3ω12 ω 22 + iγ 1 (2ω 22 − ω 2 )ω 2
(10.17)
with ωi(i = 1,2) defined through κ = miω i2 . The amplitude of the laser-cooled ion is enhanced resonantly near ω+ and ω− with maxima shifted from ω+ and ω− by an amount that increases with γ1. The resonance width is approximately γ1. The in-phase amplitude has a dispersive shape, as for identical ions, and when only the laser-cooled ion is subject to a driving force (excitation by laser modulation) the amplitude has a zero-crossing at ω+ and ω−. For electrode excitation, where both ions are driven, the zero-crossing is offset from ω+ or ω− by an amount that, for γ1/ω+,− << 1 is of the order of (γ1/ω+,−)2. Intuitively, such an offset is present because, when both ions are driven, the driven motion of the non-cooled ion perturbs the phase of the laser-cooled ion. In contrast, when only one ion is driven, the other ion is ‘pulled’ or ‘pushed’ into phase with the driven ion. 10.3.4.1 Amplitude Detection In the amplitude method, the motion of the ions over many oscillation periods (typically 104 periods) is integrated such that it is the modulation frequency-dependent amplitude that is observed. The most precise determinations of the normal mode frequencies are found through fitting the relevant expression for z1(ω) (that is Equation 10.14 for identical ions and Equation 10.16 for non-identical ions) to the measured amplitudes of the laser-cooled ion as a function of the drive frequency, as shown in
Sympathetically-Cooled Single Ion Mass Spectrometry
305
Figure 10.6a for two identical ions. For many purposes, however, it is sufficient simply to determine the modulation frequency where the amplitude is maximal, which occurs at
ω max ≈ ω + ,−
(10.18),
because γ/ω+,− is typically ca 10−3−10−2. The relative error in the measurement of ω+,− is thus often smaller than 10−4. In Figure 10.4c and d, images of two laser-cooled atomic ions and one atomic and one molecular ion (CaO +) are presented when applying a driving force of a frequency close to ω + . The smearing of the fluorescence as compared to Figure 10.4a and b, shows clearly that the ions are excited motionally. 10.3.4.2 Phase Detection In the phase-detection method, the image intensifier is gated such that the camera integrates only light emitted in a short time interval in phase with the driving force. Hence it is the amplitude of the component of the ion motion which is in phase with the driving force that is recorded, that is, z1(ω)cos(ϕ). This in-phase amplitude shows the characteristic dispersive profile observed in Figure 10.6c and g and features a sharp zero-crossing of the in-phase amplitude near or at ω+,−. With laser light modulation, the voltage pulse for the gating signal can be overlapped temporally with the light pulse by detecting the latter with a photo diode. The adjustment of the detection phase is slightly more difficult when using the electrode excitation driving force, because the circuit used to couple the driving voltage to the electrodes leads to a frequency-dependent phase shift of the signal originating from the frequency synthesizer.
10.4 EXPERIMENTAL RESULTS In this section, some recent experimental results obtained in this laboratory using the SCSI-MS technique are described. In the first part, Section 10.4.1, experiments with atomic ions involving either two laser-cooled 40Ca+ ions or one laser-cooled 40Ca+ ion trapped together with another Ca+ isotope or a Mg+ ion are considered. These experiments serve to test the accuracy of the SCSI-MS technique. In Section 10.4.2, experiments performed with single molecular ions are presented: (1) the reaction Ca + + O2 → CaO + + O; (2) the reaction Mg + + HD → MgD + (MgH +) + H(D), in which a strong isotope effect was observed; and (3) experiments that demonstrate trapping and sympathetic cooling of single more complex ions, namely aniline ions, together with a study of their consecutive photofragmentation. As explained in Section 10.2, a non-destructive mass measurement can be performed by measuring the characteristic oscillation frequencies of the COM and/or the BR mode of a two-ion crystal. These frequencies can be determined by exciting the COM and the BR mode motion with a periodic driving force (see Section 10.3.3). In Figure 10.6 are shown the detected amplitude and the in-phase amplitude signals
306
Position [µm]
(a)
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (b)
80
100
60 40
50
20 0
92
94
96
98
Position [µm]
80
Amplitude [µm]
94
96
98
90
92
94
96
98
92
94
96
98
92 94 96 Drive frequency [kHz]
98
100
40
50
20 92
94
96
98
0
10
(f) 10
5
5
0
0 92
In−phase amplitude [µm]
92
60
0
(g)
90
(d)
(c)
(e)
0
94
96
98
90 (h) 10
5
5 0
0 −5
−5
92
94 96 Drive frequency [kHz]
98
−10 90
Sympathetically-Cooled Single Ion Mass Spectrometry
307
as a function of drive frequency for electrical excitation of the COM mode as well as the BR mode of two 40Ca+ ions. While the amplitude and the in-phase signals are well described by the basic theory given in Section 10.3 (Equations 10.13 and 10.14) for the COM mode, the BR mode signals show strong asymmetries. The origin of this latter asymmetry is discussed further below as well as in Section 10.5 [25].
10.4.1 Experiments with Atomic Ions 10.4.1.1 Two Calcium Ions Because the amplitude of the driven COM mode motion observed in Figure 10.6a is well described by Equation 10.14, ω + can be determined accurately from a fit to the data. Equation 10.13 is exact for the COM mode in the sense that no expansion of the Coulomb interaction term was needed for identical ions; hence, the solution in Equation 10.14 should be valid for arbitrary amplitudes. In order to verify this condition, the amplitude was recorded as a function of drive frequency for different driving forces. All of the amplitudes could be fitted to Equation 10.14 and, as shown in Figure 10.7, the center frequencies near 95.625 kHz are identical within 20 Hz, thus showing good agreement with the basic theory. In contrast, Equation 10.13 for the BR mode is correct for small amplitudes only. As shown in Figure 10.8, even at moderate amplitudes a significant asymmetry of the amplitude profile is evident with a sharp drop in amplitude on the high frequency side of the resonance. As discussed in more detail in Section 10.5, the shape of the amplitude profile is caused by a non-linear response (‘frequency-pulling’ [26]) because the oscillation frequency is increasing with the amplitude. While frequency-pulling is an interesting non-linear phenomenon, in the present context it leads, unfortunately, to difficulties in making an accurate determination of the ‘true’ BR mode frequency, that is, the eigenfrequency at small amplitudes. 10.4.1.2 One Calcium and One Magnesium Ion For the case of two non-identical ions, the COM mode frequency is also amplitude dependent. In order to investigate this effect, the amplitude was recorded as a function of drive frequency for one 40Ca+ ion and one 24Mg+ ion. As shown in Figure 10.9, FIGURE 10.6 (Opposite) The position-resolved ion fluorescence along the ion-trap axis as a function of the drive frequency of the modulation voltage. Each gray scale contour plot is composed of axial projections of the ion fluorescence intensity in images recorded during the frequency scans (step size: (a,c) 25 Hz/image, (b,d) 100 Hz/image). (a) and (c) show an excitation of the COM mode with the image intensifier in normal and gated modes, respectively, while (b) and (d) show an excitation of the breathing mode (BR) also with the image intensifier in normal and gated modes, respectively. The dark lines in the fluorescence track in (a) and (c) are due to shelving of the 40Ca+ ion in the non-fluorescing 3d 2D5/2 state [24]. (e through f): Amplitude of the ion motion as a function of the drive frequency. The amplitude is found from (a through b) as the difference between the HWHM of the fluorescence signal at a given drive frequency and the HWHM of the non-excited ion (that is, at a far-off resonant drive frequency). (g through h): In-phase amplitude of the ion motion as a function of the drive frequency. The in-phase amplitude is found from (c through d) as the center of the ion fluorescence track.
308
Resonance frequency [kHz]
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
95.64 95.63 95.62 95.61 10
15
20
25
30
35
40
Amplitude[µm]
FIGURE 10.7 Measured oscillation frequency for the COM mode of two laser-cooled 40 Ca+ ions at different amplitudes. The dotted line indicates the mean value of 95.625 kHz.
Amplitude [µm]
12
8
4
0 76
78 80 Drive frequency[kHz]
82
FIGURE 10.8 Amplitude vs. drive frequency for the BR mode of two laser-cooled 40Ca+ ions at different driving forces. The vertical line indicates the expected BR resonance frequency assuming no non-linear frequency pulling.
only at large amplitudes can a significant asymmetry of the amplitude profile be observed. The small asymmetry observed at low amplitudes can be explained by the slightly asymmetric shape of the amplitude predicted by Equation 10.16; such an asymmetry is predicted also by Equation 10.14 where it can be recognized as being due to the damping coefficient γ. The BR mode motion for two non-identical ions shows an asymmetry similar to that shown in Figure 10.7 for two identical ions. Figures 10.6 through 10.9 indicate that the ‘true’ COM mode frequency can be determined more easily and more accurately than the ‘true’ BR mode frequency. While this statement is true generally, there may be, on occasion, good reason to make use of the BR mode as is discussed further in Section 10.5 [27].
309
Sympathetically-Cooled Single Ion Mass Spectrometry
Amplitude [µm]
60
40
20
0 61
62 63 Drive frequency [kHz]
64
FIGURE 10.9 Amplitude vs. drive frequency for the COM mode of one laser-cooled 40 Ca+ ion and one 24Mg+ ion at different driving forces. The solid lines indicate fits to a model similar to that developed by Evoy et al. [27]. The fits yield the same resonance frequency for all of the five scans. 132240
40Ca+ – 40Ca+
132220 132200
Frequency (Hz)
132180 130600
40Ca+ – 42Ca+
130580
ν1=132208±10 Hz
ν–=130573±2 Hz
130560 130540 129000 128980
40
Ca+ – 44Ca+ ν–=128964±3 Hz
128960 128940
Time
FIGURE 10.10 Measured oscillation frequencies for three combinations of Ca+ isotopes. The stated resonance frequencies are the statistical averages of the respective data points [3]. The horizontal position of the data points reflects the time order of the experiments. (Reproduced from Drewsen, M.; Mortensen, A.; Martinussen, R.; Staanum, P.; SØrensen, J.L., Phys. Rev. Lett. 2004, 93, 243201. With permission from the American Physical Society.)
10.4.1.3 Precision Mass Measurements of Calcium Isotopes In order to test the accuracy of the SCSI-MS technique, the masses of the calcium isotopes 42Ca+ and 44Ca+ were measured with reference to the mass of the 40Ca+ ion; one of the isotopes was trapped together with a 40Ca+ ion and the COM resonance frequency was determined by the phase-detection method (see Ref. [3]). As a reference, the COM frequency for two 40Ca+ ions was determined also by the phase-detection method. The experimental oscillation frequencies from a series of measurements
310
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
are presented in Figure 10.10. Using Equation 10.10, mass ratios of µ40/42 = 0.9526(3) and µ40/44 = 0.9095(3) can be deduced from the results. Within the uncertainties of 3 × 10 –4 these results are in agreement with the expected values of µ40/42 = 0.952441 and µ40/44 = 0.909174 [28].
10.4.2 Experiments with Molecular Ions Molecular ions can be produced, trapped and cooled by the application of a variety of techniques. Within this laboratory, we can use essentially three methods: (I) reaction of trapped and laser-cooled atomic ion species with neutral molecules in a gas leaked into the vacuum chamber of the ion-trap apparatus; (II) electron impact ionization, using an electron gun, of neutral molecules leaked into the ion-trap center (see Figure 10.3); and (III) photoionization of leaked-in neutral molecules. In the following discussion, attention is focused on experiments using methods I and III only, because method II in our current experimental arrangement leads to unwanted high charging of insulating material close to the ion-trapping zone. This problem could be solved, however, by utilizing a more complex linear ion-trap configuration that permitted the movement of ions from an ‘ion production-region’ to a ‘mass-spectrometry-region’. Alternatively, a molecular ion could be introduced from an ion beam either by opening and closing rapidly the trapping potential along the z-axis or by introducing temporarily a buffer gas into the ion-trapping region while loading a molecular ion [29]. 10.4.2.1 CaO + Ions As a simple test of reaction method I, the reaction Ca + (4p) + O2 → CaO + + O, was chosen because it gives rise to a relatively large mass difference between the reactant atomic ion and the product molecular ion. Upon application of the amplitude detection method together with on-line monitoring of the amplitude of the ions’ motions on the CCD images when scanning the frequency in steps of 100 Hz, it was possible to obtain not only a mass resolution better than 102 but to distinguish that the CaO + ion shown in Figure 10.4b and d was, indeed, a 40Ca16O + ion [3]. Furthermore, this reaction made it possible to monitor a subsequent CaO + + CO → Ca + + CO2 reaction by leaking CO gas into the chamber. This latter reaction was observed readily by fluorescence from the Ca+ ion appearing as the product ion outcome of the reaction [30]. 10.4.2.2 MgH + and MgD + Ions In another series of experiments, the reaction between Mg+ ions in the 3p excited state and hydrogen molecules and its isotopologues was studied [31,32]. Attention was focused specifically on the outcome of the reaction between Mg + (3p) ions and HD molecules. From only ca 250 single reaction events, the branching ratio between MgD+ and MgH + formation was found to be greater than five [32]. The results are illustrated in Figure 10.11 where the ratio of the numbers of MgD+ ions and MgH + ions observed is plotted as a function of the pressure ratio between HD and the unavoidable H2 in the vacuum chamber. At high values of this ratio, where reactions with HD dominate, a clear preference for MgD + formation is seen. The solid line in
311
Sympathetically-Cooled Single Ion Mass Spectrometry
NMgD+/NMgH+
4
3
2
1
0 0
1
2
3 4 PHD/PH2
5
6
7
FIGURE 10.11 The ratio of the number of MgD + and MgH + ions produced in reactions Mg + + HD/H2 → MgD/H + + H/D at different ratios of the pressures of HD and H2, (PHD/PH2) where each P is italized but not emboldened. The solid line indicates a fit to the model described in Ref. [32]; the dashed line indicates the prediction of the model in the absence of isotope effects. (Reproduced from Staanum, P.F.; Højbjerre, K.; Wester, R.; Drewsen, M., Phys. Rev. Lett. 2008, 100, 243003. With permission from the American Physical Society.)
Figure 10.11 indicates the fit to a simple model that is described in Ref. [32], while the dashed line indicates the prediction of the model in the absence of isotope effects. The observation of this strong isotope effect provides insight into the details of the reaction dynamics in a reaction that should be a simple test case for theoretical modeling. Further insight was gained from additional measurements with H2 and D2 ions, which showed that the probability of forming either a MgH + or a MgD + ion from any of the intermediate MgH2+, MgHD +, or MgD2+ complexes is the same [32]. 10.4.2.3 The Aniline Ion and its Photofragments A series of experiments with the aniline ion (C6H5NH2+) illustrates the application of the SCSI-MS technique to complex molecular systems and demonstrates that it is possible to study consecutive photodissociation of molecular ions at the single ion level in an ion trap [33]. In addition to providing detailed information on fragmentation processes, such consecutive dissociation processes provide also a probabilistic way of preparing a wealth of single molecular ions that could be objects for further study or serve as reactant ions in other experiments. The aniline ion was formed through photoionization into a series of ro-vibrational states. From certain states further dissociation is possible through absorption of a single photon from the cooling laser [33]. By continuously scanning a broad range of masses, such consecutive dissociation reactions were detected. A reaction sequence where the ions C6H6 +, C6H5 +, and finally, C3H3+ were produced and detected consecutively is presented in Figure 10.12 [33]. Identification of the fragment ions was facilitated by a complementary set of experiments with fully deuterated aniline. Furthermore, it was found that the C3H3+ ion was produced in its cyclic configuration,
312
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Mass to charge ratio [m/z] 93
C6H5NH+2 C5H6+
66 65
C5H5+ C3H3+
39 0
1
2
3
Time [min]
FIGURE 10.12 A consecutive photodissociation sequence observed in an experiment with a single aniline molecular ion being the initial reactant. (Reproduced from Højbjerre, K.; Offenberg, D.; Bisgaard, C.Z.; Stapelfeldt, H.; Staanum, P.F.; Mortensen, A.; Drewsen, M. Phys. Rev. A. 2008, 77, 030702(R). With permission from the American Physical Society.)
that is, the cyclopropenium ion, as tested through a series of reaction experiments between C3H3+ ions and C2D2 molecules [33].
10.5 ACCURACY OF THE SCSI-MS TECHNIQUE In this section, various effects that can limit the accuracy of the SCSI-MS technique are considered. In particular, deviations from the idealized case considered in Section 10.3.4, due to finite size motional amplitudes and the effects of the Doppler cooling force will be discussed.
10.5.1 Laser-Cooling Force Doppler laser-cooling is an essential ingredient in the SCSI-MS technique. First, it provides the necessary damping force to cool directly and sympathetically the atomic and molecular ions, respectively, such that a cold and strongly-coupled twoion system is formed. Second, it gives rise to the fluorescence photons used in the detection process. Third, the radiation pressure force can be modulated to excite the common motion of the ions. In Section 10.3.4, the Doppler cooling force was approximated by a frictional force characterized by a friction (or cooling) rate γ1. This approximation can be shown to be valid whenever the velocity v of the cooled ion is much smaller than the absolute value of the detuning of the laser cooling light, |δ|, divided by the wavenumber k = 2π/λ of the light driving the cooling transition [22]. For the experiments involving 40Ca+ and 24Mg + , the cooling rate γ1 needed was as low as 2π × 100 Hz when applying a strong radial cooling. This condition required
Sympathetically-Cooled Single Ion Mass Spectrometry
313
detunings of the order of ca –10 Γ, where Γ is the cooling transition linewidth; such detunings can be applied readily but they lead to constraints on the velocity, v, of the ion such that vmax ca 80 m/s. For typical COM or BR mode resonance frequencies of ω+,– of ca 2π × 100 kHz and excitation amplitudes of ca 10 µm, the maximal ion velocity is ca 6 m/s, and hence is much smaller than the value for vmax stated above. Because the ion velocity is less than vmax, the frictional force approximation can be met readily. For the results presented in Figure 10.7 with large excitation amplitudes, the magnitude of the laser detuning was only 1–2 Γ, and so the frictional force approximation was fulfilled for over only part of the ion trajectory. The measured oscillation frequency was (within the error of the measurements), however, essentially independent of the maximum amplitudes because the γ1 dependence of the amplitude is weak for γ1 << ωz (see Equation 10.14). By taking measurements while using larger detuning, the relative shift in the resonance frequencies due to the non-linear friction force can be expected to be smaller than 10 –5. For measurements with non-identical ions, another laser-cooling related effect to be considered is the intensity imbalance of the counter propagating beams. Such an imbalance, which gives rise to a constant force induced by the radiation pressure, will push the laser-cooled ion away from its equilibrium position established with balanced beams; in addition, the imbalance will lead to a shift in the mass-dependent trap frequency used in Equation 10.10 [24]. For presently typical parameters (ω1 = 2π × 100 kHz, γ1 = 2π × 100 Hz, δ = –Γ) and a very conservative assumption of 10% intensity imbalance, the relative frequency shift is, at most, 1 × 10−5 for the COM mode motion of a 40Ca+ ion trapped together with an ion that is up to three times lighter or heavier. For the BR mode, the shift is much higher but less than 4 × 10−4. By minimizing the difference between the resonance frequencies for the two possible ion configurations (that is, with the laser-cooled ion located either to the left or to the right of the ion-trap center), the intensity imbalance effect can probably be reduced to a level of < 10−6.
10.5.2 Non-Linearity in the Coulomb Interaction In Section 10.3.4, the dynamics were treated of two harmonically-confined ions having small amplitudes of oscillation around the equilibrium positions of the ions. More specifically, in deriving Equations 10.10 and 10.13 (for z = z−) it was assumed that the change of the ion distance is small compared to the equilibrium distance ∆z, such that the Coulomb interaction energy can be approximated by a harmonic potential. For large changes in the ion–ion distance, additional terms in the Coulomb energy lead to an anharmonic interaction and, hence, to an amplitude-dependent oscillation frequency. For the COM mode oscillation of identical ions, the ions move in the same direction with the same amplitude, corresponding to a normalized eigenvector [9,10] of (1,1)/ 2 , which means that the inter-ion separation is always equal to the equilibrium distance. Therefore, the Coulomb interaction energy is simply a constant term in the potential and cannot give rise to an amplitude dependence for the oscillation frequency. Indeed, Figure 10.7 above shows that the oscillation frequency is independent of amplitude. For the BR mode, in contrast, the ions move away from each
314
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
6
Center-of-mass mode Breathing mode
|q2-q1|/|q1|
5 4 3 2 1 0
0.5
1.0
1.5 2.0 Mass ratio µ
2.5
3.0
FIGURE 10.13 The ratio between the difference of the eigenvector coordinates and the amplitude of coordinate 1 for the laser-cooled ion as a function of mass ratio, µ. The larger relative difference for the breathing mode leads to a larger amplitude dependence of the oscillation frequency.
other (the eigenvector is now (1,–1)/ 2 ) such that the inter-ion separation is different generally from the equilibrium distance. Hence, for amplitudes that amount to a significant fraction of the equilibrium distance, the higher-order terms of the Coulomb interaction give rise to an amplitude-dependent oscillation frequency. In Figure 10.8, the shape of the amplitude as a function of drive frequency is, indeed, a result of an amplitude-dependent oscillation frequency as is discussed in more detail below. For different ions, the two coordinates of the eigenvector of the COM mode generally have a different magnitude. Therefore, even for the excitation of the COM mode, the ion distance differs from the equilibrium distance. Hence, for both COM and BR mode motion, higher-order terms of the Coulomb interaction give rise to an amplitude-dependent oscillation frequency. The two coordinates of the eigenvectors, q1 and q2, are mass dependent [10]. In Figure 10.13, the quantity |q2–q1|/|q1|, that is, the change in ion distance relative to the amplitude of ion 1, which is the observed laser-cooled ion, is shown as a function of the mass ratio µ. Clearly, the change in ion distance is much smaller for the COM mode than for the BR mode and, consequently, the COM mode oscillation frequency is much less amplitude dependent than is the BR mode frequency. In the limit of µ→1 (identical ions), the amplitude dependence vanishes completely for the COM mode. In a harmonic potential where ωz = 2π × 100 kHz for a 40Ca+ ion, the equilibrium distance, ∆z, equals 26.0 µm, while typically applied amplitudes in the experiments described in Section 10.4 are about 5 µm. In this case, it is necessary to consider the effect of higher-order terms in the Coulomb interaction for the BR mode and, in some cases, also for the COM mode for two different ions. This effect could be reduced, but not eliminated, if much smaller amplitudes could be detected by using a state-of-the-art imaging system with almost diffraction-limited resolution
Sympathetically-Cooled Single Ion Mass Spectrometry
315
of ca 1 µm; however, there is the cost of additional experimental complexity to be considered because lenses must be mounted inside the vacuum chamber [34]. An amplitude-dependent oscillation frequency implies that the frequency dependence of the driven ion motion becomes non-linear because the resonance condition for the frequency of the driving force is amplitude dependent. The effect of this non-linearity is seen most clearly in Figure 10.6b and d and in Figure 10.8. When one follows the amplitude from the low-frequency side in Figure 10.8, the amplitude continues to rise even after the ‘true’ small-amplitude resonance frequency has been passed. This behavior is an example of frequency pulling where the effective resonance frequency gets shifted to higher values due to the increased amplitude. Jumps between large and small amplitudes appear around the point where the drive frequency is so far from the small-amplitude resonance frequency that the driving force is able, only marginally, to sustain non-linear oscillations. As indicated in Figure 10.8, the small-amplitude BR mode frequency is found on the low frequency side of the resonant structure. A ‘trained eye’ may be able to estimate this frequency with ca 1 kHz precision from a series of scans with varied driving force, such as are presented in Figure 10.8. For the COM mode excitation of different ions, such non-linear effects at 5 µm amplitude are not seen; only for very large amplitudes can be observed the effect of an amplitude-dependent oscillation frequency, as is shown in Figure 10.9. In conclusion, at practically-applied amplitudes of ca 5 µm, the breathing mode oscillation frequency is perturbed significantly by higher-order terms in the Coulomb interaction. Although the amplitude dependent shift prevents very accurate mass measurements based on the BR mode frequency, a measurement of the BR mode frequency can be useful in some situations, as is discussed in Section 10.5.7 below. In contrast, the amplitude-dependent shift of the COM frequency is much smaller than that for the BR mode and, at ca 5 µm amplitude, will lead typically to a relative shift in the resonance frequency of about 10 –4 or less. For ions of almost the same mass, the shift is particularly small.
10.5.3 Ion-Trap Imperfections So far we have considered the one-dimensional motion in a perfectly-harmonic potential. In this section, the various unavoidable effects that give rise to deviations from the ideal case will be discussed. 10.5.3.1 Anharmonicity In reality, the trapping potential contains terms other than the harmonic one. Higher order even terms (z4, z6, etc.) in the potential are unavoidable for practically-feasible shapes of the electrodes. Although odd terms are absent in a perfectly-symmetric trap, the symmetry is, in practice, broken by contact potentials [35] due to material deposited on the electrodes, by surrounding conductive or dielectric elements as well as by small misalignments and imperfections of the electrodes. Hence odd terms proportional to z, z3, etc., can be expected to be present also.
316
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
A term linear in z leads only to a change in the equilibrium positions of the ions but not in oscillation frequency. All terms higher than second order will, however, influence the oscillation frequency. The effect of the nth higher-order term can be estimated by assuming the trapping potential to be of the form ½mω+,−2z2[1 + (z/z0) n–2], and by calculating the oscillation period for a given amplitude by integration of the instantaneous velocity. We have carried out such an estimation for n = 3 and n = 4 with z0 equal to the half-length of the center electrodes (2.7 mm). This choice of z0 is a worst-case scenario; in reality, higher-order terms are significantly smaller and, in fact, the fourth-order term cancels when the center electrode length is chosen appropriately [24]. We find that for both n = 3 and n = 4 the relative change of the oscillation frequency is less than 10 –5 for amplitudes up to 5 µm and less than 10 –4 for amplitudes up to 20 µm. Hence, at a relative frequency accuracy of 10 –5, anharmonic effects can be neglected completely. The effects of anharmonic terms have been sought experimentally by measuring the COM mode frequency at different amplitudes for two 40Ca+ ions. Because the non-linearity of the Coulomb force does not affect the oscillation frequency in this case, any amplitude dependence should be due to either anharmonicity or the nonlinear velocity dependence of the Doppler cooling force as discussed in Section 10.5.1. As shown in Figure 10.7, at amplitudes between 15 µm and 35 µm, the measured COM frequencies near 95.625 kHz differ by less than 20 Hz and are equal within the error bars. Hence, at least at the level of 2 × 10 –4 anharmonic effects can be ignored. Furthermore, if anharmonic effects were significant, a frequencypulling effect should have been apparent, but no such effect was observed. 10.5.3.2 Trapping Field Imperfections The three-dimensional harmonic potential ½m(ωx2 + ωy2 + ωz2) gives rise to six normal modes for two ions [36]; the purely axial COM and BR modes, two purely radial modes, and two mixed modes that involve both radial and axial motion. Because the modes are normal at small amplitudes, it is justifiable to consider separately the axial modes, the COM and the BR mode. If, however, the potential contains undesired terms which couple the axial and the radial coordinates, that is to say, xz or yz terms, the coupling will change the axial eigenfrequencies when the ions are not confined to the z-axis. For ωx,y > ωz, two ions are aligned, on average, along the ion-trap axis [16], however, due to the finite ion temperature (ca 5 mK), they will move slightly off the ion-trap axis by up to ca 1 µm. Furthermore, when the ions move off the ion-trap axis, heating of the radial motion due to micromotion may be transferred to the axial motion either due to coupling terms in the potential or due to the mode mixing. Spurious RF fields along the ion-trap axis (for example, due to slightly different RF amplitudes on the 12 electrodes) can give rise to an unwanted confining potential along the ion-trap axis. By minimizing the micromotion along the ion-trap axis, the RF voltages can be matched to within about one percent by adjusting individually the capacitive loads of the electrodes (see Figure 10.5). Let us consider the magnitude of a residual effective RF voltage equal to one percent of a typical applied V of 400 V at Ω = 2π × 4 MHz. The value of the qz parameter is 0.004 for 40Ca+ and the corresponding secular frequency is 2π × 6 kHz which should be added to the axial ion-trap frequency in quadrature to obtain the total ion-trap frequency. Thus, for a
Sympathetically-Cooled Single Ion Mass Spectrometry
317
typical value for ω1 of 2π × 100 kHz, the axial frequency is changed by a few parts in a thousand. Because the secular frequency is mass dependent, this small change of axial frequency can lead to a correspondingly small error in the mass measurements when the mass is derived from ω1 and ω + (see Equation 10.10). For ions with masses close to that for the laser-cooled ion, the systematic relative error may be as small as 10 –4. Finally, the stability of the source for the DC voltage U applied to the end-cap electrodes is obviously an important parameter because the ion-trap frequency is related directly to U through Equations 10.1 and 10.5. With the DC supply at hand, the relative drift of the ion-trap frequency is below 10 –4 over approximately 15 minutes (see Figure 10.15 below). As discussed in Section 10.5.7, a slow drift is noncritical for many experiments where a reference measurement and a measurement of ω+ can be made within a short time interval. 10.5.3.3 Residual Magnetic Fields Residual magnetic fields with a component in the radial plane will couple the radial and axial motions of the ions. The effect of the cyclotron motion on the resonance frequencies ω+,− can be judged by considering the simpler single-ion situation. Here, the coupled equations of motion read
z + ω 2z z − ω c r = 0
(10.19)
r + ω r2r + ω c z = 0
(10.20)
with ωz and ωr being the axial and radial ion-trap oscillation frequencies, respectively, and with ωc being the cyclotron frequency. In the limit where ωc << ωz and ωc << ωr and ωc << |ωz –ωr| the modified axial frequency ω ′ is given by z
ω ′z ≈ ω z +
1 ωc2 2 ωz
(10.21).
For a ‘typical’ residual magnetic field of ca 1 Gauss (if, for example, the Earth’s magnetic field is not compensated for), the corresponding cyclotron frequency ωc of the ions is ca 2π × 100 Hz. Hence, with ωz /2π being typically of the order of 100 kHz, the relative change in frequency is only of the order of 10 –6. A similar result can be found for ω+,−. In this laboratory, the amplitude of the residual magnetic field can be reduced readily to 0.1 Gauss, meaning that the effect of the magnetic field can be reduced to a level of 10 –8.
10.5.4 Ion Loading When loading the ion trap with atomic ions, at least two undesirable effects may lead to uncontrolled changes in the trapping frequencies and, hence, may compromise the mass measurements. The effects are, (1) creation of contact potentials [35]
318
Relative shift
(a)
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (b)
0.003 0.002 0.001 0.000 −0.001
0
80
160
0
80
160
Scan number
FIGURE 10.14 Relative shift of the measured ion-trap frequency in a longer series of scans for a single 40Ca + ion. For each scan, the shift is defined as ω/ω0 –1, where ω is the trap frequency measured in the scan and ω0 = 2π × 95.51 kHz is the frequency measured in the first scan of part (a). (a) A series of scans where the neutral calcium beam is blocked (white background columns) and unblocked (gray background columns) sequentially. (b) Trap frequency measurements where the UV photoionization beam is blocked (white background columns) and unblocked (gray background columns) sequentially. The change of trap frequency when the UV beam is ‘on’ is most likely due to charging of insulating parts near the ion-trap region.
due to contamination of the ion-trap electrodes by neutral atoms from the effusive atomic beam (see Section 10.3.2), and (2) charging of insulating parts close to the ion-trapping region through photoelectron emission due to stray UV light. In order to investigate the possible influence of the atomic beam, a series of single Ca + ion oscillation frequency measurements was carried out while the atomic beam was blocked and unblocked sequentially by a shutter (see Figure 10.3). During these measurements, the UV photoionization beam was blocked. Similarly, the influence of the photoionization beam was investigated in another series of measurements where the UV photoionization beam was blocked and unblocked sequentially while the atomic beam was blocked. The measured single ion oscillation frequencies are shown in Figure 10.14 for both series of measurements. While no evidence for a systematic effect was found due to the atomic beam flux, the UV photoionization beam gives rise to oscillation frequency changes in the 100 Hz range. The rapid initial rise in the oscillation frequency followed by a slower rise in subsequent ‘on’ periods indicates that the charging of the affected insulating parts reaches a steady state eventually. The source of the “problematic” UV light is currently unknown, but it is likely either weak reflections from the exit window in the vacuum chamber (even though it is anti-reflection coated at 272 nm) or weak halo-like structures surrounding the main ionization beam. When the source is identified, its effect can probably be reduced dramatically by proper beam shaping and positioning. The effects observed here for calcium will most likely be observed also for other ions such as beryllium, magnesium, and mercury, which can be laser cooled; for these ions, photoionization requires UV light at 235, 285, and 185 nm, respectively. Furthermore, the laser cooling light
Sympathetically-Cooled Single Ion Mass Spectrometry
319
at 313, 280, and 194 nm for beryllium, magnesium, and mercury, respectively, may cause charging effects. To avoid the charging problem in the special case of calcium, photoionization can be achieved using two lasers in succession at the wavelengths 423 nm and 390 nm, respectively [37]. This arrangement would probably significantly reduce the problem observed in the experiments with light of 272 nm. Another more complex, but general, solution to the photoionization loading problem, would be to employ a linear ion trap with multiple trapping zones, loading ion(s) into one zone, a ‘loading zone’, and transferring the trapped ion(s) into another ‘experiment zone’ [38].
10.5.5 Background Gas The main effect of background gas (residual gas or reactant gas) is due to ion/neutral collisions of background gas with the cold atomic and molecular ions. Some collisions are manifested by a sudden disappearance of fluorescence light because the ions acquire sufficient kinetic energy to move away from the ion-trap axis. They are, however, typically not expelled from the pseudo-potential well of depth ca 1 eV and, after a while, they become laser-cooled sufficiently to be re-aligned along the iontrap axis and they resume fluorescing. In this laboratory, the base pressure in the vacuum chamber is ca 10 –10 mbar, at which collisions are extremely rare, that is to say about one in several minutes. For reaction studies, a gas pressure up to 10 –9–10 –8 mbar is applied typically. In these cases, it is possible to carry out scans of the drive frequency wherein few, if any, ionheating events occur. At much higher pressure, collision events can be observed so frequently that the analysis of a scan becomes unfeasible. Another potential effect issuing from the introduction of a gas is the creation of contact potentials due to contamination of the electrodes. As a test of this effect, the COM mode frequency was measured for two 40Ca+ ions, before and after O2 gas was leaked into the vacuum chamber at a pressure of 10 –7 Torr for a few minutes. Between the measurements, 40Ca+ ions were reloaded. Only a small shift at the level of 1 × 10 –4 in the COM mode frequency was observed; this shift may be related to the reloading procedure (see Section 10.5.4). Finally, it can be speculated as to whether there are other pressure-related effects, such as damping introduced by close or distant collisions that may influence the mass determination. However, the oscillation frequency of two 40Ca+ ions has been measured at pressures up to 10 times higher than the base pressure without observing any systematic effects.
10.5.6 Photon Detection 10.5.6.1 Spatial Resolution The position of a laser-cooled ion is measured by imaging the fluorescence light emitted during the Doppler laser-cooling process onto a CCD camera. There are two fundamental factors that limit the resolution: the spatial resolution of the imaging system and the finite ion temperature.
320
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
While the resolution of the imaging system is independent of the trapping parameters, the thermal excursion of the laser-cooled ion at a fixed temperature scales as ωz–1. From the simple physics point of view of being able to monitor small motional energy excitations, one would choose typically an ion-trap frequency ωz such that the thermal excursions of the ion are roughly the same as the spot size resolution of the imaging system. The better the resolution of the imaging system, the higher can be the ion-trap frequency, and the more precise can be the measurements expected. However, for a particular experimental arrangement, the effect of the non-linearity in the Coulomb interaction (see Section 10.5.2) must be taken into account when choosing ωz. In this laboratory, the resolution is dominated by the spot size resolution of the imaging system which is about 3 µm HWHM, while the amplitude for a 40Ca+ ion is less than 2 µm at a temperature of 5 mK and an ion-trap frequency of 2π × 100 kHz. 10.5.6.2 Scattering Rate and Collection Efficiency The number of collected photons from an ion is proportional to the photon scattering rate, the collection efficiency, and the CCD exposure time Texp. Through the imaging system, the number of collected photons is converted to a digital signal for each pixel on the CCD chip. In Figure 10.6a through d, these pixel values are projected onto the z-axis for each drive frequency and the projected value is indicated by the grey scale coding. In the following discussion, the (z-dependent) projected value of the signal is termed S. The signal-to-noise ratio, S/N, that is, the signal divided by its standard deviation, limits the precision with which the frequency ω+ can be determined. In order to determine ω+ with a precision ∆ω from either the amplitude or the in-phase amplitude as a function of drive frequency (see Figure 10.6e and g), the uncertainty of the measured amplitude must be less than the variation of the amplitude over the frequency interval ∆ω. For the amplitude method, this variation is simply ∆ω2 times the curvature ca A0/γ12, where A0 is the maximal amplitude. The uncertainty on the amplitude is of the order of A0(S/N) –1, from which it follows that ∆ω ≈ γ1/ S / N ∝ Texp –1/2. For the phase method, the variation of amplitude over ∆ω is ∆ω times the slope A0/γ1 which implies that ∆ω ≈ γ1/(Sg/Ng) ∝ Texp –1, where the subscript ‘g’ indicates the S/N ratio when the image intensifier is gated. In Figure 10.15 are shown the running average and its statistical uncertainty for a series of measurements of the COM mode frequency for two 40Ca+ ions where either the amplitude method (Figure 10.15a) or the phase method (Figure 10.15b) is applied. The statistical uncertainty for each measurement is in reasonable agreement with the estimates of ∆ω given above. The limiting value of the average, with a statistical uncertainty of only ca 2 Hz, is reached with fewer phase-method measurements than with the amplitude method measurements. The limiting value is reached, however, in approximately the same measurement time, because the amplitude method scans were five times faster than were the phase method scans. In the estimates of ∆ω above, it has been assumed tacitly that the damping coefficient γ1 is much larger than is the inverse exposure time that, indeed, is the case in these experiments. When this assumption is not fulfilled, the width of a resonance has an additional contribution from the inverse integration time. In the extreme case of γ1 << 1/Texp, it is found that ∆ω ∝ Texp –3/2 for the amplitude method and ∆ω ∝ Texp –2 for the phase method.
321
Sympathetically-Cooled Single Ion Mass Spectrometry
Resonance frequency [kHz]
(a)
(b) 95.65
95.64
95.63 0
10
20
30
40
0
10
20
Number of frequencies included in average
FIGURE 10.15 Running average and statistical uncertainties of the COM mode frequency for two 40Ca+ ions obtained (a) with the amplitude method and (b) with the phase method. For the amplitude method, the CCD exposure time was 100 ms and S/N ca 17, while for the phase method the exposure time was 500 ms and Sg /Ng ca 6. γ1 is 2π × 113 Hz in both cases, yielding Δω ca 2π × 27 Hz for the amplitude method and Δω is ca 2π × 19 Hz for the phase method. These values are in reasonable agreement with the statistical uncertainty of about 2π × 7 Hz for both methods. The total measurement times for the amplitude measurements and the phase measurements were 13 minutes and 23 minutes, respectively.
10.5.7 Reference Measurements In order to determine the mass of an unknown ion from a measurement of the COM mode frequency using Equation 10.10, a reference measurement of ω1 is needed. Alternatively, from a measurement of ω+ and ω−, µ can be determined provided that ω1 is known sufficiently well to determine whether ω + < ω1 or ω + > ω1. The best choice of the reference measurement depends on the specific use of the SCSI-MS technique as well as the uncertainty on the mass derived from the frequency measurements. In reaction experiments as described in Section 10.4.2.2, where two laser-cooled ions are trapped and one of them reacts to form an unknown molecular ion species, a reference measurement of ω1 with two laser-cooled ions can be carried out easily in appropriate time intervals before the reaction occurs. In this way, ω1 is measured shortly before ω+ such that a slow drift of ω1, due to changing DC voltage for example, has negligible influence. Similarly, in experiments where a single lasercooled ion is trapped and a second unknown ion is loaded, a fragment of a complex molecular ion [33] or a super-heavy ion species for example [39–41], a reference measurement of ω1 can be made with the single ion. In a longer reaction sequence, for example either of bi-molecular reactions as described in Section 10.4.2.1, or photoreactions as described in Section 10.4.2.3, or a combination thereof, a drift of the ion-trap frequency would compromise the mass measurements for the different reaction steps. A linear drift can be corrected for by measuring the ion-trap frequency before and after the reaction sequence, provided that the necessary trap loading after the sequence does not give rise to significant shifts of the ion-trap frequency. Measurement of ω+ and ω− during the
322
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
reaction sequence can effect a supplementary, self-referenced mass measurement. Since ω+ and ω− can be measured within a short time interval, the self-referenced measurement is unaffected by slow drifts. The main drawback of this measurement is the reduced precision with which ω− can be determined due to the strong amplitude dependence. In combination, the two measurements should enable accurate mass measurements over long time scales even in the presence of a small drift of the ion-trap frequency. Finally, we recall from Section 10.5.3.2 that, in case there is a significant induced secular potential due to an RF field along the ion-trap axis, the mass determination from ω1 and ω+ will contain a small error, while the mass determined from ω+ and ω− will not be influenced by this error source. The uncertainties on the measured frequencies, either ω1 and ω+ or ω+ and ω−, translate to an uncertainty on the mass m2 of the unknown ion determined from Equation 10.10. As shown in Figure 10.16, the relative uncertainty depends on the chosen reference measurement as well as the mass ratio µ. For a mass ratio larger than 1.25, it appears to be advantageous to determine the mass from ω+ and ω−, however, the difficulties in making a precise determination of ω− are sufficiently severe that a mass measurement based on ω1 and ω+ is generally preferable, provided that error sources, which are more severe for mass measurements based on ω1 and ω+ , for example, drift of the ion-trap frequency and RF-induced secular potentials, are sufficiently small.
100 ω1 and ω+
(δm2/m2) / (∆ω/ω)
ω– and ω+
10
1
1
2 Mass ratio µ
3
FIGURE 10.16 The ratio between the relative uncertainty on the unknown ion mass and the relative uncertainty of the frequency measurement (Δω/ω = Δω1/ω1 = Δω+ /ω+ = Δω–/ω) as a function of mass ratio μ. The two curves represent only COM mode measurements (One reference measurement of ω1 and one measurement of ω + ) and combined COM and BR mode measurements, respectively.
Sympathetically-Cooled Single Ion Mass Spectrometry
323
10.5.8 Choice of Parameters Regardless of whether the COM mode frequency or the BR mode frequency is measured by either the amplitude or the phase method, it is advantageous to be able to measure low amplitudes in order to reduce, as much as possible, amplitudedependent shifts due to the Coulomb interaction. Furthermore, a weak cooling force along the ion-trap axis (z-axis) leads to a narrow resonance and, hence, to a more precise frequency determination. Because a strong radial cooling force is advantageous, the optimum situation is to be able to change the detunings independently of the radial and axial laser beams. However, as proven by the experimental results presented in this chapter, an acceptable compromise can be found normally using fixed detuning and different light intensities. The choice of ion-trap frequency is a compromise between several requirements: on the one hand a high relative precision (ω+,− /∆ω) and a small ion spot width requires a large ion-trap frequency; on the other hand amplitude-dependent frequency shifts are reduced for a large equilibrium ion distance, that is, for a small trap frequency. It was found that with the relatively poor image resolution (spot size of ca 3 µm HWHM) in our setup ω+,− ca 2π × 100 kHz and γ1 ca 2π × 100 Hz is a convenient set of parameters that enables accurate mass determination at the level of 10 –4. From the discussion in Section 10.5.6, it is clear that the amplitude method and the phase method have their advantages and disadvantages. The phase method features a sharp zero crossing which enables an accurate frequency determination. The price to pay is the long integration time required in order to obtain an adequate S/N ratio, because light is collected only during a fraction of the oscillation period. The amplitude method is much faster, but the profile of the amplitude is not so sharp. Ultimately, for large Sg/Ng, the phase method is, therefore, the more accurate one but, with the present error sources at the level of a few times 10 –4, the amplitude method is superior owing to the shorter measurement time.
10.6 CONCLUSION AND OUTLOOK In this Chapter, we have presented a mass spectrometric technique, the SCSI-MS technique that relies on the measurement of the normal mode frequencies of stronglycoupled two-ion systems. The strongly-coupled regime is reached by sympathetic cooling of the ion of interest (the unknown ion) through the Coulomb interaction with a simultaneously-trapped and laser-cooled atomic ion. The technique is inherently a single ion mass spectrometric technique, utilizing a non-destructive detection step. This latter feature of the SCSI-MS technique is suitable particularly for monitoring mass changes by, for example, either photofragmentation (see Section 10.4.2.3) or multiplestep reactions. The single ion aspect of such studies enables one not only to obtain an ensemble-averaged result, but presents the opportunity for comparing individual ion “histories”. In particular, in experiments that involve complex molecular ions with complex internal structures, such individual ion “histories” may be essential for the identification of important processes.
324
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Though our current experimental arrangement has not been optimized for SCSI-MS, we have proven that a mass resolution of m/∆m of ca 104 can be obtained (see Section 10.4). Based on the discussion in Section 10.5 of the various effects that can limit the mass resolution, there seems to be room for improvement in the measurement accuracy by a factor of 10–100 and, hence, mass measurement with an accuracy of m/∆m ≈ 105–106 should be within reach. In this respect, the main error sources to be minimized are charging effects during loading of ions (Section 10.5.4) and imperfections in the RF-field configuration (see Section 10.5.3.2). By a modest optimization of our current experimental arrangement, we expect to be able to achieve a relative mass measurement accuracy of < 10 –4, and thus be able to discriminate between various mass doublets (for example, 24MgH + and 25Mg + ). In the future, we are also planning to exploit other advantageous features of SCSI-MS, namely, the excellent optical access to the ions in the trap and the fact that the ion of interest is both translationally cold and spatially very well localized. Currently, we plan to cool MgH + ions rotationally by irradiating the molecular ion with infrared light that drives a single vibrational transition [42,43]. This experiment can serve as a preparation step for doing cold chemistry. It is, furthermore, very compelling to exploit the extreme spatial localization of a molecular ion to study phase-sensitive coherent processes induced by laser light [44,45]. Finally, as mentioned briefly also in Section 10.5.7, the SCSI-MS technique may possibly find applications in the studies involving exotic ions, such as superheavy element ions [39] or rare isotope elements, because the technique can lead to quantitative results using only a relative small number of ions as is illustrated in Section 10.4.2.2.
Acknowledgment This work is supported by QUANTOP – Danish Natural Research Foundation Centre for Quantum Optics. PFS acknowledges support from the Danish Natural Science Research Council.
REFERENCES
1. Dempster, A.J. A new method of positive ray analysis. Phys. Rev. 1918, 11, 316–325. 2. March, R.E.; Todd, J.F.J. Practical Aspects of Ion Trap Mass Spectrometry I–III. CRC Press, Boca Raton, 1995. 3. Drewsen, M.; Mortensen, A.; Martinussen, R.; Staanum, P.; Sørensen, J.L. Nondestructive identification of cold and extremely localized single molecular ions. Phys. Rev. Lett. 2004, 93, 243201. 4. Blauth, E.W. Dynamic Mass Spectrometry. Elsevier, New York, 1966. 5. March, R.E. Quadrupole ion trap mass spectrometry: a view at the turn of the century. Int. J. Mass Spectrom. 2000, 200, 285–312. 6. Wollnik, H. Time-of-flight mass analyzers. Mass Spectrom. Rev. 1993, 12, 89–114. 7. Comisarow, M.B.; Marshall, A.G. Fourier transform ion cyclotron resonance mass spectroscopy. Chem. Phys. Lett. 1974, 25, 282–283.
Sympathetically-Cooled Single Ion Mass Spectrometry
325
8. Rainville, S.; Thompson, J.K.; Pritchard, D.E. An ion balance for ultra-high-precision mass measurements. Science 2004, 303, 334–338. 9. James, D.F.V. Quantum dynamics of cold trapped ions with application to quantum computation. Appl. Phys. B. 1998, 66, 181–190. 10. Morigi, G.; Walther, H. Two-species Coulomb chains for quantum information. Eur. Phys. J. D. 2001, 13, 261–269. 11. Drewsen, M.; Jensen, L.; Lindballe, J.; Nissen, N.; Martinussen, A.; Mortensen, A.; Staanum, P.; Voight, D. Ion coulomb crystals: a tool for studying ion processes. Int. J. Mass Spectrom. 2003, 229, 83–91. 12. Stick, D.; Hensinger, W.K.; Olmschenk, S.; Madsen, M.J.; Schwab K.; Monroe, C. Ion trap in a semiconductor chip. Nature Physics 2006, 2, 36–39. 13. Powell, H.F.; Segal, D.M.; Thompson, R.C. Axialization of laser cooled magnesium ions in a Penning trap. Phys. Rev. Lett. 2002, 89, 093003. 14. Prestage, J.D.; Dick, G.J.; Maleki, L. New ion trap for frequency standard applications. J. Appl. Phys. 1989, 66, 1013–1017. 15. Raizen, M.G.; Gilligan, M.J.; Bergquist, J.C.; Itano, W.M.; Wineland, D.J. Ionic crystals in a linear Paul trap. Phys. Rev. A. 1992, 45, 6493–6501. 16. Drewsen, M.; Brøner, A. Harmonic linear Paul trap: stability diagram and effective potentials. Phys. Rev. A. 2000, 62, 045401. 17. Ghosh, P. Ion Traps. Clarendon Press, Oxford, 1995. 18. Wuerker, R.F.; Shelton, H.; Langmuir, R.V. Electrodynamic containment of charged particles. J. Appl. Phys. 1959, 30, 342–349. 19. Denison, D.R. Operating parameters of a quadrupole in a grounded cylindrical housing. J. Vac. Sci. 1971, 8, 266–269. 20. Kjærgaard, N.; Hornekær L.; Thommesen, A.M.; Videsen, Z.; Drewsen, M. Isotope selective loading of an ion trap using resonance-enhanced two-photon ionization. Appl. Phys. B. 2000, 71, 207–210. 21. Mortensen, A.; Lindballe, J.J.T.; Jensen, I.S.; Staanum, P.; Voigt, D.; Drewsen, M. Isotope shifts of the 4s2 1S0→4s5p 1P1 transition and hyperfine splitting of the 4s5p state in calcium. Phys. Rev. A. 2004, 69, 042502. 22. Metcalf, H.J.; van der Straten, P. Laser Cooling and Trapping. Springer-Verlag, New York, 1999. 23. Larson, D.J.; Bergquist, J.C.; Bollinger, J.J.; Itano, W.M.; Wineland, D.J. Sympathetic cooling of trapped ions: a laser-cooled two-species non-neutral ion plasma. Phys. Rev. Lett. 1986, 57, 70–73. 24. Staanum, P. Quantum optics with trapped calcium ions. PhD. diss., University of Aarhus, Denmark, 2004. http://www.phys.au.dk/main/publications/PhD. 25. Staanum, P.; Jensen, I.S.; Martinussen, R.G.; Voigt, D.; Drewsen, M. Lifetime measurement of the metastable 3d 2D5/2 state in the 40Ca + ion using the shelving technique in a few-ion string. Phys. Rev. A. 2004, 69, 032503. 26. Cross, M.C.; Zumdieck, A.; Lifshitz, R.; Rogers, J.L. Synchronization by nonlinear frequency pulling. Phys. Rev. Lett. 2004, 93, 224101. 27. Evoy, S.; Carr, D.W.; Sekaric, L.; Olkhovets, A.; Parpia, J.M.; Craighead, H.G. Nanofabrication and electrostatic operation of single-crystal silicon paddle oscillators. J. Appl. Phys. 1999, 86, 6072–6077. 28. Audi, G.; Wapstra, A.H. The 1993 atomic mass evaluation: (i) atomic mass table. Nucl. Phys. A. 1993, 565, 1–65. 29. Ostendorf, A.; Zhang, C.B.; Wilson, M.A.; Offenberg, D.; Roth, B.; Schiller, S. Sympathetic cooling of complex molecular ions to millikelvin temperatures. Phys. Rev. Lett. 2006, 97, 243005. 30. Drewsen, M.; Hornekær, L.; Kjærgaard, N.; Mølhave, K.; Thommesen, A.M.; Videsen, Z.; Mortensen, A.; Jensen, F. Ion Coulomb crystals and some applications.
326
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Non-Neutral Plasma Physics Conference IV, AIP Conference Proceedings. 2002, 606, 135–144. 31. Mølhave, K.; Drewsen, M. Formation of translationally cold MgH + and MgD + ions in a linear Paul trap. Phys. Rev. A. 2000, 62, 011401(R). 32. Staanum, P.F.; Højbjerre, K.; Wester, R.; Drewsen, M. Probing isotope effects in chemical reactions using single ions. Phys. Rev. Lett. 2008, 100, 243003. 33. Højbjerre, K.; Offenberg, D.; Bisgaard, C.Z.; Stapelfeldt, H.; Staanum, P.F.; Mortensen, A.; Drewsen, M. Consecutive photodissociation of a single complex molecular ion. Phys. Rev. A. 2008, 77, 030702(R). 34. Schlosser, N.; Reymons, G.; Protsenko, I.; Grangier, P. Sub-poissonian loading of single atoms in a microscopic dipole trap. Nature 2001, 411, 1024–1027. 35. Ashcroft, N.W.; Mermin, N.D. Solid State Physics, Saunders College Publishing, Fort Worth, 1976. 36. Rohde, H.; Gulde, S.T.; Roos, C.F.; Barton, P.A.; Leibfried, D.; Eschner, J.; SchmidtKaler, F.; Blatt, R. Sympathetic ground-state cooling and coherent manipulation with two-ion crystals. J. Opt. B: Quantum Semiclass. Opt. 2001, 3, S34–S41. 37. Gulde, S.; Rotter, D.; Barton, P.; Schmidt-Kaler, F.; Blatt, R.; Hogervorst, W. Simple and efficient photo-ionization loading of ions for precision ion-trapping experiments. Appl. Phys. B. 2001, 73, 861–863. 38. Guthöhrlein, G.R.; Keller, M.; Hayasaka, K.; Lange, W.; Walther, H. A single ion as a nanscopic probe of an optical field. Nature 2001, 414, 49–51. 39. Drewsen, M. Cooling, identification and spectroscopy of super-heavy element ions. Eur. Phys. J. D. 2007, 45, 125–127. 40. Schädel, M. Super-heavy element chemistry at GSI – status and perspectives. Eur. Phys. J. D. 2007, 45, 67–74. 41. Haba, H.; Akiyama, T.; Kaji, D.; Kikunaga, H.; Kuribayashi, T.; Morimoto K.; Morita, K.; Ooe, K.; Sato, N.; Shinohara, A.; Takabe, T.; Tashiro, Y.; Toyoshima, A.; Yoneda, A.; Yoshimura, T. Startup of superheavy element chemistry at RIKEN. Eur. Phys. J. D. 2007, 45, 81–86. 42. Vogelius, I.S.; Madsen, L.B.; Drewsen, M. Blackbody-radiation-assisted laser cooling of molecular ions. Phys. Rev Lett. 2002, 89, 173003. 43. Vogelius, I.S.; Madsen, L.B.; Drewsen, M. Rotational cooling of heteronuclear molecular ions with 1Σ, 2Σ, 3Σ, and 2Π electronic ground states. Phys. Rev A. 2002, 70, 053412. 44. Shapiro, M.; Brumer, P. Coherent control of atomic, molecular, and electronic processes. Adv. At. Mol. Opt. Phys. 2000, 42, 287–342. 45. Brixner, T.; Damrauer, N.H.; Gerber, G. Femtosecond quantum control. Adv. At. Mol. Opt. Phys. 2001, 46, 1–54.
Trap: A Versatile 11 Ion Tool for the Atomic Clocks of the Future! Fernande Vedel Contents 11.1 11.2 11.3 11.4
Introduction.................................................................................................. 328 Ion Atomic Clock: Why and How?.............................................................. 329 Atomic Clock Devices................................................................................. 331 The Best Traps for the Best Clocks.............................................................. 334 11.4.1 Traps for Microwave Frequency Standards................................... 334 11.4.1.1 Properties of Linear Traps............................................ 334 11.4.1.2 Quadrupole Linear Trap at Jet Propulsion Laboratory (JPL)........................................................... 337 11.4.1.3 Linear Ion Trap at National Institute of Standards and Technology (NIST)................................................ 337 11.4.1.4 Multi-Pole Traps at Jet Propulsion Laboratory (JPL).............................................................................. 339 11.4.2 Traps for Optical Frequency Standards..........................................340 11.4.2.1 Single Ion Preparation and Clock Signal Building.......340 11.4.2.2 Miniature Ion Trap Design............................................343 11.4.2.3 An Example of an End-Cap Electrodes Trap................345 11.4.2.4 Ring Traps..................................................................... 347 11.5 Ion Clocks: Current Research...................................................................... 352 11.5.1 The Jet Propulsion Laboratory (JPL) LITS (Linear Ion Trap Standard) Project at 40.5 GHz....................................................... 352 11.5.2 Ion Optical Clocks..........................................................................355 11.5.2.1 National Institute of Standards and Technology (NIST) Mercury Projects..............................................356 11.5.2.2 The Logic Atomic Clock...............................................356 11.5.2.3 729 nm Frequency Metrology with Ca+ in Marseille....... 357 11.5.2.4 State of the Art of Optical Frequency Ion Clocks........ 358 11.6 Conclusion....................................................................................................360 References...............................................................................................................360
327
328
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
11.1 INTRODUCTION Since the birth of ion trapping techniques in the 1960s, physics and chemistry applications have evolved fruitfully in concert, leading to the award of one-quarter of the 1989 Nobel Prize in Physics to each of W. Paul [1] and H.G. Dehmelt [2]. Although radiofrequency ion traps were used largely for high resolution atomic spectroscopy, spin exchange experiences, and lifetimes of atomic levels [3,4], the so-called ‘QUISTOR’[5] (QUadrupole Ion STORe) opened a new vista in mass spectrometry not only for component analysis but also for studies of chemical reaction paths also. Atomic excitations that are employed in lasers and the subsequent manipulation of such atomic lasers have stimulated the realization of the huge possibilities presented by ion trapping. Laser cooling, which has permitted the attainment of extreme accuracy in atomic spectroscopy measurements, has provided the means for precise internal and external ion energy manipulation thus opening the way to novel and powerful applications, such as quantum information and atomic clocks in the optical domain [6]. In general, lasers afford quite extraordinary means for both the formation of monatomic and polyatomic ions in specific excitation states and for probing such species in excited states. Recent progress in the development of a wide range of lasers, including the advent of laser diodes, offers myriad opportunities for new kinds of investigations. Another important research direction was opened by the liberties taken with the hyperboloidal geometry of the quadrupole ion trap. Enhanced understanding of the confinement properties of quadrupole ion traps allowed the exploitation of new geometries that were easier to construct, such as cylindrical ion traps that permitted ready access for experimentation, exploited new characteristics such an non-linear modes of excitation, and introduced a stretched geometry that led to improvement in mass resolution. Recently, miniaturization of cages [7] or traps on micro-chips [8] have been exploited in atomic quantum physics as well as in chemistry analysis.* It is worthy to note that the extent of progress made in each of these domains was realized due to the curiosity of researchers and their desire to learn from the results of the experiments carried out. The purpose of this chapter is to expound upon the specific topic of atomic clocks utilizing ion traps and the new challenges engaged presently for the measurement of time with extremely high precision. Today, ion atomic clocks are designed for various systems (that is, the ion–trap–oscillator ensemble) that could offer new opportunities for the design of new ion trapping applications. This chapter is organized as follows. First, the need for still further enhancement in the precision with which time is measured will be justified, and the concept of atomic clocks and their properties will be described in detail. Then, the properties required for such an atomic system suitable for time and frequency metrology will be developed as well as the conditions necessary to attain them, following schemes involving either ions or neutral atoms. For the utilization of ions in atomic clocks, the well-known technique of ion trapping is used. The next part of the discussion will be devoted, therefore, to both a panoramic view of the ion trap geometries used * See Volume 4, Chapter 2: Ion Traps for Miniature, Multiplexed and Soft Landing Mass Spectrometers
by R. Graham Cooks, Zheng Ouyang, M. Fico, Q. Song, L. Gao, and C. C. Mulligan.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
329
during recent years and a justification of the different research directions that are being followed in attempts to achieve the twin goals of the preparation of a pure atomic system and the necessary control of the environment of that system. Finally, some recent examples of projects or realizations of atomic clocks will be described. Two of these atomic clocks are concerned with the mercury ion: the first one, which operates in the microwave frequency domain, has been studied at the Jet Propulsion Laboratory (JPL) of the California Institute of Technology for deep space applications; the second one, which operates in the optical domain at 282 nm, has been studied at the National Institute of Standards and Technology (NIST) in Boulder, CO. These projects demonstrate forcefully the potential for developing the best atomic clocks. Another project, currently under study in Marseille and using the calcium ion, will be presented together with new schemes for clocks operating in the terahertz (THz) frequency domain.
11.2 ION ATOMIC CLOCK: WHY AND HOW? All scientific inquiries must overcome the inescapable requirement of measure. Thus, a major part of physics is dedicated permanently to the enhancement of measurement and the quantification of phenomena. In fact, such a goal requires progress in instrumentation that cannot be fulfilled without better definitions of the fundamental units. Among them, the time unit is, and has been for a long time, one of the most crucial units necessary for the advancement of knowledge. For this reason, the time unit was the first unit for which the definition put aside any material systems. Greenwich Mean Time (GMT) was defined in 1884, assuming one second equal to 1/86,400 of the mean solar day [9]. While this definition introduced a measure of order for commerce and permitted the introduction of train schedules both within and among times zones, the utilization of a mean solar day sowed the seeds of the demise of this definition because the solar day itself was subject to variation. Thanks to the progress realized by using the beating of a quartz crystal instead of a ‘classic’ pendulum, it was possible to propose locking the second with an internal atomic oscillation, which is inherently stable over long periods of time and great distances. The first atomic clock, built in 1949 by the National Burean of Standards-NBS, now renamed NIST, involved excitation of the ammonia (NH3) molecule in the microwave domain made possible by the development of radar before and during World War II. However, this clock proved to offer little advantage over existing standards, and attention was turned to the use of cesium atoms. The first practical cesium atomic frequency standard was built at the National Physical Laboratory (NPL), at Teddington, in England in 1955, and, in collaboration with the U.S. Naval Observatory, the frequency of the cesium reference standard was established relative to astronomical time. In 1967, the natural frequency of the cesium atom was recognized formally as the new international unit of time; the second was defined as exactly 9,192,631,770 oscillations or cycles of the transition between two hyperfine levels of cesium-133 (133Cs). Nowadays, local time is defined or delivered by national services, which link local time to International Atomic Time (TAI) fabricated from a set of atomic clocks based in different countries and weighted according to their individual performance.
330
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
In fact, transportation, communications, power stations, stock exchanges, etc., require accurate clocks. As scientific research and new applications are becoming more demanding and continue to ask for a still more accurate clock, a further definition of the second will be one of the new challenges. Yet it should be possible to improve dramatically the precision in the secondary standards, as well as to furnish new and easier secondary standards for length because length corresponds to the path traveled by light in vacuum during a time interval of 1/299,792,458 of a second, thus setting in 1983 the value of the velocity of light as a new universal constant. Moreover, the needs of fundamental physics for quantum electrodynamics or general relativity require the precise measurement of time. Most of the fundamental constants can be related either to a time or to a frequency measurement. The quest for the detection of the smallest possible time variation in these constants [10], that is, the attainment of a time variation measurement of these constants at the 10 −17–10 −18 level, could produce answers for such fundamental problems as the local position invariance. Highly accurate clocks would be welcomed in astronomy because, thanks to deep, baseline interferometry, it could be possible to detect gravity waves which, would become a new tool for observing the general relativity theory in the universe. Such clocks could permit observation of the so-called ‘gravitational red-shift’, induced by massive stars in our universe, in order to determine the Earth’s geo-potential to a height equivalent of 1 cm, which is used largely now in meteorology models. Finally, ground-positioning system (GSP)-Galileo systems require a panoply of atomic clocks located in satellites as well on the Earth’s surface. All atomic clocks are based on the same servo-loop scheme (Figure 11.1). An internal atomic oscillator at ωAt is used to lock an external or local atomic oscillator at frequency ω0. The local oscillator is used to probe the atomic transition at ωAt, and Clockwork device
Oscillator
Atomic reference
ω0 ωAt Detector Servo signal
Servo control Error signal
FIGURE 11.1 A schematic diagram of an atomic clock. The atomic reference (ωAt) is provided by an atomic transition of an atom or a molecule. The fluorescence of the transition is detected by the oscillator (the local oscillator), which excites the atomic system at ω0. Before each interrogation, to make ω0 as close as possible to ωAt, the servo loop adjusts the ω0 fluorescence signal. A part of the signal of the local oscillator at ω0 delivers the clock signal.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
331
to deliver the clock signal. In order to make ω0 as close as possible to ωAt, a correction signal adjusts ω0 from the signal height delivered by the atomic oscillator before probing again the atomic oscillator. The more narrow the atomic frequency line, the more precisely will work the servo loop. In this chapter, we will focus upon the efforts made to prepare the atomic oscillator based on stored ions in such a way that the spectral profile of the atomic transition will be as sharp as possible, as stable as possible, and with the lowest possible degradation. Some parameters will be introduced here because the device must be assessed and evaluated. For a primary standard, which defines the unit, the precision (that is, the dispersion of the measurements) and the stability (as it relates to the temporal properties) must be as close to being impeccable as is possible. When a secondary standard is used, it is then necessary to estimate its accuracy, that is, the dispersion of the output signals relative to the primary standard. The system shall be reproducible also; that is, two or more devices should produce the same frequency, within a given confidence interval. In the case of atomic clocks it is assumed that, ideally, the frequency transition holds one fixed value. The output signal obtained from a given transition, because of systematic effects, will be more or less accurate. In any case, precision and accuracy are evaluated from appropriate measurements and models. With respect to the stability, a special tool called the Allan Variance [11] was defined because the frequency fluctuations are not stationary. This variance σ2(τ), at τ, is obtained from the square of the two fractional frequencies, ∆ f /f, averaged over two different integration times that differ from τ. The temporal evolution of the Allan Variance reflects the quality of the oscillator and is indicative of the nature of the major frequency noise at a given time. An analytic expression of the Allan deviation [12] that can be obtained from a statistical formulation is
σ (τ) =
∆f
1
f S/N
Tc τ
(11.1),
where S/N is the signal-to-noise ratio, ∆f is the range of measured frequencies about the mean frequency f, and Tc is the time of the cycle measurement. In order to convey some appreciation of the magnitude of these parameters, the short-term frequency stability of the first atomic fountain at Laboratoire national de métrologie et d’essais, Système de références temps espace (LNE-SYRTE ), in Paris, was evaluated in 1995 to be equal to 2 × 10 − 13τ − 1/2 with an accuracy of less than 10 − 14, where τ is the integration time.
11.3 ATOMIC CLOCK DEVICES Since the first Cesium-atomic clock based on an atomic beam, different approaches have been developed, justified by the need to reduce the global uncertainty of the system. The search for an appropriate atomic transition is guided by seeking its potentiality to have the narrowest spectral transition possible. Atomic calculations have assisted in making this choice. At the beginning of the atomic time era,
332
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
t ransitions of interest were located in the radiofrequency domain that is not ‘too far’ from the Hertz range (and, thus, not too far from the time unit, the second), and for which radiation sources are easily attainable. Hyperfine transitions of odd isotopes of elements (that is, where the arithmetic sum of the numbers of protons and neutrons in the nucleus is an odd number) offer good quality factors, Q, where Q = ∆ f/f and ∆ f defines the natural line width, especially for heavy elements. The clock transition interrogation process must be executed under optimum and well-characterized environmental conditions. Generally, the main cause of degradation of a spectral profile arises from the Doppler effect due to the motion of the atoms, followed by the effects due to prevailing magnetic and electric fields, as well as from blackbody radiation. The Ramsey technique, or Ramsey fringes, proposed by Norbert Ramsey [13] constituted a valuable approach for overcoming part of the degradation problem. The atomic population is probed in two separate parts of an atomic beam crossing an radio frequency (RF) cavity; this procedure is very effective in eliminating the first-Doppler effect because the central fringe concerns only atoms at quasi-zero velocity. Because the limitation came from the relative brevity of the interrogation by the microwave field, laser cooling of atoms was used to prepare the so-called ‘atomic fountain’ where cold atoms are pushed up to an excited state and then fall back. Thus, atoms cross the RF cavity at the same place but at different times (that is, the temporal Ramsey excitation scheme rather than separated field interferometry). The lengthening of the atom/ electromagnetic field interaction leads to a drastic reduction in the relativist or second-Doppler effect. A limitation resides in gravity, which limits the time for interrogation; however, zero-g experiments have been carried out with the Projet d’Horloge Atomique par Refroidissement d’Atomes en Orbite (atomic clock project using laser atom cooling in orbit) or PHARAO project [14]. At the same time, various laboratories proposed new atomic clock schemes involving trapped ions [15]. Trapped ions move periodically and this property acts as a frequency modulation, which induces a particular shape for the first-Doppler spectral profile that is formed as a central band with lateral bands each separated from its neighboring bands by the secular motion frequencies. An example of a simulated first-Doppler spectral profile is shown in Figure 11.2. The lateral bands can be resolved when the motion amplitude has the same or a lower magnitude than the radiation wavelength. Different atom species, for example, ytterbium and barium, were considered. Mercury offered the best capability, such that RF linear traps with mercury ions are the only competitors of cesium atomic fountains. A project to place an ensemble of atomic clocks in space, including atomic fountains and clocks involving trapped ions, is being undertaken currently. However, the frequency area involved in the design of an atomic clock was changed recently in a radical way. Equation 11.1 shows that the performances of atomic clocks can be improved when the nominal frequency is increased toward the optical domain. Forbidden optical transitions (with a very weak excitation probability) have narrow natural line widths (ca 1 Hz), resulting in a high Q-factor (ca 1015, at frequencies of several hundreds of terahertz). During the past ten or so years, many projects for optical atomic clocks have been initiated. Laser cooling
333
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future! 0.5 ωx
0.45
ωz
0.4 0.35 0.3 0.25 0.2 ωz–ωx
0.15
ωx+ωz
0.1
2ωx
0.05 0
–1.8–1.6–1.4
–0.9–0.7
–0.2 0 0.2 ∆ (MHz)
0.7 0.9
2ωz
1.4 1.6 1.8
FIGURE 11.2 Calculated Doppler spectrum of a dipole-forbidden transition of a single trapped ion oscillating at frequencies ωx/2π = 0.7 MHz and ωz/2π = 1 MHz and cooled down to the Lamb-Dicke regime such that the Lamb-Dicke parameters are 0.1 in both directions (the laser wavevector has the same projection along the coordinates axes Ox and Oz). (Figure courtesy of Caroline Champenois, PIIM, Marseille.)
permits us now to approach more closely the realization of an ideal system, that is to say, the attainment of atoms at rest and kept in a pure or perfectly-controlled environment. The application of the simple laser cooling technique to stored ions is facile, highly efficient, and permits resolution of the band structures of the Doppler profile in the optical domain. It is even possible to put the ions in the first quantum levels of the potential well, thus leading to a profile with only few bands (the Lamb–Dicke regime, Figure 11.2). In the Lamb–Dicke regime, the amplitude of ion motion in the propagation direction of the state-manipulating radiation is much less than λ/2π, where λ is the radiation wavelength. In other words, the Lamb–Dicke limit establishes functionally a ‘temperature’ for the ions that are to be manipulated, which can be determined from the amplitude of successive lateral bands. The theoretical limit of the temperature is the photon recoil limit, which can be attained once good experimental conditions are obtained. It is worth noting that, due to the ion/ion interactions, only ‘one’ single ion can reach the lowest levels. Along with ultra cold atoms in an optical lattice, single-ion clock schemes possess all the desirable properties necessary to reach the 10 − 18 uncertainty level. However, the precision of the local optical oscillator must be consistent; this condition requires local lasers emitting around 400 THz to be as narrow as 1 Hz, and that is a rather formidable challenge! To summarize the above paragraph, trapped ions can serve as exceptional tools for new and functioning atomic clocks, in both the microwave and optical frequency domains.
334
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
11.4 THE BEST TRAPS FOR THE BEST CLOCKS In this section, different and various geometries used for atomic clocks will be presented. It is essential that the efficiencies of these devices be sufficient to yield a pure system under well-controlled environmental conditions. Different routes have been followed in order to achieve the optimum signal-to-noise ratio, S/N. As mentioned above, two classes of atomic clocks exist; these classes are either microwave or optical frequency standards, and the constraints imposed by each upon the trap geometry are different.
11.4.1 Traps for Microwave Frequency Standards Possible candidates for a microwave frequency standard were identified by measuring the hyperfine structures of heavy elements, such as the odd isotopes of barium, ytterbium, and mercury, while the candidate elements were confined in an ion trap. The measurement involved double optical-pumping techniques. Thompson has presented a review of these early works [16]. These early experiments involved the usual hyperboloidal ion traps, though the RF field was perturbed to various degrees by large holes in the electrodes; the holes permitted both the passage of laser beams and the detection of fluorescence while the microwave frequency was being swept. These traps offered a volume of almost 1 cubic cm and an ability to store ca 105–106 ions in a potential depth of some tens of electron volts [17]. However, the ion number required to obtain an adequately high signal defines the ion cloud volume, the size of which results from the balance between Coulomb repulsions and ion kinetic energy (KE) that, in turn, depends on the amplitude of the RF drive voltage applied to the trap. This KE infers frequency shifts and frequency widening in the Doppler spectrum due to the second-Doppler effect. In order to suppress this effect, following the scheme of a race track (that is, a mass spectrometer bent to close the ion path), Dehmelt [18] proposed a linear trap with an electrode at each end and upon which a DC voltage is applied; this protocol, which permits the storage of only ions of the same charge sign, is applicable to ion clocks because one ion species is stored. This geometry favors a linear ion cloud strung out along the axis with a very small radial expansion, thus the micro-motion amplitude is weak. The confinement in the radial direction is identical to that of a quadrupole mass filter, but it is weakened by the addition of the static potential due to the static voltage on the electrodes at each end of the rod assembly. In the following section, the main properties of the linear ion trap and their effect upon the ion dynamics, particularly with respect to the potential distribution, will be described, followed by descriptions of some devices that have been realized. 11.4.1.1 Properties of Linear Traps The first idea was to build a trap from the linear quadrupole mass filter structure but, rapidly, the properties of multi-pole potentials were exploited also. For a quadrupole linear trap, the RF electric field [19] is transverse to the z-axis of the ion trap; near this axis, the time potential, ϕ, in the x- and y-directions can be expressed by
φ=
V ( x 2 − y2 ) cos Ω t 2r02
(11.2),
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
335
where r0 is the distance from the axis to the surface of the trap electrodes, V is the zero-to-peak amplitude of the RF drive potential of radial frequency Ω, and t is time. The related harmonic pseudopotential well depth in the radial direction, Dr, is of the form
Dr =
eV 2 2m 2 2 ( x 2 + y2 ) = ω r ( x + y2 ) 2 4 e mΩ r0
(11.3),
where e is the electronic charge, m is the mass of the ion, and ω r = eV / 2 m Ω r02 is the oscillation frequency in the radial direction. Confinement along the z-axis of the trap is assured by a static voltage applied to the end sections, near the axis of the trap. The shape of the ion cloud results from a balance between the viscosity contributions (laser cooling, buffer gas collisions), the coulombic interactions, and the confinement potential [20]. The shape of the ion cloud can be controlled by the shape and positions of the end-section electrodes, as well as by the magnitudes of the DC voltages applied to them. In order to reach the Lamb–Dicke regime, the ion KE, which depends upon both the macro and micro-motion kinetic energies (KEs), should lead to a motion amplitude less than the wavelength of the clock atomic transition. The macro-motion KE depends on the working point and on the thermostat (provided by collisions with a buffer gas, laser cooling, external circuit coupling, etc.). Although of low magnitude, the micro-motion KE should be added to the macro-motion KE. This micro-motion KE cannot be avoided, because it arises from the RF power supply. In a more general case, for a given distance from the trap center, the magnitude of this amplitude depends on the shape of the potential well; the flatter the shape of the potential well, the weaker is the amplitude. Multi-pole ion traps, which have been studied rigorously and extensively by Gerlich, provide an attractive vehicle for the establishment of such a flat-bottomed well [21]. The advantage of such structures resides in the property that their pseudopotential well shape becomes flatter as the pole number increases. A multi-pole structure described as a ‘k-pole’ trap consists of 2k rods; to each rod is applied an RF voltage of the same amplitude with the polarity of the amplitude reversed for alternate rods. For example, a quadrupole mass filter having four rods is a 2k-pole trap, where k = 2, and where the instantaneous polarities of the RF voltage are plus, minus, plus, and minus. The cross-section of the pseudopotential well becomes flatter as the value of 2k increases, moving from a champagne flute profile to that of a beaker. In practice, 2k can assume values upto 16, 22, … The choice of the value for k depends on a compromise between the advantage accruing from a very flat pseudopotential well, of the form ca r2k − 2 (with 2k rods), and a weak pseudopotential well depth, ca r 2, as shown in Figure 11.3. For such trap, the radial potential, ϕ(r), at a distance r from the axis is given by
φ(r ) =
k 2 e 2V02 2 k − 2 r 16 mΩ 2r02
(11.4),
336
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Well depth normalized
(a)
(b)
Well depth
1 0.9
4-pole 12-pole
0.7
16-pole
0.6
V(rmax) [u.a]
V(r) [u.a]
0.2
8-pole
0.8
20-pole
0.5 0.4
0.1
0.3 0.2 0.1 0
0
0.5 r [u.a]
1
0
4
8
12 2k-pole
16
20
FIGURE 11.3 Shape (a) and depth (b) of multipole potential wells versus the pole number.
where k is equal to half the number of rods and r is the radial distance from the axis of the ion trap. The amplitude of the radial motion, R, will not exceed a value Rmax given by [21]
0.3 mΩ 2r02 Rmax = r0 k ( k − 1) eV
1/k − 2
(11.5).
The pseudopotential well depth decreases as V increases; this behavior is contrary to that of the quadrupolar case. When k increases, the well depth can be approximated by
Dr ,k >>1 ( Rmax ) =
mΩ 2 r02 η2max , 16 k2
(11.6)
where η is the adiabicity parameter; it quantifies how the change of the RF field over the oscillation distance is smaller than the field itself, with the result that η can be used as a stability parameter. The longitudinal motions of different ions are coupled, thus giving rise to common oscillation modes. In quantum terms, this property is used to transfer information on the internal state of the atom from one ion to another [22].
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
337
FIGURE 11.4 The details of the DC end-cap needle electrodes used to prevent ions escaping along the longitudinal axis. (Reproduced from Prestage, J.D., Dick, G.J.; Maleki, L., J. Appl. Phys. 1989, 66, 1015–1017. With permission.)
11.4.1.2 Quadrupole Linear Trap at Jet Propulsion Laboratory (JPL) Prestage et al. have described an example of a quadrupole linear ion trap [19,23]. The first linear ion trap, designed to store mercury ions, consisted of four molybdenum rods spaced equally about a cylinder of radius ca 1 cm. The ‘end-cap’ electrodes, located at each end of the rod assembly shown in Figure 11.4, were separated by a distance of ca 75 mm. The end-cap electrodes were shaped like needles so as to achieve the electrostatic effect required to realize an ‘infinitely’ long ion cloud. The diameter of the needle-shaped end-cap electrodes, which was much smaller than that of the ion trap, imposed the same magnitude on the diameter of the ion cloud yet perturbed only slightly the RF trapping field. The needles, to which a DC bias was applied, were made of oxygen-free high conductivity (OFHC) copper. The four molybdenum rods localize Coulomb interactions in the axial region to within a length of the same order of the radius, ca 1 cm. The RF voltage was 180 Vp−p at a frequency of 2π × 500 krads s−1. These conditions lead to a secular frequency for Hg + of 2π × 50 kHz. In such a geometry, the ion density n 0 and the ion number N per unit length L are given, respectively, by
n0 = 2ε 0 mω 2 /e 2 and N / L = n0 πr02
(11.7)
where ε0 is the permittivity of free space and it is assumed that the secular frequency of the radial motion and that of the axial motion are each equal to ω. 11.4.1.3 Linear Ion Trap at National Institute of Standards and Technology (NIST) The first linear trap [24] at NIST consisted of four parallel rods, with circular crosssections in the plane perpendicular to the axis of the trap, see Figure 11.5. Each of the circular trap rods was divided into two sections of unequal length, so as to form three parts: two regions where the end segments of the two rods overlap the central
338
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V ∆U3
V0 cos Ωt + ∆U1
U0 + ∆U5 ∆U4
V0 cos Ωt + U0 V0 cos Ωt + ∆U2
z
y x
FIGURE 11.5 A schematic diagram of the segmented trap showing the potentials applied to trap segments. (Reproduced with the permission of the NIST Ion Storage group.)
Berylium copper Insulating spacer
2.5 mm 1.588 mm 12.5 mm
FIGURE 11.6 The electrode design of the segmented trap. (Reproduced with the permission of the NIST Ion Storage group.)
segments of the other two, and a central region where all four longer segments overlap. The RF potential was applied to two diagonally opposite rods. The segments of each rod are RF common so that each central segment can be biased. In order to confine positive ions axially, the remaining rods were grounded. Anharmonicity can be minimized by the selection of the ratio of the rod radius to r0 close to 1.1468 [25] then to 1.14511 [26], so as to compensate for using round rods in place of hyperbolic rods. The actual ratio value used here was 1.03 for which the anharmonicities remain weak. DC potentials are applied to each end of the rod assembly. Practically, the longer sections are grounded, and the shorter sections are held at a DC potential, U, of 1 V or less. Finally, the static potential provides a minimum in the central region of the trap, while adding greater anharmonicity. The rods, which are made of beryllium copper, are 12.6 mm in length, 1.588 mm in diameter, and the value of r0 is 1.563 mm, the shorter segment is 5 mm long and the longer 7.5 mm. The central region of the trap where the four segments overlap is 2.5 mm long, see Figure 11.6. These segments are isolated electrically using spacers of alumina and
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
339
machinable ceramic, in order to allow independent DC potentials to be applied. The insulating spacers are designed to inhibit accumulation of surface charge. To compensate for contact and patch potentials on the trap electrodes, small static bias potentials can be applied to the central segments of the individual rods and to one pair of each segment. The three voltages required are supplied from a helical resonator that can attain 700 Vp-p at 12.7 MHz with ca 5 W, and a DC voltage of up to 1 kV. Under these conditions, the axial potential well (400 V) is weaker than the radial pseudopotential one and the conditions to reach the Lamb–Dicke regime should be fulfilled. This trap can stored a long string of laser-cooled mercury ions and has permitted observation of the microwave atomic transition. A new trap was then designed [27] to limit the KE of the maximum ion number possible. Indeed, ion laser cooling in a cryogenic environment was employed to prevent extra Doppler shift. These conditions explain the choice of smaller dimensions, that is, a quasi-miniature ion trap (see Section 11.4.2.2). The linear RF quadrupole trap with four rods (0.20 mm radius) for radial confinement, centered at 0.64 mm from the axis, is configured as in an RF mass analyzer (Ω/2π = 13 MHz and V = 50 V). A static axial positive potential for longitudinal confinement, allows the confinement of many ions along the axis, that is, the zero-field node line. The static DC potentials of 25 V can be applied at the ends of the trap onto biased rings, pins (more or less sharp) or onto split sections in the trap rods. Resulting motional frequencies are: ωr/2π = 350 kHz and ωz /2π = 25 kHz. The ring electrodes located at each end of the rod assembly and biased for axial confinement are drilled at an oblique angle to permit laser access to the confined ions. The shape and dimensions were chosen so as to allow light collection from the ion trap center, and to define instability for unwanted contaminant ions. The electrodes are arranged in order to form two traps in tandem, one for capturing ions in the initial loading, and the other one, closed by the rings separated from 4 mm, for receiving the loaded ions for subsequent experimentation. This design should avoid the effects of contact potentials or electric charge build up. To avoid collisions with background gas pressure, the vessel containing the trap is maintained at liquid helium temperature, leading to a gas pressure of ca 10 − 14 Pa at 4 K that would permit ion confinement for a period of several days. 11.4.1.4 Multi-Pole Traps at Jet Propulsion Laboratory (JPL) In the first evolutionary stage in the development linear traps, two traps resembling quadrupole mass filters were used to create an extended linear trap [23]. In order to combine the advantages of different geometries, a new double structure was envisaged [28] that consisted of a quadrupole ion trap combined with a multi-pole trap. The projected disposition of this structure was to define two zones, each of which would be devoted to a unique function in the attainment of a clock signal. The first zone is similar to the trap described in Section 11.4.1.2. The second zone was designed to permit a stronger reduction of the second-Doppler effect, in the absence of the light pressure effect, then to store a very stable number of ions with the weakest possible KE. This linear extended ion trap (LITE) called also Shuttle trap [28] consists then of two separate regions, a magnetically-shielded region for microwave interrogation, and a separate region for ion loading and internal state preparation/detection. Ions are shuttled back and forth between these two regions
340
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Ion loading and Fluorescence trap region (10 cm) Transition zone (10 cm) Resonance trap region (45 cm)
Solenoid magnet for resonance trap
FIGURE 11.7 The ‘one liter trap’ or linear trap ensemble at JPL. (Reproduced from Tjoelker, R.L., Prestage, R.L.; Maleki, L., Proc. Precise Time and Time Interval, PTTI. 1996, 26, 235–243. With permission.)
FIGURE 11.8 The hexapole trap at JPL. (Reproduced from Tjoelker, R.L., Prestage, R.L.; Maleki, L., Proc. Precise Time and Time Interval, PTTI. 1996, 26, 235–243. With permission.)
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
341
using only DC potentials; the transfer is effected by dropping the potential barrier between the traps and increasing simultaneously the potential of the opposite electrode from whence the ions came. The static DC potentials on the end-cap electrodes are of the order of 10 V. The radiofrequencies of the traps must be incommensurate in order to avoid phase correlation. In Section 11.5.1, the application of this trap as a microwave frequency clock will be examined together with an assessment of the quality of the principle of this design in frequency time metrology. A one-liter ion clock is now designed for use in space [29]. The trap (Figures 11.7 and 11.8) is made of rods brazed into three alumina rings on each end and at the junction between the quadrupole and the 16-pole regions. The overall length is about 17 cm and the outside diameter about 1.5 cm.
11.4.2 Traps for Optical Frequency Standards 11.4.2.1 Single Ion Preparation and Clock Signal Building As explained above, moving time/frequency standards into the optical domain opens the way to a new generation of standards with ultimate performances. In fact, trapped ions can be good challengers, when the scheme is addressed with but one single ion. Only with this configuration can the environmental conditions be established to reduce sufficiently the causes of spectral widening, provided the ion lies in a quasirest condition. In other words, the optimum position for the ion is to be in the center of the trap in which the ion can remain confined for ultra long periods of time. Due to the working principle of an ion trap, the only way in which this condition may be fulfilled is to cool the particle drastically. Much experience has been gained in the successful technique of laser cooling, which can be applied readily in a Paul trap by exciting a transition with a slightly off-resonance laser, red shifted. Because absorption occurs in a fixed direction and photon emission is isotropic, at each absorption/emission process, momentum conservation leads to a slowing down of the ion, in the direction of the incident laser, by a quantity k, where is equal to Planck’s constant h divided by 2π, and k is the wave vector of the incident radiation. For the particular ion species used, each elementary process reduces the velocity by a few centimeters per second. With a dipolar transition, almost 109 absorption/emissions events will occur in one second. In addition, the ever present small quantity of anharmonicity of the trap will couple the three motion components resulting in a slowdown of the ion, not only in the direction of the laser but in all directions, until a temperature that is of the order of millikelvins is attained; this is limited by the natural spectral width of some megahertz. It should be noted that ion temperature is defined as the averaged KE in three dimensions. The technique of sub-Doppler laser cooling can be applied as well to ions as to neutral atoms; however, for cooling ions, thanks to their pendulum motion, only one laser along one component of the motion is required. The principle of laser cooling itself affords a possibility to detect the ion. The usual electronic methods of detecting image currents or measuring the time-of-flight are not sufficiently sensitive to permit the detection of a single particle. Fluorescence techniques are required here. Moreover, fluorescence techniques give essential information on the internal state that is related itself to the local oscillator interrogation. To detect fluorescence with an adequate S/N, it must be detected from a strong
342
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
transition, such as a resonance transition. In fact, the fluorescence is detected from the cooling transition itself. A photomultiplier placed behind a window will receive only a small part of the global photon number emitted in 4π steradians. Usually, the expected signal is sufficiently high to show the presence of the ion. Optical detection can be carried out also by imaging the ion cloud. Usually, a beam splitter allows the observation of two different signals and a suitable lens can be used to optimize the collection of the light. When initiating an experiment, it is far from obvious how one might succeed in storing only one ion. Generally, ions are produced by electron ionization of the components of an atomic beam emitted from an oven. However, in practice, many ion species can be prepared in situ, that is, within the ion trap by the technique of photoionization [30], which permits both creation of the correct or desired isotopes, when not separated previously, and facile control of the ion number. When the heating of the oven is reduced, the atomic beam will be less intense but this action will not guarantee the confinement of only one ion. It is easier to produce a cloud of ions, then to reduce the ion cloud size by heating ions with the cooling laser shifted continuously to the blue side, and/or by varying the depth of the potential well. Then, it is possible to obtain a small cloud of ions. Monitoring the magnitude of the fluorescence signal permits control of the size of the cloud. With very few particles, it is possible to evaluate the contribution of one ion thanks to the possibility of performing quantum jumps. Assuming it is possible, still in the presence of the excitation laser, to prepare the internal atomic levels in such a way that interrupts for one (or more than one ion) the absorption/emission cycle (involving atomic levels ⎪1 > and ⎪2 >), the height of the fluorescence signal varies in a discrete way, and the quantum of this variation corresponds to the fluorescence of one ion. This technique of quantum jumps is used effectively to control the ion number through the fluorescence signal until only one ion is confined. One way in which to initiate the occurrence of quantum jumps is by collisions, for instance with residual gases, or by implanting a specific optical pumping scheme; one can prepare the atomic ion in another atomic level ⎪3 >, which is not involved in the laser excitation process. Then, no more resonant absorption/emission of photons from the ion will exist and the level of fluorescence will decrease proportionately to the reduction of photon number emitted by one ion during a time depending on the lifetime of the ⎪3 > level (generally a long-lived or metastable state) as shown in Figure 11.9. From the height of the signal and the number of the steps of the upper level in the figure (here there is only one step corresponding to one ion), it is then possible to know exactly the number of atoms. A fine detuning on the red side, while observing the fluorescence permits one to heat and then to eject ions one by one. Once one has succeeded in trapping but a single ion, then the ion must be placed in the lower vibration levels (n = 1, 2, 3…) in the trap. First, micro-motion reduction is required. The amplitude of this micro-motion can be controlled thanks the so-called ‘RF photon correlation’. This procedure consists of modulating the output signal from the photomultiplier at the frequency of the RF drive potential. The amplitude of the micro-motion is then revealed by the amplitude of the fluorescence which depends on the first-order Doppler shift. It is, therefore, relatively easy to use precisely either laser detuning or to modify the potential voltages so as to reach the maximum reduction of the amplitude that corresponds to the lowest KE (Figure 11.10).
343
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future! 4P1/2
3D5/2
Signal (1000cts/200ms)
4S1/2
80
100
t (s)
120
140
FIGURE 11.9 An example of a quantum jump signal. The upper signal corresponds to the fluorescence emitted at the 4P1/2 → 4S1/2 transition; the transition rate is ca 10 ns. The lower signal indicates that the ion is in the 3D5/2 state; the lifetime is ca 1 s. The inset shows the spectral terms involved.
To carry out successfully a frequency metrology experiment, the initial step is to eliminate the first-Doppler effect, which consists here of resolving the lateral bands as in the microwave case (see Figure 11.2). Two conditions must be satisfied. First, the laser width used to probe the clock transition must be either narrower or of the same order as the natural width of the clock transition; this condition is already a prerequisite for the correct functioning of the clock servo loop. Second, this natural width of the clock transition must be narrower than the secular frequency of the ion; note that this criterion is not fulfilled with the cooling transition, which is, for the major part of the level configurations used for such atomic clocks, largely wider, for example, ca 0.1 MHz compared with some Hertz! This second condition is bound directly to the ion ‘temperature’, or rather its KE, because the bands are resolved when the amplitude of the motion is of the order or smaller than the wavelength of the clock transition, that corresponds for optical transitions to a temperature of the order of some microkelvins. The frequency of the probe laser is then corrected from an error signal arising from the number of jumps recorded at two different frequencies from each side of the central frequencies. 11.4.2.2 Miniature Ion Trap Design From these remarks, it is now possible to have a better idea of a suitable trap. For a given system (that sets also the clock wavelength), it will be easier to resolve the spectral profile when the bands are spaced far apart. Then, the trapping device will have to work with higher secular frequencies. Because the secular frequency depends on the RF drive frequency, this latter frequency must be so high as possible. Combining this requirement with the stability motion conditions infers very high alternating
344
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Fluorescenceintensity
(a)
–2
–1
0
Time
Detuning doppler shift (b)
(c)
70
Compensation off
50 40 30
200 cps
Signal (arb.u.)
60
20 10 0
0
20
70
40 60 Time (ns)
–500 –400 –300 –200 –100 0 Laser detuning (MHz)
100
Compensation on
50 40
100 cps
Signal (arb.u.)
60
80
30 20 10 0
0
20
40 60 Time (ns)
80
–500 –400 –300 –200 –100 0
100
Laser detuning (MHz)
FIGURE 11.10 Minimization of the micro-motion with the correlation method. (a) principle scheme; (b) observed efficiency on the micro-motion motion; (c) observed effect on the velocity reduction through the Doppler profile.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
345
electric fields that require RF voltage amplitudes that are not attainable easily in the context of a centimeter-size trap. For this reason, the dimensions of traps for storing single ions must be on the millimeter scale. In any case, a trap constructed on the millimeter scale is advantageous because it is easier to confine a few ions, even one ion in a small volume, rather than to trap a few ions in a large volume. The suitable range of the RF drive frequency is ca 10–50 MHz. It is not easy to machine traps of such a small design and to retain the ideal hyperboloidal form. In addition, optical beam access becomes very restricted, when the trap size is diminished. Fortunately, millimeter-scale ion traps are well suited to store the particle in a very small volume (that is, of the dimension of the wavelength). In the region close to the center of the device, the conditions to obtain a quasi-pure quadrupole field are not too constraining. The goal consists essentially to realize a small trap, the geometry of which enables the creation of a suitable confinement field, and with a structure sufficiently open to permit illumination of the ion in the trap and efficient collections of photons emitted by the single atom. A variety of trap shapes have been proposed. In a first category, they are almost faithful reproductions of a classic trap, with electrodes approximating a ring and two end-cap electrodes. Some others have fewer electrodes; their geometries are derived essentially either from a pseudo-ring or from end-cap electrodes. In all the cases, the shapes of the electrodes are easy to machine. These electrodes can be tested by appropriate software as SIMION [31] or other finite-element based programs. The suitability of the electrodes can be evaluated also by calculating their ability to store ions, in comparison with the ideal quadrupolar hyperboloidal geometry [32], to which the designers of micro traps aspire, and by dedicated experiment. The first ‘small’ trap designed for storing reduced numbers of laser-cooled ions was created in Heidelberg in 1978 [33]. This trap, which stores some 10 to 20 ions, was made up of a torus of semi-circular cross-section of 0.32 cm internal diameter and two hemispherical caps. An RF drive frequency of 2.6 MHz at 200 Vp-p results in a well depth of 10 eV and an axial secular frequency of 370 kHz for the Ba + ion. Before laser cooling, the ion cloud diameter was ca 250 μm in the presence of viscous drag cooling. A NIST ‘traditional’ [34] trap for the 199Hg + ion, and constructed recently of molybdenum, consists of a ring of diameter 0.6 mm and two cylindrically-shaped end-cap electrodes, see Figure 11.11. When an RF potential of 1000 Vp-p oscillating at 12 MHz was applied to the ring electrode, secular frequencies of ca 1 MHz were obtained. At the German National Institute, the Physikalisch-Technische Bundesanstalt (PTB) in Braunschweig, a miniature trap was designed to store 171Yb + with a ring of 1.3 mm diameter. The oven and the filament act as compensation electrodes to control ion micro-motion in three dimensions [35]. 11.4.2.3 An Example of an End-Cap Electrodes Trap The NPL uses for its projects with 88Sr + (and 171Yb +), an end-cap electrodes trap made from tantalum [36].
346
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V RF
12 MHz, I kV
Secular frequency: νr ≈ 1 MHz
FIGURE 11.11 Cryogenic spherical RF (Paul) trap at NIST for Hg+. The small white spot shows the fluorescence of the single Hg+ ion. (Courtesy of NIST.)
V1 O
Vac cos Ωt
V2 O
FIGURE 11.12 A schematic diagram of the end-cap electrodes trap employed at the National Physical Laboratory. (Figure courtesy of Helen Margolis, National Physical Laboratory, UK.)
The NPL (Figure 11.12) is based on the trap described by Schrama et al. [32] The outer electrodes (very fine hollow tubes of tantalum of 1 mm inner diameter and 2 mm outer diameter) are grounded, but they can receive a small DC potential. The inner electrodes, also in tantalum, are ca 0.5 mm diameter and separated from the outer electrodes by 0.56 mm. The amplitude of the RF potential is 260 Vp−p at a frequency of 17.8 MHz, resulting in secular frequencies for strontium ion of 1.8 MHz (radial) and 3.0 MHz (axial). The electrodes are isolated from each other with a ceramic tube and the trap is mounted on a structure of machinable macor. Two compensation
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
347
FIGURE 11.13 Paul-Straubel Trap designed at PIIM (Marseille).
electrodes, orthogonal to each other and orthogonal to the trap axis, are placed to correct the micro-motion. Finally, three sets of compensation voltages of but a few volts each and directed along three axes maintain the ion at the center of the trap. 11.4.2.4 Ring Traps The geometry of the first miniature trap used in Marseille, see Figure 11.13 [37], was of the Paul-Straubel type selected for a good visualization of the confined particle and easy access for the laser beams [32]. The trap consisted of a cylindrical ring with an inner diameter of 2r1 = 1.4 mm and an overall height of 2z2 = 0.85 mm. Two compensation electrodes were formed from flat rings having an outer diameter of 13 mm; the inner diameter was 10.5 mm and the open center of the ring was covered with a micrometric wire mesh having an optical transmission of 86%. These electrodes were placed at a distance of 5.5 mm from the center of the trap and they screen the inner volume of the device from stray electrical fields. The electrodes (mesh and bulk electrodes) were made from molybdenum to avoid observable magnetic effects. The application of a constant DC potential to the compensation electrodes permitted axial compensation of the potential within the trap. In order to superimpose the center of the electrical field upon the geometric center of the trap, positioning electrodes were set in the ring plane. Each of the two flat electrodes was tapered to a point and the two electrodes were positioned perpendicularly to each other in the x and y-directions. A DC voltage of a few volts could be imposed upon each of the two flat electrodes. Working conditions were defined with the compensation electrodes grounded. The RF drive frequency was chosen to be Ω/2π = 11.6 MHz and the amplitude of the RF voltage was 1980 Vp-p; the resulting potential well depth was then ca 9.2 eV. The ambitious goal of putting one ion at rest requires a perfect knowledge of the confinement capability of this trap. In an attempt to achieve this goal, two different methods were employed. The first method consisted of applying a tickle voltage, that is, a supplementary RF potential of low amplitude to one of the end-cap electrodes with the other end-cap electrode held at ground potential [38]. When sweeping the tickle voltage
348
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
frequency, resonances will occur such that, when the frequency of the tickle voltage approaches one of the ion motion frequencies, ion motion will increase in amplitude. In this manner, the ion cloud is heated leading to a decrease in the fluorescence signal; the decrease arises from (i) the increase in ion cloud size due to the heating effect such that, with the accompanying reduction in ion density, some ions are no longer located within the beam of the laser, (ii) heated ions remaining in the laser beam are no longer ‘in tune’ with the laser due to variation of the Doppler effect, and (iii) some heated ions will be neutralized on the trap walls. Scanning the tickle frequency from 0 to Ω/4π Hz permits precise observation of all the frequencies of the real motion including frequencies of radial and axial motion components; the radial and axial motion frequencies are close to but differ from those values calculated from the Mathieu equation. Observed also are those frequencies that arise from radial/axial and macro/micro-motion couplings [39] due to additional non-quadrupolar terms and to ion/ion interactions. The second method involved the application of a DC potential to the connected end-cap electrodes and varied adiabatically, that is, the DC potential was varied slowly such that the ion cloud was not disturbed by this variation. The variation of the fluorescence as a function of the adiabatically-varied DC voltage, in the absence of any other external excitation, reveals two important properties of trapped ion dynamics; first, the boundaries of the stability diagram and, second, the intensity of possible black canyons (see Chapter 3 in Vol. 1 of this series). The frequency spectrum gives information on the dynamic properties of the ion motion and the storage capability of this home-made trap. The amplitude of the tickle was chosen to be sufficiently large to reveal the main couplings that existed without losing any ions. Figure 11.14 shows a series of frequency spectra formed by inverted peaks in the variation of the collected fluorescence. The loss in the fluorescence signal can reach up to 80% of the overall signal indicating, in this case, a resonance of high strength. Indeed, at the occurrence of a resonance, the energy deposition onto the ions in resonance leads to an increase of the Doppler broadening and fewer ions are in resonance with the laser excitation. Moreover, larger trajectories of heated ions inflate the cloud size and produce a decrease in the ion density. Both effects reduce instantaneously the fluorescence intensity. When the tickle frequency is swept past a resonance, the fluorescence signal is restored immediately to its value prior to the resonance, indicating a persistent cooling of the cloud. We can deduce then that the ions do not leave the laser beam during resonance episodes. Occasionally, the breakdown of fluorescence can be attributed to the effective loss of ions from the trap. As the unipolar mode excitation is equivalent to simultaneous excitation of dipole and quadrupole excitations, unipolar mode excitation can reveal parametric excitations in the macro-motion as well as radial/axial resonances. The most intense non-linear resonances, ωz /2, 2ωx, 2ωz, 4ωx, were assigned with a precision greater than 99%. The frequency combination ωz /2 + ωx, was found also; this coupling can be explained by the cylindrical shape of the trap. Only in the central region of the ring electrode is the field within the device described here similar to that of an ideal Paul trap. Hence, in order to produce the same potential well depth as in the ideal Paul trap, the amplitudes of the voltages applied to the cylindrical miniature trap must be greater. In order to maintain the
349
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
583 Vrms 634 Vrms 685 Vrms
4000 cps
737 Vrms
0.5
1.0
1.5
2.0
w (MHz)
FIGURE 11.14 Fluorescence at 397 nm as a function of the applied tickle frequency. The graph shows part of the frequency spectrum of a small Ca+ -ion cloud for different values of V and with U = 0. Once the ions are in resonance with the excitation frequency, the energy absorption leads to a sudden decrease of the fluorescence.
definitions of the parameters of stability, a and q, of the Paul-Straubel trap compatible with those of the ‘traditional’ case, a correction factor L must be introduced in the usual relations [32]
az =
−8eU , mr12Ω 2 L
qz =
4eV mr12Ω 2 L
(11.8).
Experimental and theoretical studies have shown that for the standard Paul ion trap the maximum cloud size can be found at the working parameters around qz = 0.55 and az = −0.03 [40]. Taking into account these values in the measurements of the amplitude of fluorescence as a function of the applied RF voltage V allows us to make a rough estimation of the correction factor for our trap to be Lz = 6.8 ± 0.5. A more precise technique for calculating this correction factor for a given working point, consists of measuring the fundamental frequencies of the ion motion. Equation 11.8 allows an estimation of L using the adiabatic approximation along the qz-axis of the stability diagram where U = 0. The correction factor L was evaluated from the measured values of ωx and ωz. A correction factor was introduced for each direction. As a result of this operation, the usual equations have to be replaced by
az Lz = −2aX LX and qz Lz = −2q x Lx
(11.9).
The main contribution to L is the loss of trapping efficiency of the cylindrical trap compared to an ideal Paul trap due to the discrepancy between the geometries, which
350
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
are not identical in the radial and axial directions. Patch potential effects due to parasitic solid deposition may build up over time mainly on the ring electrode. These effects would lead to a distortion in the rotational symmetry of the confinement potential and then to a possible difference between the x and y-directions. These effects can vary slightly over time. Loss factors depend on the ion cloud size. The larger the cloud, the greater is the number of ions that escape and the greater is the discrepancy from the quadrupolar potential. Also the values of Lz and L x are subject to variation. It has been observed experimentally that, for RF potentials V increasing from 90 to 100 V, Lz decreases from 8.0 to 7.6, whereas L x increases very slowly from 7.0 to 7.1. For a very high RF potential, due to the increase of the well depth, the two correction factors should become even closer because the ion cloud is concentrated in the center of the trap, where the potential shape is less sensitive to deviations from the quadrupole case. These loss factors will be used to calculate the beta parameters, βx and βz, of this trap* that define the secular frequencies. In order to show the effective boundaries of the stability diagram, the following experimental procedure was carried out. First, an ion cloud (ca 500 ions) is created at an RF voltage V with a zero potential U applied to the end-cap electrodes. Then, a positive (or negative) potential U is applied to the end-cap electrodes (which is equivalent to the application of–U on the ring) and is increased slowly from 0 to 150 V. During this scan, the overall fluorescence of the ion cloud is recorded. The boundaries of the stability diagram are clearly illustrated by the disappearance of any fluorescence signal due to the complete loss of the ions. A new cloud is created for each scan at U = 0. These measurements have been carried out for V varying between 250 and 500 V. For a clear visualization, the effective iso-β lines must be drawn and thus βx and βz can be calculated. Using the correction factors Lz and L x found above, Equation 11.8, the measured boundaries from (U, V) coordinates are transformed into (az, qz ) coordinates as shown in Figure 11.15. In the experiment, U = 150 V corresponding to az = 0.14 and V = 500 V leads to qz = 0.66. The resulting measured limits of the stability diagram reproduce the form of the theoretical stability region, except for its right-hand side boundary, βz = 1. Yet, a more precise analysis of the experimental stability diagram, taking into account the ω z measurements, shows that the observed right-hand side boundary corresponds to βz = 1/2. The non-linear resonances occurring along this line are so strong that no ions can stay in the trap. Moreover, on the high-qz side of the βz = 1/2 canyon, a multitude of close-lying canyons can exist. Another characterization of the ion cloud consists of locating its motional resonances in a (V, U) diagram from the absences in the fluorescence signal for given values of (V, U). Knowing the position of these canyons is very important in order to select the best trapping parameters. The fluorescence emitted by the cloud at fixed values of the RF potential V, while slowly scanning the potential U from 0 to 150 V exhibit non-linear resonances, as shown in Figure 11.16. Assignment of these resonances gives a projection of the inner shape of the stability diagram, see Figure 11.17. * See Volume 4, Chapter 1: An Appreciation and Historical Survey of Mass Spectrometry by Raymond E. March and John F.J. Todd.
351
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
0.1
az
0.0
–0.1
–0.2
0.0
0.1
0.2
0.3
0.4
qz
0.5
0.6
0.7
0.8
0.9
FIGURE 11.15 Limits of the stability diagram as observed from the fluorescence of a small Ca + -ion cloud as compared to the stability diagram of an ideal Paul trap (solid lines). The observed right-hand-side limit corresponds to a βz = 1/2 canyon; no ion could be confined beyond this boundary. α
VAC=789 Vrms
α 5000 cps
VAC=737 Vrms
α
–75
–50
–25 0 UDC (V )
25
VAC=685 Vrms
50
FIGURE 11.16 Cross-section of the stability diagram for different values of V. The canyon α, which corresponds to βz = 1/3, can be tracked for various working points.
352
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Most of these canyons can be assigned with an uncertainty lower than 7%. The assignment of the canyons βx = 2/9 and βx = 1/6 can be made only to the 7% level. In Figure 11.17, these canyons can be seen to have a slightly deviated direction. Finally, the βz = 1/3 canyon is found with a 10% uncertainty; this resonance has been observed to be large. The assignment has been confirmed by consideration of the inter-canyon distance. All of these canyons show the existence of couplings between the macro-motion along the axial and the radial directions with the RF alternating potential [39]. The resonances assigned to the βz = 2/5 canyons are observed as doublets; this kind of doubling could be explained by a defect of the axial symmetry of the trap. The assignations of the observed canyons inside the stability region are consistent with the previous assignation of the βz = 1/2 canyon. The present experimental stability diagram shows that, for the characterization of such a device as the miniature trap, the stability diagram boundaries alone do not provide sufficient reliable information for optimum use of the trap in the experiments.
11.5 ION CLOCKS: CURRENT RESEARCH In this Section, we will describe briefly the most recent projects of atomic clocks involving/based on ion traps as described above. The first part concerns microwave clocks, while the one following will be dedicated to optical frequency clocks. Performances of atomic standards can be evaluated only by comparison (frequency beatings) with another devices. When a new atomic standard can be presumed to out-perform the norm, it can be evaluated only from the comparison with a second system, which must be build in a similar way. It is worth noting that performances of each scheme depend on the local oscillator; a quartz (eventually, cryogenic) oscillator for the microwave range, and a laser for the optical one.
11.5.1 The Jet Propulsion Laboratory (JPL) LITS (Linear Ion Trap Standard) Project at 40.5 GHz The Joint Physical Laboratory with the National Aeronautics and Space Administration (NASA) at Pasadena is now the leader in the preparation of an extremely high performance clock in the microwave domain. Today, mercury is the only candidate to compete with cesium due to the facility to excite the hyperfine clock transition (⎪1 > → ⎪2 >) of ground level of 199Hg + at 40.5 GHz with 202Hg + (Figure 11.18). This transition, which occurs without variation of the sublevel magnetic quantum number, mF, such that ∆mF = 0, is not dependant on the magnetic field at low field strength. The attractiveness of the 199Hg+ /202Hg + system is due to the fortuitous coincidence of transitions of each of the 199Hg + and 202Hg + species. Through the use of a gas-discharge lamp containing the even isotope, it is possible to excite the first two resonance lines (194 nm) involving the upper level of the clock hyperfine transition (⎪2 >), followed by equi-probable relaxation of levels ⎪1 > and ⎪2 > . Continuous pumping at 194 nm of the remaining population in ⎪2 > will empty rapidly the ⎪2 > level and will reduce to zero the
353
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future! βz=1/2
150
βz=2/5
100
βz=1/3
0 –50
βx=1/6 βx=1/5
UDC (V )
50
–100
βx=2/9
–150
βx=1/4 500
600
700
800
900
–200 1100
1000
VAC [Vrms]
FIGURE 11.17 ‘Black canyons’ in a part of the stability diagram. Points of low confinement coincide with lines of constant rational βx or βz indicating a coupling of the macro- and micromotions. The lines are a guide to the eye, and indicate points of common β-value. The datum points () and (•) represent the iso-βz and iso-βx canyons, respectively. The measured limits of the stability diagram are displayed by (+); for clarity, the error bars on these values have been omitted in this graph. A few strong resonances, indicated by (X), could not be assigned with a common β-value at the 5%-level.
2P
1/2
F=1 F=0
6,9 GHz
194 nm
2S
1/2
F=1, mF=0 F=0, mF=0
40,5 GHz Hgll
2> 1>
FIGURE 11.18 First atomic levels involved of singly-ionized mercury (HgII) for the ion atomic clock at 40.5 GHz.
fluorescence level of the resonance lines. Application of a resonant microwave at 40.5 GHz on the ⎪1 > → ⎪2 > transition will cause the fluorescence to reappear. Probing the fluorescence of the resonance lines, will then permit determination of the efficiency of the excitation by the microwave source at 40.5 GHz. Tuning of this source, by a servo loop, will permit accumulation of the error signal. The evolution of the traps and their environment has been guided by the objective to reduce at each step the major systematic effects. This system has been exploited for a relatively long time [41,42], at different laboratories, including the HewlettPackard-Austin, Laboratoire de l’Horloge Atomique-Paris, etc. At the end of the last century, only NIST and JPL continued with this endeavor. At NIST, the traps
354
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
FIGURE 11.19 Prototype of the Hg + ion clock at 40.5 GHz, prepared for space utilization. (Reproduced from Prestage, J.D.; Chung, S.; Le, T.; Maleki, L., Proc. 2005 IEEE International Conference, Vancouver, BC, Canada. 2005. 472–476, With permission.)
discussed above were designed to prepare ions to reach a very high degree of accuracy (3 × 1015) at 40.5 GHz within a stability of 3.3 × 10 − 13 τ − 1/2 (τ < 2 h) [43]. However, at NIST, investigations in the optical domain were accorded priority (see Section 11.5.2.1). Sustained efforts have been made to improve the performance of hybrid traps in order to obtain the best results possible. Linear ion trap standard (LITS) 1 and LITS 2 were prepared for continuous high stability operation for ground-based application such as the NASA Deep Space network, while LITS 3 and LITS 4 were dedicated to the U.S. Navy Observatory (USNO) time scale. Further, the LITE (linear ion trap extended) was designed to separate the ion preparation zone and the ion interrogation zone. These traps, equipped with different models of local oscillators, were compared among themselves and with different masers. Linear ion trap standards with space flight specifications are designed to be included in the GPS clocks. Reduction of the second-Doppler effect is a determining criterion in the selection of a linear trap. The use of multipole LITS (Section 11.4.1.4) [44] reduces all ion number dependent effects resulting through the second-order Doppler effects. This effect depends indeed on the ion number, because its expansion contribute to the ‘ion temperature’, which is kept constant at 1% via a control loop, enabling the reproducibility of the working conditions. Collisions with neutrals, even at the 10 − 10 mbar vacuum level, must be controlled because they induce pressure effects. The pressure frequency shifts due to the buffer gas are minimized when neon [45], introduced with a capillary leak, is employed; neon produces a relative shift of 7.1 × 10 − 8 per Torr. Reduction of the Zeeman effect requires magnetic shielding and degaussing coils. In a recent publication [46], it was announced that the short-term stability is less than 4 × 10 − 14τ1/2 (1 µHz d − 1) and the residual systematic effects are less than 6 × 10 − 17, with a microwave-line-Q-factor of 5 × 1012 at room temperature. This clock is now used as the frequency reference for the JPL geodetic receiver.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
355
While still improving the earth-clock, the JPL group prepared a clock for u ltra-stable deep-space applications. A breadboard ion-clock package, based once more on Hg ions shuttled between a quadru and a 16-pole RF trap, has attained an accuracy of 2 × 10 − 13 at the 1 s (10 − 15 d − 1) stability level. The system is in a sealed vacuum configuration of about 2 L in preparation for flight, see Figure 11.19 [47].
11.5.2 Ion Optical Clocks In the field of ion optical clocks, the ‘bons mots’ are, indeed, ‘optical frequency standard based on one trapped single ion’. There are at this time very many projects of this kind. Each of them differs first by the ion species, then by the trap design, then by the optical cavity-laser ensemble and, finally, by the coupling between the atomic transition and the laser system, that is, the clock transition interrogation protocol. Only by practical realization of these projects will the best system be identified. The design of these traps is crucial because of a number of points. Systematic effects will depend not only on the environment, but also on the sensitivity of the atomic transition to its environment. The critical single ion must be placed precisely at the geometric center of the trap, which should be superimposed with the potential minimum, where these systematic effects are at a minimum. The characteristics of the environment must be well known, especially the ambient electric and magnetic fields (static and alternating). The power supplies, particularly the RF power supply, must generate a voltage of a pure frequency unencumbered by parasitic voltages at other frequencies. The ion production should be clean and controlled as much as possible so as to avoid patch potentials, which can disturb the shape of the potential even though compensation electrodes can reduce this effect. Finally, the geometry must permit the passage of at least two laser beams; one laser plays the role of the local oscillator (the pendulum), and a second laser is employed to cool the ion. Eventually, a third laser could be required in order to pump the ions back into the cooling-atomic-levels system. Two kinds of ion species are involved depending on their atomic level properties. One has two optical/peripheral electrons, such as Al + , In + , where the clock transition is based on a dipolar electric transition, and the other has only one optical electron, such as Ca + , Hg + , Sr + , and Yb + , for which the clock transition is based on either a quadrupolar or an octopolar dipole electric transition. With the first kind of ion, the cooling transition is cycling wherein 100% of the atoms relax to the lower level, while the cooling transition (nS to nP) of the second kind relaxes to two different-orbital lower levels: the fundamental (nS) and one metastable level ((n-1) D). The value of the relaxation branching ratio between the nS and metastable (n-1) D levels is such that a significant fraction of ions will populate the metastable (n-1)D level. Thus, another laser is required to pump the atomic ions from the (n-1)D level back to the optically excited state nP. New schemes are being developed currently, in addition to the usual type of schemes where the probe and the read out are done on the same ion target. This latter type of scheme takes the advantage of the progress made in quantum information with trapped ions, which offers, thanks to the quantum entanglement between ions, to provide information on the internal state of one ion from the observation on the internal state of the other one.
356
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
All the projects benefit hugely from the successful development in the frequency measurements techniques thanks to the optical frequency comb technique [48]. 11.5.2.1 National Institute of Standards and Technology (NIST) Mercury Projects This project [34], started 20 years ago [49], and it is still the ‘number one’ in the world! A frequency laser at 563 nm is doubled and servo-locked to the 282 nm 5d106s2S1/2(F = 0, mF = 0) → 5d96s2 D5/2(F = 2, mF = 0) electric quadrupole transition in a single laser cooled 199Hg + ion, which is held in a cryogenic trap (see Section 11.4.1.3). The clock laser is narrowed at 0.2 Hz onto a F = 200,000 high finesse (or quality factor) cavity constructed from an Ultra Low Expansion, ULE, spacer of 25 cm, isolated thermally and seismically; the short-term frequency lock of the laser leads to a drift rate of less than 1 Hz [50]. The cooling transition is the strong 5d106s2S1/2(F = 1) → 5d106p2P1/2(F = 0) transition at 194 nm (1064.7 THz). The state preparation and detection of the clock state are executed via quantum jumps after each probe of the clock transition by the 282 nm clock laser. Thanks to highly efficient magnetic shields with computer-controlled coils, periodic magnetic field measurements reveal an uncertainty of only ± 20 mHz. The quadrupole shift is eliminated by averaging the frequency measurements on three orthogonal orientations, resulting in a fractional uncertainty of 0.5 × 10 − 17. Estimation of all of the systematic effects (electric fields, gravity, second-Doppler effect, etc.) have yielded a correction of 1.727 Hz, with a total fractional uncertainty of 7.2 × 10ˉ17. The absolute frequency of the transition was measured with a frequency comb vs cesium to be 1064,721,609,899,144.94 Hz, with a statistically limited total fractional uncertainty of 9.1 × 10ˉ16, which is the most accurate absolute measurement of an optical frequency to date. 11.5.2.2 The Logic Atomic Clock This new scheme demonstrates the strong potentiality to implement quantum entanglement in high-resolution spectroscopy [51]. Quantum entanglement allows two ions to share or to have common state, in such a way that one atom is used as a spectroscopic source because it exhibits a promising transition. The second is used as a logic atom; it is used to cool the first atom and, due to quantum entanglement, to give the population probability of the upper level of the clock transition while it is probed by the laser clock. There are two main advantages of this scheme. The first advantage implies a sympathetic cooling technique.* Until now, the choice of the atomic transition was limited to atoms that exhibited transitions to which laser cooling could be applied. Moreover, cooling lasers must be available within the range of these transitions. Now, the metrologic ion is cooled by the other one, which increases the number of possible clock transitions. The second advantage is that the interrogation of the clock transition does not depend on the spectroscopic atom and so it is possible to choose a very weak, thus, ultra-narrow, transition.
* See Volume 5, Chapter 10: Sympathetically-Cooled Single Ion Mass Spectrometry by Peter Frøhlich Staanum, Klaus Højbjerre and Michael Drewsen.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
357
One single aluminum ion and one single beryllium ion are confined closely in a linear electromagnetic trap and slowed by laser cooling to a temperature close to absolute zero temperatures. The clock transition is the 1S0 → 3P0, transition in 27Al+ at 267 nm. This transition is extremely narrow, 8 mHz, and is insensitive to magnetic fields and electric field gradients. Information on the population distribution of the upper |∙ > and the ground |¯> levels of the spectroscopic atom (Al+) transition is transferred onto the population distribution of the (v = 1 or 0) trap vibrational levels. Because the two ions are entangled, a coherent pulse from a suitably-tuned red laser will transpose this information onto the population distribution of the (v = 1 or 0) trap vibrational levels of the second (logic) ion (Be+), which can be read finally after a projection with a suitable laser pulse onto the upper and lower levels of the concerned transition of the logic atom. For a given laser frequency, the transition probability is determined by repeating the probe many times. Sweeping the laser frequency allows one to obtain its ‘resonant’ or characteristic frequency, which can be determined with extreme accuracy. The frequency was measured with a fractional uncertainty of 5 × 10ˉ15 [52]. This story continues very nicely. Recently, and for the first time, two frequencies delivered by two independent atomic clocks (the Hg+ one described in Section 11.5.2.1, and the Al+ one) were compared using successively two independent frequency combs, locked on the NIST atomic clock. The ratio of frequencies was measured with an uncertainty of 4.3 × 10 -17 (a systematic uncertainty of 1.9 × 10 −17 for Hg+ and 2.3 × 10 −17 for Al + ). These measurements were repeated throughout a period of one year and yielded a preliminary constraint of −(1.6 ± 2.3) × 10 −17 yr−1 on the relative temporal variation of the fine-structure constant α [53]. 11.5.2.3 729 nm Frequency Metrology with Ca+ in Marseille A single Ca+ ion is cooled in a miniature radiofrequency trap (see Section 11.4.2.4) [37]. The electric quadrupole transition between the ground state and the upper level of the first metastable state (32 D5/2) of a single Ca+ ion is an attractive choice for a frequency standard in the optical domain (see Figure 11.20). This transition at 729 nm has a natural line-width of 140 mHz corresponding to a quality factor of 2 × 1015. Laser cooling is carried out on the resonance transition at 397 nm. Due to the branching ratio from the 4P-level of ca 20:1 in favor of the 4S-level relative to the 3D-level, a back pumping laser is required at 866 nm to excite the 3D population back to the 4P-level. The wavelengths of all these lasers lie in the infrared-visible domain and can be reached with solid-state lasers. The main theoretical contribution to the systematic uncertainty of this kind of frequency standard is the quadrupole shift that can be measured precisely. Then, the systematic uncertainty is expected to reach 3 × 10ˉ16, limited by the precision of the quadrupole cancellation and the Zeeman effect [54]. Stability is limited by the quantum projection noise and it is expected to reach 2.5 × 10−15 τ −1/2. The Ca + ion frequency standard is expected to out-perform the best microwave atomic frequency references, and to be competitive with other atomic optical frequency references. Probing the clock transition is carried out using quantum jump statistics that require cycle times of several seconds. The local oscillator is a laboratory-built titanium-sapphire laser pumped with 5 W of single mode laser radiation at 532 nm. Ultra-high stabilization using a high finesse ULE cavity (F ≥ 100,000) placed in a highly isolated environment (the
358
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 42P3/2 850 nm
42P1/2
854 nm
32D5/2
866 nm 393 nm 397 nm
32D3/2 729 nm
732 nm
42S1/2 Call
FIGURE 11.20 First atomic levels involved of singly-ionized calcium (CaII) for the ion atomic clock at 729 nm (light continuous line). One of the bold continuous line (397nm) indicates the transition used for the laser cooling process and the second other one (866 nm) permits the back pumping from the D3/2 toward the cooling cycle.
cavity has three thermal shields and is set in ultrahigh vacuum). The ensemble is isolated thermally and is set in an aluminum box. This box is set on an optical breadboard surface, which is isolated from mechanical vibrations. An interesting scheme is underway at this time in an effort to produce a frequency standard in the terahertz range [54]. The ultra-narrow quadrupole transition, combined with the two cooling lasers, can be used also to create three-photon-coherentpopulation trapping in a dark state. The objective of this experiment is to create a coherent superposition of the two metastable states. The resulting dark line can be employed for frequency metrology in the terahertz domain in a robust set-up with the interrogation of a relatively large ion cloud due to the first-order Doppler-free geometry of the laser beams. The objective for frequency stability lies in the 10−14 range. The referenced terahertz signal can be propagated over long distances, with the useful information being carried by the relative frequency of the three optical photons. 11.5.2.4 State of the Art of Optical Frequency Ion Clocks Naturally, in addition to the experiments described above, there are a number of other projects that are underway. While this final section is not exhaustive and does not report on all of the efforts and leading projects, it will show that many other approaches are the subjects of active inquiry. It is simple to know which schemes demonstrate the better performances however, it not so obvious to predict which scheme will be the one best suited to fit with the constraints of a frequency standard and will achieve the necessary levels of repeatability and reproducibility. Optical frequencies lie in the one hundred to one thousand terahertz range. These performances require a fine analysis of systematic effects and the estimation of their uncertainty [55]. Systematic frequency shifts can arise from several different
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
359
effects: (i) Zeeman shifts due to the external magnetic field; (ii) blackbody radiation; (iii) electric quadrupole shift; (iv) second-order Doppler shifts; (v) Stark shifts due to thermal motion; (vi) the micro-motion of the ion; (vii) applied light fields; (viii) blackbody radiation; and (ix) gravitational redshift (10ˉ16 mˉ1). Moreover, some of projects, depending in part on the ion system but also on the available laser sources for this ion species, are more or less robust with respect to the potential applications, especially those applications in space. Fundamental physics tests, such as parity violation experiments (with heavy ions), can be checked more easily with specific species [56]. Last, but not least, some systems may present less-than-impressive performances, but can be developed at very low costs, thus making such systems available widely for new kinds of experiments yet to be imagined that could involve large networks of clocks disseminated around and encircling the globe! The stability of the clocks depends on the ion, on the stability of the environment because it is the environment that is responsible of the long-term stabilization of a clock, and on the stability of the working conditions of the ion trap. The shortterm stability of the clocks depends on the short-term stability of the probing laser that is locked on the high finesse cavity. New designs for high finesse cavities with near careful total elimination (see Section 11.5.2.3) of vibration and temperature variations are required (for a cavity of 25 cm, 1 Hz corresponds to a variation of the optical length of 1 nm). However, this difficult task is beyond the scope of this monograph. Because the metrology system has the capability to overcome the performances of existing clocks, its properties (precision and stability) can be evaluated only by comparing the beating of frequencies with another identical system. Finally, the absolute frequency is then calibrated with a frequency comb [48]. It is worth noting that the rapid development of the frequency combs, their improvement, and the increasing understanding of this technique is transforming atomic and molecular spectroscopy and, almost certainly, will be used extensively in the future in new mass spectrometric ion trap devices [57]. At NPL, two projects exist. The first project concerns the strontium ion, 88Sr [58], which is studied also at the National Research Council (NRC) [59], in Ottawa. The clock transition is at 674 nm with a line width of 0.4 mHz. The line has been explored already at a resolution of 5 Hz and the frequency has been measured to within 1.7 Hz. The second project is based on an octopole electric transition (S → F), 467 nm, studied also at PTB. The lifetime of such a transition has been estimated as of the order of one year, inferring a spectral width of 10−9 Hz. At this time, the frequency is measured with 11Hz in a line width estimated at 40 Hz. In addition to this exceptionally high Q-factor transition, both groups have explored the quadrupole electric transition at 436 nm of 3.1 Hz, measured at 2.2 Hz in a line width of 15 Hz. The 237 nm 115Indium line (Max Planck Institute at Garching and Erlangen University in Germany) [60,61] of 0.8 Hz natural width is expected to have an accuracy of 10−17, see Figure 11.21. The clock transition measured with an experimental line width of 170 Hz (1.3 × 10−13), has an accuracy of 230 Hz. Calcium experiments, in addition to those being carried out in Marseille are underway in Japan at CRL [62], and in Blatt’s group in Innsbruck [63]. There, the 729 nm (0.14 Hz) frequency was measured in a linear quadrupolar trap at 0.9 Hz with a 8.4 Hz line width.
360
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Excitation probability
0.6
0.3
0.0 –40
–20
20 0 Detuning at 436 nm (Hz)
40
FIGURE 11.21 An example of a clock atomic transition. The excitation probability of the clock transition (the atomic oscillator) is measured through the quantum jump number vs. the laser tuning of the local oscillator. Each probe pulse is of 90 ms duration, and twenty probe cycles were performed for each value of the detuning. (Reproduced with the permission of the Physikalisch-Technische Bundesanstalt.)
11.6 CONCLUSION In this chapter, the versatility of ion trapping, especially in small devices, and its potential for modern atomic clocks has been demonstrated. Among the successes of these projects is the proposal for a new definition of the second [64]. Several imaginative projects involving small ion traps and mass spectrometry are underway. Perhaps the next technique in high-resolution mass spectrometry will be based on the study of one single ion, resulting in a weak probability production, confined in an ion trap thanks to sympathetic cooling!
References
1. Paul, W. Electromagnetic traps for charged and neutral particles. Rev. Mod. Phys. 1990, 62, 531−540. 2. Dehmelt, H.G. Experiments with an isolated subatomic particle at rest. Rev. Mod. Phys. 1990, 62, 525−530. 3. Schuessler, H.; Fortson, E.T.; Dehmelt, H.G. Hyperfine structure of the ground state of 3He+ by the ion-storage exchange-collision technique. Phys. Rev. 1969, 187, 5−38. 4. Werth, G. Lifetime measurement of metastable states in ions. In Frequency Standards and Metrology, Ed. A. DeMarchi, Springer, Berlin, 1989, 293−299. 5. Lawson, G.; Bonner, R.F.; Todd, J.F.J. The quadrupole ion store (QUISTOR) as a novel source for a mass spectrometer. J. Phys. E Scient. Instrum. 1977, 36, 357−362.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
361
6. Wineland, D.J.; Bergquist, J.C.; Berkeland, D.; Bollinger, J.J.; Cruz, F.C.; Itano, W.M.; Jelenkovic, B.M.; King, B.E.; Meekhof, D.M.; Miller, J.D.; Monroe, C.R.; Tan, J.N. Application of laser-cooled ions to frequency standards and metrology. Proc. Fifth Symp. on Frequency Standards and Metrology, Woods Hole, MA, 1995, October 16−19, 11−19. http://tf.nist.gov/ion/freqstd/pubs.htm. 7. Yu, N.; Dehmelt, H.G.; Nagourney, W. Trapped individual ion at absolute zero temperature Proc. Nat. Acad. Sci. USA 1989, 86, 5671. 8. Stick, D.; Hensinger, W.K.; Olmschenk, S.; Madsen, M.J.; Schwab, K.; Monroe, C. Ion trap in a semiconductor chip. Nature Physics 2006, 2, 36−39. 9. http://wwp.greenwichmeantime.com/info/conference.htm. (Accessed 2009 06 03). 10. Fortier, T.M.; Ashby, N.; Bergquist, J.C.; Delaney, M.J.; Diddams, S.A.; Heavner, T.P.; Hollberg, L.; Itano, W.M.; Jefferts, S.R.; Kim, K.; Levi, F.; Lorini, L.; Oskay, W.H.; Parker, T.E.; Shirley, J.; Stalnaker, J.E. Precision atomic spectroscopy for improved limits on variation of the fine structure constant and local position invariance. Phys. Rev. Lett. 2007, 98, 070801 (4). 11. Allan, D.W. Statistics of atomic frequency standards. Proc. IEEE, 1966, 54, 221−230. 12. Wineland, D.J.; Itano, W.M.; Bergquist, J.C.; Bollinger, J.J.; Dietrich, F.; Gilbert, S.L. High accuracy spectrometry of stored ions. In Frequency Standards and Metrology, Ed. A. DeMarchi, Springer, Berlin, 1989, 71−77. 13. Ramsey, N.H. Experiments with separated oscillatory fields and hydrogen masers. Rev. Mod. Phys. 1990, 62, 541−554. 14. Laurent, Ph.; Abgrall, M.; Jentsch, Ch.; Lemonde, P.; Santarelli, G.; Clairon, A.; Maksimovic, I.; Bize, S.; Salomon, Ch.; Blonde, D.J.; Vega, F.; Grosjean, O.; Picard, F.; Saccoccio, M.; Chaubet, M.; Ladiette, N.; Guillet, L.; Zenone, I.; Delaroche, Ch.; Sirmain, Ch. Design of the cold atom PHARAO space clock and initial test results. Appl. Phys. B, Lasers and Optics 2006, 84, 683−690. 15. Major, F.G. Microwave resonance of field-confined mercury ions for atomic frequency standard for atomic frequency standard. NASA Report, Goddard Space Flight Center, X-512-69-167, 1969. 16. Thompson, R.C. Spectroscopy of trapped ions. Adv. Atom Mol. Opt. Phys. 1993, 31, 63−136. 17. Vedel, F. On the dynamics and energy of ion clouds stored in an RF quadrupole trap. Int. J. Mass Spectrom. Ion Processes 1991, 106, 33−61. 18. Dehmelt, H.G. Coherent spectroscopy on a single atomic system at rest in free space III. In Frequency Standards and Metrology, Ed. A. DeMarchi, Springer, Berlin, 1989, 15−19. 19. Prestage, J.D.; Dick, G.J.; Maleki, L. New ion trap for frequency standard applications. J. Appl. Phys. 1989, 66, 1015−1017. 20. Hornecker, L.; Drewsen, M. Formation process of large ion Coulomb crystals in linear Paul traps. Phys. Rev. A 2002, 66, 013412-12. 21. Gerlich, D. Inhomogeneous RF fields: A versatile tool for the study of processes with slow ions. Adv. Chem. Phys. 1992, 82, 1−176. 22. Monroe, C.; Leibfried, D.; King, B.E.; Meekhof, D.M.; Itano, W.M.; Wineland, D.J. Simplified quantum logic with trapped ions. Phys. Rev. A 1997, 55, R2489−2491. 23. Prestage, J.D. Extended linear ion trap frequency standard apparatus, US Patent 1995, 5,420,549. 24. Raizen, M.G.; Gilligan, J.M.; Bergquist, J.C.; Itano, W.M.; Wineland, D.J. Ionic crystals in a linear Paul trap. Phys. Rev. A 1992, 45, 6493−6501. 25. Denison, D.R. Operating parameters of a quadrupole in a grounded cylindrical housing. J. Vac. Sci. Technol. 1971, 8, 266−269.
362
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
26. Reuben, A.J.; Smith, G.B.; Moses, P.; Vagov, A.B.; Woods, M.D.; Gordon D.B.; Munn, R.W. Ion trajectories in exactly determined quadrupole fields. Int. J. Mass Spectrom. Ion Processes 1996, 154, 43−59. 27. Poitzsch, M.E.; Bergquist, J.C.; Itano, W.M.; Wineland, D.J. Cryogenic linear ion trap for accurate spectroscopy. Rev. Sci. Instrum. 1996, 67, 129−134. 28. Tjoelker, R.L.; Prestage, R.L.; Maleki, L. The JPL Hg+ extended linear ion trap. Frequency standard: Status, stability, and accuracy prospects. Proc. Precise Time and Time Interval, PTTI 1996, 26, 235−243. 29. Prestage J.D.; Chung, S.; Le, T.; Beach, M.; Maleki, L.; Tjoelker, L. One-liter ion clock: New capability for space craft applications. Proc. Precise Time and Time Interval, PTTI 2004, 35, 427−433. 30. Kjærgaard, N.; Hornekær, L.; Thommesen, A.M.; Videsen, Z.; Drewsen, M. Isotope selective loading of an ion trap using resonance-enhanced two-photon ionization. Appl. Phys. B 2000, 71, 207−210. 31. Simion programme, available at www.simion.com. Dahl, D.A. SIMION for the personal computer in reflection. Int. J. Mass Spectrom. 2000, 200, 3−25. 32. Schrama, C.A.; Peik, E.; Smith, W.W.; Walther, H. Novel miniature ion traps. Opt. Comm. 1993, 101, 32−36. 33. Neuhauser, W.; Hohenstatt M.; Toschek, P.E.; Dehmelt H.G. Visual observation and optical cooling of electrodynamically contained ions. Appl. Phys. 1978, 17, 123−129. 34. Oskay, W.H.S.; Diddams, A.; Donley, E.A.; Fortier, T.M.; Heavner, T.P.; Hollberg, L.; Itano, W.M.; Jefferts, S.R.; Delaney, M.J.; Kim, K.; Levi, F.; Parker, T.E.; Bergquist, J.C. Single atom optical clock with high accuracy. Phys. Rev. Lett. 2006, 97, 020801(4). 35. Tamm, C.; Engelke, D.; Bühner V. Spectroscopy of the electric-quadrupole transition 2S (F = 0) → 2D (F = 2) in trapped 171Yb + . Phys. Rev. A 2000, 61, 053405 (9). 1/2 3/2 36. Sinclair, A.G.; Wilson, M.A.; Gill, P. Improved three-dimensional control of a single strontium ion in an endcap trap. Opt. Comm. 2001, 190, 193–203. 37. Champenois, C.; Knoop, M.; Herbane, M.; Houssin, M.; Kaing, T.; Vedel, M.; Vedel, F. Characterization of a miniature Paul-Straubel trap. Eur. Phys. D 2001, 15, 105−111. 38. Vedel, F.; Vedel, M. Non linear effects in the detection of stored ions. Phys. Rev. A 1990, 41, 2348−2351. 39. Vedel, M.; Rocher, J.; Knoop, M.; Vedel, F. Evidence of radial-axial motion couplings for an r.f. stored ion cloud. Appl. Phys. B 1998, 66, 191−196. 40. Iffländer, R.; Werth, G. Optical detection of ions confined in a rf quadrupole trap. Metrologia 1977, 13, 167−170. 41. Major, F.G.; Werth, G. High-resolution magnetic hyperfine resonance in harmonically bound ground-state 199Hg + ions. Phys. Rev. Lett. 1973, 30, 1155−1158. 42. Major, F.G.; Werth, G. Magnetic hyperfine spectrum of isolated (mercury-199) + ions. Appl. Phys. B 1978, 15, 201−208. 43. Berkeland, D.; Miller, J.D.; Bergquist, J.C.; Itano, W.M.; Wineland D.J. Laser cooled mercury ion frequency standard. Phys. Rev. 1998, 80, 2089−2092. 44. Prestage, J.D.; Tjoelker, R.L.; Dick, G.J.; Maleki, L. Progress report on the improved linear ion trap physics package. Proc. of the 1995 IEEE International Frequency Control Symposium, San Francisco, May 31–June 03 1995, 82−85. 45. Chung, S.K.; Prestage, J.D.; Tjoelker, R.L.; Maleki, L. Gas frequency shifts in microwave mercury ion clocks. IPN Progress Report, 2004, 42–159, 1–8. 46. Burt, E.A.; Tjoelker, R.L. Sub-10-6 frequency stability in multipole linear ion trap standards. IPN Report 2006, 42–166, 1−7. 47. Prestage, J.D.; Chung, S.; Le, T.; Lim, L.; Maleki, L. Liter sized ion clock with 10-13 stability. Frequency Control Symposium and Exposition, Proc. 2005 IEEE International Conference, Vancouver, BC, Canada, August 29–31 2005, 472−476.
Ion Trap: A Versatile Tool for the Atomic Clocks of the Future!
363
48. Udem, T.; Reichert, J.; Holzwarth, R.; Hänsch, T.W. Absolute optical frequency measurement of the caesium D1 line with a mode-locked laser. Phys. Rev. Lett. 1999, 82, 3568−3571. 49. Bergquist, J.C.; Itano, W.M.; Wineland, D.J. Recoilless optical absorption and Doppler side bands of a single trapped ion. Phys. Rev. A 1987, 36, 428−430. 50. Bize, S.; Diddams, S.S.; Tanaka, U.; Tanner, C.E.; Oskay, W.H.; Drullinger, R.E.; Parker, T.E.; Heavner, T.P.; Jefferts, S.R.; Hollberg, L.; Itano, W.M.; Wineland, D.J.; Bergquist J.C. Testing the stability of fundamental constants with the 199Hg+ single-ion optical clock. Phys. Rev. Lett. 2003, 150802-4. 51. Schmidt, P.O.; Rosenband, T.; Langer, C.; Itano, W.M.; Bergquist, J.C.; Wineland, D.J. Spectroscopy using quantum logic. Science 2005, 309, 749−752. 52. Rosenband, T.; Schmidt, P.O.; Hume, D.B.; Itano, W.M.; Fortier, T.M.; Stalnaker, J.E.; Kim, K.; Diddams, S.A.; Koelemeij, J.C.; Bergquist J.C.; Wineland, D.J. Observation of the 1S0 → 3P0 clock transition in 27Al + . Phys. Rev. Lett. 2007, 98, 220801-4. 53. Rosenband, T.; Hume, D.B.; Schmidt, P.O.; Chou, C.W.; Brusch, A.; Lorini, L.; Oskay, W.H.; Drullinger, R.E.; Fortier, T.M.; Stalnaker, J.E.; Diddams, S.A.; Swann, W.C.; Newsbury, N.R.; Itano, W.M.; Wineland D.J.; Bergquist J.C. Frequency ratio of Al + and Hg + single-ion optical clocks; metrology at the 17th decimal place. Sciences 2008, 319, 1808−1812. 54. Champenois, C.; Hagel, G.; Houssin, M.; Knoop, M.; Zumsteg, C.; Vedel, F. Terahertz frequency standard based on three-photons coherent population trapping, Phys. Rev. Lett. 2007, 99, 013001 (4). 55. Champenois, C.; Hagel, G.; Houssin, M.; Knoop, M.; Zumsteg, C.; Vedel, F. Evaluation of the ultimate performance of a Ca + single ion frequency standard. Phys. Lett. A 2004, 331, 298−311. 56. Sherman J.A.; Koerber, T.W.; Markhotok, A.; Nagourney, W.; Fortson, E.N. Precision measurement of light shifts in a single trapped Ba+ ion. Phys. Rev. Lett. 2005, 94, 243001/1-4. 57. Thorpe, M.J.; Hudson D.D.; Moll, K.D.; Lasri, J.; Ye, J. Cavity-ringdown molecular spectroscopy based on an optical frequency comb at 1.45-1.65 μm, Optic. Lett. 2007, 32, 307–309. 58. Margolis, H.S.; Barwood, G.P.; Huang, G.; Klein, H.A.; Lea, S.N.; Szymaniec, K.; Gill, P. Hertz-level measurement of the optical clock frequency of a single 88Sr + . Science 2004, 306, 1355−1358. 59. Madej, A.A.; Bernard, J.E; Dubé, P.; Marmet, L.; Windeler, R.S. Absolute frequency of the 88Sr + 5s 2S1/2 → 4d 2D5/2 reference transition at 445 THz and evaluation of the systematic shifts. Phys. Rev. A 2004, 70, 012507-13. 60. Von Zanthier, J.; Eichenseer, M.; Nevsky, A.Y.; Okhapkin, M.; Schwedes, Ch.; Walther, H. A single indium ion optical frequency standard. Laser Physics 2005, 15, 1021−1027. 61. Peik, E.; Lipphardt, B.; Schnatz, H.; Schneider, T.; Tamm, Chr.; Karshenboim, S.G. Limit on the present temporal variation of the fine structure constant. Phys. Rev. Lett. 2004, 93, 170801 (4). 62. Matsubara, K.; Hayasaka, K.; Li, Y.; Hiroyuki, I.; Nagano, S.; Kajita, M.; Hosokawa, M. Frequency measurement of the optical clock transition of Ca + ions with an uncertainty at 10-14 level. Appl. Phys. Express 2008, 1, 067011-3. 63. Chwalla, M.; Benhelm, J.; Kim, K.; Kirchmair, G.; Monz, T.; Riebe, M.; Schindler, P.; Villar, A.S.; Haensel, W.; Roos, C.F.; Blatt, R.; Abgrall, M.; Santarelli, G.; Rovera, G.D.; Laurent, P. Absolute measurement of the 40Ca + 4s2S1/2 → 3d2D5/2 clock transition. Phys. Rev. Lett. 2009, 102, 023002 (4). 64. Gill, P.; Barwood, G.P.; Klein, H.A.; Huang, G.; Webster, S.A.; Blythe, P.J.; Hosak, K.; Lea, S.N.; Margolis, H.S. Trapped ion optical frequency standards. Meas. Sci. Technol. 2003, 14, 1174–1186.
Part IV Practical Applications
12 Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT) Francesco L. Brancia, Luca Raveane, Alberto Berton, and Pietro Traldi Contents 12.1 Introduction.................................................................................................. 367 12.1.1 Resonant Excitation........................................................................ 368 12.1.2 Non-Resonant Excitation................................................................ 369 12.1.3 Boundary-Activated Dissociation (BAD)....................................... 369 12.2 Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)........ 374 12.2.1 Basic Aspects.................................................................................. 374 12.2.2 The Digital Ion Trap (DIT) Stability Diagram............................... 375 12.2.3 Boundary Activation....................................................................... 382 12.3 Conclusion.................................................................................................... 385 References............................................................................................................... 385
12.1 INTRODUCTION Quadrupole ion traps (QITs) are nowadays very versatile mass analyzers due to their ease of use, the low-price/performance ratio, and overall, the high-efficiency with which multi-stage tandem mass spectrometry (MSn) via collision-induced dissociation (CID) is performed. Tandem mass spectrometry yields information on the genealogy of successive generations of product ions from a given isolated precursor ion and, when product ions are observed at high-mass resolution, the elemental compositions of these product ions can often be identified unambiguously. Tandem mass spectrometry permits the selection of structurally-diagnostic product ions suitable for qualitative and quantitative analysis. In the Quadrupole ion trap (QIT), two-stage tandem mass spectrometry, Mass Spectrometry/Mass Spectrometry (MS/MS), is carried out by mass selection of the precursor ion of interest, resonant excitation of the precursor ion using a supplementary Radio Frequency (RF) voltage applied on the two end-cap electrodes so as to induce collisional dissociation of the precursor ion, and subsequent analysis of the product ion population. It is abundantly clear that the energy deposition phenomenon active in the QIT differs substantially from those operating in low-energy collisions performed in triple-stage quadrupole instruments (QQQ) and in high-energy 367
368
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
collisions in beam-type multi-sector mass spectrometers, for example, BEqQ where B, E, and q represent a magnetic sector, an electrostatic sector, and a quadrupole collision cell, respectively. Consequently, product ion mass spectra obtained with QIT, QQQ, and BEqQ instruments reflect the energy deposition phenomenon specific to each type of instrument, which is manifested by variation between product ion mass spectra with respect to both product ion species formed and the relative ion signal intensities of the product ions. Product ion mass spectra generated in a QIT are characterized by a relatively lower number of product ion species. Confronted by this disparity of product ion information obtained by the above three techniques, scientists have investigated alternative activation methods for MS/MS analysis.
12.1.1 Resonant Excitation The resonant excitation activation method is surely the one most utilized for structural characterization in a QIT [1,2]. It requires the use of a supplementary power supply, which generates an additional RF voltage applied in dipolar mode to the two end-cap electrodes of the ion trap. An ion exhibits well-defined secular frequencies of motion both in the axial direction and in the radial plane; the magnitudes of these frequencies depend upon the values of the trapping parameters, qz and qr, respectively. Let us consider ion axial motion only. In order to isolate an ion, a combination of voltages, generated by the supplementary RF field, with broad-band frequencies higher and lower than the secular frequency of ion axial motion of the selected ion species, is superimposed upon the main RF trapping quadrupole field. Ion excitation, ergo ion ejection, is obtained when the frequency of the supplementary RF field matches the ion secular frequency for a given qz -value. At this stage, the ion is brought into resonance, its motion becomes unstable and, in the limit, the ion is ejected from the ion trap. By this method all the ion species, save that of interest, are ejected from the ion trap leaving the ion of interest (a potential precursor ion) isolated in the ion trap. The frequency of the supplementary RF field is then changed to match that of the isolated ion of interest. Under these conditions, the motion of an ion is elongated in the z-direction and maintains, generally, the same frequency; consequently, the kinetic energy of the ion increases. Subsequent multiple collisions with an inert gas (helium or argon) present in the ion trap as quenching (or buffer) gas, lead to conversion of part of the newly-acquired kinetic energy to internal energy deposited in the ion. However, this phenomenon differs significantly from that occurring in triple-stage quadrupole instruments and under high-energy collisions conditions in sector instruments. It has been shown that the internal energy deposition by resonant excitation in an ion trap is a step-by-step phenomenon [3]. To clarify this aspect from a didactic point of view, the example reported in Figure 12.1 can facilitate the reader’s understanding. We can consider the activation energies of the different fragmentation pathways as being represented by heights of the holes placed at different levels in a glass. In the case of QQQs and high-energy collision sector mass spectrometers, the water corresponding to the energy imparted to the ion is poured into the glass completely in essentially one go (that is, in a single collision) so that it can flow out from all the holes at the same time, ergo all decomposition channels are activated contemporaneously (Figure 12.1 left-hand side). In the case of ion traps operating in the resonant excitation mode,
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
Critical energy
369
Critical energy
FIGURE 12.1 A didactical view of the difference existing between tandem mass spectrometric experiments performed by QQQ (left side) or by resonant excitation in an ion trap (right side).
internal energy is deposited into the ion via multiple collisions. This process can be visualized as when the water is poured into the glass drop by drop (Figure 12.1 righthand side). The resulting effect is that water flows out of the glass only through the holes positioned at lower levels; that is, the energy imparted to the ion can activate only the fragmentation pathways of lower activation energy.
12.1.2 Non-Resonant Excitation In order to circumvent these limitations, alternative activation methods, such as photodissociation (PD and Laser PD) [4,5], the use of broad-band frequencies [6], sweep frequencies [7], pulsed axial activation [8], and the use of fast direct current (DC) pulses [9] have been proposed. Of these methods, only the last method is available commercially [10], and leads, in some cases, to a larger number of product ions. To understand this phenomenon better, we can consider again the cartoon depicted in Figure 12.1. Experiments indicate that the DC pulse method is able to generate greater energy deposition and, thus, to activate multiple decomposition channels of similar activation energy. An alternative view of this model is that pulsed DC corresponds to a pulsed jet of water going into the glass and then re-bounding upward so that some of it reaches some of the higher-level holes. Apart from the PD methods, all the other proposed methods are based on the transfer of ion kinetic energy to internal energy at higher amplitudes of the RF potential: the main RF field is responsible for ion acceleration and for the related increase of ion kinetic energy.
12.1.3 Boundary-Activated Dissociation (BAD) As an alternative to the methods above described, a further ion activation method was proposed in 1991 [11–14]. In this method, the working point (that is, the point (az, qz) on a QIT stability diagram defined by the magnitudes of the trapping parameters az and qz) is moved close to one of the boundaries of the stability diagram; this method can be realized with the combined effect of suitable DC and RF potentials applied to the ion trap
370
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
electrodes. Under these conditions, ion dissociation can be induced, hence this method was called boundary-activated dissociation (BAD). Parenthetically, early commercial versions of the QIT (ITD™ and Saturn™, manufactured by Finnigan Corporation and Varian Inc., respectively) were fitted with DC power supplies that permitted ready manipulation of the working point of an ion within the stability diagram. This situation differs from current practice, where most ion trap operations are carried out with the working points of ions confined to the qz axis, that is, with az = 0. The magnitudes of the DC (UDC) and RF (VRF) potentials are varied to effect both ion selection and ion activation. A BAD experiment can be described by reference to the stability diagram shown in Figure 12.2. The first step consists of moving the working point of the ion of interest to a convenient qz-value on the az = 0 axis (by application of a suitable VRF and keeping UDC = 0). Then a short DC pulse (ca 50–200 ms) is applied to the ring electrode, so that the working point of the ion is moved to the upper (βr = 0) or lower (βz = 0) boundaries of the stability diagram. The diagram in Figure 12.2 shows the movement of the working point of the selected ion from point A to point B (on the (βz = 0) boundary) by application of a positive DC pulse; a positive DC pulse yields a negative value for az because az is proportional to-UDC. At point B, ion oscillation increases in the z-direction. The locus of working points for product ions formed subsequently lies on the line BC and extends beyond C to the βz = 1 boundary. The removal of the DC voltage restores the working point of the remaining precursor ion population to point A, while the working points of all product ions formed having qz-values lying between B and C are moved to the az = 0
0.4
0.2
z Stability
βr = 0
q cut-off
A
–0.2
–0.4
–0.6
1.5
1.0
az
0 B
qz
βz = 1 C
r Stability
βz = 0
βr = 1
FIGURE 12.2 Modulation of az and qz -values to perform boundary-activated dissociation (BAD) experiments.
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
371
line (see Figure 12.2). Product ions can then analyzed by a mass-selective axial instability scan of the amplitude of VRF. Those product ions whose initial working points lay beyond C have qz values in excess of 0.908, that is, qcut-off, and their trajectories become unstable when the DC pulse is removed; such ions are ejected from the QIT. The mass/ charge ratio of product ions with a working point at C represents the low-mass cut-off of the mass analyzer. The ‘boundary effect’ was described first by Paradisi et al. [11] in a study dedicated to the effect of ions stored with different working points of the stability diagram, that is, in qz≠0 and az≠0 conditions by adjustment of the UDC and VRF voltages applied to the ring electrode of the ion trap. The molecular ion, M+•, of 2-(2′-hydroxybenzoyl)-benzoic acid, C14H10O4 +• was chosen as a model. Under collisional conditions, this molecular ion gives rise to a primary water loss (see Scheme 12.1), followed by additional neutral losses of H• and CO. Previous investigations indicated that H 2O loss is a low-critical energy process; consequently M+• can be employed as a sensitive probe for determining the energy uptake by collision necessary to surmount the energy of activation barrier for the process of water loss. The scan function employed for the experiments is shown in Figure 12.3. A scan function is a pictorial representation of the temporal variation of potentials applied to various electrodes during an ion trap operation. The ionization phase (a) is followed by a two-step ion isolation step (b, c). These stages are followed by step (d) where VRF is decreased such that qz is ca 0.4 and step (h) where a DC voltage, UDC, is applied to the ion trap. By adjusting the values of UDC and VRF, the ion is stored at various working points of the stability diagram. Extensive fragmentation is observed with the working point located in the
OH
HO O
O C
+•
C
1
[C14H10O4]+• m/z 242 –H2O [C14H8O3]+• m/z 224 –H•
[C14H7O3]+ m/z 223
–CO [C13H7O2]+• m/z 196
SCHEME 12.1 Fragmentation pattern of 2-(2'-hydroxybenzoyl)–benzoic acid.
372
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V RF level c g
b d e
a
DC
f
a Ionization time b, c Separation phase d Setting qz coordinate e Storage time f Setting scan start g Acquisition phase h Setting az coordinate Time
h
FIGURE 12.3 Scan function employed for BAD experiments performed on the molecular ion of 2-(2′-hydoxybenzoyl)-benzoic acid.
proximity of both the upper (βr = 0, upon application of UDC) and lower (βz = 0, upon application of UDC) boundaries of the stability diagram. Such behavior can be explained by considering that the stability and instability regions in a,q space are separated by a quasi-stable region in which ions are subjected to a large field that is responsible for the marked increase of ion kinetic energy. Enhanced ion kinetic energy results in more energetic collisions leading to enhanced deposition of internal energy. Vachet and Glish observed that when BAD is performed in the presence of heavy gases, more extensive dissociation is observed in the resulting product ion mass spectra than with helium alone [15]. In the analysis of peptide ions, the presence of heavy gases increases the efficiency with which the precursor ion is converted into product ions [15]. Moreover, Paradisi and coworkers demonstrated that when the working point of an isolated ion is positioned close to the boundaries of the stability diagram, CID takes place [11,12]. In order to record a mass scan of the product ions, VRF is adjusted to an amplitude corresponding to the desired low-mass cut-off, that is, the low-mass limit of the mass scan (f), whereupon the amplitude of VRF is scanned (g) to generate a mass spectrum. Therefore, in an ion trap, tandem mass spectrometric experiments can be performed without the necessity of the resonant supplementary ‘tickle’ voltage applied across the two end-cap electrodes. The result obtained is supportive of the general interpretation concerning the effect of the resonant tickle voltage. However, the effective ion acceleration in these experiments is not due to ions being in resonance with the supplementary RF field, but the acceleration force originates from the field derived from the main RF drive potential. The supplementary voltage is used only to move the ion cloud into a region in which the main RF field is stronger. Collisional activation by boundary effects relies on the same physical phenomenon; in this case, the main RF field is augmented by the presence of a DC voltage of appropriate magnitude. Additional investigations have been carried out by BAD and they have been reviewed extensively [13]. The simultaneous use of BAD and tickle activation in an ion trap allows unambiguous discrimination between consecutive and competing CIDs. The experiment is carried out using DC and RF voltage modulation, as shown in Figure 12.4. As
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
373
described above, the ionization phase (a) is followed by a two-step ion isolation step (b, c). These stages are followed by step (d) where VRF is decreased such that qz is 0.4 ca and step (h) where a DC voltage, ± UDC, is applied to the ion trap to move the working point of the isolated ion species to the vicinity of either the βz = 0 or βr = 0 boundaries, respectively, of the stability diagram. At the same time, an RF tickle voltage of but a few volts is applied at a frequency corresponding to the frequency of the isolated ion species at a known value of qz in the vicinity of qz = 0.4. At this point, that is, at the commencement of step (e), the isolated ion species is subjected simultaneously to BAD and to resonant excitation. By appropriate adjustment of the amplitude of the RF tickle voltage, a selected product ion species can be ejected rapidly from the trap. At the conclusion of step (e), a mass scan is carried out, steps (f, g) as before. Using this approach, in the case of consecutive and/or competing decomposition channels, as those shown in Scheme 12.2, it is easy to determine the possible intermediacy of A+ in the formation of B+ from the precursor ion, M+•. In fact, when a tickle voltage, of frequency corresponding to A+, is applied after the stage of isolation of M+• and leads to the disappearance of B+, this observation implies that A+ is the real precursor of B+ and that the formation of B+ by tickling M+• is due to two different sequential processes [11]. RF level c g
b d e
a
DC
f
Time h
RF
FIGURE 12.4 Scan function employed for discriminating between consecutive and competing collisionally-induced dissociations.
M+•
B+
A+
SCHEME 12.2 Consecutive and competing decomposition pathways.
374
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
12.2 BOUNDARY-ACTIVATED DISSOCIATION (BAD) IN A DIGITAL ION TRAP (DIT) 12.2.1 Basic Aspects The fundamental theory of operation of a digital ion trap (DIT) has been described extensively elsewhere in this series.* In the DIT mass spectrometer, the trapping electric field and resonant excitation are provided by high-voltage switching circuits. In the steady trapping operation, a periodic rectangular wave voltage, generated between a high-voltage level V1 and a low-voltage level V2, is applied to the ring electrode of an ion trap. Under these conditions, the Mathieu equation cannot be used to determine the ion pathways inside the DIT and the matrix transform method must be utilized [16,17]. However the Mathieu parameters au and qu, defined by Equations 12.1 and 12.2
qz =
4eV mr02Ω 2
(12.1)
az =
8eU mr02Ω 2
(12.2)
can still be employed for the description of the DIT theoretical stability diagram, considering the U and V values as the average values of the DC and alternating current (AC) components of the rectangular wave voltage applied to the intermediate electrode, defined as
U = dV1 + (1 − d )V2
(12.3)
V = 2(V1 − V2 )(1 − d )d
(12.4),
where e is the electronic charge, m is the mass of the ion, r0 is the radius of the ring electrode, and Ω is the radial frequency of the rectangular wave. As shown in Figure 12.5, the duty cycle d is described as the ratio between τ and the total period T of the rectangular wave. Unlike the sinusoidal wave, the rectangular wave can be generated with different pulsing times for V1 and V2 produced by the digital circuitry. In a DIT, the az value is a function of both the DC offset and the duty cycle d. In fact, the DC component U can be generated either by an imbalance between V1 and V2 or by variation of the duty cycle d. A typical scan function, or scan table, for a forward mass scan performed by the DIT is shown in Figure 12.6. The scan function differs substantially from that used in QIT experiments. Time is reported on the abscissa, while the period (T) of the * See also Volume 4, Chapter 4: Rectangular Waveform Driven Digital Ion Trap (DIT) Mass Spectrometer: Theory and Applications by Francesco Brancia.
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
375
Voltage V2 0
Time
V1 τ
d=τ T
T
FIGURE 12.5 Definition of the duty cycle, d, in relation to the operation of the digital ion trap.
square wave is reported on the ordinate. For the QIT, the value of the RF amplitude, V0–p is reported. The DIT scan function to obtain a complete mass spectrum is based on four separate steps: (i) a standby time at high-T values, followed by; (ii) a stage devoted to ion introduction inside the trap; (iii) a field adjusting phase; and (iv) the analytical mass scan.
12.2.2 The Digital Ion Trap (DIT) Stability Diagram In order to evaluate the DIT performance for BAD experiments, the real shape of the stability diagram must be determined [17]. For mapping the stability diagrams, the values of the switching voltage levels V1 and V2 are kept at –500 V and 500 V, respectively, so that the DC offset is zero and only the duty cycle needs to be changed in to determine the stability diagram boundaries. In order to map the stability diagram, the scan table has been modified by the addition of ‘jumping, mapping, and cooling’ steps, as shown in Figure 12.7. Once again, ions are injected into the trap; after cooling with helium, the ion is moved to a specified a,q working point for the ‘mapping’ phase. Firstly, a duty cycle is fixed and then a period of waveform is applied. The numbers of ions that remain trapped are determined by running a mass scan. When the duty cycle is 50%, the DC quadrupole component is absent and ion motion can be described using only the qz -axis. When mass analysis is performed with the DIT, using the mass-selective instability scan mode, the normal ion ejection point is qz = 0.712 with az = 0. However, when a duty cycle different from 50% is used, the DC quadrupole component generated, as a consequence of this difference, must be taken into account. Under these conditions, a new straight line passing through the origin of the (a,q) diagram must be considered, in accordance with the duty cycle selected so as to determine the working point of the ion of interest. Each straight line depends on the magnitude of the DC component used that is a function of d. The slope of straight line, which is given by the ratio between 2U and V, can be described by the equation:
a 2U 2d − 1 = = q V 2d (1 − d )
(12.5),
FIGURE 12.6 A typical scan table for a forward mass scan of the digital ion trap.
376 Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
FIGURE 12.7 A scan table customized for a forward mass scan of the digital ion trap in order to map the stability diagram.
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT) 377
378
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V az d = 0.45 d = 0.48
d = 0.50
qz
d = 0.53
FIGURE 12.8 Various az /qz lines obtained by modulation of the rectangular waveform duty cycle. The gray segments indicates those parts of the lines that fall within the stability region for a given duty cycle.
Figure 12.8 shows several a/q lines obtained using different duty cycle values for the trapping rectangular waveform: the gray parts represent those parts of the lines that fall within the region of stability for a given duty cycle; the black parts represent those parts of the lines that fall beyond the βz = 1 stability boundary. For ion traps driven by a sinusoidal waveform, the superimposition of a DC potential on the main trapping field requires the use of an additional DC power supply; see Section 12.1.3. However, in the DIT, the DC component can be generated easily by varying the duty cycle of the rectangular waveform through appropriate variation of the parameter values entered into the control software of the mass spectrometer. In order to determine experimentally the boundaries of the stability diagram, the Mathieu parameters az and qz of an ion are chosen such that its working point is located close to a boundary of the stability diagram. In this experiment, a duty cycle of 0.5% (see Figure 12.8) is used for ion introduction and cooling. Subsequently, the duty cycle corresponding to the value predicted theoretically for crossing a boundary of the first stability region is employed. When the (az, qz) working point calculated for that specific duty cycle is correct, the ion of interest is lost from the ion trap and no signal is recorded in the mass spectrum. Figure 12.9a shows the electrospray mass spectrum of the isotopomers of the isotopic envelope of the doubly-protonated molecule of bradykinin; the working point for each ion species lies within the stability diagram but in the vicinity of the βz = 1 boundary of the stability diagram. The singly-protonated molecule of leucine encephalin was used to map the βz = 1 boundary. The mass spectrum shown in Figure 12.9b was obtained with the working point of the monoisotopomer (protonated molecule of lowest mass/ charge ratio, and hence the highest qz -value among all ions belonging to the isotopic envelope) on or beyond the βz = 1 boundary. The ion signals due to the monoisotopomer has been reduced to ca one-third due to ion ejection, yet the ratio of the ion signals for the 13C1-containing ions and the 13C2-containing ions remains unchanged
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
Relative abundance
(a)
379
100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 528.0 528.5 529.0 529.5 530.0 530.5 531.0 531.5 532.0 532.5 533.0 533.5 534.0 534.5 535.0 535.5 536.0 Mass/charge
100 95 90 85 80 75 70 65 60 55 50 45
Relative abundance
(b)
40 35 30 25 20 15 10 5 528.0 528.5 529.0 529.5 530.0 530.5 531.0 531.5 532.0 532.5 533.0 533.5 534.0 534.5 535.0 535.5 536.0 Mass/charge
FIGURE 12.9 Leucine encephalin [M+H]+ mass spectra: (a), inside the stability diagram; (b), the monoisotopomer of the molecular ion cluster is located on the βz=1 boundary of the stability diagram.
from that shown in Figure 12.9a. Thus, it is concluded that the working point for the monoisotopomer coincides with (or is extremely close to) the βz = 1 boundary of the stability diagram such that ion ejection occurs of the monoisotopomer only. Furthermore, under these conditions, the working points for the 13C1-containing
380
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
ions and the 13C2-containing ions were located within the stable trajectory region of the stability diagram. A description of the ion species used and their charge states is given in Table 12.1. A similar approach has been utilized to determine the remaining three boundaries (βr = 0 and 1, and βz = 0) (see Volume 4 of this series, Chapter 4, Figure 4.4) [18]. For each ion, the boundaries of the stability diagram were defined using 50 points. Each measurement was repeated five times, showing standard deviations in the range 1–2%. The data so obtained are shown in Figures 12.10 through 12.12, respectively. The solid or continuous lines represent the calculated βz = 0 and βz = 1 boundaries of the stability diagram (see Figure 12.2); the dashed lines represent the βr = 0 and βr = 1 boundaries of the stability diagram (see Figure 12.2); the data points () and TABLE 12.1 Identification of the Ions and Their Charge States Used in the Experimental Determination of the Boundaries of the Stability Diagram for the Digital Ion Trap (DIT) Ion Number Ion 1 Ion 2 Ion 3 a
Parent Compound
Degree of Protonation
Mass/Charge Ratio
Bradykinin Bradykinin fragmenta Leucine encephalin
[M+2H]2+ [M′+H]+ [M″+H]+
m/z 530 m/z 572 m/z 556
This fragment contains residues 1–5 of the normal nine residues of bradykinin.
0.2 0.1 0 –0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
az
–0.2 –0.3 –0.4 –0.5 –0.6 –0.7 –0.8
qz
FIGURE 12.10 Computed and experimentally-determined stability diagram obtained by using [M+2H]2+ ions of bradykinin. The theoretical diagram has been obtained by the method described by Konenkov, N.V.; Sudakov, M.; Douglas, D.J. J. Am. Soc. Mass Spectrom. 2002, 13, 597–613.
381
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT) 0.2 0.1 0 –0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
11
12
–0.2 az
–0.3 –0.4 –0.5 –0.6 –0.7 –0.8
qz
FIGURE 12.11 Computed and experimentally-determined stability diagram obtained by using [M+H]+ ions of RPPGF bradykinin fragment 1–5, that is, containing five residues only. The theoretical diagram has been obtained by the method described by Konenkov, N.V.; Sudakov, M.; Douglas, D.J. J. Am. Soc. Mass Spectrom. 2002, 13, 597–613.
0.2 0.1 0 –0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
11
12
az
–0.2 –0.3 –0.4 –0.5 –0.6 –0.7 –0.8
qz
FIGURE 12.12 Computed and experimentally-determined stability diagram obtained by using [M+H]+ ions of leucine encephalin. The theoretical diagram has been obtained by the method described by Konenkov, N.V.; Sudakov, M.; Douglas, D.J. J. Am. Soc. Mass Spectrom. 2002, 13, 597–613.
382
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
the corresponding gray filled square refer to experimental determinations of the βz = 0 and βr = 0 boundaries, respectively, of the stability diagram; and the data points ( ) and ( ) refer to experimental determinations of the βz = 1 and βr = 1 boundaries, respectively, of the stability diagram. As displayed in these figures, the experimental data are in good agreement with the calculated values for the βr = 0, βz = 0, and βz = 1 boundaries of the stability diagram; the agreement between experimental data and calculated values for the βr = 1 boundary is somewhat less than satisfactory. However, on the right-hand side of the diagram there is a clear discrepancy between the theoretical [16] and observed experimental values; this discrepancy is marked particularly for the doubly-protonated bradykinin 1 (m/z 530, Figure 12.11).
12.2.3 Boundary Activation For the peptide ions used in this investigation, when a large DC component is employed, corresponding to negative values of az such that the working points lie below the az = 0 axis, the graphs obtained from empirical data (Figures 12.8 through 12.10) differ from those calculated theoretically. For duty cycles corresponding to values qz > 0.9 and az between –0.3 and –0.8, the kinetic energy imparted to the ion by the field is so high that ion behavior in the proximity of the stability boundary differs from that expected on the basis of theoretical calculations due to resonant excitation/ejection. For instance, in the case of 3 (Figure 12.10), ions were ejected due to trajectory instability at the working point az = –0.500 and qz = 0.938, calculated from Equations 12.2 and 12.1, respectively. Thus the working point (0.500, 0.938) defines this point on the experimentally-determined βz = 1 boundary of the stability diagram. The qz -value of the computed βz = 1 boundary of the stability diagram at az = –0.500 is 0.958. Thus, at az = –0.500, the difference in qz -values between the experimentally-determined βz = 1 boundary of the stability diagram and the computed βz = 1 boundary of the stability diagram is 0.020. When 1 is analyzed (Figure 12.8), ions were ejected due to trajectory instability at the working point az = –0.520 and qz = 0.900, calculated again from Equations 12.2 and 12.1, respectively. Thus the working point (0.520, 0.902) defines this point on the experimentally-determined βz = 1 boundary of the stability diagram. The qz -value of the computed βz = 1 boundary of the stability diagram at az = –0.520 is 0.966. Thus, at az = –0.520, the difference in qz -values between the experimentallydetermined βz = 1 boundary of the stability diagram and the computed βz = 1 boundary of the stability diagram is 0.064. In fact, regardless of the duration of the cooling time, the application of a DC component brings ions into a region of the ion trap in which the RF field is larger and the ions are accelerated to higher kinetic energies. Because of this effect, ions undergo energetic collisions with the background gas and depletion of the monoisotopic ion signal intensity is observed due to unwanted fragmentation. Closer examination of the product ion mass spectra indicates the presence of product ions produced by collisions with the buffer gas. The results suggest that variation of the waveform duty cycle, which is achieved at the software level by entering different values in the scan table, can result in a relatively simple approach for generating BAD without the necessity of an additional power supply. The behavior of 1 can be explained considering the different charge state of the ions used as probes. Ion 1, which displays the largest discrepancy between the
383
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
computed βz = 1 boundary and the experimentally-determined βz = 1 boundary, is a doubly-charged species, with the m/z-value close to those for 2 and 3. Its interaction with the trapping field is consequently stronger: in other words, the doubly-protonated species is able to increase its kinetic energy at (az, qz)-values that are lower than those necessary to excite similarly 2 and 3 to achieve a comparable enhancement of kinetic energy. If this hypothesis is true, the loss of ion 1 in the right-hand side of the stability diagram is not due to its ejection from the ion trap and/or its discharge on the trap walls, rather it is caused by the activation of decomposition channels due to the increase of kinetic energy and effective collisions with helium. To verify this hypothesis, some experiments were performed by using different (az, qz)-values, corresponding to the points (a)–(d) shown in Figure 12.13. The mass spectra obtained under these conditions are shown in Figure 12.14. Singly-charged ions, at m/z-values higher than the doubly-charged precursor ion, are observed, providing evidence on the occurrence of boundary-activated chargeseparation dissociations. Among all the various experimental conditions examined, those leading to the best result (with respect to signal-to-noise ratio) correspond to point (b), which is reasonable considering the larger qz -range available that permits the storage of high-mass ions. However, when loci (discussed below) are drawn from the origin through the working points (a)–(d), it is found that the lengths of the loci lying within the stability diagram decrease in the order (a)–(d). Inspection of Figure 12.13 yields the qz-values for the intersection of these loci with the βz = 0 boundary from which the highest m/z-value for product ions remaining confined in the DIT can be calculated. The product ion mass/charge ratio reaches its greatest value at m/z 1730 for (a); m/z 1200 for (b); m/z 765 for (c); and m/z 672 for (d). As shown by the values for the high-mass/charge limit for points (a)–(d), m/z 555 is in no danger of being ejected and so it is not surprising that its intensity is sensibly constant. For the 0.2 0.1 0 –0.1 –0.2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8 a 0.9
1
11
qz
b
az
–0.3 –0.4 –0.5
c d
–0.6 –0.7 –0.8
FIGURE 12.13 Selected working points employed to perform BAD experiments on [M+2H]2+ ions of bradykinin; (a) (–0.140, 0.750); (b) (–0.260, 0.815); (c) (–0.450, 0.895); and (d) (–0.570, 0.926).
384
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Normalized accumulated intensity
(a) 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5
[M+2H]2+ b5
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
Mass/charge
(b) Normalized accumulated intensity
100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5
b5
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
Mass/charge
Normalized accumulated intensity
(c) 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5
[M+2H]2+
b5
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
Mass/charge
Normalized accumulated intensity
(d) 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5
[M+2H]2+
b5
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
Mass/charge
FIGURE 12.14 Product ion mass spectra obtained using BAD of [M+2H]2+ of bradykinin at the following selected working points identified in Figure 12.13: (a) (–0.140, 0.750); (b) (–0.260, 0.815); (c) (–0.450, 0.895); and (d) (–0.570, 0.926).
Boundary-Activated Dissociation (BAD) in a Digital Ion Trap (DIT)
385
point (d), the loss of some high-mass ions is expected. The locus of (az, qz)-values for all product ions is a straight line connecting point (d) (in Figure 12.13) to the origin of the stability diagram; it is seen that part of this locus falls outside the βz = 0 boundary of the stability diagram such that the trajectories of a range of product ions are rendered unstable and the ions are lost. These data indicate that BAD implemented in the DIT represents a realistic alternative for MS/MS experiments.
12.3 CONCLUSION Among all non-resonant activation techniques, BAD has been shown to have unique advantages for the formation of product ions. Due to the necessity to utilize an additional power supply for generating the DC component, such an approach has not been used in any commercial mass spectrometer. Conversely, in the DIT, variation of the duty cycle of the rectangular waveform is controlled at software level and it allows readily introduction of the DC component for BAD experiments.
REFERENCES
1. Fulford, J.E.; Hoa, D-N.; Hughes, R.J.; March, R.E.; Bonner, R.F.; Wong, G.J. Radiofrequency mass selective excitation and resonant ejection of ions in a three-dimensional quadrupole ion trap. J. Vac. Sci. Technol. 1980, 17, 829–835. 2. Louris, J.N.; Cooks, R.G.; Syka, J.E.P.; Kelley, P.E.; Stafford Jr., G.C.; Todd, J.F.J. Instrumentation, applications and energy deposition in quadrupole ion trap MS/MS spectrometry. Anal. Chem. 1987, 59, 1677–1685. 3. Gronowska, J.; Paradisi, C.; Traldi, P.; Vettori, U. A study of relevant parameters in collisional-activation of ions in the ion-trap mass spectrometer. Rapid Commun. Mass Spectrom. 1990, 4, 306–314. 4. Louris, J.N.; Brodbelt, J.E.; Cooks, R.G. Photodissociation in a quadrupole ion trap mass spectrometer using a fiber optic interface. Int. J. Mass Spectrom. Ion Processes 1987, 75, 345–352. 5. Lifshitz, C. Dissociative photoionization in the vacuum UV region with ion trapping. Int. J. Mass Spectrom. Ion Processes 1991, 106, 159–173. 6. McLuckey, S.A.; Goeringer, D.E.; Glish, G.L. Collisional activation with random noise in ion trap mass spectrometry. Anal. Chem. 1992, 64, 1455–1460. 7. Julian, R.K.; Cox, K.; Cooks, R.G. Proc. 40th ASMS Conference on Mass Spectrometry and Allied Topics. Washington, DC, 31 May–5 June 1992, p. 943. 8. Pannell, L.K.; Pu Q.L.; Mason, R.T.; Fales, H.M. Fragment pathway analysis using automated tandem mass spectrometry on an ion-trap mass spectrometer. Rapid Commun. Mass Spectrom. 1990, 4, 103–107. 9. Lammert, S.A.; Cooks, R.G. Pulsed axial activation in the ion trap: A new method for performing tandem mass spectroscopy (MS/MS). Rapid Commun. Mass Spectrom. 1992, 6, 528–530. 10. Varian, Walnut Creek, CA, USA, Technical literature. 11. Paradisi, C.; Todd, J.F.J.; Traldi, P.; Vettori, U. Boundary effects and collisional activation in a quadrupole ion trap. Org. Mass Spectrom. 1992, 27, 251–254. 12. Paradisi, C.; Todd, J.F.J.; Vettori, U. Comparison of collisional activation by the ‘boundary effect’ vs. ‘tickle’ excitation in an ion trap mass spectrometer. Org. Mass Spectrom. 1992, 27, 1210–1215.
386
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
13. Traldi, P.; Catinella, S.; March, R.E.; Creaser, S.C. Boundary excitation. In Practical Aspects of Ion Trap Mass Spectrometry, eds. R.E. March and J.F.J. Todd, Vol. 1, Chapter 7, pp. 299–341. CRC Press, Boca Raton, 1995. 14. March, R.E.; Todd J.F.J. Quadrupole Ion Trap Mass Spectrometry. 2nd Edn. John Wiley & Sons, Hoboken, NJ, 2005, pp. 280–290 (references cited therein). 15. Vachet, R.W.; Glish, G.L. Boundary-activated dissociation of peptide ions in a quadrupole ion trap. Anal. Chem. 1998, 70, 340–346. 16. Ding, L.; Sudakov, M.; Kumashiro, S. A simulation study of the digital ion trap mass spectrometer. Int. J. Mass Spectrom. 2002, 221, 117–138. 17. Konenkov, N.V.; Sudakov, M.; Douglas, D.J. Matrix methods for the calculation of stability diagrams in quadrupole mass spectrometry. J. Am. Soc. Mass Spectrom. 2002, 13, 597–613. 18. Berton, A.; Traldi, P.; Ding, L.; Brancia, F.L. Mapping the stability diagram of a digital ion trap (DIT) mass spectrometer by varying the duty cycle of the trapping rectangular waveform. J. Am. Soc. Mass Spectrom. 2008, 19, 620–625.
Study of Ion/ 13 The Molecule Reactions at Ambient Pressure with Ion Mobility Spectrometry and Ion Mobility/Mass Spectrometry Gary A. Eiceman and John A. Stone Contents 13.1 Introduction................................................................................................. 388 13.2 The Ion Mobility Spectrometer and a Mobility Measurement.................... 389 13.2.1 The Profiles of Ion Mobility Spectra Used to obtain Thermodynamic and Kinetic Data................................................ 391 13.2.1.1 Type 1. Equilibrium A + + B = AB + Exists Throughout the Source and Drift Region..................... 393 13.2.1.2 Type 2. A + + B→AB + in the Drift Region................... 393 13.2.1.3 Type 3. A+ and AB + are Formed in the Source Region and Neither A Nor B is Present in the Drift Region........................................................................... 394 13.2.2 Examples Where Thermodynamic and Kinetic Data have been obtained from Ion Mobility Spectra..................................... 394 13.2.2.1 Type 1. Ions in Equilibrium Showing a Single Peak in the Mobility Spectrum.............................................. 395 13.2.2.2 Type 2. Reaction Rate Constant Measurements...........400 13.2.2.3 Type 3. Dissociation of Adduct Ions.............................403 13.2.3 The Kinetics of Thermal Electron Capture and Thermal Electron Detachment.....................................................................406 13.2.3.1 Electron Capture...........................................................406 13.2.3.2 Thermal Electron Detachment......................................409
387
388
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
13.3 Concluding Remarks................................................................................... 411 References............................................................................................................... 412
13.1 INTRODUCTION In the Prefaces to Volumes 4 and 5 of Practical Aspects of Trapped Ion Mass Spectrometry the Editors have explained that, in defining the scope of these publications, it is considered that “an ion is ‘trapped’ when its residence time within a defined spatial region exceeds that had the motion of the ion not been impeded in some way.” Ion mobility spectrometry (‘IMS’) operated at atmospheric pressure, which falls clearly within this definition, involves the determination of the time taken for the components of a packet of ions to move through a defined distance under the influence of a specified electric field gradient and against the counter current flow of a drift gas at ambient pressure. Ion mobility spectrometers have been utilized for monitoring hazardous or controlled substances in a range of venues on land, in flight, in space, and underwater in submarines, and are ubiquitous at airports for detecting explosives in carry-on articles [1–18]. Such uses have arisen from the pragmatic attractions of these analyzers including ruggedness, small size, and affordability. Measurements by IMS are based on the production and determination of gaseous ions derived from a sample and are made commonly at ambient pressure so vacuum systems and associated pumps are unnecessary [19–22]. These features account for the portability of IMS analyzers and, along with the convenience of use and speed of operation, often make mobility spectrometers the instrument of choice for in-field, routine, measurements. An early term for IMS was plasma chromatography from the presence of a plasma, that is, both positive and negative ions in the ion source, and from the separation of ions in the supporting medium, air, or nitrogen. As with gas or liquid chromatograms, ion mobility spectra alone lack the facility for unequivocal identifications because the relationship of an ion mobility measurement to the structure or identity of an ion is under-developed. Consequently, mass spectrometers were combined with mobility drift tubes early in the development of IMS to provide ion identities through mass analyses. The combination of mobility and mass measurements can also permit the study of reactions between gaseous ions and molecules at ambient pressure in air, or other gases, and the measurement of both kinetic and thermodynamic values. Thermodynamic data that are suitable for tabulation include standard enthalpies, entropies, and free energies and can be regarded as universally applicable for systems at specified temperature when all participants are at thermal equilibrium. Though such data can also be obtained without thermal equilibrium, compensating experiments, or mathematical corrections are required, sometimes creating difficulties in practice and/or interpretation. A chemical system in the gas phase can reach thermal equilibrium, at a defined temperature, when a sufficient number of intermolecular collisions produce a Boltzmann distribution of energies in all modes, electronic, vibrational, rotational, and translational. In measurements made with an ion trap instrument or Fourier Transform Ion Cyclotron Resonance (FT-ICR) spectrometer at low pressure, hot ions must be cooled, commonly with a pulse of buffer
The Study of Ion/Molecule Reactions at Ambient Pressure
389
gas, and time allowed for thermalization, with the number of thermalizing collisions directly proportional to pressure. Such techniques are unnecessary with an ion mobility spectrometer at ambient pressure because each ion experiences more than 1010 collisions per second, mainly with neutral atoms or molecules of the supporting gas atmosphere. When the concentration of a sample compound is 1 ppm by volume, for example, then, on an average, an ion undergoes one collision with a molecule of the sample compound for every million collisions with the molecules of the gas atmosphere. Thus, there are ca 104 ion/sample-molecule collisions per second for possible reactions, under thermalized conditions for well-defined temperatures. Additionally, the residence time of an ion in an ion mobility spectrometer operating at atmospheric pressure is ca 5–50 ms, which allows the study of the interactions of ions with molecules at very low concentrations. A further advantage with IMSbased thermochemical determinations is that the available temperature range, from sub-ambient to more than 500 K, is far greater than that available with many other experimental methods. In spite of the advantages cited above, ion mobility spectrometers operating at atmospheric pressure have been used infrequently to obtain physical chemical data, kinetic and thermodynamic, in the study of ion/molecule chemistry. In this chapter, an overview is given on the type of information obtainable from ion mobility studies at atmospheric pressure and the variety of experimental methods employed in such studies. The data obtained under well-defined conditions agree favorably with those from other more frequently used methods, for example: (i) pulsed high pressure mass spectrometry (PHPMS), which is operated at well-defined temperatures but at pressures ca 200 times lower than IMS; and (ii) FT-ICR and ion trap mass spectrometers, which are operated under vacuum.
13.2 THE ION MOBILITY SPECTROMETER AND A MOBILITY MEASUREMENT An ion mobility spectrometer is a simple device with three essential elements: a source region, a drift region, and a detector. The drift tube usually is cylindrical with an overall length from 5 to 10 cm and an internal diameter from 1 to 5 cm. Metal rings separated by insulating material, for example Teflon®, provide a uniform electrostatic drift field when the source end is at high potential and the detector is essentially at ground potential. The high potential is positive for the detection of positive ions and negative for negative ions. The source region is separated from the drift region, as shown in Figure 13.1, by an ion shutter composed of two closelyspaced sets of interdigitated wires, grids 1 and 2 in Figure 13.1 [23]. When the grids are set at the same potential and consistent with their position in the spectrometer, ions pass unhindered from source to drift region; here the ion shutter is open. An imposed potential difference of ca 50–100 V between the two closely-spaced grids creates an electrostatic field far greater than the drift field that is typically around 220 V cm–1, so that ions are drawn to the grid wires and discharged: the ion shutter is now closed. Ions are formed commonly in the source region using a radioactive nickel foil, though other sources, including ultra violet (UV) discharge lamps and corona discharge, have been described. Ions are gated into the drift region in
390
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
S
M 63Ni
I
D
G1 G2
F A
T
P
B
H
FIGURE 13.1 Schematic diagram of an ion mobility spectrometer: A, amplifier; B, metal shell; D, drift gas inlets; F, Faraday plate; G1, G2, grids constituting the ion shutter; H, heating tape; I, insulation rings; M, metal field rings; P coils for pre-heating sample and drift gases; S, sample gas inlet; and T, threaded support rod.
pulses of 30–300 μs-duration by opening and closing the ion shutter at frequencies of 20–100 Hz. The ions move under the influence of the uniform electrostatic field to the detector, a Faraday plate, where signal is generated. The amplified signal, displayed as a function of the time of arrival at the detector, that is drift time, constitutes the ion mobility spectrum. The separation of ions into individual swarms occurs according to their differences in mobilities as swarms drift from the ion shutter to the detector. The drift gas is usually either purified air or nitrogen and the flow (300–1000 cm3 min–1) of the drift gas is from the detector to the ion shutter, that is, counter to the ion drift direction. This gas flow is mixed with the sample-containing source gas (10–100 cm3 min–1) before exiting the instrument at the source end, so the entry of sample vapors into the drift region is suppressed in this configuration of gas flows. The time of drift (td) for an ion swarm in the drift region (of length, l) yields a drift velocity vd given by
vd =
l td
(13.1).
The drift velocity, when normalized for electric field strength, E, produces the mobility coefficient, K, as shown in Equation 13.2
vd = KE
(13.2).
The Study of Ion/Molecule Reactions at Ambient Pressure
391
Because vd and hence K are temperature and pressure-dependent, values for K are usually normalized to 273 K and 760 mm Hg and are reported as the reduced mobility coefficient, Ko,
273 P Ko = K T 760
(13.3),
where T is temperature in kelvin and P is pressure in mm Hg. If the character of the ion does not change Ko has a constant value but, upon change of temperature, differences in ion hydration or clustering with other ambient molecules result in nonconstant Ko values. An ion under the influence of the electric field in the drift region acquires kinetic energy, some of which is lost by collision to the surrounding gas molecules. When an ion is to be accepted as thermal, it is important that the retained kinetic energy (1/2 mvd2 + 1/2 Mvd2 ) is not significant compared with thermal kinetic energy (3/2kbT ) as found in the average kinetic energy of an ion (KEav), which is approximated by the Wannier expression:
KEav =
3 1 1 kbT + mvd2 + Mvd2 2 2 2
(13.4),
where m is the mass of the ion, M is the mass of the drift gas molecule [24]. A calculation of these energies is illustrated for protonated 2,3-dimethyl pyridine with a drift velocity of 769 cm s–1 in a field of 280 V cm–1 at 350 K and 660 mm Hg of N2; the thermal energy term for the drift gas 3/2 kbT has the value 7.3 × 10 –21 J and the retained kinetic energy 1/2mvd2 + 1/2 Mvd2 has the negligible value of 6.8 × 10 –26 J. When this condition holds, a mobility spectrometer is said to be operating in the socalled low-field region and ions are regarded as thermalized. The low-field region is assumed usually when E/N is less than 2 Td, where E is the electric field (V cm–1), N is the number density of molecules (cm–3), and Td is the townsend (10 –17 V cm2). The low-field region is accessed readily with ion mobility spectrometers operating at atmospheric pressure but it is difficult to achieve satisfactorily with instruments that operate at pressures of ca 1 mm Hg. In the low-field region, the mobility coefficient at a fixed temperature is independent of field strength, which can be 100–600 V cm–1 [25]. The lower limit for E is determined practically by radial losses of ions to the walls of the drift tube; the upper limit is defined by electrical breakdown of gases in the supporting atmosphere inside the drift tube. An important property of an ion mobility spectrometer operating in the low-field region is that ion losses to the walls by radial diffusion do not introduce mass discrimination in the collected ion signal because the ratio of the radial spreading distance to the drift distance is independent of ion mass [24].
13.2.1 The Profiles of Ion Mobility Spectra Used to obtain Thermodynamic and Kinetic Data There are two approaches that may be taken to follow the course of an ion/molecule reaction occurring in an ion mobility spectrometer. The first is through the profile of the ion mobility spectrum alone, and the second is from mass spectral ion intensities.
392
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
When possible, the first method is preferable because there is always the danger of mass discrimination, ion/neutral association, or collisional dissociation when ions pass from the high pressure of a mobility spectrometer to the vacuum of a mass spectrometer. Despite these complications, mass spectrometry has value for identifying ions because otherwise ion identity must be deduced intuitively by reference to known or anticipated reactions and ion behavior in a mobility drift tube. A single type of ion, that remains unaltered during the period following injection into the drift region to its arrival at the detector, produces a near-Gaussian peak in the ion mobility spectrum. The width of the peak is determined by a combination of the pulse width of the ion shutter and by Brownian motion [26]. A slight asymmetry in the peak arises from the increased axial diffusion time experienced by the late-arriving ions. When two ions formed in the source transit the drift region with no further change, two peaks are observed, as illustrated in Figure 13.2a for ions A+ and B + . Such spectra cannot provide quantitative physical chemical data because conditions in the ion source, such as concentration gradient of neutral molecules, electrostatic field, and reaction time, are not well defined. Early attempts to use this method to obtain proton affinity differences from the ratio of the intensities of protonated polycyclic aromatic hydrocarbons yielded only the order of proton affinities [27]. When a process occurring in the drift region relates the ions, the mobility spectrum may become either simpler or more complex than shown in Figure 13.2a. Nonetheless, when experimental conditions are controlled, the spectrum may be interpretable and, subsequently, may provide thermodynamic and/or kinetic information pertinent to the process or ion/molecule chemistry. Consider the simple (b)
8
Relative intensity
(c)
Relative intensity
A+ B+ 10
12
14
16
18
20
22
Drift time (ms) 1.2 Type 2 1 A+ 0.8 0.6 0.4 0.2 0 8
24
(d)
AB+
13
1.2 Type 1 1 0.8 0.6 0.4 0.2 0 8
18
A+and AB+
18
Type 3
Ion intensity
Ion intensity
(a)
A+ AB+ 8
10
12 14 16 18 Drift time (ms)
20
22
FIGURE 13.2 Schematic mobility spectra illustrative of ion/molecule processes: (a) Ions A+ and B+ are formed in the source and experience no reaction in the drift region; (b) equilibrium A + + B = AB+ prevails throughout the drift region; (c) A + + B → AB + in the drift region; and (d) AB + →A + + B in the drift region.
The Study of Ion/Molecule Reactions at Ambient Pressure
393
association reaction described by Equation 13.5, in which the ions have a single positive charge (this discussion applies equally to negatively-charged ions) A + + B = AB+
(13.5).
At atmospheric pressure, third-body stabilization of AB + is highly efficient and the reaction may be treated as second-order in the forward direction and first-order in the reverse direction. In addition, the concentration of B is usually much greater than the concentration of A+ , so that the reaction is pseudo-first order in the forward direction with a constant concentration of B. As detailed below, ion/molecule reactions or interactions may be investigated using several experimental designs, each of which has been demonstrated with specific features or advantages. The ion mobility spectra produced in these experimental configurations differ and will be designated Types 1, 2, and 3. 13.2.1.1 Type 1. Equilibrium A + + B = AB + Exists Throughout the Source and Drift Region In the Type 1 experiment, both A + and AB + are formed in the source region and the concentration of B is uniform throughout both source and drift regions. When the concentration of B is much greater than that of the ions, the reaction will attain equilibrium prior to the ions reaching the shutter and equilibrium will prevail in the drift region. The charge spends part of the time on A and part of the time on AB so there is only one peak in the mobility spectrum (Figure 13.2b) and its arrival time (time of maximum ion intensity) depends on the relative equilibrium concentrations of A+ and AB + . The arrival time for this composite peak is intermediate between the arrival times of A+ and AB + without interactions (Figure 13.2a) and is the weighted ion number average of the ion mobilities, which is expressed by Equation 13.6 in terms of the ion mole fractions Xi and their individual arrival times ti
td = X A+ t A+ + X AB+ t AB+
(13.6).
13.2.1.2 Type 2. A + + B → AB + in the Drift Region In this category, A + ions and some AB + ions are formed in the source region, B is present at uniform concentration in the drift region, and the reaction has not gone to completion (that is, attained equilibrium) when ions are introduced into the drift region. Discrete peaks at t A+ and t AB+ are present for A + and AB + that pass unchanged through the drift region. Some AB + ions are formed as A + travels to the detector, the rate of the association reaction being given by:
−
d[A + ] = k[ A + ][ B] dt
(13.7),
where k is the rate constant. The rate of reaction is greatest near the shutter where the concentration of A + is highest, and AB+ ions formed here will arrive at the detector at a time close to t AB+ .
394
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
The rate of reaction is least near the detector, where the concentration of A+ is the smallest, and ions AB + formed here will have spent most of their transit time as A + and will arrive at times close to t A+ . The ‘fill in’ intensity between the normal peaks for A + and AB + is lowest at t A+ and rises exponentially to t AB+ , as illustrated by Figure 13.2c. The same type of mobility spectrum, as is shown in Figure 13.2c for the association reaction, Equation 13.5, is observed when the reaction is a charge exchange, as in Equation 13.8, where B is again present at uniform concentration throughout the drift region A + + B → A + B+
(13.8).
13.2.1.3 Type 3. A + and AB + are Formed in the Source Region and Neither A Nor B is Present in the Drift Region The dissociation of AB + , formed in the source region and injected into the drift region, will be observable at a temperature consistent with the activation energy for dissociation; the higher the activation energy, the higher is the required temperature. Significant reaction in this first-order process can be observed when the rate constant is of the order of the reciprocal of the drift time, which is usually of ca 100 s–1; however, the drift time can be varied to a limited extent by changing the electrostatic field. The mobility spectrum will show two distinct peaks, at time t A+ for A + and at time t AB+ , the drift time for AB + without decomposition within the drift region. Ions arriving at the detector at intermediate times have spent the first part of their drift time as AB + and the second part as A + . The rate of reaction, given by Equation 13.9, is greatest at the entrance to the drift region, where the concentration of AB + is highest, and least at the detector. The plot in Figure 13.2d illustrates the resulting exponential ‘fill in’ of the spectra between the two discrete peaks with maximum intensity at t A+ . −
d [ AB+ ] = k[ AB+ ] dt
(13.9),
where k is the rate constant.
13.2.2 Examples Where Thermodynamic and Kinetic Data have been obtained from Ion Mobility Spectra Though IMS and ion mobility spectrometry/mass spectrometry (IMS/MS) methods may not be recognized widely for determining values for enthalpy, entropy, and kinetic constants, significant experience in the study of reactions at ambient or elevated pressures exists. In the discussion below, examples are drawn from gas-phase reactions for associations and displacements of ions using either combined mobilitymass spectrometry or, in some instances when the chemistry was well known, a drift tube alone. The order of presentation follows that used in the prior section. In a later discussion, reactions with electrons are described.
The Study of Ion/Molecule Reactions at Ambient Pressure
395
13.2.2.1 Type 1. Ions in Equilibrium Showing a Single Peak in the Mobility Spectrum In dried air or nitrogen with ionization using a 63Ni source, the initial ionization event produces N + , N2+ , O + , and O2+ which lead, through a series of rapid ion/molecule reactions with the ubiquitous trace amounts of water, to the hydronium ion (H2O)nH + [20]. The subscript n denotes a range of values that depends on both the water concentration and the temperature. The equilibrium of the hydronium ion is shown in Equation 13.10,
(H 2O)n H + + H 2O = (H 2O)n+1H +
(13.10).
For example, using data from reference [28] and assuming equilibrium conditions with 1.0 ppmv (parts per million by volume) water at 300 K, the populations for n = 2, 3, 4, and 5 are 0, 21, 76, and 3%, respectively. At 400 K, the respective populations are 47, 53, 0.1, and 0%. Proton transfer from one or more of these hydrates to a molecule, present at a concentration much lower than that of water, which is present usually at a concentration of 1–10 ppmv, is often the initial step in forming an ion of interest. Though different experimental methods have been employed in the study of physical chemistry of the proton hydrate, it is not surprising that investigations have been made also with IMS given the importance of the hydronium ion as a reactant in IMS. An early attempt was made by Kim et al. to link the reduced mobility of the hydrated proton to the known range of n for (H2O)nH + in moist atmospheric air [29]. An ion mobility spectrometer interfaced to a mass spectrometer was used with a constant concentration of water throughout the drift tube of the mobility spectrometer, thereby ensuring a constant ratio of the equilibrium concentrations of the various hydrates in the drift region. In the absence of sample molecules, the hydrates constitute the major peak in the mobility spectrum, the reactant ion peak (RIP), whose drift time varies as the value of n changes with change of temperature. An ion/molecule association reaction is always exothermic, so that raising the temperature favors smaller clusters and vice versa. A larger cluster ion has lower mobility than does a smaller one, and so the drift time of the RIP decreases with increase of temperature while a lower temperature leads to a longer drift time. Each hydrate ion has a unique mobility and the drift time of the single peak in the spectrum should give a measure of the equilibrium distribution. The mobility, and hence the reduced mobility, increased as the temperature was raised, consistent with a reduction in the average value of n as the distribution shifted to the lighter, more mobile hydrates. The water concentration was not measured but was calculated by Equation 13.11 from the mass spectral intensities In and In–1 (presumably at one temperature, although this is not stated) together with the equilibrium constants Kn–1,n from the thermodynamic data of Kebarle et al. [30].
PH2O =
In I n−1K n−1,n
(13.11).
The distribution diagram for the proton hydrates was calculated over the whole temperature range with this water concentration and the equilibrium constants. The
396
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
measured ion mobility over the whole temperature range is in good agreement with that obtained from the formula in Equation 13.12, which relates the mobility coefficient K to the ion/molecule interaction potential Ω, the measure of closest approach of ion and molecule rm, and their reduced mass μ [1],
K=
3 2π 16 N µ kbT
1/2
1 π rm2Ω
(13.12).
The same studies were reported for the hydrates of NO + and NH4 + , which are minor ions in the same system. When equilibrium exists in an ion/molecule reaction then the relative peak heights of the different hydrate ions, sampled from a mobility spectrometer through a small orifice into a mass spectrometer, give a direct measure of their equilibrium concentrations, provided there is no mass discrimination in the sampling process. Gheno and Fitaire applied this method to obtain thermodynamic data for the proton hydrate equilibria over the temperature range 300–473 K with N2 drift gas containing 3 ppmv water [31]. Van’t Hoff plots yielded the standard enthalpy and entropy values of Table 13.1 that show excellent agreement with National Institute of Standards and Technology (NIST) values [28] of both the enthalpy (–ΔrHo) and entropy (–ΔrSo) changes for the formation of (H2O)3H + but less so for (H2O)4H + . The reduced agreement with NIST values for (H2O)4H + is due probably to dissociation in the interface to the mass spectrometer of the higher, more fragile n = 4 hydrate. The enthalpy change for the association of N2 with NO + and to a lesser extent the entropy change are in good agreement with the values obtained by PHPMS at ca 4 mm Hg [32]. The data for such weakly-bound clusters as N2·NO + must always be treated with some skepticism because there is usually a danger that the ions detected are clustered with drift gas molecules during adiabatic cooling in the free-jet expansion between the mobility drift tube and the mass spectrometer [33]. However, it would be surprising if exactly the same relative ion intensities in the mass spectrum occurred in expansion from atmosphere pressure and also from 4 mm Hg, and the equality of the thermodynamic data for the formation of the N2 · NO + complex validates the results, the IMS data, and the method. Preston and Rajadhyax who employed a hand-held ion mobility spectrometer, similar to those used in military applications, studied the correlation between the reduced mobility for ion processes at equilibrium and the arrival time of the single peak in the mobility spectrum [34]. Equilibrium constants were determined as a function of temperature for the formation of proton-bound dimers MZH + as shown in Equation 13.13, where each of M and Z can be pyridine, 3-(3-methoxypropoxy) propanol (DPM), acetone, or water,
MH + + Z = MZH +
(13.13).
At equilibrium, the arrival or drift time td of the single peak is the number-weighted average of the drift times of the two constituent ions as described by Equation 13.6. When the times spent as the individual ions are t MH+ and t MZH+ , respectively, it follows that
397
The Study of Ion/Molecule Reactions at Ambient Pressure
TABLE 13.1 Thermodynamic Data for Hydration Reactions from IMS/MS Determinations Reaction (H2O)2H + + H2O = (H2O)3H + (H2O)3H + + H2O = (H2O)4H + NO + + N2 = N2 · NO +
–ΔrHo (kJ mol−1)
–ΔrSo (J K−1 mol−1)
86 ± 4 (84 ± 5)a 58 ± 5 (73 ± 4)a 20 ± 4 (19 ± 1)b
109 ± 17 (94 ± 20)a 78 ± 17 (118 ± 8)a 58 ± 13 (71)b
Source: Gheno, F.; Fitaire, M. J. Chem. Phys. 1987, 87, 953–958. a Evaluated data in reference NIST. NIST Chemistry Webbook. 1998. b PHPMS data from reference Hiraoka, K.; Yamabe, S. J. Chem. Phys. 1989, 90, 3268–3273.
X MH+ a = t MH+ /td
and
X MZH+ = t MZH+ /td
(13.14),
and the measured reduced mobility for a peak at any temperature is related to the individual reduced mobilities by
K o = X MH+ K o MH+ + X MZH+ K o MZH+
(13.15).
When Equations 13.14 and 13.15 are combined, the equilibrium constant is given by Equation 13.16 in which Po is the standard pressure, 101 kPa, and PZ is the partial pressure of Z
K=
X MZH+ Po K MH+ − K o Po ⋅ = o ⋅ X MH+ PZ K o − K o MZH+ PZ
(13.16).
As the concentration of the neutral molecule Z was increased, the reduced mobility attained a limiting minimum value, calculated from the arrival time of the single peak. This value was taken to be K o MZH + , the concentration of MH + presumably being negligibly small. The term K o MH + was the reduced mobility in the absence of Z, and equilibrium constant measurements over a range of temperature yielded the results presented in Table 13.2. The standard enthalpy changes for acetone–water and pyridine–pyridine are in fair agreement with literature values but there is no agreement for pyridine–water. Some of the standard entropy changes are impossibly high and show little agreement with literature values. Although the method is sound, the instrument used for this study was not suited particularly to the determination of reliable thermodynamic data because the drift region was only 3.7 cm long. Giles and Grimsrud described an instrument designed specifically for the study of ion/molecule reactions [35]. The cylindrical drift tube was large, 40 cm long and 9 cm in diameter, and the moveable ion source allowed facile change in drift length.
398
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
TABLE 13.2 Thermodynamic Data for the Reaction MH + + Z = MZH + as Measured by Mobility Spectrometry M Acetone Pyridine Pyridine DPM
Z
–ΔrHo (kJ mol–1)
–ΔrSo (J K–1 mol–1)
Water Pyridine Water Acetone
99±10 (84) 139±25 (103) 44±9 (97) 67±7
226±50 (109) 260±80 (124) 50±50 (116) 45±6
Source: Preston, J.M.; Rajadhyax, L. Anal. Chem. 1988, 60, 31–34. Note: Values in brackets are PHPMS data from reference NIST. NIST Chemistry Webbook. 1998.
Ion mobility spectra were recorded using a Faraday plate with a small central orifice that permitted passage of ions for identification by a quadrupole mass spectrometer. Ion/molecule reaction rate constants and reaction enthalpies and entropies were determined using this IMS/MS instrument; one application followed the pioneering work of Preston and Rajadhyax [34] to measure the equilibrium constant for the association of CHCl3 and Clˉ. In this work, ions were formed in the source region and passed through a counter flow of CHCl3 in the drift region. The arrival time of the single peak, initially due to Clˉ, increased with increasing concentration of chloroform, suggesting the formation of Clˉ(CHCl3), as in Equation 13.17:
Cl − + CHCl 3 = Cl − (CHCl 3 )
(13.17).
The identity of this adduct ion was confirmed by the mass spectrum that contained both Clˉ and Clˉ(CHCl3). A limiting arrival time of the peak defined the concentration of CHCl3 at which essentially all the Clˉ was found in the adduct ion. When X Cl − and X Cl − ( CHCl3 ) are the equilibrium mole fractions of the two ions and tobs is the arrival time of the composite peak, which changes from to for Cl– to t1 for Clˉ(CHCl3), then:
X Cl − + X Cl − (CHCl3 ) = 1
(13.18)
X Cl − to + X Cl − ( CHCl3 ) t1 = tobs
(13.19)
X Cl − ( CHCl3 ) X Cl − K=
=
tobs − to [Cl − (CHCl 3 )] = [Cl − ] t1 − tobs
1 tobs − to ⋅ t1 − tobs [CHCl 3 ]
(13.20)
(13.21)
The Study of Ion/Molecule Reactions at Ambient Pressure
399
so that
1 1 1 = + tobs − to K (t1 − to )[CHCl 3 ] t1 − to
(13.22).
A plot of (tobs–to) –1 vs [CHCl3]–1 yields a straight line with slope [K(t1–to)]–1. The mobility coefficient was determined at three different temperatures and ΔrHo = –75.7 kJ mol–1 and ΔrSo = –91.2 J K–1 mol–1 were obtained from the van’t Hoff plot; these values are in good agreement with listed values that range from –75.7 to –81.6 kJ mol–1 and –61.9 to –103 J K–1 mol–1, respectively [28]. An experiment with Brˉ forming Brˉ·CHCl3 found ΔrHo = –66.1 kJ mol–1 and ΔrSo = –88.2 J K–1 mol–1, the former value being appropriately smaller than for the Clˉ analog. The chloride anion is known as a useful reactant ion for IMS, particularly in the detection of explosives [1,2,36,37], as it forms complexes with molecules by bonding via hydrogen(s) rendered acidic by the presence of NO2 groups. Such complexes were not studied for explosives that in general have very low volatility, but rather for compounds of high volatility that are added in low concentration as chemical taggants. Information on the relative strength of binding between Clˉ and several potential taggants, 1,4-dinitrobenzene (DNB), 2,3-dimethyl-2,3-dinitrobutane (DMNB), and 2,3-dimethyl-2,4-dinitropentane (DMDNP) was obtained by Lawrence et al. from equilibrium measurements by IMS/MS [38]. The Clˉ reactant ion was formed in the source region by the incorporation of trace amounts of CH2Cl2 into the ultra high-purity nitrogen source gas; samples of each nitro compound (M) produced the Clˉ(M) adduct (Equation 13.23), which was sufficiently stable to traverse the drift region to the detector:
Cl − + M = Cl − (M)
(13.23).
A measure of the stability of each adduct was obtained by raising the temperature of the whole instrument to determine the highest temperature at which the adduct was observed in the mobility spectrum. The order of stability determined, Clˉ(DNB) > Clˉ(DMDNP) > Clˉ(DMNB), is the same as the number of H atoms in a position α to the NO2 groups, that is, 4, 1, and 0. Further studies were made with DMNB because it has suitable properties, in particular vapor pressure, as a taggant. The standard enthalpy and entropy changes for the association of Clˉ with DMNB were determined by forming Clˉ in the source region and adding DMNB, in increasing and known concentrations, to the drift gas that flowed counter to the ions. A peak at time to when no DMNB was present was due to Clˉ. The single peak seen in the mobility spectrum with DMNB in the drift gas stream was identified as a mixture of Clˉ and Clˉ (DMNB) using an IMS/MS instrument. When the DMNB concentration was increased further, the peak shifted to a longer drift time as shown in Figure 13.3. As concentration was increased further, the drift time attained a constant and maximum value, t1; the sole component of the peak was mass-identified as Cl– (DMNB). Mass spectra for the composite peak at the lower concentrations of DMNB demonstrated the equilibrium of Equation 13.23. The interpretation of the results follows that of Giles and Grimsrud where a plot of (tobs –to) –1 vs [DMNB]–1 yielded a straight line with slope of [K(t1–to)]–1. The
400
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
50 ppm DMNB
Signal intensity (arbitrary units)
9.3 ppm 2.7 ppm 0.7 ppm 0.2 ppm 0 ppm
5
10 15 Ion drift time (ms)
20
FIGURE 13.3 Ion mobility spectra as functions of the concentration of DMNB at 443 K. The peak at 0 ppm DMNB is the chloride ion alone and peaks with DMNB are due to the Clˉ(DMNB) adduct where drift time is dependent upon vapor concentration of DMNB in the drift tube. (Reproduced from Lawrence, A.H., et al., Int. J. Mass Spectrom. 2001, 209, 185–195. With permission from Elsevier.)
equilibrium constant K was determined at different temperatures and, from a van’t Hoff plot, ΔrHo = –92.1±3.1 kJ mol–1 and ΔrS o = –92.1±7.4 J K–1 mol–1 were obtained. 13.2.2.2 Type 2. Reaction Rate Constant Measurements Spectra of Type 2 were used for kinetic studies by Giles and Grimsrud who determined the rate constants for the SN2 displacement of Brˉ by Clˉ from methyl-, ethyl-, isopropyl-, and n-butyl-bromide at 398 K as per Equation 13.24 [35],
Cl − + RBr → RCl + Br −
(13.24).
The chloride ion was formed by dissociative electron capture by CCl4 in a 63Ni ionization source; n-alkyl bromide, at a known concentration, was present in the drift region. The traces in Figure 13.4 are ion mobility spectra obtained at three concentrations of methyl bromide. Figure 13.4a shows the Clˉ peak in addition to two small peaks, each marked with an asterisk, that arise from impurities; the small
401
The Study of Ion/Molecule Reactions at Ambient Pressure 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0
(b)
0.06 0.04 0.03 0.02 0.01 0
(c)
(d)
Ion intensity (nA)
(a)
0.05 0.04 0.03 0.02 0.01 0 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 20
25
30
35
Drift time (ms)
FIGURE 13.4 Ion mobility spectra for the reaction of Clˉ with CH3Br at concentrations of (a) none, (b) 1.29 × 10l2, (c) 2.60 × 10I2, (d) 5.27 × 10l2 molecules cm–1. (Reproduced from Giles, K.; Grimsrud, E.P. J. Phys. Chem. 1992, 96, 6680–6687. With permission from the American Chemical Society.)
peaks were invariant with methyl bromide concentration and can be ignored. The intensity for Clˉ in the mobility spectra decreases as the Brˉ intensity increases with increasing concentration of CH3Br, as shown in Figure 13.4b through d. The contribution of each ion to the shape of the mobility spectrum was determined by concurrent mass spectrometry to obtain the relative mobility spectral area Ai assignable to each ion as a function of alkyl bromide (RBr) concentration. The pseudo-first order reaction rate constant for reaction 13.24, k’ = k24[RBr], is given by Equation 13.25 in which td is the arrival time, that is the reaction time, of Clˉ:
402
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
k′ =
1 ACl − + ABr − ln td ACl −
(13.25).
The graph of k’ vs [RBr] is a straight line of slope k24 and the value for k24 was duplicated in a different experiment involving mass spectrometric detection with the drift tube acting as a reactor with the shutter open continuously. The intensities of Clˉ and Brˉ were measured as a function of RBr concentration with reaction time the same as for the pulsed mode; a plot of ln( I Cl − /I Cl − + I Br − ) vs [RBr] yielded a straight line of slope –k24td. The results obtained for the four alkyl bromides by the area method and the mass spectrometric method are compared in Table 13.3 with those obtained from a PHPMS study completed at ca 4 mm Hg [39]. All the rate constants are much lower than the calculated average dipole orientation (ADO) collision rate constants, which are of the order of 2 × 10 –9 cm3 molecule –1 s –1, consistent with the inhibiting central barrier described by the double-well potential theory [40]. There is excellent agreement between the PHPMS and IMS results for ethyl bromide (EtBr) and i-PrBr, but there is a significant difference for methyl bromide (MeBr), which suggests that the difference is not an experimental artifact but might be due to the different pressure regimes. Less efficient stabilization of the initially formed Clˉ···CH3Br complex at the lower pressures could be the explanation. To confirm that the difference is real, Knighton et al. compared the IMS/MS rate constants at ambient pressure, 640 Torr, with PHPMS rate constants at 3 Torr over the temperature range 308–423 K [41]. A significantly-higher rate constant was found at the higher pressure, although it was still far below the collision rate. This result, interpreted in terms of the double-well potential theory, suggests that the higher pressure leads to increased stabilization of the short-lived entrance channel intermediate (MeBr)Clˉ*. From the IMS/MS results, the estimated depth of the central barrier below the incoming channel, 2.2 kcal mol–1, is identical with the value obtained from a high level calculation [42].
TABLE 13.3 Rate Constants (cm3 molecule –1 s –1) at 398 K by IMS for the Reaction Cl– + RBr→RCl + Br– Method RBr MeBr EtBr i-PrBr n-BuBr
Area
MS
3.4 × 10 1.1 × 10–11 8 × 10–13 2.2 × 10–11 –11
PHPMSa
3.4 × 10 1.3 × 10–11 7.6 × 10–13 2.2 × 10–11 –11
8.8 × 10–12 9.7 × 10–12 6.2 × 10–13 2.0 × 10–11
Source: Giles, K.; Grimsrud, E.P. J. Phys. Chem. 1992, 96, 6680–6687. a
Caldwell, G.; Magnera, T.F.; Kebarle, P. J. Am. Chem. Soc. 1984, 106, 959–966.
The Study of Ion/Molecule Reactions at Ambient Pressure
403
The IMS/MS result cannot confirm that the high-pressure limit of kinetic behavior has been attained, but it demonstrates clearly that low-pressure experiments for such reactions cannot provide definitive information regarding the problem. There is no difference between the results for the reaction of Clˉ with EtBr and n-BuBr at 640 Torr and at 3 Torr, suggesting that the intermediates in these reactions have longer lifetimes than that of their methyl analog and that their high-pressure limit is attained even at 3 Torr. The upper pressure range of the mobility spectrometer was extended to enable a further study in the reaction of Clˉ with MeBr from 300 Torr to 1100 Torr N2 [43]. Over this range, the reaction rate constant increased by ca 25% demonstrating that the high-pressure limit was not attained even at 1100 Torr. The nascent collision complex must have a lifetime toward back dissociation that is much less than the ca 40 ps between stabilizing collisions. When the nitrogen drift gas was replaced by methane, the rate constant increased further. Better quenching of the intermediate (MeBr)Clˉ* by the more complex drift gas molecule is the most likely explanation for the increase. 13.2.2.3 Type 3. Dissociation of Adduct Ions Rate constants for the first-order dissociation of symmetrical proton-bound dimers, M2H + → MH + + M, have been determined for organophosphorus compounds (M = 2,4-dimethylpyridine (DMP) and dimethyl methylphosphonate (DMMP)), where the shapes of the mobility spectra are of the form shown in Figure 13.2d [44]. Some proton-bound dimers decompose in the time taken for the ions to travel between the shutter and the detector plate, and this residence time was varied by changing the electrostatic drift field strength. Typical ion mobility spectra obtained at different field strengths are shown in Figure 13.5 and peaks were mass identified as: first peak, H + (DMP), the protonated monomer and second peak H + (DMP)2, the proton-bound dimer. The raised baseline between the peaks was due entirely to (DMP)H + , from the decomposition of the proton-bound dimer as in Equation 13.25
H + (DMP)2 → H + (DMP) + DMP
(13.25).
The reaction rate constant was determined in the following manner. The distance x from the shutter at which decomposition occurs at time tx is given by Equation 13.26, in which L is the shutter-detector plate distance, td is the drift time of (DMP)2H +, and tm is the drift time of (DMP)H + t − tm x = L td − tm
(13.26).
The value of tx is then given by
tx =
x xtd t − tm = = td td − tm vd L
(13.27).
404
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Intensity (counts)
3.E+05
100 V cm–1
2.E+05 1.E+05 0.E+00
0
5
10
15
20
25
30
35
Drift time (ms)
Intensity (counts)
2.E+06
200 V cm–1
1.E+06
0.E+00
0
5
10
15
20
25
Drift time (ms)
Intensity (counts)
6.E+06
280 V cm–1
3.E+06
0.E+00
0
5
10
15
20
25
Drift time (ms)
FIGURE 13.5 Mobility spectra for DMP at three electric field strengths obtained in air at 350.4 K and 5 ppm moisture. (Reproduced from Ewing, R.G.; Eiceman, G.A.; Harden, C.S.; Stone, J.A., Int. J. Mass Spectrom. 2006, 255, 76–85. With permission from Elsevier.)
The concentration of ions from proton-bound dimer not decomposed at time tx is proportional to the area of the mobility spectrum from tx to the end of the proton-bound dimer peak. For the first-order decomposition of M2H + , a plot of the logarithm of this area vs tx is a straight line of slope –k, where k is the reaction rate constant. Examples of plots for the decomposition of (DMP)2H + from measurements obtained at 350 K and different field strengths are shown in Figure 13.6. The sudden drop in ion signal in each graph signifies the end of the proton-bound dimer peak. The figure demonstrates that the drift time of (DMP)2H + varies inversely with the field strength, and that the slope is independent of field strength. Values of k obtained at different temperatures enabled the determination of an activation energy Ea and
405
The Study of Ion/Molecule Reactions at Ambient Pressure 18
In [(DMP) 2H* remaining]
17.5
280 260 240 220 200
17 16.5 16
180 160
15.5
140
15 14.5
120 100
14 13.5
0
5
10
tx (ms)
15
20
25
FIGURE 13.6 Plots of ion intensity for (DMP)2H + remaining at time ts at different field strengths E (V cm–1) at 349 K and 5 ppm moisture in the supporting atmosphere of the drift tube. (Reproduced from Ewing, R.G.; Eiceman, G.A.; Harden, C.S.; Stone, J.A., Int. J. Mass Spectrom. 2006, 255, 76–85. With permission from Elsevier.)
TABLE 13.4 Arrhenius Activation Energies and Pre-exponential Factors for the Reaction M2H + → MH + + M in the Presence of Different Water Concentrations M DMP DMP DMMP DMMP DMMP
Water (ppmv)
T range (K)
Ea (kJ mol–1)
Log [A (s–1)]
5 2 × 103 5 5 × 102 5 × 103
338–58 311–42 478–98 478–98 478–98
94±2 31±5 127±3 130±2 115±1
15.9±0.4 6.3±0.7 15.6±0.3 15.3±0.4 14.5±0.3
Source: Ewing, R.G.; Eiceman, G.A.; Harden, C.S.; Stone, J.A. Int. J. Mass Spectrom. 2006, 255, 76–85. With permission.
pre-exponential factor A for the dissociation from an Arrhenius plot of ln k vs 1/T. Because some water vapor is always present in mobility drift gases, the experiments presented a favorable opportunity to examine any effects that water vapor may have on reaction rates. The results obtained with different water vapor concentrations are shown in Table 13.4. Of note is the relatively narrow temperature range over which measurements could be made due to the high activation energies for reaction, coupled with the small dynamic range of the mobility spectrometer. The narrow temperature range does not, however, preclude obtaining results with good precision. The pre-exponential factor of ca 1015 s–1 is the maximum expected for unimolecular decompositions [45].
406
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
A measure of the accuracy of the results can be obtained with the knowledge that there should be little or no reverse activation energy for the decomposition, and the activation energy will differ by only RT from the enthalpy of dissociation. The value of Ea obtained for DMP at 5 ppmv water is in excellent agreement with the expected ca 92 kJ mol–1 for symmetrical N2–base dimers [28]. Similarly, the activation energies for DMMP are consistent with the expected ca 134 kJ mol–1 for symmetrical oxygen-base dimers. The activation energy and pre-exponential factor for the decomposition of (DMP)2H + at high water concentration are anomalously low and suggest the influence of a displacement reaction
H 2O + (DMP)2H + → (DMP)H + (H 2O) + DMP
(13.28),
in which the symmetric proton-bound dimer becomes an asymmetric protonbound dimer ((DMP)H + H2O), the activation energy being the difference in bonding enthalpy of DMP and water to (DMP)H + . Lower reduced mobility values for (DMP) H + at the higher water concentration of 2 × 103 ppmv implies, and mass spectra confirmed, that (DMP)H + was indeed hydrated in the experimental temperature range. Data from PHPMS [46] suggest a difference of ca 36 J K–1 mol–1, and the difference of 63 J K–1 mol–1 in Table 13.4 may imply that more than one water molecule is involved in the reaction. The much lower entropy change expected for a displacement reaction compared with unimolecular dissociation is also consistent with the lower pre-exponential factor. The dissociation of (DMMP)2H + occurred at a much higher temperature than that of (DMP)2H + , and the slight effect of water vapor on the kinetics shows that the reaction was mainly unimolecular even at 5 × 103 ppmv water.
13.2.3 The Kinetics of Thermal Electron Capture and Thermal Electron Detachment Ion sources operating at atmospheric pressure in N2 or air provide an abundant source of thermalized electrons. In the absence of molecules with suitable electron affinity, and with an ion mobility spectrometer operating in the negative ion mode, electrons are injected into the drift region to arrive at the detector about 100 times faster than any ion. Suitable molecules present in the drift region may capture electrons, leading to a decrease in electron signal and a negative ion signal at much longer times. Anions formed in the source region by attachment of thermalized electrons, and injected into the drift region in the absence of attaching molecules, may thermally lose the electron at suitable temperatures and produce an interpretable mobility spectrum. Mobility spectrometers operating at atmospheric pressure have been employed to obtain reaction rate constants for both electron capture and for thermal electron detachment studies and these methods are discussed separately below. 13.2.3.1 Electron Capture Spangler and Lawless [47] measured the rate constant for dissociative electron capture by chlorobenzene by monitoring the production of Clˉ. Electrons traveling down the drift tube to the detector encountered chlorobenzene molecules from an exponential dilution
The Study of Ion/Molecule Reactions at Ambient Pressure
407
flask whose concentration varied in a well-defined manner. The Clˉ ion signal was related to the chlorobenzene concentration and a rather complex mathematical modeling of the system gave a rate constant of 7.1 (±3.1) × 10 –11 cm3 s–1 at 473 K, in agreement with experimental data from the more conventional swarm beam technique [48]. Mayhew and co-workers constructed a mobility spectrometer–mass spectrometer combination that has been used extensively to determine the kinetics of the attachment of low energy electrons to halogen-containing molecules at atmospheric pressure [49–56]. Attachment rate coefficients have been determined in two ways. One method is by monitoring the attenuation of a pulse of electrons passing through the drift region containing a known concentration of attaching molecules, and the second by determining the axial distribution of the resulting anions. Excellent agreement between the results obtained by the two methods and with results from the literature was found for SF6 [56]. A pulse of electrons passing through the drift region is attenuated exponentially by radial scattering to give a detector signal Io. In the presence of a constant concentration of an attaching molecule M the signal is further attenuated to a value I, as in
I = I oe − αL [ M ]
(13.29),
where α is the density-normalized electron attachment coefficient and L is the drift length. A plot of ln(I/Io) vs the concentration of M has slope –αL. The rate constant for electron attachment k is obtained from αL with the substitution L = wt where t is the electron arrival time and w is the mean electron drift velocity. The value of w, which is a function of E/N, is different for each drift gas and is determined theoretically (values for nitrogen are available in the literature [37]). The electron attachment rate constant for SF6 in nitrogen at ambient temperature and pressure showed a smooth decline with increasing E/N over the range of 0.39– 0.78 Td [56]. As shown in Figure 13.7, the results obtained by IMS agree closely with those obtained by the well-established high-pressure swarm technique [57]. A further series of experiments with E/N from 0.05 to 0.9 Td confirmed this excellent agreement between the two methods [55]. Both the mobility and the swarm experiments showed that the electron energy distribution in nitrogen is not thermal, even at E/N < 0.1 Td; however, when N2 was replaced by CO2 as the drift gas, the rate constant was determined as 2.38 (±0.15) × 10 –7 cm3 s–1 and this value was independent of E/N at values up to 1.6 Td, the limit of experimentation. Furthermore, the value of the rate constant in nitrogen as E/N→0, extrapolated to the same value [49]. The demonstrated ease of use and convenience of IMS for obtaining electron capture rate constants would appear to favor it over the more conventional high-pressure swarm technique. Ideally, the instantaneous concentration profile of anions, formed by electron capture along the drift region immediately after the pulse of electrons passes through, is exponential, with a maximum value at the shutter. Collisions with drift gas molecules stabilize very rapidly the nascent anions on a time scale far less than their arrival time and the mobility spectrum has maximum intensity at the longest time. With the probable formation of some anions in the source region before the gate, the mobility spectrum
408
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 2
k (10–7 cm3 s–1)
1.5
1
0.5
0
0
5
10
E/N (10–18 V cm2)
15
FIGURE 13.7 A diagrammatic comparison of the dependence on E/N of the rate constant for electron capture by SF6 using data (solid squares) obtained by ion mobility spectrometry. Liu, Y.; Mayhew, C.A.; Peverall, R. A new experimental approach to investigate the kinetics of low energy electron attachment reactions. Int. J. Mass Spectrom. Ion Processes 1996, 152, 225–242. with data (open squares) obtained by the swarm method. (From Hunter, S.R.; Carter, J.G.; Christophorou, L.G., Low energy electron attachment to sulfur hexafluoride in nitrogen, argon, and xenon buffer gases. J. Chem. Phys. 1989, 90, 4879–4891.)
will be of the form shown in Figure 13.2c. The zero of the time scale for anion formation is at the position of peak A + , which represents electrons that reach the detector without reacting, and the location of peak AB + represents the arrival time of anions that formed in the source region. The upward-sloping fill-in region is due to anions formed in the drift region. Analysis of the arrival time spectrum of SF6ˉ in N2 at E/N = 0.67 Td is in excellent agreement with the value obtained by monitoring the electron signal [56]. Tabrizchi and Abedi have constructed a mobility spectrometer, using corona discharge Ionization, with which they have measured the rate constants for electron attachment to CCl4, CHCl3, and CH2Cl2 by monitoring the arrival time of anions formed in N2 in the drift region [58]. The two mobility spectra in Figure 13.8, permit a comparison of the introduction of CCl4 in the source but not in the drift region with that for introduction in the drift region. Introduction of CCl4 into the source produces two discrete peaks with no baseline ‘fill in’ intensity between them, graph (a). The very narrow peak at close to zero time is due to electrons and the second, broader peak at 10 ms is due to Cl− produced by dissociative electron capture. Introduction of CCl4 into the drift region produces the same two peaks but now with an increase in intensity throughout the drift time period between them, graph (b). The slope of this intervening line, plotted on a logarithmic scale, is directly proportional to the electron capture rate constant, and was found to be independent of electric field strength from 247 to 626 V cm−1. The determined electron attachment rate constants for CCl4, CHCl3, and CH2Cl2 are in good agreement with those reported in the literature [59].
409
The Study of Ion/Molecule Reactions at Ambient Pressure 250
Cl– ions produced in the ionization region
Electrons
Intensity (arbitrary unit)
200
150
100
Cl– ions produced in the drift region
50
0
(b) (a) 0
2
4
6
8
10
Drift time (ms)
12
14
tp
Figure 13.8 Mobility spectra for CCl4 when introduced into the ion source region or into the drift region of the mobility spectrometer. (Reproduced from Tabrizchi, M.; Abedi, A. J. Phys. Chem. A. 2004, 108, 6319–6324. With permission from the American Chemical Society.)
13.2.3.2 Thermal Electron Detachment Sahlstrom et al. [60] showed that the thermal detachment of an electron from an anion is observed readily in an ion mobility spectrometer. At an appropriate temperature, anions formed in the source decompose by thermal electron detachment in the drift region. The electrons move rapidly in the electrostatic field to the detector plate and their intensity at arrival time is a measure of the number of anions disappearing at that time. The resulting spectrum, of the form of Figure 13.2d, shows an elevated baseline that has a maximum at zero time, that is, for electron detachment at the shutter where the anion concentration is highest, and terminates at the peak for survivor anions. Examples of the mobility spectra obtained for thermal electron detachment from the azulene anion at different temperatures are shown in Figure 13.9. The Clˉ peaks in the spectra are due to background ions formed in the source and do not interfere with the analysis. The exponential decay of the elevated baseline is described by
ln I = − kt + ln I o
(13.30),
in which I is the electron intensity at time t, Io is the initial intensity, and k is the rate constant for electron detachment. Plots of ln I vs t are linear with slope –k and, although the magnitude of k restricted measurement to the temperature range 398– 498 K, an excellent Arrhenius plot was obtained. The results obtained by the same authors at three different pressures, below and above atmospheric, are compared in Table 13.5 with a PHPMS value obtained at 4 mm Hg [61].
410
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (a) Az–
Relative intensity
(b) Az– Cl–
(c) Cl–
0
20
40 60 Time (msec)
80
100
FIGURE 13.9 Thermal electron detachment spectra for the azulene anion (Az–) at different temperatures: the Cl– ions are background ions formed in the source. (A) 375K, (B) 387 K, (C) 427 K. Drift field 144 V cm–1, P = 740 Torr. (Reproduced from Sahlstrom, K.E.; Knighton, W.B.; Grimsrud, E.P., Int. J. Mass Spectrom. 1998, 179/180, 117–127. With permission from Elsevier.)
TABLE 13.5 Arrhenius Parameters for the Thermal Electron Detachment from the Azulene Anion Method
P (mm Hg)
A (s–1)
PHPMS IMS IMS IMS
4 300 740 1,100
1.14 × 10 3.4 × 1013 2.1 × 1013 1.3 × 1014
a
Ea (kJ mol–1) 11
65.7±4.2 89.1±8.4 87.9±8.4 95.0±8.4
Source: Sahlstrom, K.E.; Knighton, W.B.; Grimsrud, E.P. Int. J. Mass Spectrom. 1998, a
179/180, 117–127. Grimsrud, E.P.; Chowdhury, S.; Kebarle, P. J. Chem. Phys. 1985, 83(8), 3983–3989.
The Study of Ion/Molecule Reactions at Ambient Pressure
411
The large difference between the results at the high pressures and the one at low pressure is significant. The authors suggested that an association of one or more drift gas molecules to the ion might explain the increase in both the activation energy and the preexponential factor. Such an association would lead to an activation energy higher by the binding energy of the associated molecule(s) and to a higher pre-exponential factor due to the increased entropy of dissociation. This topic certainly merits further investigation.
13.2 CONCLUDING REMARKS An ion mobility spectrometer operating at atmospheric pressure is a simple instrument that has seen extensive use in chemical measurements for detection of compounds in very low concentrations for aviation security and military preparedness; however, little use has been made of IMS in experiments of value for physical chemistry. A great advantage of IMS methods and experimental conditions is that the high pressure in an IMS drift tube ensures that the measurements are more realistic to events and processes occurring at or near ambient pressure, because thermal conditions always prevail for ions in the lower electrostatic field condition. There is also a large range of temperatures over which reactions of positive and negative ions can be studied from below ambient to at least 600 K. Such advantages have motivated some to use IMS instruments for elementary studies such as ion/molecule association and displacement reactions, where results for standard enthalpy and entropy measurements are found to be in excellent agreement with those obtained by more usual methods. Reaction rate constant measurements for association and displacement reactions have been determined and, in the case of the displacement of Br– from CH3Br by Clˉ, an effect of pressure on the rate constant has been observed. Electron association and detachment reactions are studied more easily in an ambient pressure ion mobility spectrometer than the more conventional swarmbeam method and we suggest that this will be the method of choice in future studies. Because there is no well-developed predictive capability between mobility coefficients and exact identities of ions, the lack of ion identity of a stand-alone mobility spectrometer must be regarded as a major drawback of simple configurations of drift tubes. The combination of IMS and MS analyzers eliminates this problem but the relative intensities of ions of interest may be distorted due to mass discrimination in the interface. Distortion of the mobility spectrum may also arise from the dissociation of weakly-bound adduct ions and the reverse, the clustering of neutrals with ions in the free-jet expansion. Trace amounts of water in the drift gas pose a problem that may be solved, in part, by either extensive preparation of the gas supply or by operating drift tubes at elevated temperatures. Despite such considerations, there is often sufficient information contained in the mobility spectrum as described above to obtain excellent kinetic and/or thermodynamic information. Consequently, further exploration of the capabilities and refinement of this versatile instrument is warranted. Apart from the determination of rate constants and thermodynamic values, a key, and growing, use of mobility methods includes the study of ion structure where impact in chemistry and biochemistry is occurring and can be anticipated to grow in importance [62]. The interface between these largely separated developments should be a promising subject for future studies.
412
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
REFERENCES
1. Karasek, F.W.; Denney, D.W. Detection of 2,4,6-trinitrotoluene vapours in air by plasma chromatography. J. Chromatog. 1974, 93, 141–147. 2. Lawrence, A.H.; Neudorfl, P. Detection of ethyleneglycol dinitrate vapors by ion mobility spectrometry using chloride reagent ions. Anal. Chem. 1988, 60, 104–109. 3. Lawrence, A.H. Characterization of benzodiazepine drugs by ion mobility spectrometry. Anal. Chem. 1989, 61, 343–349. 4. Bollan, H.R.; West, D.J.; Brokenshire, J.L. Assessment of ion mobility spectrometry for monitoring monoethanolamine in recycled atmospheres. Int. J. Soc. Ion Mobility Spectrom. 1998, 1, 7–12. 5. Gan, T.H.; Corino, G. Selective detection of alkanolamine vapors by ion mobility spectrometry with ketone reagent gases. Anal. Chem. 2000, 72, 807–815. 6. Asbury, G.R.; Klasmeier, J.; Hill, H.H. Analysis of explosives using electrospray ionization: Ion mobility spectrometry (ESI:IMS). Talanta 2000, 50, 1291–1298. 7. Wu, C.; Siems, W.F.; Hill, H.H. Secondary electrospray ionization ion mobility spectrometry/mass spectrometry of illicit drugs. Anal. Chem. 2000, 72, 396–403. 8. Palmer, P.T.; Limero, T.F. Mass spectrometry in the U.S. Space Program: Past, present, and future. J. Am. Soc. Mass Spectrom. 2001, 12, 656–675. 9. Ewing, R.G.; Atkinson, D.A.; Eiceman, G.A.; Ewing, G.J. A critical review of ion mobility spectrometry for the detection of explosives and explosive related compounds. Talanta 2001, 54, 519–529. 10. Bollan, H.R.; Brokenshire, J.L. The development and sea trials of a prototype portable ion mobility spectrometer for monitoring monoethanolamine on board submarines. Int. J. Soc. Ion Mobility Spectrom. 2001, 4, 7–12. 11. McGann, W.J.; Haigh, P.; Neves, J.L. Expanding the capability of IMS explosive trace detection. Int. J. Ion Mobility Spectrom. 2002, 5, 119–122. 12. Neves, J.L.; Haigh, P.E.; Wu, C.; McGann, W.J. ITMS-MS analysis of smokeless powder. Int. J. Ion Mobility Spectrom. 2003, 6, 1–3. 13. Marr, A.J.; Groves, D.M. Ion mobility spectrometry of peroxide explosives TATP and HMTD. Int. J. Ion Mobility Spectrom. 2003, 6, 59–62. 14. Steiner,W.E.; Clowers, B.H.; Haigh, P.E.; Hill, H.H. Secondary ionization of chemical warfare agent simulants: Atmospheric pressure ion mobility time-of-flight mass spectrometry. Anal. Chem. 2003, 75, 6068–6076. 15. Johnson, P.V.; Kim, H.I.; Beegle, L.W.; Kanik, I. Electrospray ionization ion mobility spectrometry of amino acids: Ion mobilities and a mass-mobility correlation. J. Phys. Chem. A 2004, 108, 5785–5792. 16. Kanua, A.B.; Haigh, P.E.; Hill, H.H. Surface detection of chemical warfare agent simulants and degradation products. Anal. Chim. Acta 2005, 553, 148–159. 17. Dong, C.; Wang, W.; Li, H. Atmospheric pressure air direct current glow discharge ionization source for ion mobility spectrometry. Anal. Chem. 2008, 80, 3925–3930. 18. Zimmermann, S.; Barth, S.; Baether, W.K.M.; Ringer, J. Miniaturized low-cost ion mobility spectrometer for fast detection of chemical warfare agents. Anal. Chem. 2008, 80, 6671–6676. 19. Carr, T.W. Ed. Plasma Chromatography. Plenum Press, New York, 1984. 20. Eiceman, G.A.; Karpas, Z. Ion Mobility Spectrometry. 2nd Ed. CRC Press, Boca Raton, FL, 2005. 21. Collins, D.C.; Lee, M.L. Developments in ion mobility spectrometry-mass spectrometry. Anal. Bioanal. Chem. 2002, 372, 66–73. 22. Kanu, A.B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill, H.H. Ion mobility-mass spectrometry. J. Mass Spectrom. 2008, 43, 1–22.
The Study of Ion/Molecule Reactions at Ambient Pressure
413
23. Eiceman, G.A.; Nazarov, E.G.; Rodriguez, J.E.; Stone, J.A. Analysis of a drift tube at ambient pressure: Models and precise measurements in ion mobility spectrometry. Rev. Sci. Instrum. 2001, 72, 3610–3621. 24. Mason, E.A.; McDaniel, E.W. Transport Properties of Ions in Gases. Wiley-Interscience, New York, 1988. 25. Mason, E.A. Ion mobility: Its role in plasma chemistry. In: Plasma Chromatography, ed. T.W. Carr, 43–93, Plenum Press, New York, 1984. 26. Spangler, G.E. Space charge effects in ion mobility spectrometry. Anal. Chem. 1992, 64, 1312. 27. Vandiver, V.J.; Leasure, C.S.; Eiceman, G.A. Proton affinity equilibria for polycyclic aromatic hydrocarbons at atmospheric pressure in ion mobility spectrometry. Int. J. Mass Spectrom. Ion Processes 1985, 66(2), 223–238. 28. NIST. Mass Spectra. www.webbook.nist.gov/chemistry accessed June 21, 2009. 29. Kim, S.H.; Betty, K.R.; Karasek, F.W. Mobility behaviour and composition of hydrated positive reactant ions in plasma chromatography with nitrogen carrier gas. Anal. Chem. 1978, 50, 2006–2012. 30. Kebarle, P.; Searles, S.K.; Zolla, A.; Scarborough, J.; Arshadi, M. The solvation of the hydrogen ion by water molecules in the gas phase. Heats and entropies of solvation of individual reactions: H + (H2O)n-1 + H2O → H + (H2O)n. J. Am. Chem. Soc. 1967, 89, 6393–6399. 31. Gheno, F.; Fitaire, M. Association of molecular nitrogen with ammonium ion and oxonium and mono-, di-, and trihydrates. J. Chem. Phys. 1987, 87, 953–958. 32. Hiraoka, K.; Yamabe, S. How are nitrogen molecules bound to nitryl ion and nitrosyl ion (NO2 + and NO + )? J. Chem. Phys. 1989, 90, 3268–3273. 33. Spangler, G.E. Characterization of the ion-sampling pinhole interface for an ion mobility spectrometer/mass spectrometer system. Int. J. Mass Spectrom. 2001, 208, 169–191. 34. Preston, J.M.; Rajadhyax, L. Effect of ion/molecule reactions on ion mobilities. Anal. Chem. 1988, 60, 31–34. 35. Giles, K.; Grimsrud, E.P. The kinetic ion mobility mass spectrometer: Measurements of ion/molecule reaction rate constants at atmospheric pressure. J. Phys. Chem. 1992, 96, 6680–6687. 36. Su, C.; Babcock, K. The effect of sampling materials on the formation of different clusters during the ion mobility spectrometry detection of secondary high explosives. Int. J. Ion Mobility Spectrom. 2002, 5, 55–58. 37. Matz, L.M.; Tornatore, P.S.; Hill, H.H. Evaluation of suspected interferents for TNT detection by ion mobility spectrometry. Talanta 2001, 54, 171–179. 38. Lawrence, A.H.; Neudorfl, P.; Stone, J.A. The formation of chloride adducts in the detection of dinitrocompounds by ion mobility spectrometry. Int. J. Mass Spectrom. 2001, 209, 185–195. 39. Caldwell, G.; Magnera, T.F.; Kebarle, P. SN2 reactions in the gas phase. Temperature dependence of the rate constants and energies of the transition states. Comparison with solution. J. Am. Chem. Soc. 1984, 106, 959–966. 40. Olmstead, W.N.; Brauman, J.I. Gas-phase nucleophilic displacement reactions. J. Am. Chem. Soc. 1977, 99, 4219–4228. 41. Knighton, W.B.; Bognar, J.A.; O’Connor, P.M.; Grimsrud, E.P. Gas-phase SN2 reactions of chloride ion with alkyl bromides at atmospheric pressure. Temperature dependence of the rate constants and energies of the transition states. J. Am. Chem. Soc. 1993, 115, 12079–12084. 42. Schmatz, S.; Botschwina, P.; Stoll, H. Coupled cluster calculations for the SN2 reaction Cl- + CH3Br → ClCH3 + Br-. Int. J. Mass Spectrom. 2000, 201, 277–282.
414
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
43. Sahlstrom, K.E.; Knighton, W.B.; Grimsrud, E.P. Rate constants for the gas phase reaction of chloride ion with methyl bromide over the pressure range, 300 to 1100 Torr. J. Phys. Chem. A 1997, 101, 5543–5546. 44. Ewing, R.G.; Eiceman, G.A.; Harden, C.S.; Stone, J.A. The kinetics of the decompositions of the proton bound dimers of 1,4-dimethylpyridine and dimethyl methylphosphonate from atmospheric pressure ion mobility spectra. Int. J. Mass Spectrom. 2006, 255, 76–85. 45. Levine, I.N. Physical Chemistry. 3rd Ed. McGraw-Hill, New York, 1988. 46. Meot-Ner, M. The ionic hydrogen bond. I NH + ...O, NH + ...N, and OH + ...O bonds. Correlations with proton affinity. Deviations due to structural effects. J. Am. Chem. Soc. 1984, 106, 257–264. 47. Spangler, G.E.; Lawless, P.A. Measurement of electron capture rates for chlorobenzene with negative ion plasma chromatography. Anal. Chem. 1978, 50, 290–294. 48. Christophorou, L.G.; Compton, R.N.; Hurst, S.G.; ReinHardt, P.W. Dissociative electron capture by benzene derivatives. J. Chem. Phys. 1966, 45, 536–547. 49. Mayhew, C.A.; Critchley, A.D.J.; Howse, D.C.; Mikhailov, V.; Parkes, M.A. Measurements of thermal electron attachment rate coefficients to molecules using an electron swarm technique. Eur. Phys. J. D 2005, 35, 307–312. 50. Mayhew, C.A.; Critchley, A.; Jarvis, G.K. Electron attachment to and anion reactions with SF5Cl: Electron-swarm and selected ion flow tube studies. Int. J. Mass Spectrom. 2004, 233, 259–265. 51. Kennedy, R.A.; Mayhew, C.A. A study of low energy electron attachment to trifluoromethyl sulphur pentafluoride, SF5CF3: Atmospheric implications. Int. J. Mass Spectrom. 2001, 206, i–iv. 52. Jarvis, G.K.; Kennedy, R.A.; Mayhew, C.A. Investigations of low energy electron attachment to ground state group 6B hexafluorides (SF6, SeF6, and TeF6) using an electronswarm mass spectrometric technique. Int. J. Mass Spectrom. 2001, 205, 253–270. 53. Williamson, D.H.; Mayhew, C.A.; Knighton, W.B.; Grimsrud, E.P. Effect of pressure and temperature on the competition between nondissociative and dissociative electron attachment to POCl3. J. Chem. Phys. 2000, 113, 11035–11043. 54. Jarvis, G.K.; Mayhew, C.A.; Singleton, L.; Spyrou, S.M. An investigation of electron attachment to CHCl2F, CHClF2 and CHF3 using an electron-swarm mass spectrometric technique. Int. J. Mass Spectrom. Ion Processes 1997, 164, 207–223. 55. Jarvis, G.K.; Peverall, R.; Mayhew, C.A. A novel use of an ion-mobility mass spectrometer for the investigation of electron attachment to molecules. J. Phys. B At. Mol. Phys. 1996, 29, 1713–1718. 56. Liu, Y.; Mayhew, C.A.; Peverall, R. A new experimental approach to investigate the kinetics of low energy electron attachment reactions. Int. J. Mass Spectrom. Ion Processes 1996, 152, 225–242. 57. Hunter, S.R.; Carter, J.G.; Christophorou, L.G. Low energy electron attachment to sulfur hexafluoride in nitrogen, argon, and xenon buffer gases. J. Chem. Phys. 1989, 90, 4879–4891. 58. Tabrizchi, M.; Abedi, A. A novel use of negative ion mobility spectrometry for measuring electron attachment rates. J. Phys. Chem. A 2004, 108, 6319–6324. 59. Christodoulides, A.A.; Christophorou, L.G. Electron attachment to brominated aliphatic hydrocarbons of the form n-CNH2N + 1Br (N = 1-6, 8, 10). I. An electron swarm study. J. Chem. Phys. 1971, 54, 4691–4705. 60. Sahlstrom, K.E.; Knighton, W.B.; Grimsrud, E.P. Thermal electron detachment of the molecular anion of azulene at elevated pressures by ion mobility spectrometry. Int. J. Mass Spectrom. 1998, 179/180, 117–127.
The Study of Ion/Molecule Reactions at Ambient Pressure
415
61. Grimsrud, E.P.; Chowdhury, S.; Kebarle, P. Thermal energy electron detachment rate constants. The electron detachment from azulene and the electron affinity of azulene. J. Chem. Phys. 1985, 83(8), 3983–3989. 62. Pringle, S.D.; Giles, K.; Wildgoose, J.L.; Slade, S.E.; Thalassinos, K.; Bateman, R.H.; Bowers, M.T.; Scrivens, J.H. An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/travelling wave IMS/oa-ToF instrument. Int. J. Mass Spectrom. 2007, 261, 1–12.
Role of Trapped 14 The Ion Mass Spectrometry for Imaging Timothy J. Garrett and Richard A. Yost Contents 14.1 Introduction to Imaging Mass Spectrometry (IMS).................................... 417 14.1.1 Overview........................................................................................ 417 14.1.2 Imaging Mass Spectrometry (IMS) with Quadrupole Ion Traps.................................................................... 418 14.1.3 3D Quadrupole Ion Trap and 2D Quadrupole Ion Trap................ 419 14.1.4 Sample Preparation and Image Creation....................................... 421 14.2 Aspects of Imaging Mass Spectrometry (IMS) particular to Ion traps....... 422 14.2.1 Space-charge Control for Tissue Analysis.................................... 422 14.2.2 Tandem Mass Spectrometry.......................................................... 425 14.2.2.1 Structural Characterization of Compounds.................. 425 14.2.2.2 Isobaric Ion Identification............................................. 427 14.2.2.3 Full-Scan MS Imaging and Tandem MS Imaging....... 429 14.2.2.4 Multiple-Stage Tandem Mass Spectrometry (MSn)...... 430 14.2.3 Intermediate-Pressure Matrix-Assisted Laser Desorption Ionization (MALDI)...................................................................... 430 14.3 Quantification.............................................................................................. 432 14.4 Conclusions.................................................................................................. 436 References............................................................................................................... 436
14.1 INTRODUCTION TO IMAGING MASS SPECTROMETRY (IMS) 14.1.1 Overview We describe here the important role that quadrupole ion traps (both the classic 3D ion trap and the linear ion trap) have played, and will continue to play, in imaging mass spectrometry (IMS) (which is not to be confused with Ion Mobility Spectrometry (IMS)). The pulsed mode of operation of an ion trap is an ideal match to a pulsed laser desorption event. Even more important, quadrupole ion traps provide Mass spectrometry/Mass spectrometry MS/MS (and MSn), which is critical for the mapping of individual compounds in complex samples such as tissue, and for 417
418
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
the identification of these compounds. We describe the principles and instrumental approaches, and then provide real-world examples of how the capabilities of ion traps are being used in modern IMS. IMS offers the ability to analyze directly thin tissue sections and to create chemically-selective images of intrinsic chemical distributions. This technique not only offers the ability to characterize known compounds from a variety of tissues, but also the capability of identifying unknown chemical signatures for a variety of studies such as disease progress or pharmaceutical studies. While IMS does not provide the spatial resolution of some other imaging techniques (spot sizes are typically 50–150 µm), the chemical specificity provided by mass spectrometry cannot be paralleled. IMS using matrix-assisted laser desorption/ionization (MALDI) provides the ability to characterize and localize compounds within tissue sections, identifying potential but unknown markers of diseases such as cancer, or to determine where an administered drug (and its metabolites) has localized in a tissue section. An interesting aspect that mass spectrometry provides to the field of imaging is the identification of non-targeted compounds. In a typical fluorescence experiment, a fluorophore must be administered that binds to the desired compound (and ideally only to that compound), thus permitting the identification of that specific compound. With IMS, the targeted compound can be localized and other compounds that may localize with the targeted compound can be identified, thus providing a more complete understanding of the chemical signature of the specific state under investigation.
14.1.2 Imaging Mass Spectrometry (IMS) with Quadrupole Ion Traps IMS studies have been performed on most widely-used mass analyzers. The primary requirement for performing imaging on a particular mass spectrometer is that the instrument incorporates an ionization source capable of localized desorption ionization such as MALDI [1], secondary ion mass spectrometry (SIMS) [2], or even more recent techniques such as desorption electrospray ionization (DESI) [3]. The only additional qualities needed to produce images with such a source are: (1) the ability to step the sample or the desorption source (laser or ion beam, typically) in a defined and reproducible distance with respect to each other; and (2) software capable of extracting individual ion signals with respect to position and thus to generate images. Although there are many instruments capable of IMS, the purpose of this chapter is to discuss the application of quadrupole ion trap mass spectrometry to tissue samples using MALDI. The quadrupole ion trap, while not a widely-used mass analyzer for MALDI studies primarily because of the interest in protein identification, offers an ideal match for MALDI due to the pulsed nature of the technique and the pulsed nature of ion trapping. The quadrupole ion trap is capable of trapping and storing ions produced from a given laser pulse, and then scanning out the trapped ions rather than scanning continuously, such as with a quadrupole mass filter. Therefore, relatively few ions desorbed from the tissue specimen are lost as the timing of the laser pulse (or pulses) is synchronized with the opening of the ion trap. This arrangement provides for multiple experiments to be performed on the same tissue specimen, as has been shown by our group [4]. However, in some cases,
The Role of Trapped Ion Mass Spectrometry for Imaging
419
it may be desired to discard the first few laser pulses due to the excess matrix signal that is observed. Current quadrupole ion traps, including the 2D ion trap, are limited to a mass range of 4000 Th [4], although, on occasion ions have been analyzed successfully up to m/z 100,000 [5]. Thus, ion traps do not compete with time-of-flight (TOF) mass analyzers for imaging proteins [6]. Therefore, the primary interest in the use of ion trap technology in imaging studies is the application of tandem mass spectrometry (MS/MS) and MSn for small molecule analysis and identification. Thus the analysis of drugs [7], peptides [8], lipids [4,9–11], and even metabolites [7] have been performed with quadrupole ion traps using tandem mass spectrometry to identify the compound and to generate specific images using product ions. Typically, MALDI has been performed under low pressure conditions (ca 10−6 Torr). In recent years, intermediate pressure (IP, 10−2 to 1 Torr) [10,12–14] and atmospheric pressure (AP) MALDI [15,16] systems have been developed to analyze more labile molecules and for tissue analysis [10]. Atmospheric pressure matrix-assisted laser desorption/ionization (AP-MALDI) offers the ability to couple a MALDI source to most commercial instruments, including quadrupole ion traps, because the source uses existing ion optics and vacuum transfer lines of commercial instruments to introduce the ions generated by MALDI. A brief discussion of IP-MALDI with ion trap technology is presented, but the primary focus is on the instrumentation aspect of ion trap technology, in particular linear ion trap technology, for direct tissue analysis and imaging MS studies.
14.1.3 3D Quadrupole Ion Trap and 2D Quadrupole Ion Trap The first ion trap capable of tissue analysis using MALDI was developed by our group [17,18] in 1999, based on a 3D quadrupole ion trap. The laser beam (N2, 337 nm or CO2, 10.6 µm) was directed to the tissue surface so as to impinge perpendicularly to the sample. Ions leaving the source were redirected through 90º toward the quadrupole ion trap with a DC turning quadrupole. The instrument was capable of both MALDI and laser desorption/chemical ionization (LD/CI). A diagram of this instrument is shown in Figure 14.1. This instrument was designed to probe specific areas of the tissue rather than to create images. The first studies utilizing this instrument were for the analysis of small pharmaceutical compounds in tissue sections. In one study, the analysis, by both MALDI and LD/CI, of paclitaxel in an ovarian tissue sample containing a tumor detected an estimated level of 290 pg of paclitaxel in the tumor region only of the tissue. Critical to this level of detection was the utilization of tandem mass spectrometry, as the parent, or precursor, ion was not detected from the tissue in a mass scan experiment [18]. With the development of the 2D quadrupole ion trap (linear ion trap or LIT) [19,20], the application to tissue specimens was explored further due to the increased storage capacity, increased injection efficiencies, and increased trapping efficiencies; the use of higher MALDI pressures for tissue samples was explored also. The instrument employed for most of the discussion following was described in detail in a recent publication [4]. A diagram of this instrument is shown in Figure 14.2. Briefly, the instrument is a linear ion trap fitted with a MALDI ion source that operates
420
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
UV or IR laser
1 × 10–6 torr
Turning Quad
Ion trap Laser beam in
Tissue sample
Ion source
Ions out
FIGURE 14.1 Diagram of the first 3D quadrupole ion trap instrument designed for the analysis of tissue by MALDI, LD, or LD/CI.
Laser beam
AUX rods
MALDI modification q00
q0
Lens 0
Octopole
Lens 1
3 section linear ion trap
Skimmer
Sample plate
FIGURE 14.2 Diagram of the LTQ with vMALDI source. (Reproduced from Garrett, T.J.; Prieto-Conaway, M.C.; Kovtoun, V.; Bui, H.; Izgarian, N.; Stafford, G.C.; Yost, R.A., Int. J. Mass Spectrom. 2007, 260, 166–176. With permission from Elsevier.)
at intermediate-pressure (Finnigan LTQ with vMALDI from Thermo Electron Corporation, San Jose, CA), where LTQ is a trademark term for the description of a linear trap quadrupole, LIT. The MALDI source (vMALDI), where v denotes intermediate vacuum as opposed to AP-MALDI, consists of an N2 laser (337 nm) directed to the source by fiber optics at an incident angle of 32º and a modified set of quadrupole rods (q00) added to the front of the LTQ to accommodate the laser entrance and the camera for sample viewing. The spot size of the laser is typically 120 µm. The source operates at a pressure of 0.17 Torr (22.6 Pa) with N2 gas.
421
The Role of Trapped Ion Mass Spectrometry for Imaging
14.1.4 Sample Preparation and Image Creation A basic IMS experiment follows the workflow shown in Figure 14.3. A tissue sample, from freshly-frozen tissue, is sectioned to a desired thickness, typically less than or equal to 10 µm. That tissue section is dried sufficiently and then coated with a matrix if MALDI is to be performed. The appropriate MALDI matrix is dependent on the types of analytes desired for analysis and will vary accordingly. In addition, additives to the matrix solution can be chosen to produce the desired ion, such as the addition of cationic salts for production primarily of cation adducts. Many coating methods have been employed, including airbrushing [4], inkjet [21], acoustic wave [22], electrospraying [1], and sublimation [23] to deposit the matrix onto the sample. The most important issue for matrix coating is the deposition of a uniform layer of matrix across the entire section at a sufficient concentration for ionization. Once the matrix is applied, the sample is dried again and then placed into the instrument for data acquisition. Typically, the tissue is rastered at step sizes equal to the laser spot size, but step sizes smaller than the laser spot size (oversampling) or even larger than the laser spot
Section the tissue
100 90 80 70 60 50 40 30 20 10 0
782.8
Extract to Generate image
756.9 753.8 723.9 751.9 720
740
798.7 772.7 769.8 760
780 m/z
810.7
Position specific mass spectrum
828.6 826.7 832.7 848.7 800
820
840
Send to MS
Place on microscope slide
Coat with matrix
Raster laser across tissue in MS
FIGURE 14.3 Imaging mass spectrometry (IMS) work flow. The tissue is first sectioned to 10 µm, placed on a microscope slide, dried, and then coated with a MALDI matrix. The sample is then analyzed by mass spectrometry where position-specific mass spectra are recorded. Post-processing of the data generates ion images corresponding to individual ion signals. When MS/MS is employed, ion images can be generated from specific product ion signals.
422
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
size (undersampling) can be employed as well [24]. The raster-step size is based on the desired information to be obtained from an individual sample. Oversampling can be employed to improve the spatial resolution, but must involve extinguishing analyte over the entire spot before stepping the laser to the next location [25]. Undersampling can be a more rapid way to collect data from the tissue section when high spatial resolution is not necessary. After data collection, images are generated typically by post-processing the data using a program that extracts the intensity for individual ions from the data file. These images can be normalized to the total ion intensity at a given spot or can be used without further processing. Future research in IMS will help to solidify the best approach for image generation and post-processing, but probably it will be dependent on the tissue type and research study (for example, comparison of disease state to control). An added benefit of using the ion trap for IMS studies is that non-conductive plain glass slides can be employed, since this mass analyzer does not rely on the initial kinetic energy of the ions. In an initial study to compare the use of indium–tin oxide coated glass (conductive) to plain glass slides, identical mass spectral correlation and image production for phospholipids were observed [4]. This result may offer the ability to analyze archived samples that probably would have been prepared on plain glass microscope slides and to reduce the costs of sample preparation.
14.2 Aspects of Imaging mass spectrometry (IMS) particular to Ion traps 14.2.1 Space-charge Control for Tissue Analysis With the use of ion traps for tissue analysis, the deleterious effects of space charge can be a significant issue. Space-charging occurs when the ion trap is overfilled with ions and results in peak broadening and mass shifting, affecting the accuracy of the m/z-values reported [26,27]. In the direct analysis of tissue samples, the number of possible ions generated from a particular area can be overwhelming. One particular spot will contain lipids, metabolites, peptides, and proteins along with the matrix employed; therefore, proper control of the number of ions introduced in to the ion trap is critical. Automatic gain control (AGC), used typically for controlling space charge in liquid chromatography/mass spectrometry (LC/MS) and LC/MSn on ion traps, employs a pre-scan that evaluates the ion current and then fills the trap to keep the number of ions approximately constant [28]. This feature uses ion injection time as a variable for filling the trap, employing longer injection times for low ion currents and shorter injection times for high ion currents. For MALDI, the number of ions generated is primarily based on two parameters: (1) the laser power; and (2) the number of laser shots. Generally, the laser power is adjusted manually to provide sufficient ionization with the lowest power possible to reduce photo-induced fragmentation. On the system described here, AGC determines the appropriate number of laser shots to fill the trap based on a pre-scan using a manually-selected laser power [4]. Because each area on the tissue that the laser interrogates can be considered a unique microenvironment, the number and variety of compounds present in a given spot can vary and, therefore, the number of ions or,
The Role of Trapped Ion Mass Spectrometry for Imaging
423
rather, the population of ions generated can vary. Thus, for tissue analysis employing AGC, the number of laser shots required for each individual spot is likely to vary. The primary issue with this type of control concerns the importance of the change in ion signal from one microenvironment to the next across the tissue. By varying the number of laser shots from spot to spot, one may lose valuable information about the specific tissue sample and uniformity of the matrix coating. Thus, in some cases, the use of AGC should be avoided. However, without using AGC, space-charging must be controlled by setting manually the appropriate laser power and number of laser shots. This operation can be performed by interrogating manually multiple spots from the tissue surface to set the power. Typically, the number of laser shots is set to a chosen number and the power is adjusted to get a strong signal (usually around a total ion count (TIC) of 1 × 105 counts). Usually, the laser is set to fire 10 laser shots or less when using a low repetition rate laser because the number of laser shots, the step size chosen, and the size of the sample are the three primary factors which determine the length of a given experiment. For a rat brain (typically 10 × 14 mm), the duration of an experiment with a step size of 100 µm could range from ca 2 h to more than 4 h depending of the number of laser shots at each step. Another option for controlling space charge is to collect mass spectra from a small area (ca 10 spots) on the tissue with AGC turned on, and then review the data to determine the average number of laser shots and set the number of laser shots at that value [9]. These two controls of space charge maintain the number of laser shots constant over the entire tissue section; however, there is still the possibility that some areas will yield a highlyconcentrated signal that would cause space-charging. Those areas may be of interest for future studies, and a researcher could thus return to that area for further investigation. In either case (AGC on or AGC off), the tissue sample must be interrogated first to determine the appropriate laser power that should be employed, unless multiple samples are to be employed and have been coated with matrix in a similar manner. Figure 14.4 shows an example where two mass spectral images for m/z 810.6, phosphatidylcholine (PC) with 18:0 at the sn-1 position and 18:1 at sn-2, were detected as a sodium ion adduct. The shorthand description of this ion is [PC (18:0,18:1) + Na] + , where the two (or three) capital letters refer to the specific phospholipid and the set of numbers in parentheses refers to the fatty acyl chains at the sn-1 and sn-2 positions of the glycerol backbone, respectively: the length of each fatty acyl chain is defined by the number before the colon and the degree of unsaturation is defined by the number following the colon. Signals due to the sodium ion adduct were acquired initially from a tissue section (10 µm rat brain section) with AGC turned off (laser shots at 11), and then from the same tissue sample with AGC turned on. The observations were obtained using a Finnigan LTQ equipped with a vMALDI ion source. The data were extracted for a mass range of 810.0–811.0 Th, and were normalized to the total ion current. Both images show a similar distribution in the tissue, but the intensity appears to be higher and more uniform in the areas of expression, indicated in Figure 14.4, with the AGC-on image, while the AGC-off image shows a relatively higher intensity in the forceps major of the corpus callosum. There are many factors that may be involved also in this analysis such as the order in which the files were obtained; therefore, a more complete study should be conducted to evaluate the use of AGC with tissue studies.
424
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Forceps major of the corpus callosum
Dentate gyrus
Cortex
10000 8000 6000 4000 2000
AGC off
0
Periaueductual gray Substantia nigra
AGC on
Medial geniculate body Cerebral peduncle
FIGURE 14.4 Mass spectral images for m/z 810.6, [PC(18:0,18:1) + Na] + , acquired with AGC-off (left) and with AGC-on (right) on the same 10 µm rat brain tissue section. The images show a similar distribution across the tissue, but the AGC-off image appears to indicate a higher intensity in the forceps major of the corpus collosum. The AGC-off file was collected prior to the AGC-on file; differences in the images could be associated with the order of analysis. Both images are normalized to the total ion current. The number of laser shots for the AGC-off file was maintained at 11.
The use of the 2D linear ion trap instead of the traditional 3D quadrupole ion trap offers a substantial increase in the number of ions that can be trapped before space-charge effects become evident [19]. An abundant supply of ions is critical in the analysis of tissue specimens because of the large number of compounds and the accompanying wide range of concentrations that can be found at each individual spot. For tandem mass spectrometry experiments, the probability of space-charging is reduced significantly and, thus, the primary concerns here should be placed on storing sufficient ions to generate product ions signals of high intensity, while paying close attention to the duration of the experiment. A particular characteristic inherent in the direct analysis of tissue samples is the wide concentration range of compounds present in tissue samples. For a brain, 50% by weight consists of lipids, with proteins, peptides, metabolites, and DNA makes up the other 50%. In addition, trace components in some tissues can be present at concentrations in the low ng mL −1 to pg mL −1 level. Even in the class of lipids, concentrations vary, with PC representing 10% of lipids and phosphatidylinositols (PI) representing less than 1% [29]. In those classes of lipids, the variation in fatty acid tails extends further the range of concentrations of compounds detected. In addition, using MALDI, matrix ions are present at very high concentration in relation to the sample, which adds further complexity to the sample because of the multiple matrix ions that are now present. Given this information, it is important to have a mass analyzer capable of handling such a wide-dynamic range. The expanded volume of the LIT affords the ability to detect compounds present at very high concentrations
The Role of Trapped Ion Mass Spectrometry for Imaging
425
and those present at low concentrations. In addition, the wide-dynamic range of the LIT is evident in tandem mass spectrometry experiments as will be described later in this text.
14.2.2 Tandem Mass Spectrometry 14.2.2.1 Structural Characterization of Compounds Ion traps offer the ability to perform multiple stages of tandem mass spectrometry (MSn) while also providing full-scan tandem mass spectrometric data, allowing for structural characterization of a desired ion signal using the multiple fragments produced from the ionized compound of interest. In the analysis of phospholipids, for example, product ion mass spectra are helpful in determining the type of ion present (either protonated or cationized) because the fragmentation pathway is unique; PC species lose 59 Da (choline) when they are cationized, but fragment to produce m/z 184 (phosphocholine) when they are protonated [10,30,31]. In addition, the dynamic range of the linear ion trap provides for greater confidence in the assignment of product ions formed in low abundance; thus identification of the fatty acid tails and their location on the glycerol backbone can be made with confidence. The analysis of phospholipids from tissue sections, primarily nerve tissues, using ion traps has been of considerable interest because of the capabilities of tandem mass spectrometry for the identification of this class of compounds. Understanding phospholipid profiles can help understand brain development [32], aging [33], and disease studies [34–36]. The general mass range for most phospholipids is from m/z 650 to m/z ca 900. An example of the phospholipid mass region is shown in Figure 14.5, for ions collected from a rat brain tissue section coated with 2,5-dihydroxybenzoic acid (DHB) containing sodium acetate. The mass spectrum is composed of ion signals from a wide variety of substances. Using tandem mass spectrometry, researchers have been able to classify correctly most of the ionized compounds in this mass region as corresponding to PC and sphingomyelin (SPM) species. An example of phospholipid characterization is shown in Figure 14.6 for the identification of [PC (18:0,18:1) + Na] + . The data were collected on a 10 µm rat brain tissue section (Sprague-Dawley rat, a model system for the general study of human health and disease) coated with DHB, isolating m/z 810.6 with a 1.5 Th-wide window and 30% CID. The percentage of CID represents the fraction of the maximum voltage (AC) applied across the end-cap electrodes to induce dissociation. Typically, the percent CID is chosen so as to leave less than 5% of the precursor ion remaining. The entire tissue was analyzed under these same conditions, generating 10,528 individual mass spectra with a step size of 120 µm (same as the laser spot size), with 10 laser shots at each point and AGC turned off. The neutral loss (NL) of 59 Da (to yield m/z 751.4) corresponds to a loss of choline; the neutral losses of 183 Da (to form m/z 627.5, phosphocholine) and of 205 Da (to form m/z 605.5, phosphocholine plus sodium) confirm the identification of a PC species ionized as an [M + Na] + ion. To identify not only the compound, but also the type of ion, is critical when looking directly at tissue because the tissue medium may contain
426
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 782.73
100 90 80
Relative abundance
70 756.84
60 50 40
753.90
30
772.73
20
723.86 697.86
10 0
798.66 810.69
672.88
713.77 734.93
832.62 848.63
880.53
660 680 700 720 740 760 780 800 820 840 860 880 900 m/z
FIGURE 14.5 A mass spectrum from a brain tissue section showing the phospholipid signature in the positive ion mode.
cations and other species that can cause adduction. A significant aspect of this mass spectrum is the presence of ions at m/z 526.3 and 528.3 that correspond to losses of the fatty acid tails; m/z 526.3 ([M + Na−C18H36O2] + ) corresponds to the loss of stearic acid while m/z 528.3 ([M + Na−C18H34O2] + ) corresponds to the loss of oleic acid. These product ions are over 100 times less abundant than the most abundant product ion at m/z 751.4; however, they are critical in identifying the correct location of the fatty acid tail on the glycerol backbone for PC species. Previous studies using MS/MS have shown that the loss of the fatty acid at the sn-1 occurs preferentially to the loss of the fatty acid at the sn-2 position [30,37]. For this example, this means that the loss of oleic acid (corresponding to m/z 526.3) occurs more readily, thus producing a higher intensity product ion, indicating that oleic acid is located at the sn-1 position. The correct identification of this m/z 810.6 ion is thus, PC (18:0,18:1) as stated earlier. Using tandem mass spectrometric data, the phospholipids can be characterized almost completely and mapped in the tissue as has been shown previously [4,9,23,38]. The remarkable utility of MS/MS and MSn to identify and to characterize compounds structurally from intact tissue is an important advantage of ion traps for IMS.
427
The Role of Trapped Ion Mass Spectrometry for Imaging 751.4
90,000
NL 59 N(CH3)3
80,000 70,000 60,000 526.3 50,000 40,000 30,000 20,000
0
528.3
500 400
506.4
300 200 100
10,000
NL 284 C 18:0 NL 282 C 18:1
600
0
463.3 264.5 349.4
385.0
350
250 300
467.3
NL 205
441.3
m/z 400
450
550 500 m/z
NL 183
627.5 605.5 600
767.1
656.0 700
800
FIGURE 14.6 Average product ion (MS2) mass spectrum (10,528 mass spectra) of m/z 810.6 at 30% CID and 1.5 Th-wide isolation window. The mass region from m/z 250–550 is expanded (inset) to include the less abundant ions in the characterization. The neutral loss (NL) of 59 Da (to form m/z 751.4) corresponds to a loss of choline from a PC species; the neutral losses of 183 Da (phosphocholine) and 205 Da (sodium phosphocholine) confirm the identification of a PC in the form [M + H]+ and [M + Na] +. The ions at m/z 526.3 and 528.3 result from the losses of the fatty acid tail at the sn-1 and the sn-2 position, respectively.
14.2.2.2 Isobaric Ion Identification A unique feature of the ion trap is that full-scan product ion spectra over a specified mass range from a desired precursor ion are collected in the same amount of time as monitoring a single product ion transition, as in selected reaction monitoring on a triple quadrupole. Observation of a complete product ion mass spectrum is advantageous for direct tissue analysis in the identification of isobaric species as well as for the detection of targeted species. The primary advantage is that unique fragmentation pathways can be used to differentiate the isobars; when an MS/MS experiment (see below) is conducted across the entire tissue section, one can map each isobar from the specified MS/MS isolation event and can acquire full-scan product ion mass spectra. Creating an image from a product ion provides for a more specific image than creating it from a particular ion from full-scan mass spectrometric data. An example of isobaric identification is shown in Figure 14.7 for the characterization
0
1000
2000
3000
4000
5000
0
2000
4000
6000
8000
10,000
0
30,000
60,000
20,000
30,000
120,000 40,000 90,000
150,000
0
10,000
544.3‡
0
800
1600
2400
620
623.4*
300
400
500
700
600 m/z
Area expanded
660
700
740
800
900
810.0
785.1‡
769.3*
692.9◊ 741.3†
674.9◊
652.9◊ 645.4*
◊DHB Cluster ion
‡PE (18:0, 20:4) [M–H+Na+K]+ 828.5
800 600 400 524.3‡ 572.3‡ 200 0 440 600 520 Area expanded
828.5
†PS (16:0, 18:1) [M–2H+3Na]+
*PC (16:0 22:6) [M+Na]+ 828.6
FIGURE 14.7 Average product ion (MS2) mass spectrum of m/z 828.6 (10,000 mass spectra collected over the entire tissue). Specific MS2 images for three product ions (m/z 769.4, 652.9, and 544.3) are shown as insets. The use of MS2 and MS3 enabled the identification of four different isobaric ions appearing at m/z 828.6 (*PC, †PS, ‡PE, and, ◊DHB cluster ion). For m/z 828.6, an isolation window of 1.5 Th was used. The product ions relating to each isobar are shown using symbols. The images generated from the MS/MS product ions of the isobars show a different distribution for each isobar. The classification of each isobar is shown at the top of the figure. Two mass ranges are expanded and displayed as insets to show the intensity of these low-abundance product ions with respect to the overall mass spectrum. (Reproduced from Garrett, T.J.; et al. Int. J. Mass Spectrom. 2007, 260, 166–176. With permission from Elsevier.)
m/z 544.3‡
m/z 652.9◊
m/z 769.4*
Intensity
428 Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
429
The Role of Trapped Ion Mass Spectrometry for Imaging
of m/z 828 as consisting of four different isobaric ions − [PC (16:0,22:6) + Na]+, [phosphatidylethanolamine, PE, (18:0,20:4)−H + Na + K] +, [phosphatidylserine (PS) (16:0,18:1)−2H + 3Na] + , and a DHB cluster ion [4]. The four different ions were identified based on MS/MS fragmentation pathways for each given phospholipid and utilizing MS3 to elucidate the structure. In the small molecule mass range, the presence of an enormous variety of metabolic products in the tissue (in addition to MALDI matrix cluster ions at many m/z-values) increases considerably the possibility of isobaric interferences. A critical aspect to point out is that the signal for the major PS fragment in Figure 14.7 is over 100 times lower than that of the most abundant ion, showing the wide dynamic range of the linear ion trap. Using tandem mass spectrometry, the separation of these isobars in accomplished and fragmentation information is provided in the same experiment, thus allowing for confident identification. 14.2.2.3 Full-Scan MS Imaging and Tandem MS Imaging In addition to isobaric separation, utilizing tandem mass spectrometric data for image generation provides a much clearer and possibly more accurate perspective of the actual distribution of the compound in the tissue sample. Figure 14.8a and b show the ion images for m/z 837 using data from a mass scan (a) and tandem mass spectrometric data (b) from a rat brain tissue section coated with 2,5-DHB. The tandem mass spectrometric data were collected across the entire tissue section with a step size of 120 µm using 10 laser shots at each point and isolating m/z 837.6 with an isolation width of 1.5 Th. The most abundant product ion was m/z 778.5, arising from a NL of 59 Da. This ion, m/z 778.5, was identified as SPM 24:0. The intensity of this phospholipid ion is ca 5% of that of the most abundant phospholipid ion detected in the tissue section (m/z 782.6 PC (16:0,18:1), [M + Na] + ). The mass scan image in Figure 14.8a does appear to show more localization of m/z 837.6 in the white matter of the brain, but it is not well defined and is poorly resolved. On the other hand, the tandem mass spectrometric image of m/z 778.5 in Figure 14.8b (a) m/z 837.6 (SPM 24:0)
500
(b) 837.6 778.5 (NL 59)
100,000
400
80,000
300
60,000
200
40,000
100
20,000
0
0
FIGURE 14.8 Images generated from mass scan data (a) and from MS/MS data (b) for m/z 837 (SPM 24:0) in a rat brain tissue section. The intensity scales are shown at the right of each image. The mass scan data image was normalized to the total ion intensity at each point, while the MS/MS data image is the raw intensity for the transition of m/z 837m/z 778 (neutral loss, NL, of 59 Da).
430
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
shows a very distinct localization in the white matter of the brain and permits both the identification of substructures of the brain and the localization of this compound to more specific regions. For correct localization of less abundant ion species such as SPM, it is critical to utilized tandem mass spectrometric capabilities to provide a clear understanding of localization and relative concentration in the tissue section. In the analysis of phospholipids from brain tissue section, over 30 experiments were performed on a single tissue sample before re-application of matrix. Performing multiple experiments on a single tissue section provided the opportunity to analyze multiple MS/MS spectra collected from the entire tissue section; these data showed that at several m/z-values, three different phospholipid species were identified (PC, phosphtidylethanolamine (PE), and PS). Furthermore, at every single m/z-value selected for an MS/MS experiment, a DHB cluster ion was detected (primarily based on the loss of 136 Da). 14.2.2.4 Multiple-Stage Tandem Mass Spectrometry (MSn) The quadrupole in trap is the only mass analyzer capable of performing multiple stages of tandem mass spectrometry (MSn) in an efficient and reproducible manner. This ability with respect to IMS studies is critical for small molecule identification. As an example, the identification of an ion at m/z 616 from rat brain using MS5 is shown in Figures 14.9a through d and 14.10a through d. In a mass scan experiment, a strong ion signal at m/z 616 was observed in a specific location of the rat brain. The ion image, shown in Figure 14.9 (inset), indicated that the ion signal was localized to only one area, indicated by an arrowhead, of the brain. To identify this compound, all tandem MS/MS experiments were performed in this small area after the entire tissue was scanned. The product ion mass spectrum of m/z 616 showed a primary loss of 59 Da to form m/z 557 (Figure 14.9a). Given the previous discussion on phospholipids, one might identify tentatively this ion as a lysophosphatidycholine species (lyso indicates the loss of one fatty acid tail due to enzymatic activity); such ions are, typically, in the mass range m/z 500–600. However, with MS3, the ion shows a subsequent loss of 59 Da (Figure 14.9b), thus confirming that the ion is not a PC because they cannot suffer consecutive losses of 59 Da. After two additional stages of tandem MS (MS4 and MS5), the ion shows losses of 15 and 16 Da, respectively. In further discussion, heme was suggested as a possible match to this m/z-value. A standard of heme was obtained and a MALDI spectrum was generated using DHB. Comparison of the fragmentation pathway generated from the tissue specimen to one generated from the standard of heme by MALDI (Figure 14.10a through d), shows near identical losses, thus confirming that the ion at m/z 616 is heme. It is clear the MSn was necessary in this case to identify correctly this unexpected ion.
14.2.3 Intermediate-Pressure Matrix-Assisted Laser Desorption Ionization (MALDI) Increasing the pressure in the source region for MALDI experiments has been shown to decrease fragmentation of labile species occurring in this part of the instrument. The effect of increasing the pressure on reduced fragmentation for SPM 16:0 is shown in Figure 14.11 [4]. A standard of SPM 16:0 at a concentration of 10 ppm
431
The Role of Trapped Ion Mass Spectrometry for Imaging
8000 6000
(a) 100 50
4000
Tissue MS2 616→...
Relative abundance
0 200 (b) 100 50 0 200 (c) 100 50 0 200 (d) 100 50 0 200
250
2000 0
300
557 m/z 616
2000 4000 6000 8000 10,000 12,000 14,000
350
400
450
500 498
MS3 616→557→...
485 250
300
350
400
450
300
350
400
450
300
550
600
650
550
600
650
550
600
650
550
600
650
NL 59
499 500
350
500
467 NL 16 468 454 469
MS5 616→557→498→483→...
250
558
483 NL 15 482 497 468 498
MS4 616→557→498→...
250
544
498
NL 59
400
450
500
m/z
FIGURE 14.9 (a) through (d); a sequence of MS2 through MS5 mass spectra of m/z 616 from tissue from a specific region in a brain tissue section (shown in the inset, white and indicated with arrowhead) are most intense. Even after MS5, the compound still remained unidentified.
was prepared using the matrix 6-aza-2-thiothymine; the standard was analyzed under two different pressure conditions, ca 10−6 Torr (top) and 0.17 Torr (bottom) in two separate instruments. The lower-pressure mass spectrum was acquired on a custombuilt 3D ion trap [18] and the intermediate-pressure mass spectrum was acquired on a commercial 2D ion trap (Thermo LTQ with a vMALDI ion source). Four ions were detected relating to SPM 16:0: m/z 703.7 [M + H] + , m/z 725.7 [M + Na] + , m/z 184.3 (fragment from [M + H] + ), and m/z 666.8 (fragment from [M + Na] + ). The ion at m/z 184.3 is the phosphocholine head group and results from fragmentation of the [M + H]+ ion only and is likely to occur during ionization, while the ion at m/z 666.8 is a fragment from the [M + Na]+ ion only [4]. As can be seen from the mass spectra in Figure 14.11, the ion at m/z 184 is the dominant ion in each pressure region; however, when the pressure is increased, fragmentation of the [M + H]+ precursor ion is reduced, leading to a diminution of the signal intensity of m/z 184.3. Likewise, fragmentation of the sodium adduct ion species is reduced to less than 10%. Furthermore, operation of the MALDI source at higher pressure (along with tolerance of the ion trap for ions of varying kinetic energy) means that glass slides work fine for sample mounting because m/z-values are not altered using non-conductive surfaces [4].
432
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a) 100 50 0 200 (b) 100 Relative abundance
50 0 200 (c) 100 50 0 200 (d) 100 50 0 200
×20
Standard MS2 616→... 250
485 300
350
400
500 498 485
300
350
400
450
300
350
400
450
300
600
650
550
600
650
550
600
650
550
600
650
NL 59
512 500
557
350
500
468 NL 16 483 453
MS5 616→557→498→483→...
250
550
483 497 NL 15 498
MS4 616→557→498→...
250
NL 59
498 526 544
450
MS3 616→557→...
250
557
400
m/z
450
500
FIGURE 14.10 (a) through (d); a sequence of MS2 through MS5 spectra from a standard of heme (m/z 616), showing a good match in fragmentation pathways when compared to the MSn spectra acquired from the tissue section shown in Figure 14.9(a) through (d).
14.3 QUANTIFICATION Although MALDI is not considered a quantitative method, a recent study has used MALDI for tissue analysis showing the change in phospholipid signals in spinal cord and sciatic nerve samples [11]. Rather than using images, that study focused on evaluating the overall expression of phospholipids in both spinal cord and sciatic nerve samples from both rats treated with the investigational drug dichloroacetate (DCA) and control rats. In the study, 10 µm-sections were prepared in triplicate and analyzed on a linear ion trap with a MALDI source (LTQ with vMALDI). Rats treated with DCA showed a marked decrease in specific phospholipids (primarily PC and SPM). A critical note in the paper pointed out that IMS offers the ability to probe small sections of a tissue rather than extracting the lipids from the entire tissue section. In a separate experiment, the profiles of PC species in five control rat brain sections were compared to the expected mole percent; the values showed very good correlations [39]. The numbers could be improved possibly if the MS/MS data of the most abundant product ion only was considered because of the likely isobaric interferences. In addition, the expected values included the analysis of the whole brain, while the former study analyzed only 10 µm-sections from the brain, which may have caused some deviation as well. One of the major difficulties in using MALDI/MS for quantitation is irreproducible signal intensities, which can be caused by inhomogeneous crystal formation,
433
The Role of Trapped Ion Mass Spectrometry for Imaging 70,000 184.1 Fragment from [M+H]
60,000
+
50,000
Fragment from [M+H]+
Intensity
100
184.3
80 Relative abundance
427.0*
30,000
285.3
90
~10–6 Torr
40,000
20,000 143.6*
285.0* 306.6*
10,000
70
0 100
60 50
200
300
~10–1 Torr
40
666.6 725.6 [M+Na]+
400
0 100
307.3 200
300
400
600 m/z
600
700
800
900
1000
725.7 [M+Na]+
666.6 500
m/z
6-aza-2-thiothymine matrix
20 10
500
703.6 [M+H]+
Fragment from [M+Na]+
30 144.3
Fragment from + [M+Na]+ 703.6 [M+H]
700
886.5 741.7 835.6 800
900
1000
FIGURE 14.11 Matrix-assisted laser desorption ionization mass spectra acquired at a pressure of ca 10 –6 Torr (upper) and 0.17 Torr (lower). The matrix used was 6-aza-2-thiothymine for the analysis of SPM (16:0). The traditional dried droplet method was used. By increasing the pressure in the source region, source fragmentation can be reduced, although not eliminated, for the analysis of this phospholipid. Interestingly, the protonated species exhibits a greater degree of fragmentation than the sodium adduct ion species under the same conditions. (Reproduced from Garrett, T.J.; et al. Int. J. Mass Spectrom. 2007, 260, 166–176. With permission from Elsevier.)
inconsistent sample preparation, and laser shot-to-shot variability. Traditional MALDI experiments show that consideration of the ratio of the peak intensities of the analyte to those of a deuterated internal standard can improve signal reproducibility. As stated earlier, upon direct analysis of tissue sections, a variety of compounds are ionized, which can interfere with a targeted analysis of exogenous compounds such as drugs; therefore, it is critical to utilize MS/MS and MSn in order to differentiate exogenous ions from endogenous ones as well as matrix ions. One approach for combining the use of an internal standard with MS/MS would be to perform two alternating MS/MS experiments: first perform MS/MS of the analyte ion, then MS/ MS of the internal standard, and calculating the ratios of the intensities of the targeted product ion to the internal standard fragment ion. An alternative approach is to use a wide isolation window that includes both the targeted compound and internal standard ions, performing a single MS/MS experiment. In the wide-isolation window approach, a wide mass range is chosen in order to include both the precursor
434
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
ion m/z value and the internal standard m/z value in the same isolation event. The center of the isolation window is an m/z value between those of the precursor ion and the internal standard. Preliminary results have shown that using a single isolation method (the wide isolation window) compared to two alternating MS/MS experiments improves precision dramatically (10–20 times reduction in the percent relative standard deviation (% RSD)) [39]. An example of the application of the wide-isolation window in tandem mass spectrometry is the detection and quantitative imaging of cocaine in post-mortem human brain tissue [39]. In this instance, it was determined necessary to develop an MS3 wide-isolation window method because of interfering background ions. Figure 14.12 shows the mass spectrum of cocaine detected in brain tissue confirmed by matching six MS3 product ions with those from a cocaine standard. The mass spectral image for the MS3 product ion at m/z 150 and the area of tissue to which it corresponds are shown also in the figure. For the quantitation of cocaine analyzed from the tissue, the deuterated (2H3) internal standard was spiked beneath the tissue at known concentrations before matrix application to develop a calibration curve. The MS3 wide-isolation window method was then employed for the analysis of cocaine and cocaine-d3. Cocaine was analyzed successfully and quantified using this approach.
(m/z 150–151) / TIC 150.2
100
2.5 mm
Relative abundance
80 82.3
60
40
0.0
50.0
100
108.2 91.4
20
0
70.4 50
93.3 100
119.2 122.1
165.3 182.2 150
200
250
300
350
m/z
FIGURE 14.12 A product ion (MS3) mass spectrum acquired from tissue using the wideisolation method for the analysis of cocaine in post-mortem human brain tissue. The MS3 method first isolated m/z 305.8 in MS/MS and then m/z 183.5 for MS3 (isolation width was 6 Th, 30% CID). The % CID refers to the fraction of the maximum AC voltage applied across the end-cap electrodes, normalized to mass. Reich, R.F.; Cudzilo, K.; Yost, R.A. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, 2008.
435
The Role of Trapped Ion Mass Spectrometry for Imaging
In the analysis of drugs and metabolites, researchers used a combination of IMS with MALDI on an ion trap, laser capture microdissection (LCM), and high performance liquid chromatography (HPLC) combined with MS/MS to confirm the identity of crystals depositing in the spleen of rats in a toxicological drug study [7]. The deposits were detected initially as crystals under polarized light microscopy, which provided no specific information as to what was contained in the crystals. IMS was utilized to determine whether the crystals contained the pro-drug, the active drug, metabolites, or a combination, while also providing a means to correlate localization seen under polarized light with mass spectral data. Using full-scan mass spectral data, the presence of the active drug at m/z 448 was identified in the crystal-rich regions rather than the pro-drug or any of the three active metabolites, with the regions of highest analyte concentration corresponding to the crystal regions in the light microscopy image (Figure 14.13). The mass spectral image shown in Figure 14.13 is the sum of the precursor ion m/z 448, and two product ions generated in the source region, m/z 420 and 202. For further confidence in the detection, MS/MS data were collected across the entire tissue section, isolating for the active drug compound at m/z 448. In a crystal-rich region, the product ion signal was approximately an order
9.1 mm
(a)
4.6 mm
(b)
Ion signal intensity
FIGURE 14.13 A cryosectioned splenic tissue sample from a high-dosage animal is shown (a) with highlighted areas of high density of microcrystals as determined by cross-polarized light microscopy. An imaging mass spectrometric analysis from a mass scan of this sample and data mining using “marker ions” afforded the ion intensity image (b) of the active drug (BMS-X, MW 447, sum of m/z 448, 420, and 202). At least two areas of highest analyte concentrations, that is, IMS intensity, are co-localized with the microcrystals on the optical tissue image. Notably, the active drug was detected throughout the entire tissue with ca 1 order of magnitude less intensity, which might be expected in a sample from a high-dosage animal. (Reproduced from Drexler, D.M.; Garrett, T.J.; Cantone, J.L.; Diters, R.W.; Mitroka, J.G.; Prieto-Conaway, M.C.; Adams, S.P.; Yost, R.A.; Sanders, M., J. Pharm. Toxicol. Methods 2007, 55, 279–288. With permission from Elsevier.)
436
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
of magnitude higher than in a crystal-free region of the same tissue section. Further studies, isolating the crystal-rich regions using LCM and performing HPLC/MS/MS, demonstrated that the crystal-rich regions contained 177 pg/µL of the drug while the crystal-free regions contained 66 pg/µL, thus verifying the semi-quantitative results using MS/MS directly from the tissue.
14.4 CONCLUSIONS Quadrupole ion traps offer several advantages for IMS: they allow for the use of non-conductive sample surfaces for tissue analysis, increased pressure for MALDI analysis and, most importantly, they offer the ability to perform tandem mass spectrometry and, in particular, MSn. Without a pre-separation of mixture components, as would be done with conventional sample preparation and HPLC, it is critical to separate isobaric ions using characteristic fragmentation pathways and then to use those specific pathways to identify the isobaric ions. Utilizing tandem mass spectrometry is critical as smaller molecules, such as metabolites and drugs, are analyzed in tissue sections where MS3 is needed to separate the targeted ion signal from that due to the chemical background and to confirm the presence of the species. The combination of ion trapping with a source capable of imaging provides not only localization of the ion of interest in the tissue, but offers also the ability to identify other compounds that have a similar localization. Analyzing for targeted and untargeted species provides an opportunity to characterize the chemical changes that may result from the introduction of a foreign substance, such as a drug, into the tissue.
References
1. Caprioli, R.M.; Farmer, T.B.; Gile, J. Molecular imaging of biological samples: Localization of peptides and proteins using MALDI-TOF MS. Anal. Chem. 1997, 69, 4751–4760. 2. McDonnell, L.A.; Piersma, S.R.; Altelaar, A.F.M.; Mize, T.H.; Luxembourg, S.L.; Verhaert, P.; van Minnen, J.; Heeren, R.M.A. Subcellular imaging mass spectrometry of brain tissue. J. Mass Spectrom. 2005, 40, 160–168. 3. Cooks, R.G.; Ouyang, Z.; Takats, Z.; Wiseman, J.M. Ambient mass spectrometry. Science 2006, 311, 1566–1570. 4. Garrett, T.J.; Prieto-Conaway, M.C.; Kovtoun, V.; Bui, H.; Izgarian, N.; Stafford, G.C.; Yost, R.A. Imaging of small molecules in tissue sections with a new intermediatepressure MALDI linear ion trap mass spectrometer. Int. J. Mass Spectrom. 2007, 260, 166–176. 5. Kaiser, R.E.; Cooks, R.G.; Stafford, G.C.; Syka, J.E.P.; Hemberger, P.H. Operation of a quadrupole ion trap mass spectrometer to achieve high mass charge ratios. Int. J. Mass Spectrom. Ion Processes 1991, 106, 79–115. 6. Chaurand, P.; Schwartz, S.A.; Caprioli, R.M. Profiling and imaging proteins in tissue sections by IMS. Anal. Chem. 2004, 76, 86A–93A. 7. Drexler, D.M.; Garrett, T.J.; Cantone, J.L.; Diters, R.W.; Mitroka, J.G.; Prieto-Conaway, M.C.; Adams, S.P.; Yost, R.A.; Sanders, M. Utility of imaging mass spectrometry (IMS) by matrix-assisted laser desorption ionization (MALDI) on an ion trap mass spectrometer in the analysis of drugs and metabolites in biological tissues. J. Pharm. Toxicol. Methods 2007, 55, 279–288.
The Role of Trapped Ion Mass Spectrometry for Imaging
437
8. Verhaert, P.D.; Prieto-Conaway, M.C.; Pekar, T.M.; Miller, K. Neuropeptide imaging on an LTQ with vMALDI source: The complete “all-in-one” peptidome analysis. Int. J. Mass Spectrom. 2007, 260, 177–184. 9. Cha, S.; Yeung, E.S. Colloidal graphite-assisted laser desorption/ionization mass spectrometry and MSn of small molecules. 1. Imaging of cerebrosides directly from rat brain tissue. Anal. Chem. 2007, 79, 2373–2385. 10. Garrett, T.J.; Yost, R.A. Analysis of intact tissue by intermediate-pressure MALDI on a linear ion trap mass spectrometer. Anal. Chem. 2006, 78, 2465–2469. 11. Landgraf, R.R.; Garrett, T.J.; Calcutt, N.A.; Stacpoole, P.W.; Yost, R.A. MALDI-linear ion trap microprobe MS/MS studies of the effects of dichloroacetate on lipid content of nerve tissue. Anal. Chem. 2007, 76, 8170–8175. 12. O’Connor, P.B.; Costello, C.E. A high pressure matrix-assisted laser desorption/ionization Fourier transform mass spectrometry ion source for thermal stabilization of labile molecules. Rapid Commun. Mass Spectrom. 2001, 15, 1862–1868. 13. Baldwin, M.A.; Medzihradszky, K.F.; Lock, C.M.; Fisher, B.; Settineri, T.A.; Burlingame, A.L. Matrix-assisted laser desorption/ionization coupled with quadrupole/orthogonal acceleration time-of-flight mass spectrometry for protein discovery, identification, and structural analysis. Anal. Chem. 2001, 73, 1707–1720. 14. Loboda, A.V.; Krutchinsky, A.N.; Bromirski, M.; Ens, W.; Standing, K.G. A tandem quadrupole/time-of-flight mass spectrometer with a matrix-assisted laser desorption/ionization source: Design and performance. Rapid Commun. Mass Spectrom. 2000, 14, 1047–1057. 15. Moyer, S.C.; Cotter, R.J. Atmospheric Pressure MALDI. Anal. Chem. 2002, 74, 468A–476A. 16. Doroshenko, V.M.; Laiko, V.V.; Taranenko, N.I.; Berkout, V.D.; Lee, H.S. Recent developments in atmospheric pressure MALDI mass spectrometry. Int. J. Mass Spectrom. 2002, 221, 39–58. 17. Troendle, F.J. A quadrupole ion trap laser microprobe for the mapping of pharmaceutical compounds in intact tissue. Ph.D. Dissertation, University of Florida, 2000. 18. Troendle, F.J.; Reddick, C.D.; Yost, R.A. Detection of pharmaceutical compounds in tissue by matrix-assisted laser desorption/ ionization and laser desorption/chemical ionization tandem mass spectrometry with a quadrupole ion trap. J. Am. Soc. Mass Spectrom. 1999, 10, 1315–1321. 19. Schwartz, J.C.; Senko, M.W.; Syka, J.E.P. A two-dimensional quadrupole ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 2002, 13, 659–669. 20. Hager, J.W. A new linear ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 2002, 16, 512–526. 21. Baluya, D.L.; Garrett, T.J.; Yost, R.A. Automated MALDI matrix deposition method with inkjet printing for imaging mass spectrometry. Anal. Chem. 2007, 79, 6862–6867. 22. Aerni, H.R.; Cornett, D.S.; Caprioli, R.M. Automated acoustic matrix deposition for MALDI sample preparation. Anal. Chem. 2006, 78, 827–834. 23. Hankin, J.A.; Barkley, R.M.; Murphy, R.C. Sublimation as a method of matrix application for mass spectrometric imaging. J. Am. Soc. Mass Spectrom. 2007, 18, 1646–1652. 24. Todd, P.J.; Schaaf, T.G.; Chaurand, P.; Caprioli, R.M. Organic ion imaging of biological tissue with secondary ion mass spectrometry and matrix-assisted laser desorption/ ionization. J. Mass Spectrom. 2001, 36, 355–369. 25. Jurchen, J.C.; Rubakhin, S.S.; Sweedler, J.V. MALDI-MS imaging of features smaller than the size of the laser beam. J. Am. Soc. Mass Spectrom. 2005, 16, 1654–1659. 26. Eades, D.M.; Johnson, J.V.; Yost, R.A. Resonance in a quadrupole ion trap. J. Am. Soc. Mass Spectrom. 1993, 4, 917–929. 27. Guan, S.; Marshall, A.G. Space charge in an ion trap. J. Am. Soc. Mass Spectrom. 1994, 5, 757–764.
438
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
28. Stafford, G.C. Jr.; Taylor, D.M.; Bradshaw, S.C.; Syka, J.E.P.; Uhrich, M. Enhanced sensitivity and dynamic range on an ion trap mass spectrometer with automatic gain control (AGC). Proc. 36th ASMS Conference on Mass Spectrometry and Allied Topics, San Francisco, CA, June 5–10, 1988. 29. Agranoff, B.W.; Benjamins, J.A.; Hajra, A.K. Lipids. In Basic neurochemistry. Molecular, cellular and medical aspects, 6th ed.; Siegel, G.J., Agranoff, B.W., Fisher, S.K., Albers, R.W., Uhler, M.D., Eds.; Lippincott-Raven: Philadelphia, 1999, 47–67. 30. Han, X.; Gross, R.W. Electrospray ionization mass spectroscopic analysis of human erythrocyte plasma membrane phospholipids. Proc. Natl. Acad. Sci. 1994, 91, 10635–10639. 31. Hsu, F-F.; Bohrer, A.; Turk, J. Formation of lithiated adducts of glycerophosphocholine lipids facilitates their identification by electrospray ionization tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 1997, 9, 516–526. 32. Martinez, M.; Mougan, I. Fatty acid composition of human brain phospholipids during normal development. J. Neurochem. 1998, 71, 2528–2533. 33. Anderson, M.; Dawson, W.W.; Gonzalez-Martinez, J.; Curcio, C.A. Drusen and lipidfilled retinal pigment epithelium cells in a rhesus macula. Vet. Opthal. 2006, 9, 201–207. 34. Forlenza, O.V.; Schaeffer, E.L.; Gattaz, W.F. The role of phospholipase A(2) in neuronal homeostasis and memory formation: implications for the pathogenesis of Alzheimer’s disease. J. Neural Trans. 2007, 114, 231–238. 35. Pettegrew, J.; Panchalingam, K.; Hamilton, R.; McClure, R. Brain membrane phospholipid alterations in Alzheimer’s disease. Neurochemical Research 2001, 26, 771–782. 36. Ross, B.M.; Mamalias, N.; Moszczynska, A.; Rajput, A.H.; Kish, S.J. Elevated activity of phospholipid biosynthetic enzymes in substantia nigra of patients with Parkinson’s disease. Neurosci. 2001, 102, 899–904. 37. Han, X.; Gross, R.W. Structural determination of picomole amounts of phospholipids via electrospray ionization tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 1995, 6, 1201–1210. 38. Jackson, S.N.; Wang, H-Y.J.; Woods, A.S. In situ structural characterization of phosphatidylcholines in brain tissue using MALDI-MS/MS. J. Am. Soc. Mass Spectrom. 2005, 16, 2052–2056. 39. Reich, R.F.; Cudzilo, K.; Yost, R.A. Quantitative imaging of cocaine and its metabolites in postmortem brain tissue by intermediate-pressure MALDI/linear in trap tandem mass spectrometry. Proc. 56th ASMS Conference on Mass Spectrometry and Allied Topics, Denver, CO, June 1–5, 2008.
Progress 15 Technology and Application in GC/ MS and GC/MS/MS Mingda Wang and John E. George III Contents 15.1 Introduction................................................................................................. 439 15.2 Advances in gc/ms and gc/ms/ms........................................................ 442 15.2.1 Instrumentation.............................................................................. 442 15.2.1.1 Non-Linear Ion Trap with Higher Scan Speed, Higher Mass Resolution, and Extended Charge Capacity........................................................................ 442 15.2.1.2 Internal Ionization with Coated Electrodes..................444 15.2.1.3 External Ionization with Pulsed Electron Beam.......... 445 15.2.1.4 Chemical Ionization (CI)..............................................448 15.2.2 Progress in GC/MS/MS................................................................. 454 15.2.2.1 Ion Isolation.................................................................. 455 15.2.2.2 Ion Activation................................................................ 461 15.2.2.3 Low Mass Cut-Off .......................................................465 15.3 Ion Trap Applications in GC/MS................................................................466 15.3.1 Limitations of Early Ion Trap Technology for Applications.........466 15.3.2 Acceptance of Ion Traps for Use with USEPA Methods............... 467 15.3.3 Specific Applications..................................................................... 470 15.3.3.1 USEPA 8270 Contaminated Extract Analysis.............. 470 15.3.3.2 Crude Vegetable Extracts.............................................. 476 15.3.3.4 Ion Trap Analysis with Liquid Chemical Ionization (CI) Reagents: USEPA Method 521.............................. 481 15.3.3.5 Analysis of Polychlorinated Biphenyls (PCBs) by Ion Trap Mass Spectrometry......................................... 483 15.4 Summary..................................................................................................... 486 Acknowledgments................................................................................................... 486 References............................................................................................................... 487
15.1 INTRODUCTION The quadrupole mass filter and the 3D quadrupole ion trap were invented simultaneously by Paul and Steinwedel, who received a patent for these two 439
440
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
inventions in 1960 [1]. Subsequently, the quadrupole mass filter was commercialized rapidly as a mass analyzer while the 3D quadrupole ion trap was used as a storage device and chemical reactor for ion/molecule research in universities during the 1960s and 1970s. Characterization of the quadrupole ion trap during this period laid the foundations for modern quadrupole ion trap mass spectrometry. The excellent book of Quadrupole Ion Trap Mass Spectrometry by R.E. March and J.F.J. Todd [2] has provided a detailed summary on the history of the development and the basic theory of operation of the ion trap as a mass spectrometer. The development of mass-selective instability technology by Finnigan MAT (now Thermo Scientific) in 1983 [3,4] changed significantly the mode of operation of the quadrupole ion trap and led to the first commercial quadrupole ion trap-gas chromatography/ mass spectrometry (GC/MS) ion trap detector (ITD), the ITD-700™. In the ITD-700™, ions were formed inside the ion trap volume by electron impact (EI) ionization. The ion trap is operated intentionally at a relatively high pressure of approximately 10 –1 Pa (7 × 10 –4 Torr) because collisions with light buffer gas, such as helium, enhance significantly the mass resolution and sensitivity of the ion trap mass spectrometer [2]. After collisional cooling with helium buffer gas, the amplitude of the RF trapping potential was ramped up. As the amplitude was ramped up, the qz -values of the trapped ions species crossed sequentially the instability boundary, at qz = 0.908, in order of increasing mass-to-charge ratio, m/z. Thus, ions of consecutive values of m/z were ejected successively along the axial direction and were detected by an external detector. In 1989, Finnigan MAT introduced a second generation of the 3D GC/MS ion trap (ITS-40™). ITS-40™ is one of the GC/MS product names of Finnigan MAT and the research version of this instrument is the ITMS 40™; Varian’s Saturn line of GC/MS instruments is an Original Equipment Manufacturing (OEM) product of the ITS-40™. In the Finnigan MAT ITS-40™, ions were ejected resonantly and sequentially by an AC dipole electric field (commonly called axial modulation) while the amplitude of the RF trapping potential was being ramped up [5,6]. As the amplitude was ramped up, the secular frequencies of trapped ions increased until the secular frequency of each ion species coincided with the frequency of the AC dipole field. At this juncture, ions of consecutive values of m/z absorb energy successively and were ejected resonantly along the axial direction. The modified mass-selective instability technology (axial modulation or resonance ejection scan) led to a significant improvement in both the mass resolution and the charge capacity of the 3D ion trap, shown in Figure 15.1. In 1989, Varian introduced Saturn-I™, an OEM version of the ITS-40. It is the ITS-40, Saturn-I, and successive generations of these products, that have made ion trap mass spectrometry such a powerful technique that can be employed routinely with GC/MS instruments. In 1994, Varian introduced the Saturn 4D™ gas chromatography/tandem mass spectrometry-(GC/MS/MS)-ion trap mass spectrometer that was the first practical and commercial GC/MS/MS ion trap instrument. Today, GC/ MS/MS ion trap instruments are used routinely in the environmental field, as well
Technology Progress and Application in GC/MS and GC/MS/MS TIC: 3,000
Ionization Time: 5000 μs
Intensity
205 TIC: 300
441
Ionization with axial modulation off
m/z
210
Ionization Time: 500 μs
Intensity Ionization with axial modulation off
205 TIC: 4,100
m/z
210
Ionization Time: 5000 μs
Intensity Ionization with axial modulation on
205
m/z
210
Figure 15.1 The data were acquired with a Varian Saturn GC/MS ion trap instrument. Each mass spectrum was obtained from the TIC. Top: instrument is operated in the massselective axial instability ejection mode; clearly, isotopic peaks are not resolved when the duration of the ionization is 5000 µs. Middle: the duration of the ionization is reduced to 500 µs; the number of ions in the middle figure is about 10% of that in the top figure. While the isotopic peaks are resolved, the mass resolution, even at m/z 207, is still less than unit mass resolution due to column bleeding. Bottom: the instrument is operated in the resonance ejection (axial modulation) mode. The number of ions in this figure is about the same as that in the top figure. Note that unit mass resolution is achieved despite the significant increase in charge capacity with the axial modulation mode.
as in food, forensic, and drug analysis, especially for the analysis of dirty matrix samples. Theoretically, in a pure quadrupole ion trap, the RF electric field increases linearly in radial and axial directions. The axial and radial direction motions of ions are decoupled in such a pure quadrupole ion trap. In reality, it is impossible to create a pure quadrupole ion trap due to the finite sizes of the trap electrodes and orifices in the electrodes.
442
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
In modern ion trap mass spectrometry, weak higher-order multipole and/or dipole electric fields are always superimposed upon the quadrupole electric field. The weak superimposed higher-order multipole and/or dipole electric fields are found to enhance the mass spectral performance with respect to mass resolution, charge capacity, scan speed, and to chemical mass shift. In fact, all commercial ion trap mass spectrometers are non-linear ion traps [7–9]. In the Finnigan ion trap instrument, the positions of the two end-cap electrodes are stretched out from the theoretical positions to add the higher-order multipoles. In the Bruker ion trap, the profiles of the end-cap electrodes are crafted to create the superimposed higher-order multipoles. In the Varian 4000 ion trap, a dipole and higher-order multipoles are superimposed upon the quadrupole field by a switchable electric circuit [10]. Thermo Scientific has introduced the 2D (or linear) liquid chromatography/mass spectrometry (LC/MS) and liquid chromatography/tandem mass spectrometry (LC/ MS/MS) ion trap mass spectrometer [11], known as the LTQ™. The ion injection efficiency has been improved dramatically in the linear ion trap. In addition, ions in a linear ion trap are distributed along the central axis, rather than concentrated in the ion trap center for a 3D ion trap; thus, both charge capacity and sensitivity in a linear ion trap have been increased significantly. Although at the present time there is neither commercial linear ion trap/GC/MS nor commercial linear ion trap/GC/ MS/MS instruments, a commercial version of a linear ion trap/GC/MS is expected to appear in the near future.
15.2 ADVANCES IN GC/MS AND GC/MS/MS 15.2.1 Instrumentation Since the middle 1990s, most recent ion trap developments have been focused on the improvement of LC/MS instruments. However, GC/MS and GC/MS/MS ion trap mass spectrometers have matured in the same period. In particular, the GC/MS/ MS instrument has grown from a novel research instrument into a routine, widelyapplied analytical device. In this section, a brief discussion of new GC/MS and GC/ MS/MS instrumentation development is presented. 15.2.1.1 Non-Linear Ion Trap with Higher Scan Speed, Higher Mass Resolution, and Extended Charge Capacity The non-linear ion traps with stretched or modified electrode profiles have been discussed previously in detail. In 2004, Varian introduced the 4000™ instrument with a new type of non-linear ion trap [9,10], in which a dipole and higher-order multipoles are superimposed upon the quadrupole field by a switchable electric circuit, shown in Figure 15.2. In the ion trap, both dipole and quadrupole supplemental fields are applied to the two end-cap electrodes with their frequencies tuned to βz = 2/3. The tripleresonance scan function [9] results in improved mass resolution, higher scan speed, and extended charge capacity, as shown in Figure 15.3a and b. It is well known that the mass resolution of the quadrupole ion trap mass spectrometer degrades when the total number of ions inside the trap reaches the spectral charge
443
Technology Progress and Application in GC/MS and GC/MS/MS
3V
100 V ω Ω
Collector
–3 V
Amplifier
Figure 15.2 In the 4000 instrument, a trapping dipole (ca 3%) and other higher-order multipoles are superimposed on the quadrupole field through a series inductor and capacitor. The non-linear field is electrically switchable. (a)
Mass spectrum of m/z 614 at 20 kTh/s
Intensity
25000 20000 15000 10000 5000 0 612
613
614
(b) 600
615 m/z
616
617
618
576.7
Intensity
500 400
577.2
300 577.7
200 100
587.6 588.6
578.2
0 575
580
m/z
585
590
Figure 15.3 (a) Data were acquired with a modified 4000 MS. At a scan speed of 20,000 Th s–1, the isotopic peak of the ion of m/z 614 is almost baseline separated. (b) Data were acquired with a Varian 500 LC/MS in order to show the mass resolution of the doubly-charged ions (separated by 0.5 Th) in a triple-resonance instrument. The mass spectrum shows clearly that the isotopic peaks of the doubly-charged ions are almost baseline separated at a scan speed of 5000 Th s–1. Note that this excellent mass resolution is achieved with the normal scan speed of 5000 Th s–1, not in zoom scan mode.
444
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
limit. An automatic gain control (AGC) method [12] has been applied to control the total number of ions inside the trap. In AGC mode, a pre-scan is applied to survey the total ion signal created during a fixed ionization time. The total ion signal of the prescan is then used to optimize the ionization time for the next micro-scan, the analytical scan. The principle of the AGC method is to create the optimized number of ions in each analytical scan as the sample concentration is changed. AGC has improved the linear dynamic range and has enhanced the sensitivity of the ion trap. It is evident that the spectral charge limit (known as ion trap charge capacity) relates directly to ion trap performance. There is no quantitative definition, as yet, for ion trap charge capacity. Generally, this term is taken to mean the maximum number of ions within an ion trap that will yield, upon ejection, a mass spectrum of unit mass resolution. It is found that a number of factors affect ion trap charge capacity, such as ion trap geometry, the scan mode, the ratio of dipole and higher-order fields to the quadrupole trapping field, the trapping frequency, the amplitude and phase of the supplemental dipole and/or the supplemental quadrupole electric fields, and the ion trap analytical scan speed. Schwartz has reported that Thermo Scientific’s LCQ, a stretched ion trap, has a charge capacity of ca 1,000 ions at its normal scan speed [13]. Schubert et al. [14] and Mordehai et al. [15] have demonstrated that Bruker’s High Capacity Trap (HCT)* has a charge capacity of ca 10,000 ions while operating in a non-linear resonance mode at a m/z scanning rate of 26,000 Th s–1. Specht and co-workers have revealed [16] that Varian’s 500MS™ instrument has a charge capacity of ca 10,000 ions while operating in the triple-resonance mode at a scanning rate of 10,000 Th s–1. Higher charge capacity will enhance greatly an ion trap’s analytical performance, such as with respect to an increase of sensitivity. 15.2.1.2 Internal Ionization with Coated Electrodes In the internal ionization mode, a sample is introduced directly into the interior of the ion trap and subjected to EI within the confined volume by a pulsed electron beam as shown in Figure 15.4. Ions created by EI are confined in the ion trap when they have stable trajectories under the ionization conditions. After collisional cooling with buffer gas, the stored ions are ramped out sequentially and detected in order of increasing m/z-value. In the internal ionization mode, the ion trap acts as both the ionization source and the mass analyzer. The dual functions of the ion trap are fulfilled sequentially in time. Due to its simplicity and higher-ionization and ion-trapping efficiencies, internal EI ionization frequently is the first choice in GC/MS and GC/MS/MS applications. However, for an ion trap to be an ion source, where its surface area can be as large as 100 times that of the surface of a traditional EI ion source, strong surface absorption of polarized compounds on ion trap electrodes results in serious chromatographic peak tailing of these compounds [17], such that both the linearity and detection limits of those compounds are degraded significantly. Taylor et al. had observed these surface absorption and decomposition problems and, in an attempt to reduce these effects, coated the surfaces of the ion trap * See Volume 4, Chapter 13: An Examination of the Physics of the ‘High-Capacity Trap’ (HCT) by Andreas Brekenfeld, Ralf Hartmer, Desmond Kaplan, Carsten Baessmann, Jochen Franzen, and Michael Schubert.
Technology Progress and Application in GC/MS and GC/MS/MS Filament endcap
445
E.M. endcap Electron multiplier
Filament Electron gate Transfer line tip
H.E.D. Ring electrode
Figure 15.4 A schematic diagram showing typical internal EI ionization in the ion trap with an off-axis detector. In the internal ionization mode, analytes are eluted into the ion trap through a transfer line. Electrons are injected into the trap and ionize the analytes.
electrodes with chromium [18]. Chromium coating works well for many polarized compounds, particularly when the chemical concentration is around the subnanogram level and higher. In 1995, Brittain and Wang [19] introduced ‘Silchrom’ coating on the surfaces of ion trap electrodes. Because the silchrom coating is more inert to polarized compounds than is the chromium coating, both surface adsorption and chromatographic peak tailing are reduced further with silchrom coating while the detection limit is extended to the picogram level for polarized compounds, as is shown in Figure 15.5. The Silchrom coating is deposited onto the electrode surfaces by a chemical vapor deposition (CVD) process. In contrast to traditional conductive coating on electrodes, this coating is composed of a non-conductive material, particularly in the outer-most layers. Because the ion trap is operated in a pulsed ionization mode and the coating is very thin, charges on the surface are able to dissipate and to be neutralized while the RF electric field is applied to the electrodes. 15.2.1.3 External Ionization with Pulsed Electron Beam In the internal ionization mode, ions are formed, stored, and excited resonantly to the point of ejection inside the ion trap in the presence of neutral sample molecules. Primary fragment ions confined in the ion trap experience the flow of neutral sample molecules through the ion trap for a period of time that varies from ca 2 ms to ca 200 ms. The longer is this period of time, the greater is the probability of the occurrence of undesirable ion/molecule reactions; this process is known as self-chemical ionization, or self-CI [20]. Typically, self-CI will create [M + H] + ions due to proton transfer. However, adduct ion peaks may be formed when an alkyl group is transferred. The self-CI process may affect library-search results of those compounds because the majority of mass spectra in mass-spectral libraries were acquired with quadrupole mass filters or magnetic sector mass spectrometers wherein self-CI does not
446
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 1A DFC_silchr_250C_01_10-30-2007.SMS TIC
MCounts
2.0
CI
O
O
CI
1.5
N
1.0
CH2 O N
CH3
N
0.5 0.0
DFC_250C_chrom_11-5-2007 001.SMS TIC
100
kCounts
90 80 70 60 50 40 15.75
16.00 16.25 minutes
16.50
Figure 15.5 Total ion chromatograms of fungicide Difenoconazole with Silchrom (top) and Chrom (bottom) coated electrodes, respectively. It is unambiguous that the isomers of Difenoconazole are separated clearly and have higher peak intensities with Silchrom-coated electrodes.
occur. When the ion/molecule reaction time is minimized by, for example, changing the mass segment break, the extent of self-CI reaction may be reduced but not eliminated completely. Yost and co-workers have demonstrated the elimination of self-CI in an ion trap by using an external ion source [21]. The external ion source used in the ion trap (see Figure 15.6) is quite similar to those ion sources used widely in quadrupole mass filters. Usually, magnets are installed behind the filament assembly in such a manner that the magnet field direction is perpendicular to the ion beam axis. Thus, electrons travel through the opening hole of the ion source and, upon entering the ionization volume, follow tight-spiral trajectories so that the paths of the electrons through the ion source are greater than the physical dimensions of the source. Ions formed inside the ion source are extracted or pushed out of the ionization volume. After passing through a focusing system of lenses, ions are injected into the ion trap where they are confined subsequently. Ion trapping efficiency depends strongly on a number of factors, such as buffer gas pressure, atomic/molecular weight of the buffer gas relative to that of the ionic species, ion kinetic energy, and the qz -value of injected ions. It is clear that the ion trapping process will exhibit a strong mass dependence. The mass-dependent ion trapping efficiency can be improved by varying both the amplitude of the RF potential
Technology Progress and Application in GC/MS and GC/MS/MS
447
Figure 15.6 A schematic diagram showing the external ion source of the ion trap with an off-axis detector. The source is switchable between EI and CI. Analytes ions are created in the ionization chamber outside the ion trap. Subsequently, analyte ions are injected into the ion trap through lenses.
and ion kinetic energy during the ion trapping period. The Thermo Scientific ITQ GC/MS system has demonstrated library-searchable mass spectra due to the autotune capability of the instrument. The ITQ GC/MS instruments have been used for routine environmental analysis and industrial quality control. The overall trapping efficiency of the 3D ion trap is only about 5% or less for externally-created ions, and this relatively low trapping efficiency has contributed to the popularity of the 2D LC/MS ion trap among ion trap users. There is little doubt that 2D GC/MS instruments soon will be popular with practitioners of GC ion trap mass spectrometry.* Instrument detection limit, LOD, is defined by the ratio of signal-to-noise (‘S/N’), thus, the LOD increases as the background noise is reduced. It is well known that excited neutrals can strike the ion detector and generate background noise. Bier et al. [22] invented an external ion source that takes advantage of the pulsed mode of ion trap operation; they introduced a temporal variation to the electron energy inside the ionization volume. During the mass analysis scan period, the electron energy is controlled at a level below the ionization energy of helium gas. Thus, few excited helium atoms are produced during that period and noise due to neutrals striking the ion detector is reduced noticeably. Wells et al. [23] demonstrated a pulsed external ion source in which the electron beam direction is controllable by the repeller and the electron lens. With this source, the electron beam is directed into the ionization chamber during the ion injection period and then is deflected away from the ionization chamber during the analytical scan period. Noise generated by helium neutrals is eliminated almost entirely and contamination of the ion chamber and ion lens assembly is decreased significantly, thus cleaning of the ion source is required * For a detailed discussion of 2D, or linear, ion traps, see Chapters 11 to 14 in Volume 4 of this series.
448
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Figure 15.7 Left, a photograph of a filament assembly. The filament is sandwiched between repeller and electron lenses. The electron beam direction is reversible by adjusting the voltages of the repeller and electron lenses. Right, a photograph of a filament assembly with the electron lens removed so that the filament assembly design is seen to be similar to that of a normal filament.
less frequently. In Figure 15.7 is shown a photograph of the filament assembly of the external ion source. 15.2.1.4 Chemical Ionization (CI) There are several techniques that have been developed recently concerning the chemical ionization (CI) mode of operation, such as liquid CI, pulsed positive ion negative ion CI (PPNICI), and hybrid CI. A brief examination of CI in the ion trap is addressed in this section. 15.2.1.4.1 Selective-Ejection Chemical Ionization (SECI) The basic principles of CI operation with quadrupole mass filters and quadrupole ion traps are the same for each type of instrument. In brief, a CI reagent is ionized to form primary ions that react with CI reagent molecules to form, usually, protonated CI reagent molecules. The affinity of CI reagent molecules for a proton is relatively low and, upon collision with sample molecules, proton transfer occurs from which the product ion is a protonated sample molecule. However, the operating pressures and reaction times are quite different between the two ion sources. In a conventional CI ion source, ions are produced continuously and ion residence time inside the ion source is ca several 10’s to 100’s of microseconds. In order to generate a sufficient quantity of protonated sample molecules by CI and to minimize the quantity of sample ions formed by EI, the CI reagent gas pressure is ca 1 Torr in a conventional ion source. In order to effect internal CI within an ion trap, an electron beam is injected into the ion trap to ionize CI reagent gas molecules. Because the pressure of reagent gas in ion trap is ca 10 –4 –10 –5 Torr, the trapped primary reagent ions can react with sample neutrals during a period of ca 1–200 ms. Subsequently, sample CI ions, typically protonated molecules, are ramped out of the ion trap for detection. The RF level during the primary ionization period is kept low intentionally so as to suppress the
Technology Progress and Application in GC/MS and GC/MS/MS
rf
Ionization
Reaction
449
CI Background mass
2 Step rf CI scan function
rf
Ionization
SECI
Reaction
CI Background mass
Higher mass ejection wave form Saturn 2000 CI scan function
Figure 15.8 Top, the two-step RF CI scan function used in early ion trap CI mode. Bottom, the SECI scan function in which the low frequency square waveform that acts as a low mass pass filter, is used to isolate CI reagent ions.
trapping of sample ions formed by EI [24]. The RF level is stepped up during the reaction period so as to confine sample CI ions. A scan function for this operation is shown in the upper part of Figure 15.8. Nevertheless, some EI fragment ions continue to be detected during the above 2-step-RF CI mode. A number of selective-ejection chemical ionization (SECI) methods have been proposed. The basic principle of such methods is to isolate the CI reagent ions and to eject sample ions, formed by EI, prior to reaction. Strife et al. have introduced an RF and DC isolation step so as to isolate CI reagent ions [25]. In the Varian commercial Saturn 2000 3D ion trap product, a low frequency square wave is applied to the two end-cap electrodes to eject undesired sample ions formed by EI, as shown in the lower part of Figure 15.8. In the Varian 4000MS instrument, a broadband waveform, rather than a square wave, is applied as low-mass pass filter. In addition, to remove residual EI components from CI mass spectra, SECI has the ability to perform CI with mass-selected reactant ion [26,27]. For example, when methane is used as reagent gas, either proton transfer or hydride abstraction CI is performed depending on which reactant ion, CH5+ or C2H5 + , is mass selected. The potential of the mass-selective CI technique is worthy of further exploration. 15.2.1.4.2 Liquid Chemical Ionization (CI) As the operating pressure of a traditional CI source is ca 1 Torr, the choice of reagent gas is limited, usually, to permanent gases. Because the pressure of the CI reagent in the ion trap is ca 10 –3–10 –5 Torr, the possibility exists to use volatile liquids as CI reagents, thus widening the range of compounds that may be used as CI reagents. The use of a volatile liquid for CI is termed ‘liquid CI’. Proton transfer reaction is the most important of the ion/molecule reactions that occur in the CI mode. Liquid CI is able to provide a suitable reagent, for example with a proton affinity that is slightly higher that the proton affinity of the sample molecules; thus, the proton transfer reaction will be slightly exothermic and fragmentation of the nascent protonated sample
450
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
molecule will be curtailed. In this manner, it is possible to determine molecular mass as an aid to chemical identification. Currently, liquid CI is used chiefly in environmental and forensic analysis. However, liquid CI may be used in many areas, such as the elucidation of chemical structure. In order to determine an unknown chemical structure, the availability of an array of reagents can be crucial. Each reagent offers a characteristic reaction with a specific functional group. Using a combination of all reactions, it should be possible to identify the functional groups of an unknown. Liquid CI will become a useful research tool once researchers have explored thoroughly its potential. Figure 15.9 shows the design of the liquid CI system for the Varian 4000MS instrument. Clearly, liquid CI is a useful and convenient tool for routine analysis and academic research [28,29].
Liquid reservoir
15.2.1.4.3 External Chemical Ionization (CI) The ion trap external CI ionization source, as shown in Figure 15.10, is essentially the same as that of the conventional CI ion source except that the ion trap external CI ion source of ion trap is operated in a pulsed mode [22,23]. The external CI source is able to create both positive and negative analyte ions. In a CI ion source, negative
Restrictor
CI Shutoff value RestrictorCI Solenoid
To manifold
CI Needle value
Reagent vial → To roughing pump
Vial shield↑
↑ Change this restrictor to switch to gas CI reagents
Figure 15.9 Left, the plumbing diagram for liquid CI. Right, the reservoir for liquid CI reagent. Gas and liquid CI can be converted easily by switching the inlet tubing.
Sample inlet for external
Figure 15.10 A typical external CI ion source for the ion trap. It is similar to that for the quadrupole mass filter except for operating under pulsed mode.
Technology Progress and Application in GC/MS and GC/MS/MS
451
ions are formed usually by electron capture or electron attachment processes. Such process require near thermal electron energy. Because electrons do not follow stable trajectories and cannot be cooled collisionally in the normal operation of the ion trap, negative CI is not possible by internal CI ionization within an ion trap mass spectrometer. In the negative external CI mode, high energy electrons experience inelastic collision with the reagent gas and are thermalized. Then the thermalized electrons are captured by analyte and form negative analyte ions. Finally, the negative analyte ions are injected into the ion trap for mass analysis. With an external CI ionization source, the ion trap GC/MS instrument gains negative ion detection ability, which provides higher selectivity and enhanced sensitivity for many halogencontaining compounds, such as PCBs, fire retardants, and pesticides. Hunt et al. had proposed a PPNICI method for a quadrupole mass filter instrument [30]. Thermo Scientific offers an ion trap instrument (ITQ series ion trap) that incorporates an external CI ionization source; thus PPNICI may be carried out with this ion trap instrument, that is, both positive and negative ions may be detected with alternating positive chemical ionization/negative chemical ionization (PCI/NCI) scans. 15.2.1.4.4 Hybrid Chemical Ionization (CI) Hybrid CI, where hybrid refers to the injection of positive and/or negative CI reagent ions generated externally to the ion trap that react within the ion trap to form protonated molecules, is essentially a special CI experiment that can be performed in an ion trap that is configured in a highly-specific way. Figure 15.11 shows the basic hybrid CI configuration of the Varian 4000 GC/MS mass spectrometer. The steps
Sample inlet
Reagent gas inlet
Figure 15.11 A typical hybrid CI ionization source for the ion trap. The reagent gas is connected to the ionization chamber of the hybrid ion source. Sample molecules enter directly into the ion trap, which is separated from the ionization chamber. Chemical ionization reagent ions are pulsed into the ion trap through a lens assembly and react with sample molecules to form sample CI ions.
452
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
are: (a) CI reagent positive ions are created inside the external ion source by EI and, depending upon the experimental conditions, CI reagent negative ions may be created here also; (b) the CI reagent ions are transferred into the ion trap cavity through a series of lenses whereupon all species or forms of reagent ions are trapped; (c) upon the application of waveforms, specific ions are selected and stored for reaction in the ion trap; (d) sample compounds entering the ion trap cavity directly via the column effluent are allowed to react with the selected reagent ions to form protonated molecules; and (e) mass analysis occurs by RF voltage ramping and detection of ions by the conversion dynode/electron multiplier. The end result is highly-selective CI that provides a powerful ionization technique for the study of molecules. It is clear that hybrid CI is a novel technique for the quadrupole ion trap, in that it combines reagent ions generated externally with sample ions generated internally. Hybrid CI preserves the valuable features of ions generated both externally and internally, negative ion detection, high trapping efficiency of sample ions, liquid CI, and pure CI mass spectra. A simple application for this technique can be demonstrated by using FC-43 (perfluorotributylamine, PFTBA), a common tuning compound for mass spectrometers. In this example, the C3F5+ ion of m/z 131 from FC-43 is reacted with two isomeric polynuclear aromatic compounds (PNAs), phenanthrene and anthracene, in order to distinguish between them. First, however, let us consider the difficulties of distinguishing between these two isomers by other means. The EI mass spectra of phenanthrene and anthracene are shown in Figures 15.12 and 15.13, respectively; these two mass spectra are identical and distinction between phenanthrene and anthracene is not possible. The CI mass spectra of phenanthrene and anthracene obtained with methane are shown in Figure 15.14 but the CI mass spectra are identical also and, again, distinction between phenanthrene and anthracene is not possible. The mass spectrum of calibration compound, FC-43, as shown in Figure 15.15, is well-known to mass spectrometry practitioners because it is seen frequently during 178 Phenanthrene, MW 178
0
39
50
40
50
76
63 60
70
80
89
152 98
126
139
90 100 110 120 130 140 150 160 170 180 190 m/z
Figure 15.12 Electron impact mass spectrum of phenanthrene.
453
Technology Progress and Application in GC/MS and GC/MS/MS 178
100 Anthracene MW 178
50
0
39
63
50
76
89 98
126
139
152
20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 m/z
Figure 15.13 Electron impact mass spectrum of anthracene.
M+1
M+29
M+41
Both the EI and CI virtually identical
Figure 15.14 Chemical ionization mass spectra for phenanthrene (top) and anthracene (bottom). Protonated molecule and adducts are shown.
454
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 69
100
Perfluorotributylamine
F
219 50
0
F
F F
F F
F F
131
31 50 30
264 169 201 246
100 80
130
180
230
280
N F
F
F
F
F
F
F
F
414 330
380
m/z
F F
F
430
F
F
F
F
F
F F
F 466 502 480
530
580
630
680
Figure 15.15 Mass spectrum of perfluorotributylamine, showing m/z 131 selected for hybrid chemical ionization.
calibration of instruments in mass spectrometry. In this example, the ion at m/z 131, from the mass spectrum of FC-43, was stored in the ion trap prior to reaction with two specified PNAs, phenanthrene and anthracene. The hybrid CI mass spectra detected for phenanthrene and anthracene, and obtained using C3F5 + , show clear differences that are evident in Figure 15.16 and permit distinction of phenanthrene and anthracene. The hybrid CI technique is clearly advantageous in situations where structurally-similar isomers are difficult to distinguish. The current hybrid CI ion source in Varian 4000 GC/MC mass spectrometers is not an optimized design. If the above structural arrangement in Figure 15.11 was to be extended by the addition of differential pumping between the ion source and the ion trap, the performance should be enhanced. The mass-selective procedure employed for the isolation of CI reagent ions is relatively coarse and is achieved inside the ion trap. In the future, the insertion of a mass-selective device between the external ionization chamber and the ion trap will improve CI reagent ion selection and prevent possible reaction between undesirable reagent ions and analyte during the injection and isolation periods.
15.2.2 Progress in GC/MS/MS It is well known that there are two types of tandem mass spectrometers; these types are ‘tandem-in-space’ and ‘tandem-in-time’. The quadrupole ion trap mass spectrometer belongs to the latter type of instruments. One advantage of tandem-in-time is that no extra hardware is required for MSn, that is, multiple stages of tandem mass spectrometry. The three basic steps of MS/MS in a quadrupole ion trap, that is, ion isolation, ion activation, and mass analysis, were well established in the 1980s and early 1990s. However, the technique was used principally in research laboratories because of the complexity of the time-consuming procedure of tuning the MS/ MS parameters. The endless demands for improved detection limits and sensitivity, selectivity, and ease of use in the gas chromatography community provided the driving force that brought about the birth of GC/MS/MS. Because the elution time
455
Technology Progress and Application in GC/MS and GC/MS/MS
Anthracene Hybrid PCI with 131+ from FC43
[M+91]+
M+ Phenanthrene Hybrid PCI with 131+ from FC43
[M+49]+
[M+111]+
[M+131]+
Figure 15.16 Anthracene (top) and Phenanthrene (bottom) chemical ionization mass spectra obtained upon reaction of each compound with m/z 131 in the hybrid CI experiment.
for a typical GC peak width is only several seconds, GC/MS/MS method development must be automated. It should be possible to optimize the instrument, when required, in two to three injections, with minimum manual interference. The GC/ MS/MS method should be able to handle a dynamic peak range without re-tuning. As a routine analyzer, the tuning parameters should be both stable and reproducible for day-to-day operation. Recent progress in ion isolation and ion activation is described in the following sections. 15.2.2.1 Ion Isolation In the early days, many ion isolation methods were developed for use with the quadrupole ion trap [31,32]. Unfortunately, either the time to create and to tune the MS/MS program with these methods was too long for a GC/MS/MS system or the isolation efficiency and/or isolation resolution of the method was too low.
456
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
15.2.2.1.1 Two-Step Isolation Wells proposed an ion isolation method [33], which consisted of two steps for the isolation of precursor ions. The method employed a modified mass-selective instability technology (axial modulation technology) to eject sequentially ions with m/z less than that of the molecular precursor ion, M + , for example. Combined with an empirical calibration procedure, ions with mass/charge ratio > M Th are ejected resonantly with a broadband waveform. Figure 15.17 shows the scan function of this isolation method. In this method, ions with mass/charge ratio < M Th are ejected with axial modulation technology; this same technology is used in mass analytical scanning. Thus, the low-mass ejection process uses the same mass calibration table as does the mass analytical scanning process. Axial modulation ejection is well known for its ability to eject ions rapidly from the ion trap during a mass-selective scan; scanning rates of, for example, 5000–10,000 Th s–1 have been achieved with sub-unit mass resolution. This low-mass ejection process can be carried out at comparable scanning rates. After low-mass ejection, the trapping RF potential amplitude is dropped immediately to a lower level to avoid the loss of precursor ions. In this method, the low-mass isolation step will eject all ions with m/z ≤ (M-1) Th with almost no loss of precursor ions. High-mass ejection is carried out at a high qz -value with a broadband waveform while the amplitude of the trapping RF potential is either fixed or ramped down slightly. The broadband waveform consists of all frequencies corresponding to qz < 0.845. Only one fixed broadband waveform is used for all precursor ions in this method. Calibration of qz × m versus RFDAC known as the ‘trapping frequency calibration’, is one of the keys to success for the high-mass ejection technique. RF DAC Eject <M Eject >M
Isolation
WF1 WF2 AC Dipole
Selective trapping Eject >M Eject <M
Figure 15.17 The process of ion isolation is executed sequentially. First, lower masses are ejected with a modified mass-selective axial instability scan that is, essentially, the same as normal analytical scanning. Second, higher masses are ejected resonantly with a broadband waveform. The ejection order may be reversed to avoid product ions of higher m/z-values. WF1 is a notch waveform to eject unwanted ions during ionization and the post-ionization period; WF2 is a broadband waveform to eject higher mass ions during higher mass isolation period.
Technology Progress and Application in GC/MS and GC/MS/MS
457
refers to the number of digital-to-analog conversions effected, over a given range of RF voltage, to obtain a stepped RF voltage profile from a linear voltage profile. The relationship of qz × m to RFDAC can be expressed as a first-order approximation by Eq. (15.1)
q z × m = A + B × RFDAC
(15.1)
Usually, ions of m/z 69 and 414 from the calibration chemical PFTBA are used to find the values for the two parameters, A and B. The trapping frequency calibration is carried out near qz = 0.845 at a fixed RF trapping field. A linear relation between qz × m and RFDAC is true only under ideal conditions but, as a first-order approximation, it works well for this two-step isolation method in a non-ideal quadrupole ion trap. Typically, in an ion trap for which the oscillation frequency of the RF potential is 1 MHz, the frequency error of the calibration is less than 1 kHz, which corresponds to an error of < 1 Th in the high-mass isolation step. The amplitude of the broadband waveform is determined empirically by the manufacturer and can be accessed by users. The fully-automated ion isolation method and its modified versions have been used on commercial GC/MS/MS instrument since 1994. At the manufacturer’s preset (default) values of MS/MS parameters, the efficiency of ion isolation is > 90% for stable precursor ions (as discussed above), as shown in Figure 15.18. Because the broadband waveform is applied for limited time only, the frequency spectrum of the broadband is not a perfect rectangular shape, as is shown in Figure 15.19. The tailing of the high-frequency side will excite modestly the precursor ions. Some of the chemically-unstable ions, that is, ions having low fragmentation threshold energy, may gain sufficient energy to be fragmented in the higher-mass isolation process. For these unstable ions, the isolation window should be ca 3–5 Th. The wider isolation window combined with reduced amplitude of the high-mass ejection waveform can minimize the loss of unstable precursor ions. 15.2.2.1.2 Notch Isolation Application of the ‘Notch’ waveform for ion isolation is another practical technique that is employed in GC/MS/MS. As discussed above, the precision of the trapping frequency calibration is ca 1 kHz. The difference between the secular frequencies of ion motion for ions M + and [M + 1] + increases rapidly as the qz -values of the ions approach 0.908, at which the qz -axis intersects the βz = 1 boundary of the stability diagram. Thus, as a practical compromise, the notch waveform isolation is performed typically at qz > 0.8. At a qz -value of ca 0.8, unit mass of isolation resolution is achievable, as can be seen in Figure 15.18. In the isolation process, the magnitude of the RF trapping field is set to a value such that the secular frequency of the chosen precursor ions falls within the width of the frequency window of the notch waveform, as is shown in Figure 15.20. The frequency bandwidth of the notch waveform covers the secular frequencies of all unwanted ion species, typically 5 kHz–500 kHz in a 1 MHz trapping field. Thus, all unwanted ions would be ejected resonantly with the notch waveform, leaving the chosen precursor ions confined within the ion trap.
458
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Scan number: 16, Time: 0.243 min. RIC: 11541, Ion range: 256.5-270.5 m/z
Base peak: 263.9, Base amount: 4093 Ion: 2000 us, Segement: 1, C
263.9 3136
3.0
k Counts
2.5 2.0 1.5 1.0 0.5 0.0 257.5
260.0
262.5
Scan number: 8, Time: 0.116 min. RIC: 12139, Ion range: 256.5-270.5 m/z
267.5
270.0
Base peak: 263.9, Base amount: 4035 Ion: 2000 us, Segement: 1, C
263.9 3102
3.0 2.5 k Counts
265.0
m/z
2.0 1.5 1.0 0.5 0.0 257.5
260.0
262.5
m/z
265.0
267.5
270.0
Figure 15.18 Top, a pseudo-product ion mass spectrum of m/z 264 of PFTBA but without collision-induced dissociation (CID); the mass spectrum was acquired with default isolation parameters and isolation window of 1 Th. Bottom, a full scan mass spectrum of PFTBA obtained with the same duration of ionization and showing both the m/z 264 peak and the 13Cisotopomer peak at m/z 265. By comparison, loss of the m/z 264 ion during the isolation process is essentially zero for this chemically-stable ion. For chemically-unstable ions, loss of precursor ion during mass-selective isolation can be minimized by using an isolation window of 3–5 Th.
Several methods for constructing a notch waveform have been proposed [34–36]. Marshall developed originally the SWIFT technique [34] for Fourier transform ion cyclotron resonance, FT-ICR, spectrometry.* However, because the trajectories of trapped ions in each of an ICR cell and a quadrupole ion trap are characterized by secular frequencies of ion motion, the SWIFT technique can be applied also to the quadrupole ion trap. The SWIFT technique is able to generate a near perfect notch * For a detailed discussion of FT-ICR see Volume 5, Chapter 5: Fourier Transform Ion Cyclotron Resonance Mass Spectrometry in the Analysis of Peptides and Proteins by Helen J. Cooper.
459
Technology Progress and Application in GC/MS and GC/MS/MS
0.4
0.2
0
0
100
200 300 Frequency (kHz)
400
500 Expanded
0.4 0.2 0 450
460
470
480
490
Figure 15.19 The frequency spectrum of the broadband waveform used to eject higher mass ions. The expanded view of the spectrum (bottom right corner) showed clearly that the spectrum tailed into higher frequency area.
window when a quadratic phase relation is applied to its frequency components and the first quarter and fourth quarter of the time domain signals are apodized, as shown in Figure 15.21a and b. A quadratic phase relationship results in the waveform power being distributed throughout most of the time period such that the required dynamic range of the electronics will decrease. By nature, the quadratic phase SWIFT waveform is a pseudofrequency sweep waveform. The length of the SWIFT waveform is fixed. When the waveform is terminated in the middle of the length, the frequency spectrum will be distorted, as shown in Figure 15.21c. Figure 15.22 shows the time and frequency domain signals with random frequency phases. Both the waveform power and the frequency components are distributed uniformly throughout the entire time period. However, the magnitude-values that cross the frequency domain in random phase are not as uniform as those in SWIFT, as shown in Figure 15.22b. The edges of the notch window with random phase are not as sharp as those with quadratic phase SWIFT. Louris and Taylor [35] proposed a notch waveform that is a compromise between a SWIFT waveform and a random phase notch waveform; it is relatively more uniform in frequency components throughout the entire time period than is the SWIFT waveform. However, its frequency spectrum is not as flat as that of SWIFT. Generally, there is a trade-off between the dynamic range of the time signal and the flatness and notch-window sharpness of the frequency spectrum. Depending on the manner in which noise and the filter are generated, the Filtered
460
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
0.2
Move precursor ion to higher q
a
Position of precursor in stability diagram
0
–0.2 0.0
0.4 Precursor ion
Notch waveform
q
0.8
1.2
Frequency spectrum of notch waveform Amplitude Notch window
Fourier transform Frequency
Figure 15.20 The basic concept of notch waveform isolation. In the isolation process, the working point of precursor ions is moved first to a qz -value at which the secular frequency of precursor ions lies inside the window of the notch waveform. Subsequently, the notch waveform is turned on to eject all unwanted ions. Due to the limited duration of application of the notch waveform, the window is not a perfect box as is seen in the bottom picture of this figure.
Noise Field (FNF) notch waveform [36] will have a performance somewhere between that of the quadratic SWIFT notch waveform and the random phase notch waveform. Buttrill has shown that in a notch waveform, the amplitude of each frequency component should be weighted [37]. He suggested a formula, (Ai/An) = (Mn/Mi)x, in which 0.5 < x < 1.5. The quantities Ai and An are the amplitudes of the frequency components corresponding to the secular frequencies of ion species ‘i’ and ion species ‘n’, respectively, whilst Mi and Mn are the m/z-values of ion species ‘i’ and ion ‘n’, respectively. Wang et al. [38] found that the ion motion of two ions having similar m/z-values is coupled tightly due to the closeness of their secular frequencies. When a notch waveform is applied to separate [M + 1] + ions from M + ions, more power is needed to eject the [M + 1] + ions than that which is required to separate M + ions from [M + 1 + n] + ions, in which n is a positive number equal to or greater than 1. Conventionally, the application of the entire composite waveform signal at a higher average power to eject effectively the [M + 1] + ions has solved this requirement. However, the higher power reduces the effective width of the notch window and, consequently, the isolation resolution and the intensity of M + ions is reduced. In order to solve the above problem, Wang et al. generated a notch waveform in which
Technology Progress and Application in GC/MS and GC/MS/MS (a) Notch waveform of SWIFT 100 50
(b) Frequency spectrum of SWIFT notch waveform Intensity
Intensity
0.15
0
0.1
–50
0.05
100
461
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Time (ms)
0
0
50 100 150 200 250 300 350 400 450 500 Frequency (kHz)
(c) Frequency spectrum from section of SWIFT notch waveform Intensity
0.2 0.15 0.1 0.05 0
0
50 100 150 200 250 300 350 400 450 500 Frequency (kHz)
Figure 15.21 (a) Notch waveform created with SWIFT technique. (b) The frequency spectrum of the SWIFT notch waveform. The notch window has a sharp drop on both edges. (c) The frequency spectrum of a section of SWIFT notch waveform. Apparently, the frequency spectrum is distorted.
the relative amplitude of the component immediately adjacent to the notch window is ca 1.5–3.0 times greater than the amplitude of the other components, as shown in Figure 15.23. Schwartz and co-workers have proposed a modified notch isolation method that has somewhat higher resolution [39,40]. The method combines a notch waveform with slow RF resonance ejection scanning. In the first step of the method, most of the unwanted ions, except the precursor and ions of mass/charge ratio similar to that of the precursor ions, are ejected by a notch waveform. The trapping RF is then scanned slowly down to eject ions with m/z-values higher than that of the precursor ions. Next, the trapping RF is ramped slowly up to eject ions with m/z-values lower than that of the precursor ions. An isolation resolution, Δm, of ca 0.3 Th has been demonstrated with the technique. 15.2.2.2 Ion Activation Comprehensive summaries of ion activation methods can be found in Volume 3 of this series [31,32]. In all of the above methods, precursor ions are mapped at a fixed qz -value for the duration of the activation period. However, the non-ideal quadrupole field and space charge will cause the ion species’ secular frequency to deviate away from its theoretical value such that it no longer matches the frequency of the supplemental electric field calculated for the fixed value of qz. Due to the lack of a technique
462
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V (b) frequency spectrum of random phase waveform
(a) Notch waveform with random frequency phase 100
Intensity
Intensity
50
0.4
0 0.2
–50 100
0
0.5
1
1.5
Time (ms)
2
0
0
100
200
300
400
500
Frequency (kHz)
Figure 15.22 Time (a) and frequency (b) domain signals with random frequency phase. Though the power distribution is quite uniform throughout all of time period, its frequency spectrum fluctuates over all the frequency components. Tailings of notch window edges have narrowed the frequency window, which may result in ejection of precursor ions. 0.4
Intensity
0.2
0
0
100
200 300 Frequency (kHz)
400
500
Figure 15.23 In the notch waveform, the two frequency components adjacent to the notch window have higher relative amplitudes compared to the other components.
for the precise calibration of ion secular frequency for ion activation, several broadband resonance excitation methods, such as random noise excitation, SWIFT, and FNF broadband excitations, have been proposed. In these methods, the frequency spectrum of the applied supplemental electric field will cover the above-mentioned deviation of ion secular frequency from the expected value. Thus, at least one of the component frequencies will match, or nearly match, the ion secular frequency. 15.2.2.2.1 RF Modulation During the CID Period In 1994, Wells [41] proposed a new technique, RF modulation during the CID period in ion trap tandem mass spectrometry. In this method, the precursor ion secular frequency is brought into resonance with the applied supplemental AC dipole field, rather than matching the frequency of the AC dipole with the ion secular frequency as was done in early methods. The frequency modulation is realized by modulation of the
Technology Progress and Application in GC/MS and GC/MS/MS
463
0.2
a
0
–0.2 0.0
0.4
q
1.2 Time
RF Secular frequency
0.8
0
ωp
Ω/2
AC Frequency
Figure 15.24 Top, illustration of the modulation of an ion’s qz- -value and, thus, its secular frequency, by modulation of the RF trapping field. Bottom, the secular frequency of the ion oscillates about the fixed frequency of the AC field during the CID period.
amplitude of the RF trapping field, as seen in Figure 15.24. Typically, the modulation frequency is < 500 Hz. In practice, multiple periods of modulation of the RF trapping field are not necessary; one half period modulation (the RF amplitude is scanned either down or up once) is sufficient to fragment precursor ions consistently. Jackson et al. [42] have proposed another interesting CID method in which the CID process is accomplished during a mass acquisition scan. The method is very similar to that of axial modulation except that the amplitude of the AC dipole field is tuned to fragment precursor ions rather than ejecting precursor ions. They have shown this method to be fast, efficient, and highly energetic CID process. 15.2.2.2.2 Activation Energy The amplitude of the applied AC dipole is another important parameter for optimal fragmentation in GC/MS/MS experiments. The amplitude of the resonant supplemental AC dipole field to effect fragmentation depends not only on the various instrumental parameters, but on the chemical structure of precursor ions also. Schwartz and Taylor [43] described a normalized collision energy (NCE) method, in which the optimized amplitude for the resonant supplemental AC dipole is calibrated with a set of pre-selected chemicals (calibrants). They found that the amplitude of the excitation voltage required to fragment their calibrants is substantially linear with ion m/z-value for an individual instrument. With the amplitude of NCE, identical MS/MS results can be observed from instrument to instrument because the error of trapping frequency calibration with different instruments has been corrected by the NCE process.*
* For an application of the NCE method see Volume 5, Chapter 6: MS/MS Analysis of Peptidepolyphenols Supramolecular Assemblies: Wine Astringency Approached by ESI-IT-MS by Benoît Plet and Jean-Marie Schmitter.
464
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Amplitude (V)
Amplitude (V)
It is well known that fragmentation of molecules having dissimilar structures will require, almost invariably, the acquisition of dissimilar amounts of internal energy. In order to obtain optimal efficiency of CID for compounds having dissimilar structures without prior knowledge of the optimal amplitudes of resonant excitation necessary for the acquisition of dissimilar amounts of internal energy, several methods have been developed. Mulholland and Yost [44] have demonstrated a multi-level CID method to solve the above problem. In this method, the amplitude of the supplemental AC dipole is increased in steps during the CID period. The multi-level CID technique obviates the need for manual adjustment of the amplitude of the supplemental AC dipole. Later, Schwartz and coworkers [45] demonstrated a stepped normalized collision energy (SNCE) technique that can perform automated MS/MS and can eliminate individual optimization of collision energy. Brekenfeld et al. have described a method in which the amplitude of a multiple frequency AC dipolar excitation potential is ramped up, known as SmartFrag [46,47]; the SmartFrag method eliminates the need for amplitude optimization for dissimilar compounds. Figure 15.25 shows diagrammatically both SmartFrag and SNCE. In a data-dependent scan (DDS), the amplitude of the supplemental AC dipole field must be chosen automatically in an MS/MS experiment without prior knowledge of the chemical structure. Wang has shown a chemical structure-insensitive method in Varian’s turbo DDS technique [48]. In this approach, the frequency of the supplemental AC dipole field is intentionally set lower than the secular frequency of the precursor ion, by ca 2 kHz, for example. During the CID period, the RF trapping potential is ramped slowly down. Thus, the ion secular frequency will approach gradually the frequency of the supplemental AC dipole field. It is noticeable that, during the CID process, the excitation of the precursor ions starts with a near-resonance condition and moves gradually toward the resonance condition. When precursor ions approach progressively the resonance condition, they gain more kinetic energy and at an increasing rate of kinetic energy acquisition. Thus more internal energy is acquired by the precursor ions in the collision process. When the internal energy
0
CID period
30 ms
0
CID period
30 ms
Figure 15.25 A schematic diagram showing SmartFrag (Top) and SNCE (Bottom). During the CID period, the excitation voltage of the AC dipole field increases with time in both methods. Both methods will cause the dissociation of precursor ions having different fragmentation thresholds without prior optimization knowledge.
465
Technology Progress and Application in GC/MS and GC/MS/MS 0.2
a
0
–0.2 0.0
0.4
q
0.8
1.2
RF Level Time Secular frequency 0
ωp
Ω/2
AC Frequency
Figure 15.26 A diagrammatric representation of chemical structure-insensitive CID. In the technique, the secular frequency of the ion is moved toward to the frequency of the applied supplemental AC dipole field.
of a precursor ion reaches its threshold for fragmentation, the precursor ion will fragment to form product ions. In this technique, depending on their fragmentation threshold, precursor ions adjust automatically their breakdown point, instead of adjusting the amplitude of supplemental AC dipole field as shown in the above methods. For a more stable precursor ion, fragmentation occurs at the point at which the ion’s secular frequency is more close to the frequency of supplemental AC dipole field. Furthermore, this technique avoids the inconsistency associated with errors of secular frequency calibration. Figure 15.26 is a diagram of the method described above. The method may be modified with a frequency sweep in place of an RF ramp up or down. The amplitude of the supplemental AC dipole field may vary with time rather than remaining at a fixed value as stated above. 15.2.2.3 Low Mass Cut-Off During the CID period, precursor ions are excited resonantly by an AC dipole field. The energetic precursor ions experience inelastic collisions with buffer gas and gain in internal energy. When the accumulated internal energy exceeds the threshold for fragmentation, precursor ions fall apart and generate product ions. Because of this threshold, a trapping RF level, and hence the qz-value of the precursor ions, is selected to achieve a reasonable fragmentation efficiency. Typically, the qz-value of the precursor ions is between 0.2 and 0.4. Therefore, some of the product ions of low m/z-value will not be detected when their qz-values are greater than 0.908 at the trapping RF level. With the traditional CID method, the qz-value selected for the precursor ions is based on the trade-off between CID efficiency and low mass cut-off. Glish and co-workers [49] and Schwartz [50] demonstrated recently that the trade-off could be avoided with their HASTE CID and HighQ Pulsed CID techniques, respectively. The two techniques are basically identical except for the Q value (qz-value)
466
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
RF level in CID period Single frequency wave
Figure 15.27 A diagram of High-Q Pulsed CID. Precursor ions are excited at a higher qzvalue, ca 0.8, for a short period, then after a short delay time, for example 1 ms, the trapping RF level is reduced to a lower value of qz, for example, 0.1.
of the CID period. HASTE reported a Q value of ca 0.25 for CID, while for HighQ Pulsed CID a Q value of ca 0.6–0.85 is preferred. In the HighQ Pulsed CID method, precursor ions are excited at a higher qz-value, ca 0.8, for a short period then, after a short delay time, for example 1 ms, the trapping RF level is reduced to a lower value of qz, for example, 0.1, as shown in Figure 15.27.
15.3 ION TRAP APPLICATIONS IN GC/MS 15.3.1 Limitations of Early Ion Trap Technology for Applications Early ion traps could not provide a routine, powerful quantitative tool for analytical chemists in application work. Finnigan Corporation (now Thermo Scientific) introduced the first commercial ion trap in 1983 as the ITD-700. Ions formed by a fixed time-duration pulse of electrons were ejected from the ion trap by increasing linearly the amplitude of the RF voltage, causing the trajectories of ions of successive mass-to-charge ratios to become unstable in the axial direction and to exit the trap through a series of holes in one of the two end-cap electrodes. The ion trap operated at elevated temperatures, required for chromatography, which increased the rate of air and water permeation into the vacuum chamber. The enormous background of water caused ion/molecule reactions to occur that produced a mixture of EI and CI mass spectra. The relatively limited mass resolution achieved using the mass-selective instability mode of ejection, along with the CI reactions, resulted in poor quality mass spectra and incorrect mass assignments. A fixed ionization time, together with high air and water background, caused a larger-than-optimum number of ions to be stored in the ion trap, and resulted in loss of mass resolution and accuracy. During the development of the ITD-700, it was discovered that certain classes of compounds gave molecular ions with m/z-values that were shifted up or down relative to the m/z-values of ions in the calibration mass spectrum. The end result was that poor quantitation and spectral infidelity were observed. The instrument could not be used for reliable routine quantitative analysis. The Finnigan ITD-800, that was introduced subsequently, had a separate high temperature vacuum enclosure that contained the ion trap electrodes. The reduction of water and air contamination in the vacuum chamber reduced the CI content of the mass spectra to a negligible level. However, the use of early XT personal
Technology Progress and Application in GC/MS and GC/MS/MS
467
computer technology limited mass centroiding to a simple algorithm; this algorithm was unable to handle complex mass spectra with significant mass defects, and so yielded frequently improper mass assignment of the molecular ion. The development of resonant ion ejection scanning, commonly called axial modulation, increased the mass resolution to unit resolution over the entire mass range. The ITS-40 was introduced in 1989 with axial modulation scanning and a high temperature vacuum chamber. The increased power of microprocessors available at that time allowed more sophisticated mass centroiding algorithms to be applied. Varian, Inc. entered the bench-top GC/MS market in 1989 with the introduction of the Saturn I™ GC/MS system, under license from Finnigan. The Saturn I instrument was the Varian OEM version of the Finnigan ITS-40 ion trap instrument. Varian then made several quality and performance improvements to this original product and introduced, thereby, the Saturn II, III, IV, and 2000 models. Additional features and enhancements included design modifications to the RF, DC, and electron multiplier power supplies, control over production of the ion trap electrodes, production of high-performance turbo pumps, and reducing substantially the size and the cost of the ion trap (2000 MS). Software improvements included better centroiding algorithms, the ability to store selectively ions in the ion trap (Selective Ion Storage, SIS), ‘turn-key’ MS/MS with automated method development (AMD), and release of a Windows-based version of the software for both control and data handling. The Saturn 4000™ ion trap GC/MS system, introduced in 2004, contained no licensed technology or know-how. It provided a unique triple-resonance scan function that allows for enhanced storage of ions in the ion trap resulting in high mass accuracy and sensitivity. Although the most serious problems affecting the spectral quality of the ion trap as a mass analyzer have been resolved, the perception of the ion trap as a reliable mass analyzer was compromised. Early adopters found the instruments to be unstable and difficult to use. It was not possible to obtain quality full-scan mass data useful for library searches so as to identify unknowns. Quadrupole mass filters at the time were better suited for the analysis of environmental and other ‘dirty’ matrices, as they were not subject to space charging and ion/molecule reactions.
15.3.2 Acceptance of Ion Traps for Use with USEPA Methods As stated earlier, the ion traps available in the early to mid-1980s did not have the ability to control the ion population in the ion trap. Space charging and poor mass resolution resulted, and this was a disaster when chemists tried to run samples in heavy matrices. Chemists needed to rely on full scan mass spectral matches to known libraries to document the presence of a compound, and this was not possible with distorted spectral data being generated on these early ion traps. AGC [12] was a big leap forward, which allowed variable ionization times as a function of ion population. This feature controlled the space-charge effect allowing the user to generate ‘quadrupole-like’ mass spectra for good matches against known databases. Axial modulation was the next big step in allowing the chemist to tune the instrument with standard compounds that were readily available, namely bromofluorobenzene (BFB) and decafluorotriphenylphosphine (DFTPP). The use of
468
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
these two tuning compounds was a fixed requirement of the early methods in mass spectrometric examination of volatile and semi-volatile compounds. The following represents data taken on an ion trap GC/MS set up properly for EPA Method 8270 [51]. DFTPP is used to ensure that the MS is tuned properly. This tune-check must be completed every 8–12 h. A ‘quadrupole-like’ mass spectrum of DFTPP is obtained, demonstrating excellent mass accuracy and resolution. The importance of meeting this tuning requirement cannot be overstated. If any of the ion ratios are not within the limits, all of the data taken following the tune sample are to be considered invalid, even when the other quality control requirements were met. So for out-of-tune results, sample extracts would need to be re-injected and this situation resulted in a lot of repeated work by the chemists. Today’s commercial ion trap instruments can meet these tuning requirements routinely; this level of performance is essential for high-volume production laboratories. Figure 15.28 and Table 15.1 illustrate an ion trap instrument passing the EPA Method 8270 tuning requirements. Despite the significant advances made in ion trap technology, ion trap instruments were accepted only very slowly by commercial and government environmental laboratories. The reputation of the early ion traps was poor, as described above, and many potential purchasers thought that the new technology was not robust sufficiently for analysis involving heavy matrices. Even today, single quadrupole instruments, that is, quadrupole mass filters, continue to dominate in most environmental testing laboratories. However, those laboratories that have embraced quadrupole ion trap instruments have found some clear advantages over single quadrupole instruments. In the early 1990s, the Office of Water developed several USEPA methods for the analysis of analytes that were to be regulated under the Safe Drinking Water Act (SDWA) [52]. New organic analytical methods were being created to monitor more target analytes in the interest of protecting public health, and public water suppliers were required to test their source of water and finished waters (final product tap water) throughout their distribution systems.
Tune spectrum 22.620 min, scan: 1611 45:450, ion: 557 us, Ric: 143111
22.620 min BP 100%
198
75% 50% 25% 0%
51
442
77 110
50 74
107
100
255 206 200
275
m/z
443 300
400
Figure 15.28 Mass spectrum of the tuning compound DFTPP used in USEPA Method 8270. The mass intensity list for this mass spectrum is shown in Table 15.1.
469
Technology Progress and Application in GC/MS and GC/MS/MS
Table 15.1 Proscribed Tuning Criteria for USEPA Method 8270 Using Decafluorotriphenyl Phosphine m/z
Acceptance Criterion
Value
P/F
51 68 69 70 127 197 198 199 275 365 441 442 443
30–60% of m/z 198 less than 2% of m/z 69 Present less than 2% of m/z 69 40–60% of m/z 198 < 1% of m/z 198 Base peak 5–9% of m/z 198 10–30% of m/z of 198 > 1% of m/z 198 present and < m/z 443 > 40% of m/z 198 17–23% of m/z 442
43.28 0.86 42.09 0.27 41.31 0.00 100.00 6.37 12.44 5.03 39.51 60.63 17.41
Pass Pass Pass Pass Pass Pass Pass Pass Pass Pass Pass Pass Pass
Note: For each of 13 ions in the mass spectrum of DFTPP shown in Figure 15.28, the acceptance criterion and the normal ion signal intensity relative to the base peak are given. Pass/fail for each ion species relates to the acceptance criterion.
One such example was USEPA method 525.2 [53], that required the analysis of semi-volatile compounds in source and finished drinking water. The ion trap was the perfect instrument for such a method, primarily due to its excellent full scan sensitivity. Single quadrupole instruments often required selected ion monitoring (SIM) to reach the detection limits described in the early versions of USEPA Method 525. Several of the analytes, to which this method applied, were required to be monitored, by both public water suppliers and governmental agencies, at concentrations much lower than those cited in the method; these changes were necessitated by adverse health effects data obtained on certain contaminants. Endrin and heptachlor epoxide, for example, required reporting limits of 0.01 and 0.020 μg L –1 (10 and 20 parts-per-billion (ppb) in the final extract), respectively. In full mass scan mode, the ion trap was able to detect these contaminants at these levels with the required 1.0 μL injected on column. Single quadrupole instruments had to rely on SIM, thus a library-search confirmation was not possible. Another great advantage of quadrupole ion traps was the ability to store ions selectively in the ion trap and to perform CID, and to produce a pure product ion mass spectrum through MS/MS. This process of selective ion isolation permits the elimination of matrix ions, and enhances greatly the sensitivity and specificity of the analysis. The technique becomes extremely important when there is a high background present in the sample, as is illustrated in Figures 15.29 and 15.30.
470
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V TIC filtered
3
MCounts
Full scan region of chromatogram 2 MS/MS Region of chromatogram Tetraconazol
1
Pirimiphos-methyl Chlorpyrifos-methyl
0
Diazinon 12.5
15.0
Minutes
17.5
20.0
Figure 15.29 GC/MS/MS chromatogram of a crude vegetable extract. The full mass scan region contains a high concentration of low-mass ions, reducing greatly the sensitivity for target pesticides. When MS/MS is applied, ions of the target pesticides are observed clearly and become detectable.
With the MS/MS technique, the chemist now has a product ion mass spectrum that can be stored in a user library and called up readily for confirmation. Product ion intensity ratios within the product ion mass spectrum provide further specificity and confidence in the analysis. Figure 15.31 is an illustration of a product ion mass spectrum for pyrimiphos-methyl, a pesticide used on agricultural products. As a result of these advances in commercial ion trap instruments, quadrupole ion traps have been included in several USEPA methods. Examples are USEPA methods 524.2 [54], 525.2 [53], 527 [55], 528 [56], 529 [57], 521 [58], and several SW-846 wastewater and solid waste methods, such as 8260 and 8270 [59]. EPA method 521 requires specifically the utilization of MS/MS due to interferences that were encountered in sample extracts; these interferences could not be overcome by using a full mass scan and SIM mass spectrometry.
15.3.3 Specific Applications The following specific applications of the use of quadrupole ion trap instruments demonstrate clearly the power of this analytical tool. Software and hardware improvements have made the set-up and analysis of data extremely easy for users, even when performing different scanning functions in the same analysis. 15.3.3.1 USEPA 8270 Contaminated Extract Analysis EPA Method 8270 is a challenging environmental method with chemically-broad and diverse target analytes. It is well known by users of the method that many problems encountered are NOT due to the mass spectrometer (that is, the detector).
224.1
207.0
60.0
132.9
355.8 401.1 385.1
294.9
355.0
303.8
281.1
250
157.0 223.2 108.0 57.0 151.0 221.8 56.0 113.0 175.0 227.1
110.8
96.0
93.0
91.0
73.0
500
548.8 630.4
750
726.4 773.0
941.9
1000
14.577 min, scan: 2000, 50:1000, ion: 451 us, RIC: 749238, BC
Acquired range m/z
457.0
BP: 73.0 (26390=100%), 500ppb_matrix_4-19-2008_5-36-03 pm.sms
Figure 15.30 Mass spectrum taken from the full mass scan region shown in Figure 15.29. Note the presence of low mass ions, making identification of trace level components extremely difficult.
0%
25%
50%
75%
100%
Technology Progress and Application in GC/MS and GC/MS/MS 471
472
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Figure 15.31 Product ion mass spectrum for pirimiphos-methyl. The precursor ion, m/z 290.1, was isolated and fragmented using CID. The ratio of ion intensities for m/z 151.1 to m/z 262.1 can be monitored for further qualitative confirmation.
Compounds such as nitrophenols and other nitro-compounds can degrade easily on active sites in injection port liners and on analytical columns. Phthalates at low ppb concentrations are especially problematic during sample preparation due to contamination from the sample preparation apparatus and reagents; other compounds, such as pentachlorophenol, are thermally labile and/or photosensitive. Some of the target compounds are polar and are much more amenable to high-performance liquid chromatography (HPLC) or LC/MS analysis. In addition to these challenges, the types of extracts encountered range from relatively clean ground water samples to heavily-contaminated soil samples that contain low-mass hydrocarbons which add significant background interference in full scan mass spectrometry. The most significant challenge to an ion trap instrument in the performance of Method 8270 is the ability to identify and to quantitate target analytes in highlycontaminated samples. Frequently, soil samples extracted using this method contain high concentrations of low molecular weight hydrocarbons. Fuels and waste oils are often co-extracted from soil samples coming from contaminated industrial brown fields or EPA Superfund clean up sites. Figure 15.32 demonstrates that target compounds in contaminated extracts can be extracted from a total ion chromatogram (TIC) display obtained from the full scan mass spectrum. When the background subtraction is performed with software tools, a user can obtain a mass spectrum that will match the National Institute of Standards and Technology (NIST) or other mass spectral databases. Fit/Reverse fits of 900 or greater are achieved routinely on ion trap instruments even in the presence of matrix interference, as is demonstrated in Figure 15.33, in which 2,4-dichlorophenol is the example chosen.
Technology Progress and Application in GC/MS and GC/MS/MS
473
Figure 15.32 Top: TIC for a #2Diesel/10W30 extract of 3000 ppm spiked with a 5 ppm standard of 2,4-dichlorophenol. Middle: the ion profile of m/z 162 from 2,4-dichlorophenol extracted from the TIC above. Bottom: full scan mass spectrum obtained from the central part of the ion profile of m/z 162 above.
The new technology embodied in the quadrupole ion trap instruments of today provides also excellent quantitative data. A calibration curve was prepared for Method 8270 using, as the required low, mid, and high concentration levels, solutions of 5 ppm, 50 ppm, and 150 ppm; the solutions were prepared in pure solvent (dichloromethane, DCM). A TIC obtained at 5 ppm is shown in Figure 15.34. In addition, these same calibration concentration levels were prepared in DCM containing a mixture of both #2 diesel fuel oil and 10W-30 motor oil at a concentration of 3000 ppm. The corresponding TIC is shown in Figure 15.35. When the calibration curves were compared, several compounds at the low end of the calibrated concentration range were affected by components already present in the diesel/oil extract. For example, low-levels of some PNAs and phthalates, present naturally in these refined petroleum products, were detected in the unspiked diesel/oil extract. Also, some of the phenols in this dirty matrix were reactive in the injection liner; indeed, the matrix itself can passivate the liner for some target compounds. Passivation in this sense means that the liner surface becomes coated with non-volatile components, forming a barrier between the analyte and the bare, more reactive glass surface. While this issue is not related to ion trap mass spectrometry per se, it will be present in any analytical GC/MS system. As illustrated in the example below, calibration curve linearity (as represented by relative percent standard deviation, or RSDs, of the relative response factor at each calibration concentration level) and correlation coefficients for most compounds in the pure solvent were identical statistically to those prepared in the 3000 ppm diesel/oil matrix spikes, as are shown in Figures 15.36 and 15.37.
897 2,4-Dichlorophenol 873 Phenol, 2,4-dichloro868 Phenol, 2,4-dichloro865 Phenol, 2,4-dichloroPhenol, 2,6-dichloro864 Phenol, 2,5-dichloro862 Phenol, 2,3-dichloro859 Phenol, 2,3-dichloro866 843 Phenol, 2,6-dichloroPhenol, 2,4-dichloro836 830 Phenol, 2,4-dichloro826 Phenol, 2,5-dichloro819 Phenol, 2,3-dichloroPhenol, 2,6-dichloro813 804 Phenol, 2,5-dichloroPhenol, 3,5-dichloro802 Phenol, 3,5-dichloro816 Phenol, 3,4-dichloro799 Phenol, 3,4-dichloro780 Phenol, 3,5-dichloro774 Hydrazinecarbonximidamide,...841
858 815 814 812 809 806 805 804 804 794 791 785 781 774 765 760 757 745 744 743 700
61.02 61.02 61.02 61.02 12.53 11.07 10.64 10.64 12.53 61.02 61.02 11.07 10.64 12.53 11.07 2.26 2.26 1.37 1.37 2.26 0.29
162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 262
120-83-2 120-83-2 120-83-2 120-83-2 87-65-0 583-78-8 576-24-9 576-24-9 87-65-0 120-83-2 120-83-2 583-78-8 576-24-9 87-65-0 583-78-8 591-35-5 591-35-5 95-77-2 95-77-2 591-35-5 5001-32-1 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C6H4CI20 C9H12CI2N... 63.0 23313
BP: 161.9
105.0 11494
97.9 18280
163.9 201.33
63.0 498 99.0 351
98.0 709
100 R. Match: 897, F. Match: 858
0%
25%
50%
75%
164.0 632
200
300
CI
OH
400 Acquired range m/z
CI
2,4-Dichlorophenol CAS No. 120-83-2, C6H4C120, Mvv 162
10.610 min, Scan: 1040, 45:450, Ion: 1932 us, RIC: 245686, BC 161.9 3228.3
BP 162.0 (999=100%) 23 in Tutorial 162.0 100% 999
Match
0%
25%
50%
75%
100%
Figure 15.33 Mass spectral match for 2,4-dichlorophenol at 5 ppm in the #2Diesel/10W30 extract against the mass spectrum for 2,4-dichlorophenol in the NIST database.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
474 Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
475
Technology Progress and Application in GC/MS and GC/MS/MS TIC Filtered
45:450 4
MCounts
3
2
1
0
10
15
20 Minutes
25
30
Figure 15.34 A TIC of the EPA 8270 standard mixture spiked into pure dichloromethane solvent at a concentration of 5 ppm. The standard mixture contains 69 components that are analyzed commonly by this method.
TIC Filtered
45:450 4
MCounts
3
2
1
0 5
10
15
20 Minutes
25
30
35
Figure 15.35 A TIC of the same standard mixture as in Figure 15.34 but spiked with 3000 ppm of mixed diesel #2/motor oil extract. This motor oil extract is typical of the low molecular weight hydrocarbons found in contaminated soil extracts from environmental samples.
476
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 4000_dh_nodiesel.mth: 240-MS/4000.56: Fluoranthene Internal standard analysis Curve fit linear, origin: Ignore, weight: 1/n×2 Resp.Fact. RSD: 10.87%, Coeff.Det.(r2): 0.996926 y = +1.6164×–0.0235 1 1 Replicates P e 5 a k 4 s 3 i z 2 e 1 / S 0 t 0.5 1.0 1.5 2.0 2.5 3.0 d Amount/amt.std. (ng/uL)
1
3.5
Figure 15.36 Fluoranthene calibration curve obtained with the ion trap mass spectrometer. Standard solutions of fluoranthene in dichloromethane were prepared at concentrations of 5 ppm, 50 ppm, and 150 ppm. 4000_dh_diesel.mth: 240-MS/4000.56: Fluoranthene Internal standard analysis Curve fit linear, origin: Ignore, weight: 1/n×2 Resp. Fact. RSD: 3.665%, Coeff.Det.(r2): 0.999629 y = +1.5579×–0.0086 1 Replicates 1 P e 5 a 4 k s 3 i 2 z e 1 / S 0 t 0.5 1.0 1.5 2.0 2.5 3.0 d Amount/amt.std. (ng/uL)
1
3.5
Figure 15.37 Fluoranthene calibration curve obtained with the ion trap mass spectrometer under identical conditions to the curve shown in Figure 15.32 but where the standard solutions of fluoranthene were spiked with 3000 ppm of mixed diesel#2/motor oil.
15.3.3.2 Crude Vegetable Extracts In this type of analysis, a vegetable sample extract is prepared by cryo-milling followed by a micro-liquid-liquid extraction procedure known as QuEcHERS [60]. Very little sample extract clean-up is performed and, as a result, the extracts present a significant challenge to any analytical GC/MS system. Figure 15.38 shows a TIC of a pesticides’ calibration standard at 5 ppb prepared in a highly-colored (green) mixed vegetable extract. Despite the obviously large amount of matrix, pesticides
477
Technology Progress and Application in GC/MS and GC/MS/MS 45:450 8
Crude vegetable extract spiked with 5ppb various pesticides
7 6 MCounts
TIC Filtered
5 4 3 2 1 0 10
15
20 Minutes
25
30
Figure 15.38 Full mass scan TIC of a crude vegetable extract spiked with 5 ppb of a pesticides’ mixture on an ion trap instrument. At this low concentration level, the pesticide peaks are buried beneath the elevated baseline due to the presence of matrix.
Figure 15.39 Extracted ion profile taken from the TIC shown in Figure 15.34 at m/z 314 representing the compound chlorpyriphos. The full mass scan spectrum of chropyriphos is shown in the bottom part of the figure.
can be identified and quantified readily. For example, Figure 15.39 shows one of the pesticides, chlorpyriphos, at 5 ppb in the mixed vegetable extract; the extracted ion at m/z 314 is shown together with the full scan mass spectrum of chlorpyriphos for clear identification. Calibration curves are prepared typically in-matrix for this
478
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V pcb ext_dutch_final_dh.mth: 240-MS/4000.56: Chlorpyrifos Internal standard analysis Curve fit: Linear, origin: Ignore, weight: 1/n×2 Resp. Fact. RSD: 6.585%, Coeff.Det.(r2): 0.998892 y = +0.1333×–5.9803e-4 Replicates 2 1 1 1 1 P e a k s i z e / P S S t d
1
1.5 1.0 0.5 0.0 2.5
5.0 7.5 10.0 Amount/amt. std. (ppb)
12.5
Figure 15.40 Calibration curve for chlorpyriphos in a mixed fruit and vegetable matrix from 5 ppb to 500 ppb by ion trap GC/MS. The %RSD over the range is 6.6 % with r2 = 0.998.
method, because the matrix will passivate the injector liners and columns causing variable response factors between standards prepared in pure solvent and those prepared in-matrix. Figure 15.40 demonstrates the quantitative ability of the ion trap in this analysis for chlorpyrifos prepared in-matrix. Another example of excellent spectral integrity under these demanding conditions is illustrated in Figure 15.41. The ion of m/z 200 for atrazine is extracted from the TIC and the background subtracted mass spectrum is a good match to the reference mass spectrum. 15.3.3.3 Polybrominated Diphenyl Ethers (PBDEs) Another MS/MS application for ion trap mass spectrometry involves the analysis of polybrominated flame retardants (PBDEs). As a result of regulations promulgated during the past five years, such as the Restriction of Hazardous Substances (RoHS) [61] and other similar programs, there has been a determined effort to decrease the use of brominated flame retardants in the manufacture of consumer products. The chemical structure for a typical PBDE is illustrated in Figure 15.42. The challenge in this analysis lies in the efficient extraction and analysis of the complex matrices in which PBDEs are found. The compounds are used in a variety of plastic materials found in consumer products, and samples are prepared typically using either Soxhlet extraction or microwave digestion, followed by few or no clean-up steps. MS/MS can be used to isolate M + and [M-2Br] + isotopic cluster ions for accurate identification and quantitation. Three stages of the MS/ MS analysis of a PBDE are shown in Figure 15.43. Figure 15.44 shows a TIC of the common PBDE isomers that are tested for under the RoHS regulations; these common PBDE isomers are listed in Table 15.2. Excellent quantitative results
0.0
0.5
1.0
1.5
2.0
2.5
11.2
11.3
11.4
Start integration
Ions: 200.0 Merged
11.5
Start search
11.6
11.7 Minutes
11.8
End search
11.9
12.0
End integration
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
0%
25%
50%
75%
100%
138 865 201 258
202 1226
104 109
138 180
202 273
300
100
200
300
100 200 300 Raw sample spectrum for scan: 889 RT: 11.662 min. BP 200 (3915=100%) matrixcal5_.sms 200 91 3915 3376 55 2319 105 202 1552 57 1307 281 109 159 102 223 676 575 521 299
68 175
200 999
100 200 Reference spectrum for atrazine Scan: 1021 RT: 11.628 min. 500ppb-7 .sms
87 979
Sample spectrum for scan: 889 RT: 11.662 min. BC BP 200 (3865=100%) matrixca15_.sms 200 3865
341 508
373 350
400
415 574
400
400
415 358
m/z
m/z
m/z
Figure 15.41 Full scan mass spectrum of atrazine within a matrix at a concentration of 5 ppb. The left portion of the figure shows an extracted ion chromatogram at m/z 200. The top mass spectrum is the background corrected mass spectrum taken from the extract; the middle mass spectrum is the reference mass spectrum from a pure standard; and the bottom mass spectrum is that of the uncorrected raw sample.
K Counts
3.0
Technology Progress and Application in GC/MS and GC/MS/MS 479
480
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V Br Br
O
Br
Br Br
Figure 15.42 Structure of 2,2′,4,4′,6-Pentabromodiphenylether (PBDE-100). There are 209 possible PBDE congeners; each PBDE congener has been assigned a specific number (1–209).
Figure 15.43 An example of the isolated molecular ion cluster for a heptabromodiphenylether congener. Multiple resonant frequencies in the ion trap can be used to achieve dissociation of the entire cluster. The approach is highly sensitive and selective.
in matrix are achieved readily using ion trap mass spectrometry. Figure 15.45a and b represent the detection of BDE 209 (deca-brominated BDE) in an extract of acrylonitrile butadiene styrene (ABS) plastic. After 100 total injections of the matrix, the analytical system continues to show excellent reproducibility and stability. The reporting limits for this MS/MS method were 1 ppb (extract concentration) for all PBDEs, except for Deca-BDE (209) for which the reporting limit was 10 ppb.
481
Technology Progress and Application in GC/MS and GC/MS/MS TIC 406.0>200... 486.0>... 566... 644.0>400... 724.0>... 802.0>500:8... 880.0>600... 960.0>700:970 [2...
100 4
KCounts
75 50
10
5–1 5–2 6–1
3
7
6–2
8
25
9
0 7
8
9
10 Minutes
11
12
13
Figure 15.44 PBDE MS/MS Chromatogram, 50 ppb concentration (100 ppb for BDE 209). See Table 15.2 for peak names.
Table 15.2 The Analysis of PBDE Congeners in Accordance with the RoHS Regulation. The Peak Number in the Table Refers to the Peaks Labeled in Figure 15.44 Peak Number 3 4 5–1 5–2 6–1 6–2 7 8 9 10
PBDE Isomer BDE 28 BDE 47 BDE 99 BDE 100 BDE 153 BDE 154 BDE 183 BDE 205 BDE 206 BDE 209
15.3.3.4 Ion Trap Analysis with Liquid Chemical Ionization (CI) Reagents: USEPA Method 521 Ion traps can be configured readily to use liquid reagents for CI directly inside the ion trap cavity. Low vapor pressure liquids, such as acetonitrile (CH3CN) or methanol (CH3OH), can be used as CI reagents. The sequence of reactions is as follows:
CH3CN + e – → CH3CN+• + 2e –
(15.2)
482
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a)
(b)
RoHS Matrix 4, taken between 10 other matrix injections RSD 10.8% Raw Area
Peak Area
150000 100000
Series1
50000 0
1
2
3
4 5 6 7 Injection Number
8
9
10
Figure 15.45 Decabromodiphenyl ether, congener BDE-209: (a) upper, total ion chromatograph showing detection of BDE-209 in an acrylonitrile butadiene styrene (ABS) plastic extract; (a) lower, product ion mass spectrum of BDE-209 showing the loss of Br2 from each of the ions of the molecular ion cluster; (b) %RSD in the chart is obtained from raw peak area data for a standard solution run after 10 injections of the ABS extract.
CH3CN+• + CH3CN → CH3CNH+ + CH2CN•
(15.3)
CH3OH + e – → CH3OH2+• + 2e –
(15.4)
CH3OH +• + CH3OH → CH3OH2+ + CH3O•
(15.5)
where the protonated molecules CH3CNH + and CH3OH2+ are CI reagent ions. The basic steps are: (a) form and trap reagent ions in the ion trap; (b) remove unwanted ions formed by EI using applied waveforms, leaving only CI reagent ions
483
Technology Progress and Application in GC/MS and GC/MS/MS
Table 15.3 The Accuracy and Precision of the Seven Nitrosamines Listed in EPA Method 521 Using the CI/MS/MS Technique Compound Name NDMA NMEA NDEA NDPA NPYR NPIP NDBA
Meas.Conc. (ppb)
Accuracy (%)
0.934 0.986 0.997 1.056 0.904 0.924 0.896
93.4 98.6 99.7 105.6 90.4 92.4 89.6
Precision (RSD%) 7.30 11.49 3.08 10.09 5.08 4.55 5.42
Note: These data are based upon 11 injections of a standard solution of each Nitrosamine at a concentration of 1 ppb.
in the ion trap; (c) reagent ions react with sample to form protonated molecules or adduct ions; and (d) ions ejected in sequential mass order to form the mass spectrum. The technique is applicable particularly to compounds that yield multiple fragments (thus low response) under normal EI conditions. It is also useful for identifying and/ or confirming the molecular weight of a compound. USEPA Method 521 is used to determine the concentration of certain nitrosamines in source water and finished drinking water. Early health effects or toxicity data suggest that nitrosamines, in general, are powerful carcinogens [62]. Therefore, there is a need for an analytical method for the detection of nitrosamines at very low concentrations (parts-per-trillion levels or ppt) in finished drinking waters and source waters. Nitrosamines under normal EI conditions have very poor response, with multiple fragment ions being formed of low mass/charge ratio. CI provides protonated ions with much better response, and MS/MS adds specificity to the analysis. Nitrosamines provide a good example of how a specific ion trap technique can solve a tough analytical problem. These compounds are extracted from the water sample using a form of activated carbon, so matrix interference can become significant in waters that contain a high total organic carbon (TOC) content. The data shown in Table 15.3 illustrate the excellent accuracy and precision of the CI/MS/MS technique for the seven nitrosamines at a concentration of 1 ppb. These data are based upon 11 injections of a standard solution. A TIC of typical target compounds, to which the CI/MS/MS technique is applied, is shown in Figure 15.46. 15.3.3.5 Analysis of Polychlorinated Biphenyls (PCBs) by Ion Trap Mass Spectrometry Polychlorinated biphenyls (PCBs) exist as 209 individual congeners, exhibiting all the variations of position and number of chlorine substitutions in a biphenyl molecule.
484
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
250
NPYR
NMEA
TIC NDBA
NDEA
200
NPIP
NMOR
kCounts
150 100 50
NDMA NDMA-d6(Surr.)
0 15
NDPA NDPA-d14((IS)
20
Minutes
25
30
Figure 15.46 CI/MS/MS TIC of some common nitrosamines listed in USEPA 521. Extract concentration is 50 ppb for each nitrosamine.
Each PCB congener is named according to the positions of chlorine substitution on the two phenyl rings of biphenyl, and the toxicity of PCBs correlates strongly with their structures. The toxicity of PCB mixtures is due principally to a small group of non-ortho and mono-ortho-substituted congeners [63]. USEPA Method 505 [64] is used for the determination of PCBs and other pesticides in both ground and surface water. A sample is prepared by extracting a small volume, typically 40 mL, with 2 mL of hexane. The hexane layer is removed and analyzed by gas chromatography. The method uses electron capture detection (ECD) for added sensitivity for halogenated compounds. Although ECD is a very sensitive detector for this analysis, it is prone to matrix interference and can result in false positive identification due to co-eluting peaks. Mass spectrometry can overcome this problem, however the sensitivity of the technique in full mass scan or SIM can be challenging in relation to the micro-extraction sample preparation procedure described in EPA Method 505. Current ion trap mass spectrometers, using a full mass scan for many compounds, can obtain very similar sensitivity to that achievable with ECD. Figure 15.47 illustrates the detection and identification, by ion trap mass spectrometry, of PCB isomer 2,3-dichlorobiphenyl at a concentration of 0.05 ppb (1.75 pg injected on-column). The PCBs were extracted using the micro-extraction technique listed in USEPA Method 505. Tables 15.4 and 15.5 illustrate that the compounds can be detected at levels similar to those attainable with ECD with excellent precision and accuracy. A set of replicates, spiked at low concentrations in the aqueous sample (0.025 to 0.25 ppb) of polychlorinated biphenyl (PCB) congeners in reagent and surface
Technology Progress and Application in GC/MS and GC/MS/MS
485
Figure 15.47 2,3-Dichlorobiphenyl at a concentration of 0.05 ppb in the aqueous sample (0.75 pg injected on-column). The upper half of the figure is an extracted ion mass chromatogram for m/z 152 + m/z 222. The lower half of the figure is a full scan mass spectrum of the same peak.
Table 15.4 Calibration Ranges Studied by Ion Trap Mass Spectrometry PCBs mono, di and tri tetra, penta and hexa hepta and octa deca
Aqueous Concentration (ppb)
Conc. on Column (pg)
0.025–5 0.05–10 0.075–15 0.125–25
0.875–175 1.75–350 2.625–525 4.375–875
Note: The aqueous concentration refers to the final concentration of the listed PCB isomers spiked into laboratory reagent water and extracted using the procedure described in EPA method 505. The second column lists the actual amount of the PCB isomers that were injected into the analytical system from the resulting aqueous extractions.
waters, was extracted and run in EI full mass scan mode for the determination of Method Detection Limits (MDL). The MDL in USEPA Method 505 is calculated based on the standard deviation of quantitative results multiplied by Student’s t at 99% confidence level [65].
486
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
Table 15.5 Method Detection Limits (MDLs) of PCBs by Full Scan Ion Trap Mass Spectrometry in Reagent Water and Surface Water PCBs
Reagent Water (n = 8)
Surface Water (n = 16)
mono di tri tetra penta hexa hepta octa deca
0.010 0.011 0.010 0.009 0.017 0.016 0.014 0.017 0.082
0.010 0.015 0.014 0.024 0.053 0.040 0.068 0.085 0.221
Note: The units are μg L–1. A total of eight spiked samples for reagent water and 16 spiked samples for surface water were used in the statistical calculations based upon the MDL procedure outlined in EPA Method 505.
15.4 SUMMARY Since the middle 1990s, numerous efforts on fully-AMDs have made GC/MS and GC/MS/MS ion trap mass spectrometers to be practical and mature commercial products. Since then, GC/MS and GC/MS/MS ion traps have grown from novel research instruments into routine, widely-applied analytical instruments. GC/MS and GC/MS/MS ion trap instruments have reached a high level of maturity as they have grown in the directions of higher performance, such as faster scan speed, higher mass resolution, and extended charge capacity. The advent of the high-performance linear ion trap has provided the opportunity for GC/MS and GC/MS/MS to grow even further in the future. The GC/MS applications discussed above demonstrate clearly that ion traps provide excellent data for applications, despite a history of poor performance in early ion trap designs. Qualitative and quantitative analysis in heavy matrices are possible because increased ion trapping capacity and ion population control is available in modern instrumentation. The technology has been accepted for use with major USEPA methods as a routine analytical tool for challenging environmental samples. The analyzer is versatile, because scan modes such as MS/MS, liquid CI, hybrid CI, and full scan mass spectrometry can be performed on the same instrument. This instrumental versatility reduces cost and increases specificity by providing more information about the molecules under study.
ACKNOWLEDGMENTS The authors would like to thank Dr. Barbara Bolton, Dr. Kenneth Newton, and Dr. Haibo Wang for their helpful discussions and useful data.
Technology Progress and Application in GC/MS and GC/MS/MS
487
References
1. Paul, W.; Steinwedel, H. Apparatus for Separating Charged Particles of Different Specific Charges, US Patent 1960, 2,939,952. 2. March, R.E.; Todd, J.F.J. Quadrupole Ion Trap Mass Spectrometry, Second edition of Quadrupole Storage Mass Spectrometry. Chemical Analysis Series, vol. 165. John Wiley, Hoboken, New Jersey, 2005. 3. Stafford, G.C., Jr.; Kelley, P.E.; Stephens, D.R. Method of Mass Analyzing a Sample by Use of a Quadrupole Ion Trap, US Patent 1985, 4,540,884. 4. Stafford, G.C., Jr.; Kelley, P.E.; Syka, J.E.P.; Reynolds, W.E.; Todd, J.F.J. Recent improvements in and analytical applications of advanced ion trap technology. Int. J. Mass Spectrom. Ion Processes. 1984, 60, 85–98. 5. Kelley, P.E.; Stafford, G.C. Jr.; Syka, J.E.P.; Reynolds, W.E.; Louris, J.N., Amy, J.W.; Todd, J.F.J. New advances in the operation of the ion trap mass spectrometer. Proc. 33rd ASMS Conference on Mass Spectrometry and Allied Topics, San Diego, CA, 1985, 707–708. 6. Syka, J.E.P.; Louris, J.N.; Kelley, P.E.; Stafford, G.C. Jr.; Reynolds, W.E. Method of Operating Ion Trap Detector in MS/MS Mode, US Patent 1988, 4,736,101. 7. Louris, J.N.; Schwartz, J.C.; Stafford, G.C. Jr.; Syka, J.E.P.; Taylor, D.M. The Paul ion trap mass selective instability scan: trap geometry and resolution. Proc. 40th ASMS Conference on Mass Spectrometry and Allied Topics, Washington DC, 1992, 1003–1004. 8. Franzen, J. Simulation study of an ion cage with superimposed multiple fields. Int. J. Mass Spectrom. Ion Processes. 1991, 106, 63–78. 9. Splendore, M.; Marquette, E.; Oppenheimer, J.; Huston, C.; Wells, G.J. A new ion ejection method employing an asymmetric trapping field to improve the mass scanning performance of an electrodynamic ion trap. Int. J. Mass Spectrom. 1999, 190/191, 129–143. 10. Wells, G.J.; Wang, M.; Marquette, E.G. Mass Scanning Method Using an Ion Trap Mass Spectrometer, US Patent 1998, 5,714,755. 11. Schwartz, J.C.; Senko, M.W.; Syka, J.E.P. A two-dimensional quadrupole ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 2002, 13, 659–669. 12. Stafford, G.C. Jr.; Taylor, D.M. Method of Increasing the Dynamic Range and Sensitivity of a Quadrupole Ion Trap Mass Spectrometer, US Patent 1992, 5,107,109. 13. Schwartz, J.C. Do space charge effects limit LC quadrupole ion trap performance. Proc. 9th Sanibel Conference on Mass Spectrometry, Sanibel Island, FL, 1997. 14. Brekenfeld, A.; Kaplan, D.A.; Hartmer, R.; Wilson, J.; Gebhardt, C.; Schubert, M. The right place and the right time: increasing the capacity in modern 3-D ion traps. Proc. 55th ASMS Conference on Mass Spectrometry and Allied Topics, Indianapolis, IN, 2007. 15. Mordehai, A.; Miller, B.; Bai, J.; Brekenfeld, A.; Baessmann, C.; Schubert, M.; Hosea, K. Improved 3D ion trap-ion detector coupling and techniques for evaluating exact ion trap capacity. Proc. 52nd ASMS Conference on Mass Spectrometry and Allied Topics, Nashville, TN, 2004. 16. George, J.E.; III, Specht, A.; Newton, K.; Wells, G. A comparison of the spectral charge capacity between two ejection methods. Proc. 54th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, 2006. 17. Libong, D.; Bruneau, S.P.C.; Rogalewicz, F.; Ricordel, I.; Bouchonnet, S.J. Adsorptiondesorption effects in ion trap mass spectrometry using in situ ionization. Chromatog. 2003, 1010 (issue 1), 123–128. 18. Taylor, D.M.; Amy, J.W.; Stafford, G.C. Jr. Metal Surfaces for Sample Analyzing and Ionizing Apparatus, US Patent 1991, 5,055,678.
488
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
19. Brittain, R.; Wang, M. Surface Coating to Improve Performance of Ion Trap Mass Spectrometers, US Patent 1997, 5,633,497. 20. McLuckey, S.A.; Glish, G.L.; Asano, K.G.; Van Berkel, G.L. Self chemical ionization in an ion trap mass spectrometer. Anal. Chem. 1988, 60, 2312–2314. 21. Booth, M.M.; Stephenson, J.L.J.; Yost, R.A. Gas chromatography/ion trap mass spectrometry using an external ion source. Proc. 41st ASMS Conference on Mass Spectrometry and Allied Topics, San Francisco, CA, 1993, 716a–716b. 22. Bier, M.E.; Syka, J.E.P.; Taylor, D.M.; Fies, W.J. Ion Source Assembly for an Ion Trap Mass Spectrometer and Method, US patent 1998, 5,756,996. 23. Wells, G J.; Yee, P.P.; Ruport, M.A.; Huston, C.K. Pulsed Ion Source for Ion Trap Mass Spectrometer, US Patent 2001, 6,294,780. 24. Louris, J.N.; Syka, J.E.P.; Kelley, P.E. Method of Operating Quadrupole Ion Trap Chemical Ionization Mass Spectrometry, US Patent 1987, 4,686,367. 25. Strife, R.J.; Keller, P.J. Ion trap ionization mass spectrometry-RF/DC for isolating unique reactant ions. Org. Mass Spectrom. 1989, 24(3), 201–204. 26. March, R.E.; Todd, J.F.J. eds. Practical Aspects of Ion Trap Mass Spectrometry. Vol. 3, “Chemical, Environmental, and Biomedical Application”, Ch. 7, “Chemical Ionization in Ion Trap Mass Spectrometry”, p. 239–253, by Creaser, C.S. Modern Mass Spectrometry series, CRC Press, Roca Baton, FL, 1995. 27. Berberich, D.W.; Heil, M.V.; Johnson, J.V.; Yost, R.A. Mass-selection of reactant ions for chemical ionization in quadrupole ion traps and triple quadrupole mass spectrometers. Int. J. Mass Spectrom. Ion Processes. 1989, 94, 115–147. 28. Van Pelt, C.K.; Carpenter, B.K.; Brenna, J.T. Studies of structure and mechanism in acetonitrile chemical ionization tandem mass spectrometry of polyunsaturated fatty acid methyl esters. J. Am. Soc. Mass Spectrom. 1999, 10, 1253–1262. 29. Brandt, S.D.; Freeman, S.; Fleet, I.A.; Alder, J.F. Analytical chemistry of synthetic routes to psychoactive tryptamines Part III. Characterisation of the Speeter and Anthony route to N,N-dialkylated tryptamines using CI-IT-MS-MS. Analyst 2005, 130, 1258–1262. 30. Hunt, D.F.; Stafford, G.C.; Crow, F.W.; Russell, J.W. Pulsed positive negative ion chemical ionization mass spectrometry. Anal. Chem. 1976, 48, 2098–2104. 31. March, R.E.; Todd, J.F.J. eds. Practical Aspects of Ion Trap Mass Spectrometry. Vol. 3, Ch. 2, “Ion Trap as Tandem Mass Spectrometers”, pp. 27–88, by March, R.E.; Strife, R.J.; Creaser, C.S. Modern Mass Spectrometry series, CRC Press, Roca Baton, FL, 1995. 32. March, R.E.; Todd, J.F.J. eds. Practical Aspects of Ion Trap Mass Spectrometry. Vol. 3, Ch. 4, “Practical Ion Trap Technology: GC/MS and GC/MS/MS”, pp. 121–185, by Yates, N.A.; Booth, M.N.; Stephenson, J.L., Jr.; Yost, R.A. Modern Mass Spectrometry series, CRC Press, Roca Baton, FL, 1995. 33. Wells, G.J. Quadrupole Trap Improved Technique for Ion Isolation, US Patent 1993, 5,198,665. 34. Marshall, A.G.; Ricca, T.L.; Wang, T.L. Tailored Excitation for Trapped Ion Mass Spectrometry, US Patent 1988, 4,761,545. 35. Louris, J.N.; Taylor, D.M. Method and Apparatus for Ejecting Unwanted Ions in an Ion Trap Mass Spectrometer, US Patent 1994, 5,324,939. 36. Kelley, P.E. Mass Spectrometry Method Using Notch Filter, US Patent 1992, 5,134,286. 37. Buttrill, S.E. Jr. Quadrupole Ion Trap Method Having Improved Sensitivity, US Patent 1994, 5,300,772. 38. Wang, M.; Lee, D.; Newton, K.; Schachterle, S. High-Resolution Ion Isolation Utilizing Broadband Waveform Signals, US Patent 2008, 7,378,648. 39. Schwartz, J.C.; Syka, J.E.P.; Quarmby, S.T. Improving the fundamentals of MSn on 2D linear ion traps: new ion activation and isolation techniques. Proc. 53rd ASMS Conference on Mass Spectrometry and Allied Topics, San Antonio, TX, 2005.
Technology Progress and Application in GC/MS and GC/MS/MS
489
40. Salmon, K.; He, M.; Choudhary, G.; Schwartz, J.; Cho, D. Improved isolation efficiency using higher resolution isolation in an ion trap mass spectrometer. Proc. 54th ASMS Conference on Mass Spectrometry and Allied Topics, Seattle, WA, 2006. 41. Wells, G.J. Quadrupole Trap Improved Technique for Collisional Induced Disassociation for MS/MS Process, US Patent 1994, 5,302,826. 42. Jackson, G.P.; Hyland, J.J.; Laskay, U.A. Energetics and efficiences of collision-induced dissociation achieved during the mass acquisition scan in a quadrupole ion trap. Rapid Commun. Mass Spectrom. 2005, 19, 3555–3563. 43. Schwartz, J.C.; Taylor, D.M. Method of Ion Fragmentation in a Quadrupole Ion Trap, US Patent 2000, 6,124,591. 44. Mulholland, J.J.; Yost, R.A. Multi-level CID: a novel approach for improving MS/MS on the quadrupole ion trap. Proc. 47th ASMS Conference on Mass Spectrometry and Allied Topics, Dallas, TX, 1999. 45. Salmon, K.; Choudhary, G.; Schwartz, J.; Cho, D. Enhanced fragmentation of small molecules in a linear ion trap mass spectrometer using stepped normalized collision energy. Proc. 53rd ASMS Conference on Mass Spectrometry and Allied Topics, San Antonio, TX, 2005. 46. Brekenfeld, A.; Schubert, M.; Franzen, J. Fragmentation in Quadrupole Ion Trap Mass Spectrometers, US Patent 2002, 6,410,913. 47. Goodley, P.C. Technical Note, 5988-0704EN, 2000, Agilent Technologies. 48. Wang, M. Chemical Structure Insensitive Method and Apparatus for Dissociating Ions, US patent application, publication pending. 49. Cunningham, C., Jr.; Glish, G.L.; Burinsky, D.J. High amplitude short time excitation: a method to form and detect low mass product ions in a quadrupole ion trap mass spectrometer. J. Am. Soc. Mass Spectrom. 2006, 17, 81–84. 50. Schwartz, J.C. High-Q Pulsed Fragmentation in Ion Trap, US Patent 2005, 6,949,743. 51. SW-846, Method 8270D, Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, available from: National Technical Information Services, US Department of Commerce, 5285 Port Royal Road, Springfield, VA 22161. 52. SDWA: Safe Drinking Water Act, EPA 816-F-04-030, June 2004, “Understanding the Safe Drinking Water Act”. 53. 525.2 Methods for the Determination of Organic Compounds in Drinking WaterSupplement III (EPA/600/R-95-131). Citation Information Methods for the Determination of Organic Compounds in Drinking Water-Supplement III (EPA/600/R-95-131). This document is available through NTIS (http://www.ntis.gov). Alternatively, the methods from this source can be found on the following CD-ROM: EPA Methods and Guidance for Analysis of Water, Version 2.0. 54. EPA Method Guidance CD-ROM (includes MCAWW Methods, and most current EPA Methods) Citation Information EPA Methods and Guidance for Analysis of Water, Version 2.0 ’ This CD-ROM includes all EPA wastewater test methods approved at 40 CFR 136, all EPA drinking water test methods approved at 40 CFR 141, and various EPA guidance documents related to EPA’s wastewater and drinking water programs. New and revised EPA OW methods and guidance documents will be added to the CD-ROM during periodic updates. Web at: http://www.ntis.gov/product/environmental-test-methods.htm 55. Price, E.K.; Prakash, B.; Domino, M.M.; Pepich, B.V.; Munch, D.J. 2005, Determination of selected pesticides and flame retardants in drinking water by solid phase extraction and capillary column gas chromatography/mass spectrometry: U.S. Environmental Protection Agency Report EPA/815/R-05/005, Version 1.0. 56. Methods for the Determination of Organic and Inorganic Compounds in Drinking Water, Volume 1 (EPA/815-R-00-014) Citation Information Available through: NSCEP Item # 815-R-00-014, (800) 490-9198 or (513) 489-8190 or order from the Web at: http://www. epa.gov/ncepihom/
490
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
57. Munch, J.W. Revision 1.0, Sept. 2002, NERL Method 529: Determination of Explosives and Related Compounds in Drinking Water by Solid Phase Extraction and Capillary Column Gas Chromatography/Mass Spectrometry (GC/MS). 58. Munch, J.W. Revision 1.0, Sept. 2004, NERL Method 521: Determination of Nitrosamines in Drinking Water by Solid Phase Extraction and Capillary Column Gas Chromatography with Large Volume Injection and Chemical Ionization Tandem Mass Spectrometry (MS/MS). 59. SW-846, Method 8270D, Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, available from: National Technical Information Services, US Department of Commerce, 5285 Port Royal Road, Springfield, VA 22161 60. Anastassiades, M.; Lehotay, S.J.; Stajnbaher, D.; Schenck, F. Fast and easy multiresidue method employing acetonitrile extraction/ partitioning and dispersive solid phase extraction for the determination of pesticide residues in produce. (QuEChERS method). J. AOAC Int. 2003, 86, 412–431. 61. The Directive on the Restriction of the Use of Certain Hazardous Substances in Electrical and Electronic Equipment 2002/95/EC[1] http://eur-lex.europa.eu/LexUriServ/Lex UriServ.do?uri = OJ:L:2003:037:0019:0023:EN:PDF 62. Lijinsky, W.; Epstein, S.S. Nitrosamines as Environmental Carcinogens. Eppley Institute for Research in Cancer, University of Nebraska College of Medicine, www.nature.com/ nature/journal/v225/n5227/abs/225021a0.html 63. Tanabe, S. PCB Problems in the Future: Foresight from Current Knowledge. Environ. Pollut. 1988. 50, 5–28. 64. Method 505: Analysis of Organohalide Pesticides and Commercial Polychlorinated Biphenyls (PCB) Products in Water by Microextraction and Gas Chromatography, Winfield, T.W.; Munch, J.W. 1995. National Environmental Research Laboratories, Office of Research and Development, USEPA, Cincinnati, OH 45268. 65. Method 505: Analysis of Organohalide Pesticides and Commercial Polychlorinated Biphenyls (PCB) Products in Water by Microextraction and Gas Chromatography, Winfield, T.W.; Munch, J.W. 1995, pp.18–19. Note: EPA Method 505 from: Methods for the Determination of Organic Compounds in Drinking Water-Supplement III (EPA/600/R-95-131). This document is available through NTIS (http://www.ntis.gov). Alternatively, the methods from this source can be found on the following CD-ROM: EPA Methods and Guidance for Analysis of Water, Version 2.0.
Monitoring 16 Remote of Volatile Organic Compounds in Water by Membrane Inlet Mass Spectrometry Romina Pozzi, Paola Bocchini, Francesca Pinelli, and Guido C. Galletti Contents 16.1 Introduction................................................................................................. 492 16.2 Membrane Inlet Mass Spectrometry (MIMS)............................................. 493 16.3 Membrane Inlet Mass Spectrometry (MIMS) Instrumentation for Prolonged Monitoring.................................................................................. 494 16.3.1 Experimental................................................................................. 494 16.3.2 Laboratory Tests............................................................................ 496 16.3.3 Field Tests...................................................................................... 496 16.4 Results and Discussion................................................................................ 497 16.4.1 Laboratory Tests............................................................................ 497 16.4.1.1 Detection Limits (LOD)................................................ 497 16.4.1.2 Reproducibility............................................................. 498 16.4.1.3 Linearity........................................................................ 498 16.4.1.4 Matrix Effects............................................................... 499 16.4.2 Case Studies................................................................................... 499 16.4.2.1 Acrylonitrile.................................................................. 499 16.4.2.2 Comparison of MIMS with Purge-and-Trap (P&T)/Gas Chromatography (GC)/ Mass Spectrometry (MS)..............................................500 16.4.3 Field Tests...................................................................................... 502 16.5 Conclusion................................................................................................... 505 References...............................................................................................................506
491
492
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
16.1 INTRODUCTION European Union Directive 98/83 underlines the importance of determining the quality of drinking water in order to protect human health. In particular, a number of chemical compounds are listed, such as benzene, 1,2-dichloroethane, tetrachloroethylene, chloroform, and trihalomethanes, whose concentration in drinking water must be kept under well-defined thresholds. In Italy, laws DL 31/01, DM 152/99, and DM 471/99 set the norms for the concentrations of these, and of many other compounds in drinking water, wastewater, and contaminated sites, respectively. Volatile Organic Compounds (VOCs) constitute a very important class of water pollutants because of their persistence; in addition, many of them are suspected of being carcinogenic. There are about 60 VOCs, including benzene, toluene, ethylbenzene, and xylenes (‘BTEX compounds’), halomethanes, and haloethanes. The presence of some of them in water is due to anthropic activities, for example, the use of chlorinated solvents in industries and laundries, and the formation of halomethanes as by-products of water disinfectants. With respect to Italian law DL 31/01, the maximum allowable concentration (threshold) for the sum of trichloroethylene and tetrachloroethylene concentrations in drinking water is 10 ppb, whereas the minimum account for the sum of a set of four halogenated compounds, namely chloroform, bromoform, bromodichloromethane, and chlorodibromomethane must be as low as possible and must not exceed 30 ppb. Note that 30 ppb is equivalent to 30 μg L –1. A real-time, on-line, continuous monitoring system for such compounds would allow either prompt actions to be taken in order to avoid the diffusion of pollutants into the water system or to take appropriate countermeasures, thus restoring safe conditions in the case of accidental contamination. In general, only the conventional chemical-physical parameters, such as dissolved oxygen temperature, pH, conductivity, and turbidity, are monitored continuously in water [1]. VOCs are usually analysed in the laboratory by means of Purge and Trap/Gas Chromatography/ Mass Spectrometry (P&T/GC/MS) using the U.S. Environmental Protection Agency (USEPA) Method No. 8260B which sets the standard for the analysis of VOCs in water. Although the method is state-of-the-art in terms of sensitivity, reproducibility, validation of the overall procedure and has been adopted worldwide by water laboratories, it can by no means be considered an alarm tool giving rapid warning of concentration increases. For an analytical procedure to be considered a warning device, it should be rapid, simple, and able to work unattended 24-hours-a-day for several days in unmanned sites and to send remotely analytical reports. As appropriately stated by Mikkelsen and coworkers, reporting upon a robust and sensitive on-line remote monitoring system for heavy metals in natural waters, “it is a great distance from developing a method (…) for continuous outdoor measurements” [2]. To our knowledge, little research has been made on the use of the ion trap for continuous, on-line monitoring of environmental parameters. Masuyoshi Yamada et al. studied a continuous monitoring system for the determination of polychlorinated biphenyls in air; the system employed direct sampling atmospheric pressure chemical ionization (APCI)/ion trap mass spectrometry (ITMS) [3]. Direct sampling ion trap mass spectrometers with two direct sampling interfaces, developed at Oak Ridge National Laboratory, TN, USA,
Remote Monitoring of Volatile Organic Compounds in Water
493
have been tested in field studies to determine VOCs in the effluents from hazardouswaste incinerators [4]. Kurten et al. developed an ion trap mass spectrometer for the on-line chemical analysis of atmospheric aerosol particles [5].
16.2 MEMBRANE INLET MASS SPECTROMETRY (MIMS) Riter et al. applied Membrane Inlet Mass Spectrometry (MIMS) coupled to a miniature mass spectrometer equipped with a cylindrical ion trap (CIT) analyzer to monitor the flavor components directly from human breath [6]. Johnson et al. measured ethanol concentrations on-line in fermentation broths from a 9000-L fermentation reactor for a period of four days [7]. However, data reported in the above papers referred to experiments lasting no more than a few days. MIMS has been extensively studied for the determination of VOCs in various environmental matrices, especially water and air samples [8–16]. Ketola and coworkers published a review that listed 172 references of MIMS applications to water and air [17]. MIMS allows the introduction of VOCs to the mass spectrometer through a thin (some tenths of a millimeter) hollow-fiber polymeric membrane, which is selective toward organic compounds. When the membrane is in contact with the sample and an ion trap mass spectrometer is used as the detector, such as in the case here, VOCs are extracted into the membrane, concentrated in its small volume, and swept into the mass spectrometer by a gentle stream of helium carrier gas. The whole process is called pervaporation and is divided into three steps: (a) phase-partitioning equilibrium of the organic compound between the sample (water or air) and the membrane; (b) diffusion of the compound by a concentration gradient from the outer side of the membrane (in contact with the sample) to the inner side (connected to the mass spectrometer); and (c) evaporation of the compound from the inner side of the membrane [18]. Diffusion is the rate-determining step, whereas partitioning and evaporation can be considered to be instantaneous. When the membrane is exposed to a sample containing the target compounds and the ions characteristic of each target compound are detected by mass spectrometry, the inherent ion current increases up to a plateau showing that the analyte’s pervaporation rate and transport flow to the detector are equal. The pervaporation process can be described by the Fick’s equations of diffusion [18], that is,
∂ Cm ( x , t ) I m ( x , t ) = − AD ∂ x
(16.1),
∂ Cm ( x , t ) ∂ 2C m ( x , t ) = D ∂ t ∂ x2
(16.2)
where: Im = analyte flow through the membrane (mol s–1); Cm = concentration of the analyte in the membrane wall (mol cm–3); A = membrane surface (cm2); D = diffusion coefficient (cm2 s–1);
494
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
x = membrane thickness (cm); t = time (s). The characteristic ions are the qualitative information which allows identification of the analytes, while plateau height (Iss = ADCm /L, where Iss is the analyte flow and L is the membrane thickness) is the quantitative information, with sensitivity in the sub-ppb levels and dynamic range of up to four decades for many VOCs [17,19].
16.3 MEMBRANE INLET MASS SPECTROMETRY (MIMS) INSTRUMENTATION FOR PROLONGED MONITORING Although the many papers cited above have demonstrated that MIMS is a potentially excellent technique for continuous VOC monitoring given its simplicity and sensitivity, to our knowledge no account has been published of experimental attempts to demonstrate that MIMS can really be implemented in a device able to work unattended for months. Following our previous paper on MIMS upgrades [19], this present work reports on laboratory and field tests of hardware and software for MIMS instruments built in our laboratory. Four instruments were deployed in unmanned sites, where they monitored VOCs in natural waters and wastewater during a period exceeding one year for each instrument. The instruments were equipped with software that facilitated the automatic operation of each analysis, the identification and quantitation of VOCs from the raw mass spectra, and the transmission of the results to a remote control room via internet connection. In the remote control room, a personal computer with dedicated software displayed the results as bar graphs and was programed to activate alarms when set concentration thresholds were exceeded. Laboratory performance in terms of sensitivity, reproducibility, linearity tests, and comparison with P&T/GC/MS together with field performance in terms of data output, most frequent maintenance operations and technical failures, and overall stability of the four remotely-controlled instruments are discussed.
16.3.1 Experimental Table 16.1 lists the VOCs used in the present study together with their respective characteristic ions. All compounds were purchased from Sigma-Aldrich (St Louis, MO, USA). The MIMS system (Analytical Research Systems, Bologna, Italy) was equipped with a helium carrier gas cylinder (chromatography grade, SIAD, Milan, Italy), pressure regulator and a 30 m column with no stationary phase to provide a constant gas flow (1 mL min–1), sample cell, and hollow fiber membrane connected to a quadrupole ion trap mass spectrometer (Varian Inc., Walnut Creek, USA) through a fused silica column (0.32 mm ID, 5 m, Supelco) without a stationary phase. All mass spectra were acquired (5 min) from m/z 50 to 200 at a rate of 1 spectrum/5 s. The trap temperature was 170°C. The sample was kept under magnetic stirring at room temperature during the analysis. Each instrument had five sample inlets so that up to five different water streams could be analyzed. During normal operating conditions, two inlets were dedicated to blank water and calibration solutions, respectively. The device was operated by means of proprietary software able to set: (a) sampling, analytical, and data-transmission functions; (b) identification and
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Benzene Toluene Ethylbenzene Cumene Styrene 1,4-dichlorobenzene 1,2-dichlorobenzene Chloroform Trichloroethylene Tetrachloroethylene Carbon tetrachloride Bromoform Dibromochloromethane Dichlorobromomethane 1,1,1-Trichloroethane Acrylonitrile
Compound
78 91 91 + 106 77 + 105 + 120 78 + 104 146 + 148 + 150 146 + 148 + 150 83 130 + 132 164 + 166 117 + 119 173 129 83 96 + 97 52
Characteristic Ions (m/z) 0.05 0.3 0.1 9 0.2 0.2 0.2 0.03 0.03 0.08 0.1 0.20 0.1 0.1 0.09 40
LOD
MIMS
7 9 9 9 6 5 5 6 4 8 16 19 9 6 10 11
SD% 0.9996 0.9998 0.9998 0.9087 0.9922 0.9986 0.9986 0.9977 0.9984 0.9994 0.9965 0.9927 0.9986 0.9964 0.9982 1.0000
R2 0.25 0.25 0.1 0.25 0.5 0.5 0.5 0.13 0.5 0.5 1 0.5 0.5 0.5 0.5 /
LOD 3 5 6 6 6 6 6 4 4 7 5 8 5 7 7 /
SD%
P&T/GC/MS
0.9989 0.9949 0.996 0.9975 0.9981 0.9878 0.9878 0.998 0.9966 0.9951 0.9917 0.9969 0.9979 0.9977 0.9967 /
R2
0.2 0.55 0.3 0.75 0.2 0.15 0.15 0.15 0.95 0.7 1.05 0.6 0.25 0.4 0.4 /
LOD
USEPA 8260B
Table 16.1 Characteristic ions (m/z); MIMS and P&T/GC/MS Limit of Detection (LOD, ppb), Standard Deviation (SD %), and R2; USEPA Method 8260B LODs (ppb)
Remote Monitoring of Volatile Organic Compounds in Water 495
496
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
quantification functions; and (c) display and archiving of the results in the remote station. Verification of the status of the device, simple operations related to the control of the mass spectrometer, and checking of the raw results (for example, air/ water checks, tuning, view of the total ion current and mass spectra, etc.) were performed remotely by means of a commercial software package (Laplink Software Inc., Bellevue, WA, USA).
16.3.2 Laboratory Tests Limits of Detection (LOD) were determined by subsequent dilutions of standard solutions down to a signal-to-noise ratio, S/N, of ≥ 3. Signal reproducibility was determined by six replicates of analyte solutions with concentrations ten times larger than the LOD. Finally, linearity was calculated over a concentration range extending from the LOD to 20–100 times the LOD values. All analyses were performed with solutions freshly prepared immediately before use by appropriate dilution of mother solutions with organic-free triply-distilled water. In turn, mother solutions were prepared daily by dilutions of concentrated solutions of the analytes in methanol stored in a refrigerator except during the daily preparation of solutions. MIMS results were compared to those obtained by USEPA Method 8260B based on P&T/GC/MS. A Tekmar Velocity XPT Purge and Trap (Teledyne Tekmar, Mason, OH, USA) coupled to a Varian Star 3400X Saturn 2000 GC/MS (Varian, Palo Alto, CA, USA) was used under the following conditions: P&T Sample volume: 5 mL; Trap: Supelco Trap E (SP 2100/Tenax/Silica gel/Charcoal); Purge temperature: 30°C; Purge time: 11 min; Purge flow: 40 mL min−1; Desorbing temperature: 180°C; Desorbing time: 4 min; Desorbing flow: 300 mL min−1; Bake temperature: 180°C; Bake time: 10 min; Bake flow: 400 mL min−1; Transfer-line temperature: 150°C. GC/MS Column: Supelco SPB 624, 60 m x 0.32 mm ID, 1.8 µm film thickness; Injector temperature: 125°C; Oven temperature: from 35 to 50°C at 4°C min–1 holding the initial temperature for 2 min; then to 220°C at 10°C min–1 holding the final temperature for 10 min; Mass spectra: m/z 25–300 at 1 scan min–1.
16.3.3 Field Tests Four MIMS instruments were deployed for field tests in plants that produced water: one instrument was deployed in each of two plants that produced drinking water from
497
Remote Monitoring of Volatile Organic Compounds in Water
Table 16.2 MIMS Performance in Field Experiments Field Test Site A B C D
Application Ground Water Potabilization Ground Water Potabilization Surface Water Potabilization Industrial Wastewater Treatment
Total
Days On
Analyses /day
Total Analyses
Days Off
% Off
323
11
3
24
7752
492
24
5
48
23,616
510
37
7
24
12,240
526
20
4
3
1587
1,851
92
5 (aver.)
23 (aver.)
45,195
Note: A tabulation of the operational performance of the MIMS instruments with respect to functional days, non-functional days, the percentage of non-functional days, analyses/day, and total number of analyses for each of four sites.
ground water; one instrument was deployed in a plant that produced drinking water from surface water; and the fourth instrument was deployed in a plant for the treatment of industrial waters. All the plants were located in the area near Bologna. The instruments were programed to sample and to analyze water (analysis duration: 5 min; 1 scan per five seconds full scan of the mass range: m/z 50–200) with the frequency of analysis ranging from three analyses per day to two analyses per hour. Instrument performances were checked over a period ranging from 323 to 526 days (Table 16.2).
16.4 RESULTS AND DISCUSSION 16.4.1 Laboratory Tests The compounds used for the present study were chosen on the basis of the following criteria: halomethanes and haloethanes (compounds 8–15 in Table 16.1) are solvents and disinfection by-products; for compounds 8 and 2–4, the sum of the concentrations in drinking water must be less than 30 µg L –1 whereas for compounds 9 and 10, the threshold is 10 µg L –1; the remaining compounds (compounds 1–7 and acrylonitrile, Table 16.1) are of interest because they are often found in industrial wastewaters such as were used for the present study. 16.4.1.1 Detection Limits (LOD) MIMS detection limits (LOD, S/N ≥ 3) were determined by analysis of reference solutions and were compared with (a) LODs obtained by P&T/GC/MS operated as described in the previous section, and (b) LODs reported by USEPA Method 8260B (Table 16.1). MIMS LODs were (a) smaller than those obtained by P&T/GC/MS for the organohalogen compounds and for benzene by about, in some cases, one order of magnitude; (b) comparable to the other technique for toluene, ethylbenzene, styrene,
498
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
250
Counts
200 150
0.03 ppb
100 50 0 10
20
30
40 50 Minutes
60
70
80
90
FIGURE 16.1 A typical example of a temporal trace of the ion current for m/z 130-132 from trichlorethylene at a concentration of 0.03 ppb, showing the signal intensity and S/N ratio at the detection limit.
and for the two dichlorobenzene isomers (compounds 6 and 7); and (c) markedly higher for cumene. Acrylonitrile showed a high LOD probably due to its poor partitioning equilibrium in the membrane that, in turn, can be ascribed to its relatively high polarity. USEPA Method 603 reports 0.5 ppb as the detection limit for acrylonitrile, using Purge and Trap and gas chromatography with either a Porapak or a Chromosorb 101 packed column. Figure 16.1 shows the ion current (m/z 130–132) of a 0.03 ppb solution of trichloroethylene as a typical example of the signal intensity and S/N ratio at the detection limit. 16.4.1.2 Reproducibility The reproducibility (six replicates) of MIMS’ responses was compared to that obtained by the reference method. The results (expressed as standard deviation percentage, SD%, Table 16.1) were comparable for the two methods, with the exception of carbon tetrachloride and bromoform, whose MIMS standard deviations were greater than those obtained by P&T/GC/MS. The standard deviation percentage for compounds of relatively high polarity and/or low volatility (such as toluene, ethylbenzene, cumene, bromoform, and carbon tetrachloride) was relatively higher than that obtained by P&T/GC/MS; these results probably indicate an unfavorable partitioning equilibrium for these particular compounds in the membrane. 16.4.1.3 Linearity With respect to linearity (Table 16.1), MIMS and P&T/GC/MS were comparable (R2 > 0.99), with the exception of cumene, whose MIMS R2 was 0.9087, probably
499
Remote Monitoring of Volatile Organic Compounds in Water
due to the previously-mentioned lower partitioning equilibrium of this compound in the membrane; this observation was consistent with the high detection limit of this compound. Acrylonitrile (Figure 16.2) showed perfect linearity over the 50–750 ppb concentration range. 16.4.1.4 Matrix Effects Standard additions of BTEXs to industrial wastewaters showed no matrix effect. The angular coefficient of the straight line obtained by four additions in the 0.1–2 ppm range was practically identical to that of a similar calibration plot using triplydistilled water (18,954 vs 18,883), the two lines being parallel (Figure 16.3).
16.4.2 Case Studies 16.4.2.1 Acrylonitrile Acrylonitrile, a compound that is not included in the family of the VOCs, could be determined at high concentration levels (4.78 ppm) in industrial waters containing
kCounts
2.0
Ione: 52
750 ppb 500 ppb
1.5 1.0
200 ppb
0.5
50 ppb
100 ppb
0.0 25
50
Minutes
75
100
125
FIGURE 16.2 The m/z 52 ion current for acrylonitrile showed perfect linearity over the 50–750 ppb concentration range. 50,000 Signal
40,000 30,000 20,000 10,000 0
0
0.5
1
ppm
1.5
2
2.5
FIGURE 16.3 Calibration of BTEXs in triply-distilled (continuous line, y = 18,883x + 115.59, R2 = 1) and industrial water (dotted line, y = 18,954x + 7002.1, R2 = 0.9977).
500
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
a wide range of aromatic hydrocarbons. Figure 16.4 shows a MIMS full scan mass spectrum with ions at m/z 78, 104, 91, 120, 121 and 118 (typical of such aromatic substances as styrene, ethylbenzene, xylene, cumene, cumene hydroperoxide, α-methylstyrene) along with an ion at m/z 52 (acrylonitrile) of much lower ion signal intensity, whose ion currents for duplicate analyses are shown in Figure 16.5. 16.4.2.2 Comparison of Membrane Inlet Mass Spectrometry (MIMS) with Purge-and-Trap/Gas Chromatography (GC)/ Mass Chromatography (MS) A series of experiments was performed in order to compare MIMS and Head-Space Purge and Trap /GC/MS by analyzing a total of 20 industrial wastewater samples from seven different sampling points. In Figure 16.6 is shown the MIMS mass spectrum of one of the wastewater samples; in this example, seven compounds that
104
Relative intensity
100%
75%
50% 91
25% 51
63
119
78
0% 50
100
134
155 165 179 150 m/z
205 219 200
FIGURE 16.4 Full scan mass spectrum of a sample of industrial waters containing acrylonitrile (m/z 52) along with aromatic substances such as styrene, ethylbenzene, xylene, cumene, cumene hydroperoxide, and α-methylstyrene.
kCounts
25 20 15 10 5 25
Minutes
50
FIGURE 16.5 Ion current of m/z 52, acrylonitrile at 4.78 ppm, duplicate analysis of the same sample of industrial waters as was used for Figure 16.4.
50
55
65 61
60
66
75
Chloroform
79
83
100
105
Toluene 91 101 1,2 dichloroethylene 96 CCl4 117 98
125 m/z
130
134
Trichloroethylene 132
150
151
175
Tetrachloroethylene 166
200
FIGURE 16.6 Mass spectrum of industrial water used to compare MIMS and Purge-and-Trap/GC/MS. Identified compounds and ions used for quantification are reported.
0%
25%
50%
75%
100%
Remote Monitoring of Volatile Organic Compounds in Water 501
502
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V 30,00
y = 1,0671x – 0,6296 R 2 = 0,8944
ppb (P&T/GC/MS)
25,00 20,00 15,00 10,00 5,00 0,00
0
–5,00
5
10
15
20
25
ppb (MIMS)
FIGURE 16.7 Comparison of data obtained by analyzing samples with Purge-and-Trap/ GC/MS and MIMS. The concentration (in ppb) of each of toluene (m/z 91), benzene (m/z 78), 1,2-dichloroethylene (m/z 98), trichloroethylene (m/z 130 + 132), chloroform (m/z 83), and vinyl chloride and dichloroethane (m/z 62 for both compounds) in each of 20 wastewater samples determined by MIMS is plotted against that determined by Purge-and-Trap/GC/MS.
have been identified by their characteristic ions are indicated on the mass spectrum. Toluene (m/z 91), benzene (m/z 78), 1,2-dichloroethylene (m/z 98), trichloroethylene (m/z 130 + 132), chloroform (m/z 83), and vinyl chloride and dichloroethane (m/z 62 for both compounds) were detected and quantified in all 20 samples with both techniques. The concentration (in ppb) of each of the above seven compounds in each of 20 wastewater samples determined by MIMS is plotted against that determined by P&T/GC/MS as shown in Figure 16.7. The equation of the regression line calculated from these data (Figure 16.7) is y = 1.0671x–0.6296, R2 = 0.8944. The slight differences between the actual and the ideal coefficients (slope = 1, intercept = 0 and R2 = 1) are probably due to the contributions of other compounds to the abundances of the ions used for quantitation.
16.4.3 Field Tests Four instruments were deployed in different plants representative of typical cases of water treatment, namely two plants for the potabilization of ground water (A and B), the third plant for surface water potabilization (C), and, finally, a plant for industrial water treatments (D) (Table 16.2). In field test A, the instrument was deployed in a plant for the production of drinking water from ground water using chlorine dioxide as a disinfection agent. Due to past industrial activity in that area, the ground water was heavily contaminated by chloroform and trichloroethylene. Charcoal filters were used to abate the organohalogen concentration in drinking water down to 1–10 ppb levels. The instrument was located in a 2 × 3 m container maintained at room temperature. The instrument was able to identify and to quantify drinking water pollutants by means of a mass spectrum in which the ions characteristic of the individual compounds were recorded clearly, showing their
Remote Monitoring of Volatile Organic Compounds in Water
503
respective diagnostic isotopic patterns (Figure 16.8a). Hourly analyses were carried out in order to check that pollutant concentrations did not exceed the legal thresholds. The position was unmanned and the results were transmitted to the remote control room by e-mail at the conclusion of each analysis. The instrument was monitored over a period of 334 days during which it functioned for 323 days; failures and maintenance resulted in the loss of 11 days (that is, 3% of the time monitored) (Table 16.2). Almost 8000 determinations were performed corresponding to 646 hours of analysis. In plant B, both ground water and drinking water were monitored every hour, corresponding to a total frequency of one analysis per 30 minutes. Here, the contaminant was trichloroethylene. The plant was not equipped with charcoal filters, consequently the pollutant concentration was kept within the regulation limit (10 ppb) by shifting water uptake from one well to another. Figure 16.8b shows a typical full-scan mass spectrum recorded from the ground water of this location, with the characteristic trichloroethylene molecular ion isotopic quartet at m/z 130, 132, 134, and 136. As for the previous field study in plant A, the location was unmanned and the results were transmitted by e-mail. The percentage of inactivity in plant B (5%) was comparable with that of plant A (3%), despite the fact that the working period for plant B (almost 500 days) was longer than that for plant A, and the number of analyses carried out at plant B (23,626) was more than three times higher than those carried out in plant A (7,752); see Table 16.2. Field test C was an example of application to surface waters used for human consumption. Such waters were essentially uncontaminated by chemicals and needed only a conventional disinfection. Nevertheless, this plant was monitored in consideration of the fact that accidental pollution by gasoline and oil had been recorded in the past due to the proximity of an adjacent highway with heavy traffic. Figure 16.8c shows a typical mass spectrum of this instance, with no significant ions. This instrument was monitored over 547 days (12,240 analyses, 1020 hours) during which the days off were 37, that is 7% of the period (Table 16.2). Finally, industrial wastewaters (plant D) were analyzed from the outlet of a pipe connected into a municipal sewage treatment plant. In such a case, the concern was that organohalogenated compounds from industrial wastes may affect the biological treatment of urban wastewaters. The instrument recorded 3 analyses per day on those days when the industrial wastewaters were discharged. Figure 16.8d shows the typical mass spectrum of such samples. Ions characteristic of chloroform (m/z 83, 0.3 ppb), trichloroethylene (m/z 130, 23 ppb), and toluene (m/z 91, 31 ppb) were found in the mass spectrum reported in Figure 16.8d. The complex matrix of such wastes did not affect the MIMS determinations (see previous discussion, Figure 16.3). The instrument was monitored during more than 500 days (about 1600 analyses or 130 functioning hours) with a 4% of non-functioning time (Table 16.2). For all types of mass spectrometers, particularly those operated in remote locations, it is of interest to consider the frequency of mass calibration and, for those instruments that employ electron impact ionization, the frequency with which the filament must be replaced. It was found for the MIMS instruments that mass calibrations were carried out on 24 occasions and filaments were replaced on 16 occasions.
504
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
(a) 100%
164
75% Chloroform 94 96 83
50% 25%
166 Tetrachloroethylene 168
0% 50
75
100
(b) 100%
m/z
125
150
175
130 132
75% Trichloroethylene
60
50% 25%
96
134
100
150 m/z
0% 50 (c)
100%
250
45 52
75% 50%
200
41
25% 0% 50
75
100
(d)
125 m/z
150
175
200
91 Toluene
100% 75% 50%
Chloroform
25%
77 83
105
Trichloroethylene 130 132
0% 50
100
m/z
150
200
Remote Monitoring of Volatile Organic Compounds in Water
505
FIGURE 16.8 (Opposite) (a) Field test A. Mass spectrum of a water sample from a plant for the production of drinking water from ground water. Chlorine dioxide had been used as a disinfection agent. Ground water was contaminated heavily by chloroform and trichloroethylene due to past industrial activity; charcoal filters had been used to abate the organohalogen concentration in drinking water down to 1–10 ppb levels. (b) Field test B. Mass spectrum recorded from the ground water of this location. At this site, the contaminant was trichloroethylene and the mass spectrum shows the characteristic trichloroethylene molecular ion isotopic quartet at m/z 130, 132, 134, and 136. The plant was not equipped with charcoal filters, consequently pollutant concentration was kept within the regulation limit (10 ppb) by shifting water uptake from one well to another. (c) Field test C. Mass spectrum of surficial water used for human consumption. Such waters were essentially uncontaminated by chemicals and needed only a conventional disinfection. No significant ions were observed. (d) Field test D. Mass spectrum of a sample of industrial wastewaters taken from the outlet of a pipe connected into a municipal sewage treatment plant. In this case, the concern was that organohalogenated compounds from industrial wastes may affect the biological treatment of urban wastewaters. This typical mass spectrum shows ions characteristic of chloroform (m/z 83, 0.3 ppb), trichloroethylene (m/z 130, 23 ppb), and toluene (m/z 91, 31 ppb).
From the data shown in Table 16.2 concerning the numbers of operating days and the number of analyses carried out each day at each of the four sites, it is found that averages of 2825 analyses were carried out with each filament and 1883 analyses were carried out between successive calibrations. Comparing these data with those of a GC/MS instrument used presently in our laboratory and which has shown good instrumental stability and reliability, it was found that the GC/MS instrument performed 305 analyses per filament and 78 analyses per calibration. When it is borne in mind that the duration of a GC/MS analysis was 50 min while that of a MIMS analysis was 5 min, it is clear that both systems performed comparably in terms of operation time per filament.
16.5 CONCLUSION A number of laboratory tests to determine LOD, linearity and repeatability of MIMS instruments applied to the analysis of VOCs in water were performed. Data were comparable with those obtained by the classical method of VOC analysis in water (P&T/GC/MS and USEPA Method 8260B). Four MIMS instruments were tested over an extensive period of time to evaluate their on-site performance in unmanned locations. Results were remarkable: the instruments worked unchecked for long periods producing a total of more than 45.000 analyses and VOC amounts were quantified automatically and sent to a remote control room where non-expert personnel could understand the results readily. In conclusion, MIMS instruments proved to have great potential for utilization in continuous VOC-monitoring stations. These instruments are reliable, cost-effective and simple to use; they have no environmental impact because no solvent is used for the extraction of organics from water, and they can be located on-site, unattended, providing a continuous flow of data on water quality and pollution.
506
Practical Aspects of Trapped Ion Mass Spectrometry, Volume V
References
1. Irvine, K.N.; McCorkhill, G.; Caruso, J. Continuous monitoring of conventional parameters to assess receiving water quality in support of combined sewer overflow abatement plans. Water Environ. Res. 2005, 77, 543–552. 2. Mikkelsen, Ø.; Skogvold, S.M.; Schrøder, K.H. Continuous heavy metal monitoring system for application in river and seawater. Electroanalysis 2005, 17, 431–439. 3. Yamada, M.; Suga, M.; Waki, I.; Sakamoto, M.; Morita, M. Continuous monitoring of polychlorinated biphenyls in air using direct sampling APCI/ITMS. Int. J. Mass Spectrom. 2005, 244, 65–71. 4. Hart, K.J.; Dindal, A.B.; Smith, R.R. Monitoring volatile organic compounds in flue gas using direct sampling ion trap mass spectrometry. Rapid Commun. Mass Spectrom. 1996, 10, 352–360. 5. Kürten, A.; Curtius, J.; Helleisa, F.; Lovejoy, E.R.; Borrmann, S. Development and characterization of an ion trap mass spectrometer for the on-line chemical analysis of atmospheric aerosol particles. Int. J. Mass Spectrom. 2007, 265, 30–39. 6. Riter, L.S.; Laughlin, B.C.; Nikolaev, E.N.; Cooks, R.G. Direct analysis of volatile organic compounds in human breath using a miniaturized cylindrical ion trap mass spectrometer with a membrane inlet. Rapid Commun. Mass Spectrom. 2002, 16, 2370–2373. 7. Johnson, R.C.; Srinivasan, N.; Cooks, R.G.; Schell, D. Membrane introduction mass spectrometry in a pilot plant: On-line monitoring of fermentation broths. Rapid Commun. Mass Spectrom. 1997, 11, 363–367. 8. Bier, M.E.; Cooks, R.G. Membrane interface for selective introduction of volatile compounds directly into the ionization chamber of a mass spectrometer. Anal. Chem. 1987, 59, 597–601. 9. Kotiaho, T.; Lauritsen, F.R.; Choudhury, T.K.; Cooks, R.G.; Tsao, G.T. Membrane introduction mass spectrometry Anal. Chem. 1991, 63, 875A–883A. 10. Lauritsen, F.R.; Kotiaho, T.; Choudhury, T.K.; Cooks, R.G. Direct detection and identification of volatile organic compounds dissolved in organic solvents by reversedphase membrane introduction tandem mass spectrometry. Anal. Chem. 1992, 64, 1205–1211. 11. Bauer, M.; Solyom, D. Determination of volatile organic compounds at the parts per trillion level in complex aqueous matrixes using membrane introduction mass spectrometry. Anal. Chem. 1994, 66, 4422–4431. 12. Soni, M.; Bauer, S.; Amy, J.W.; Wong, P.; Cooks, R.G. Direct determination of organic compounds in water at parts-per-quadrillion levels by membrane introduction mass spectrometry. Anal. Chem. 1995, 67, 1409–1412. 13. Cisper, M.E.; Gil, C.G.; Townsend, L.E.; Hemberger, P.H. Online detection of volatile organic compounds in air at parts-per-trillion levels by membrane introduction mass spectrometry. Anal. Chem. 1995, 67, 1413–1417. 14. Mendes, M.A.; Pimpim, R.S.; Kotiaho, T.; Eberlin, M.N. A cryotrap membrane introduction mass spectrometry system for analysis of volatile organic compounds in water at the low parts-per-trillion level. Anal. Chem. 1996, 68, 3502–3506. 15. Ketola, R.A.; Mansikka,T.; Ojala, M.; Kotiaho, T.; Kostiainen, R. Analysis of volatile organic sulfur compounds in air by membrane inlet mass spectrometry. Anal. Chem. 1997, 69, 4536–4539. 16. Bocchini, P.; Pozzi, R.; Andalò, C.; Galletti, G.C. Membrane inlet mass spectrometry of volatile organohalogen compounds in drinking water. Rapid Commun. Mass Spectrom. 1999, 13, 2049–2053.
Remote Monitoring of Volatile Organic Compounds in Water
507
17. Ketola, R.A.; Kotiaho, T.; Cisper, M.E.; Allen, T.M. Environmental applications of membrane introduction mass spectrometry. J. Mass Spectrom. 2002, 37, 457–476. 18. Srinivasan, N.; Johnson, R.C.; Kasthurishnan, N.; Wong, P.; Cooks, R.G. Membrane introduction mass spectrometry. Anal. Chim. Acta. 1997, 350, 257–271. 19. Bocchini, P.; Pozzi, R.; Andalò, C.; Galletti G.C. Experimental upgrades of membrane introduction mass spectrometry for water and air analysis. Anal Chem. 2001, 16, 3824–3827.
Author Index* Ausio, J., 69 Abedi, A., 408, 409 Adamczyk, M., 88, 90 Amunugama, M., 91 Arriaga, E.A., 99 Back, J.W., 104 Badman, E.R., 8, 17 Baessmann, C., 282 Bagal, D., 227 Bartlet-Jones, M., 98 Bateman, R.H., 210 Berton. A., 367 Bier, M.E., 447 Bisgaard, C.Z., 312 Blatt, R., 359 Blom, M.N., 180 Bocchini, P., 491 Bowers, M.T., 207, 208, 219 Brancia, F.L., 367 Brekenfeld, A., 282 Brittain, R., 445 Brock, A., 140 Brodbelt, J.S., 36, 39, 46–48, 55 Bruce, J.E., 104 Burns, M.M., 180 Bush, M.F., 244, 248, 249 Caldwell, G., 397, 399–402 Carter, J.G., 408 Champenois, C., 333 Chen, X., 103 Chipuk, J.E., 36, 39, 46, 48, 55, 65 Chowdhury, S., 410 Chrisman, P.A., 8, 13 Christophorou, L.G., 408 Chung, S., 354 Clemmer, D.E., 216 Clench, M., 225 Cooks, R.G., 170, 224, 278, 328 Coon, J.J., 10, 22, 59, 63, 65, 69, 71, 72 Cooper, H.J., 121, 207 Cudzilo, K., 434 Danell, R.M., 180 Daniels, S., 98 Dehmelt, H.G., 170, 328, 334 Dey, S., 98 Dick, G.J., 337 Djdja, M.-C., 226
Douglas, D.J., 52, 53, 380, 381 Drewsen, M., 254, 291, 294, 296, 297, 300, 309, 311, 312, 356 Drexler, D.M., 435 Dryhurst, D.D., 69 Duft, D., 189, 190, 193, 194, 196, 197 Dunbar, R.C., 248, 249 Eckers, C., 228 Eiceman, G.A., 206, 387, 404 Erickson, D.E., 13 Evoy, S., 309 Ewing, R.G., 404, 405 Fenn, J., 127 Fico, M., 328 Fitaire, M., 396, 397 Fohlman, J., 131 Forbes, M.W., 239, 248, 249 Franzen, J., 263, 264, 277, 282 Froelich, J.M., 60, 83 Galletti, G.C., 491 Gao, L., 328 Gardner, M.W., 109 Garrett, T.G., 225 Garzón, I.L., 179, 180, 417, 420, 428, 433 Gauthier, J.W., 132 Ge, Y., 144 Gebler, J.C., 90 George III, J.E., 439 Gerlich, D., 245, 335 Gheno, F., 396, 397 Giles, K., 210 Glish, G.L., 6, 70, 465 Goshe, M.B., 105, 106 Grabenauer, M., 219 Griffin, T.J., 99, 100 Grimsrud, E.P., 397, 399–402, 410 Gronert, S., 43 Gunawardena, H.P., 70 Gygi, S.P., 94, 142 Haberland, H., 171 Hakansson, K., 140 Han, H., 23 Han, H.L., 64, 66 Harden, C.S., 404, 405 Hartmer, R., 282 Harvey, D.J., 227 Hattan, S., 98
* The names listed here refer only to authors whose names appear in the text and/or in the captions.
509
510 He, F., 98 Hilton, G.R., 219 Hiraoka, K., 397 Hogan, J.M., 18 Højbjerre, K., 254, 291, 311, 312, 356 Holland, R., 227 Huang, Y.N., 98 Hunt, D.F., 10, 22, 36, 37, 63, 69, 139, 451 Hunter, E.P., 38 Hunter, S.R., 408 Iavarone, A.T., 189, 190 Jackson, G.P., 463 Jacobson, A., 98 Jensen, L., 296, 297 Jockusch, R.A., 239, 248, 249 Johnson, R.C., 493 Juhasz, P., 98 Julien, R.R., 93 Julka, S., 94, 95 Jung, H.R., 228 Kaplan, D., 282 Karpas, Z., 206 Kebarle, P., 395, 402, 410 Ketola, R.A., 493 Khainovski, N., 98 Kiessel, S.E.B., 71, 72 Kim, S.H., 395 Knighton, W.B., 402, 410 Konenkov, N.V., 380, 381 Kürten, A., 493 Landman, U., 181, 182 Laskin, J., 140 Lawless, P.A., 406 Lawrence, A.H., 402 Le, T., 354 Li, L., 85 Li, S., 97 Liang, X.R., 11, 13 Lias, S.G., 38 Lindballe, J., 296, 297 Liu, J., 3, 13, 64 Liu, Y., 408 Londry, F.A., 13 Louris, J.N., 459 Lu, Y., 60, 83 Ly, T., 93 Magnera, T.F., 402 Maleki, L., 337, 340, 354 Manura, D., 273 Mao, D., 52 March, R.E., 350, 440 Marchese, J.N., 98 Margolis, H., 346 Marshall, A.G., 130, 132, 138, 144 Martin, S., 98 Martinussen, R., 294, 296, 297, 300, 309 Mayhew, C.A., 407, 408
Author Index McAlister, G.C., 59, 71, 72 McEwen, C.N., 225 McLean, J.A., 225 McLuckey, S.A., 3, 7, 8, 11, 13, 17, 18, 20, 23, 62, 64, 66–68, 72 Meany, D.L., 98, 99 Michaelian, K., 179, 180 Mikkelsen, Ø., 492 Mordehai, A., 440 Mortensen, A., 294, 296, 297, 300, 309, 312 Mulholland, J.J., 464 Mulligan, C.C., 328 Newton, K.A., 20 Nissen, N., 296, 297 Offenberg, D., 312 Olivova, P., 226 Oomens, J., 248, 249 Ouyang, Z., 328 Pappin, D.J., 98 Paradisi, C., 371 Parker, K., 98 Parks, J.H., 169, 180, 181, 189, 190, 193, 194, 196, 197 Paul, W., 262, 328, 439 Payne, A.H., 70 Peverall, R., 408 Pillai, S., 98 Pinelli, F., 491 Plass, W.R., 275–277 Plet, B., 153 Polfer, N.C., 248, 249 Pozzi, P., 491 Prestage, J.D., 337, 340, 354 Preston, J.M., 396, 398 Przybylski, M., 140 Purkayastha, S., 98 Purves, R.W., 209 Qiu, Y., 96 Rajadhyax, L., 396, 398 Ramsey, N., 332 Raveane, L., 367 Regnier, F., 94, 95 Reich, R.F., 434 Reid, G.E., 17, 60, 83, 91, 100 Reilly, J.P., 252 Riba-Garcia, I., 226 Ridenour, W.B., 225 Rizzo, T.R., 245 Roberts, K.D., 91 Roepstorff, P., 130, 131 Ross, P.L., 98 Rutherford, E., 3 Sadagopan, N., 89 Sahlstrom, K.E., 409, 410 Schmitter, J.-M., 153 Schrama, C.A., 346 Schroeder, M.J., 10, 22, 63
511
Author Index Schubert, M., 282, 440 Schwartz, J.C., 65, 440, 461, 463–465 Scrivens, J.H., 205, 219 Shabanowitz, J., 10, 22, 63, 69 Shi, X., 193, 194, 196, 197 Simon, C., 155, 164 Slade, S.E., 219 Smith, R.D., 132, 138 Soderblom, E.J., 105, 106 Song, Q., 328 SØrensen, J.L., 294, 300, 309 Spangler, G.E., 406 Specht, A., 440 Staanum, P.F., 254, 291, 294, 296, 297, 300, 309, 311, 312, 356 Stapelfeldt, H., 312 Stauber, J., 226 Steinwedel, H., 439 Stephenson J.L., 7, 63, 67, 68 Stick, D., 295 Stone, J.A., 206, 387, 404 Strife, R.J., 449 Sudakov, M., 380, 381 Summerfield, S.G., 89 Swaney, D.L., 64 Syka, J.E.P., 10, 22, 63, 65, 69 Tabrizchi, M., 408, 409 Talbot, F.O., 239 Tanaka, K., 127 Taylor, D.M., 444, 459, 463 Thalassinos, K., 205, 219, 223 Thompson, L.V., 99 Thompson, R.C., 334 Thomson, J.J., 3 Tjoelker, R.L., 340
Todd, J.F.J., 350, 440 Traldi, P., 367 Trim, P.J., 225 Turecek, F., 136 Ueberheibe, B., 69 Uetrecht, C., 221 Vedel, F., 327 Voight, D., 296, 297 Wang, H., 86 Wang, M., 445 Wang, N., 85 Watson, J.T., 89 Wells, G.J., 447, 462 Wells, J.M., 17 Wester, R., 311 Williams, E.R., 248, 249 Williams, J.P., 228 Williamson, B., 98 Wineland, D., 170 Wirtala, M., 65 Wright, P.J., 53 Wu, J., 90 Xia, Y., 11, 23, 62, 66 Xie, H., 99 Xing, X., 180, 181 Yamabe, S.J., 397 Yamada, M., 492 Yang, M., 23, 439 Yang, M.J., 13 Yoon, B., 181, 182 Yost, R.A., 225, 417, 434, 446, 464 Zeng, D., 97 Zhang, J., 47, 53 Zhou, H., 96 Zubarev, R.A., 143
Subject Index 1,4-Dinitrobenzene, 399 1,1,1-Trichloroethane, 496 1,2-Dichloroethylene, 501, 502 1,2-Dichlorobenzene, 495 1,2-Dichloroethane, 492 1,4-Dichlorobenzene, 495 115In +, 359 171Yb +, 345 18O atom, 103 18O −, 71 2 199Hg+, 345, 352, 356 2-(2′-Hydroxybenzoyl)–benzoic acid, 371, 372 2,2′,4,4′,6-Pentabromodiphenylether, 480 2,3-Dichlorobiphenyl, 484, 485 2,3-Dimethyl pyridine, protonated, 391 2,3-Dimethyl-2,3-dinitrobutane, 399, 400 2,3-Dimethyl-2,4-dinitropentane, 399 2,4-Dichlorophenol, 472–474 2,4-Dimethylpyridine, 403–406 2,4-Dinitrofluorobenzene-d 0/d3, 103 2,5-Hydroxybenzoic acid, 425, 428, 429 cluster ion of, 430 2,6-Naphthalic acid, 48 202Hg+, 352 24Mg+, 307, 312, 324 24MgH+, 324 25Mg+, 324 2D Polyacrylamide gel electrophoresis, 2D PAGE, 93 2D Quadrupole ion trap, 417, 419 2-Deoxy-5-cytidine monophosphate, 46 2-Methoxy-4,5-dihydro-1H-imidazole, 88 3-(3-Methoxypropoxy) propanol, 396 3,3′-Dithio-bis(succinimidylpropionate), DTSSP, 104 3D Quadrupole ion trap mass spectrometer, 5, 9, 7, 17, 51, 54, 62, 242, 257, 282, 417, 419, 440 40 Ca +, 300, 305, 307–309, 312–316, 318, 320, 321 40 Ca16O +, 300, 305, 310 42Ca +, 309 44Ca +, 309 4-Sulfophenyl isothiocyanate, 89 5′P-dAA, 55 5′P-Dag, 55 5′P-dGA, 55 5′P-dGG, 55 63Ni source, 207, 390, 395
6-Aza-2-thiothymine, 431, 433 Sr+, 345, 359
88
A Absorption spectroscopy, 240 AC dipole electric field, 440 Accidental contamination, 492 Accurate mass tag, AMT, 129, 140 Acetic acid, 127 Acetone, 396 Acetone–water, 397 Acetonitrile, 127, 155, 481 Acetylacetone, 88 Acid-labile isotope-coded extractants, ALICE, 96 Acquisition phase, 372 Acrylonitrile, 491, 495, 498–500 Acrylonitrile butadiene styrene plastic, ABS, 480, 482 Action spectroscopy, 240, 246, 247, 253, 282 Activated ion electron capture dissociation, AI-ECD, 137 Activation barrier, 371 Activation energy, reverse, 496 Adduct ions, 411 dissociation of, 387, 403 Adiabatic approximation, 349 Adiabatic cooling, 396 Affinity capture method, 86 Affinity tag, 105 Aging, 425 Agn+ cluster, 171, 174, 177, 178–181 Al2O3, 86 Al3+, 86 Alcohol dehydrogenase, 141, 219 Aldolase, 219 Allan deviation, 331 Allan Variance, 331 Alveolar proteomics, 140 cystic fibrosis, 140 proteinosis, 140 Alzheimer’s disease, 220 Ambient pressure, 387, 389 American Society for Mass Spectrometry, 284 Amino acid residue, modification site, 84 Ammonium acetate, 127 Ammonium hydroxide, 127 Amplitude detection, 304, 310, 320
513
514 Amplitude method, 304, 320, 321, 323 Amplitude, modulation frequencydependent, 304 Amyloid fibril, 220 Amyloidogenic protein β2-microglobulin, 220 Analysis, drug, 441 food, 441 forensic, 441 Analytical mass scan, 375, 377 Anharmonic bottleneck, 250 Anharmonicity, 315, 316, 338 Aniline ion (C6H5NH2+), 311 photofragmentation, 311, 312 Anion formation, 408 Anthracene, 452, 453, 455 Anthropic activities, 492 Apigenin, 155, 156, 162, 163 Apigenin-7-O-glucoside, 155, 156, 162, 163 Apigenin-7-O-neohesperidoside, 155, 156, 162, 163 Apomyoglobin, 64 Applied Biosystems peptide synthesizer, 156 Arabidopsis, 65 Arginine, 24, 127, 191, 247–249 Argon, 244, 368 Arrhenius parameters, 410 Arrhenius plot, 405, 409 Arrhenius rate model, 197, 198 Arrival time, 401 corrected, 218 distribution, 212 distribution profile, 214, 217, 223 ASGDI, 6, 7 ASGDI source, 8, 11 Aspartic acid cleavage, 16 Aspartic acid residue, 22 location of, 90 Astragalin, 50 Astronomical time, 329 Atmospheric pressure, 388 chemical ionization, APCI, 8, 12, 492 ion emitter, dual, 63 solids analysis probe, ASAP, 225 Atmospheric sampling glow discharge ionization, see ASGDI Atomic beam, 332, 342 Atomic clock, 327–360 schemes, 332 Atomic fountain, 331, 332 Atomic ion lifetime, 328 Atomic ion, laser-cooled, 292–295, 297, 299, 300, 304, 305, 310, 317, 318, 323 Atomic ions, 170, 299 fluorescence, 170 laser cooling, 170, 328 Atomic laser, 328 Atomic levels, 358
Subject Index Atomic oscillator, 331, 360 Atomic quantum physics, 328 Atomic spectroscopy, high resolution, 328 Atomic transition, 330, 331, 335, 339, 355 Atrazine, 478, 479 Attachment rate coefficients, 407 au, 261 AuCl2−, 70, 175, 176, 182–186 Aun− cluster, isomer structure, 184–186 Automated method development, AMD, 467, 486 Automatic gain control, AGC, 422–425, 444, 467 Auxiliary dipolar AC potential, 272 Auxiliary RF voltage, 9, 11 Average dipole orientation, ADO, 402 Averagine, 129 Aviation security, 129 Avidin affinity chromatography, 97 Axial modulation, 272, 273, 440, 456, 467 az, 261, 270, 298, 350, 369 Azobenzene radical anion, 23 Azulene anion, 410
B Ba+, 345 Backbone cleavage, 24 Background gas, 319, 339 Background noise, 447 BAD, see Boundary-activated dissociation BAPMPS, 92 Barium, 332, 334 Baseline interferometry, 330 Bath gas, 5; see also Buffer gas atomic/molecular weight, 446 helium, 42, 43, 171, 174, 175, 195 neon, 175 pressure, 10 temperature, 191 BDE 100, 481 BDE 153, 481 BDE 154, 481 BDE 183, 481 BDE 205, 481 BDE 209, 480–482 BDE 28, 481 BDE 47, 481 BDE 99, 481 Beaker profile, 335 Benzene, 36, 492, 495, 502 Benzoic acid anions, 68–70 Benzyl cation, 104 Beryllium, 319, 357 Biological molecule, 246 identification, 84, 128 structural characterization, 84, 128
515
Subject Index Biomolecular conformation, 170 Biomolecular ions, 192 trapped, 169 Biomolecule analysis, 15 Biomolecule conformational change, 186 Biomolecule folding, 255 Biomolecule protonated, photo-excitation of, 240 Biomolecule, dye-derivatized, 186–188 Biphenyl, 483, 484 BIRD, see Blackbody infrared radiative dissociation Bisuccinimidyl-succinamyl-aspartyl-proline, SuDP, 106 Bisuccinimidyl-succinamyl-aspartyl-prolylglycine, SuDPG, 106 Black hole (or canyon), 263, 348, 350–352, 353 Blackbody heating, 41 Blackbody infrared radiative dissociation, BIRD, 123, 130, 246, 253 principles, 134 protein and peptide, 134, 135 Blackbody radiation, 41, 332, 359 BMS-X, 435 Body-centered cubic (bcc) symmetry, 177 Boltzmann distribution, 388 Boltzmann sigmoidal function, 158 Boltzmann’s constant, 207 Bottom-up approach, see Proteomics, bottom-up, 101 Boundary activation, 367, 369, 382 Boundary effect, 371 Boundary-activated charge-separation dissociations, 383 Boundary-activated dissociation, 367, 369, 370, 373, 374, 385 Bovine cytochrome c charge states, 218 Bovine serum albumin, 16, 17, 141, 154 Bradykinin, 51, 52, 154 fragment (residues RPPGF), 380, 381 doubly-protonated molecule of, 378, 380, 382–384 Bragg diffraction peak, 177, 178 Brain natriuritic peptide (BNP-32), 140 BrCH2COC6H5, 91 Breakdown curve, 159, 160 Breast carcinoma cell, proteomics analysis of, 140 Breath, human, 493 Breathing (BR) mode, 294, 301–303, 305, 307, 308, 313–316, 322, 323 Brewster angle window, 187 Broad-band frequencies, 369 Bromide ion, 411 Bromodichloromethane, 492 Bromofluorobenzene, 467 Bromoform, 492, 495, 498
Broth, fermentation, 493 Brownian motion, 392 Bruker Daltonics HCTultra, 14 Bruker HCTultra/Agilent 6340 ETD ion trap, 8 Bruker ion trap, 442 BTEX compounds, 492, 499 Buffer gas, 205, 207, 245, 278, 279, 281, 310, 368, 382, 440; see also Bath gas atomic/molecular weight, 446 collision, 335, 354 pressure, 281, 446 Butanedione, 101
C C2D2, 312 C2H5+, 449 C2H5OD, 37 C60 cluster ion, 178, 179 C6D5OD, 38 C6H5CH2OD, 38 Ca+, 295, 299, 318, 349, 350, 355, 358, 359 isotope combination, 309 Cage, miniaturization of, 328 Calcium acetate, 20 Calcium beam, 318 Calibration procedure, 216, 217 California Institute of Technology, 329 Camera, 420 Canada Research Chairs Program, 284 Canadian Foundation for Innovation, 284 Cancer, 418 CaO+, 305 Carbon tetrachloride, 408, 409, 495, 498 Carbonic anhydrase, 133, 138 Carcinogens, 483 Case Studies, 499 Casein, 154 Catechin (+), 49, 155, 156, 162–164 galloylated, 49 non-galloylated, 49 Cation adducts, 421 Cation/anion complex formation, 20 Cationic salts, 421 Cavity ring-down spectroscopy, CRDS, 240, 241 CCD camera, 294, 295, 300, 301, 305, 319–321 image, 180, 310 pixel value, 178 CD3OD, 38, 43, 51–54 Center-of-mass (COM) mode, 294, 301–303, 305, 307–309, 313–316, 319–323 Central barrier, 402 Cerebral peduncle, 425 Cerebrospinal fluid, 85 Cesium atom, 329 Cesium atomic frequency standard, 329, 331, 352
516 CF3CH2OD, 38 CH3OD, 38 CH5+, 449 Champagne flute profile, 335 Charge capacity, 442 extended, 439, 442 ion trap, 444 Charge inversion, 12, 15, 19 reaction, 72 negative-to-positive, 19 positive-to-negative, 19 Charge limit, spectral, 444 Charge reduction, 16 stepwise, 19 Charge state, 15, 17, 19, 67 deconvolution, 70 manipulation, 6, 15, 16 Charge transfer, exothermicity of, 197 photoinduced, 196 Charge-coupled device, CCD, 176, 177 Charge-dependent dissociation, 16 Chemical background, 436 Chemical derivatization, 83, 85, 92 reagent, 101 strategy, 88, 89, 91, 94 fixed-charge, 101 Chemical distributions, intrinsic, 418 Chemical ionization reagents, liquid, see Ionization, chemical, liquid Chemical ionization source, 9–11, 14 Chemical ionization, see CI Chemical ionization, see Ionization, chemical Chemical mass shift, 276 Chemical modification technique, 102 Chemical reaction path, 328 Chemical signature, 418 Chemical structure elucidation, 450 Chemical vapor deposition, CVD, 445 Chemicals Warfare Convention, 229 Chemistry analysis, 328 Chicken egg white proteome, 142 Chloride ion, 400, 401, 403, 406, 407, 410, 411 Chlorine dioxide, disinfection by, 505 Chlorobenzene, 407 Chlorodibromomethane, 492 Chloroform, 408, 492, 495, 501, 502, 504 counter flow of, 398 Chloropyriphos, 477, 478 Choline, 425 Chromatographic separation, 85 Chromatographic timescale, 24 Chromosorb 101, 498 CI, see Ionization, chemical CI/MS/MS, see Ionization, chemical/tandem mass spectrometry CID, see Collision-induced dissociation CID-MS/MS, 88, 100
Subject Index Circular dichroism spectroscopy, 219 Cleavable isobaric labeled affinity tag, CILAT, 97 CLIO, 247 Clock atomic transition, 360 Clock signal, 330, 331, 339 building, 341 Clock transition, 343 Cluster ion, 175, 190, 395 symmetry, 182 (CsI)nCs+, 177, 178 Cluster ions, mass-selected, 176 CO, 310 Coating methods, 421 acoustic wave, 421 airbrushing, 421 electrospraying, 421 inkjet, 421 sublimation, 421 Coating, chromium, 445 Coating, Silchrom, 445 Cocaine, 434 Cocaine-d3, 434 Coherent motion, 125 Collision cell, 9 octopole, 208 Collision cross-section, 207, 208, 223 theoretical, 207 Collision energy, 464 normalized, 157–159 Collision frequency, 389 Collision gas, 246 Collision model, hard-sphere, 216 Collision, ion/neutral, 319 Collisional cooling, 278 Collisional dissociation, 392 Collision-induced dissociation, 18, 21, 25, 41, 49, 59–61, 64, 66, 67, 70–72, 90, 92, 96–98, 106, 108, 122, 130, 155–158, 208, 212, 226, 246, 367, 392, 458, 462, 464 beam-type, 23 chemical structure-insensitive, 465 consecutive/competing, scan function for, 373 efficiency, 465 fragmentation efficiency, 66 low energy, 104, 107 multi-level, 464 period, 464, 465 techniques, HASTE CID, 465, 466 techniques, HighQ Pulsed CID, 465, 466 Collisions, thermalizing, 389 COM mode resonance frequency, 300, 301 Combinatorial ligand library bead, 86 Compensation electrode, 346, 347 Compensation voltage, CV, 209, 347
517
Subject Index Complex mixture analysis, 15 Complex, covalently-bound, 71 Concentration gradient, 392 Concentration, maximum allowable, 492 Confinement potential, 335 Conformational dynamics, 170 Conformational family, 219, 223 Conformational fluctuations, 186, 191, 195 peptide, 195, 198 Conformational state, 220 Conformer fluctuations, rate of, 197, 198 Conformer structure, 195 Congener(s), 480, 481, 484 Contact (or patch) potential, 339, 350 Containment lenses, 12 Contaminated extract analysis, 470 Continuous monitoring system, 492 Continuous wave (CW) laser, 188 Controlled substances, monitoring of, 388 Conversion dynode, 452 Cooling rate, 312 Cooling time, 382 Cooling, 377; see also Laser cooling Copper, oxygen-free high conductivity, OFHC, 337, 338 Corona discharge ionization, 8 Corpus callosum, 423, 424 forceps major of, 425 Correction factor Lz, 349, 350 Correlated harmonic excitation fields, CHEF, 131 Correlated sweep excitation, COSE, 131 Correlation method, 344 Coulombic attraction, long-range, 24 Coulombic explosion, 127 Coulombic force, 316 Coulombic interaction, 294, 295, 299, 302, 307, 313–315, 320, 323, 335, 337 Coulombic repulsion, 241, 334 Counter current flow, 388 Creatine phosphokinase, 219 CRL, JAPAN, 359 Cross-linking reagent, cleavable, 104, 105 non-labeled, 103 stable isotope-labeled, 103 Cross-linking strategy, 102 affinity labeled, 103 solution cleavable, 103 stable isotope labeled, 103 Cross-linking, mixed isotope, MIX, 103 Cross-section, absolute, 216, 218 calibration standards of, 215 determination of, 214–216, 219, 221, 223, 227 rotationally-averaged, 208 Cross-sections, comparison of, 218, 219 normalized, 218
Crude oil analysis, 225 Crude vegetable extract, 439, 476, 477 Cryo electron microscopy, 221 Crystal formation, inhomogeneous, 432 Crystal-rich regions, 435, 436 CsI, 177, 178 cluster, 171 C-terminal residue, 19 C-trap, 14 Cu2+, 86 Cumene, 495, 498, 500 Cumene hydroperoxide, 500 Cyclotron frequency, 125, 132, 317 Cyclotron motion, 124, 293, 317 Cylindrical ion trap, CIT, 42, 328 Cysteine residue, biotinylation of, 86 Cytochrome c, 54, 141, 133 Cytochrome C peptides’ solution, 130
D D2, 311 D2O, 12, 36–55 D2S, 38, 43 Daidzein, 155, 156, 162, 163 Daidzein-7-O-glucoside, 155, 156, 162, 163 Damped harmonic oscillator, 303 Damping coefficient, 308 Danish Natural Research Foundation Centre for Quantum Optics, 324 Danish Natural Science Research Council, 324 Database search algorithm, 84 Data-dependent MS/MS method, 139 DC axial potential, 298 DC potential, 12, 341, 346–348 DC turning quadrupole, 419 DC voltage, 317, 339 DDS, see Scan, data-dependent de Broglie wavelength, 177 De novo sequencing, 84 DE50 value, 158, 159, 161–163 Debye–Scherrer rings, 176 Decabromodiphenyl ether, 482 Decafluorotriphenylphosphine, 467–469 Decomposition pathways, consecutive and competing, 373 Degrees of freedom, 157 Deinococcus radiodurans, 129 Deinococcus radiodurans proteome, 140 Dendrimer, 19 Density functional calculation, 182, 248, 251 Density functional theory, 179 Dentate gyrus, 425 Deoxyribose monophosphate nucleotide, 54 Derivatization strategy, 100 DESI, see Ionization, desorption electrospray Desorption ESI, DESI, 224
518 Detection efficiency, 320 Detection limit(s), 447, 491, 495–497, 505 Detector, 389, 394 Deuterating agent, 43, 45 gas-phase acidity and basicity of, 38 Deuteron transfer, 39 DFTPP, see Decafluorotriphenylphosphine DI, 38, 51 Diagnostic ion, 22, 87 Dialysis-related amyloidosis, 220 Dibromochloromethane, 495 Dichloroacetate, DCA, 432 Dichlorobenzene isomers, 498 Dichlorobromomethane, 495 Dichloroethane, 502 Dichloromethane, DCM, 408, 473, 475, 476 Diesel/oil extract, 473–476 Diethylpyrocarbonate, 101 Difenoconazole, 446 Differential isotopic enrichment, 98 Differential stable isotope labeling strategy, 102 Diffraction data, 177, 179, 180, 182, 184, 185 Diffraction pattern, 176, 177, 182–185 analysis, 177 calculated, 171 measured, 171 Digital ion trap, 367, 374, 275 Dimethyl methylphosphonate, 403, 406 Dipolar mode, 368 Direct current (DC) pulse, 370 activation, 369 Discharge source, 4 Disease studies, 425 Dispersed emission spectrum, 255–257 Dispersed fluorescence, 255 Displacement reaction, 411 Dissociation yield, 241 Disulfide bond, selective cleavage, 70, 137 Disulfide linkage, 21, 22 DIT, see Digital ion trap DMDNP, see 2,3-Dimethyl-2,4-dinitropentane DMMP, see Dimethyl methylphosphonate DMNB, see 2,3-Dimethyl-2,3-dinitrobutane DMP, see 2,4-Dimethylpyridine DNA, 60, 246, 424 DNB, see 1,4-Dintitrobenzene Domain, frequency, 459 Domain, time, 459 Dominant conformers, interconversion among, 195 Doppler cooling force, 303, 312, 316 Doppler effect, 301, 312, 332, 339, 348 Doppler laser cooled, 299, 312, 319 Doppler profile, 333, 344 Doppler spectrum, calculated, 333, 334 DPM, see 3-(3-Methoxypropoxy) propanol Dried droplet method, 433
Subject Index Drift cell, 205, 207, 208 IMS, DCIMS, 206, 207 ion mobility–mass spectrometry, DCIM-MS, 207, 208, 215, 216, 218, 223, 224 resolving power, 225 Drift field, electrostatic, 389 Drift gas inlet, 390 methane, 403 nitrogen, 403 Drift length, 397 Drift region, 387, 389, 390, 393 Drift time, 216, 390 Drift tube, 391 cylindrical, 397 DriftScope program, 213 output of, 214 Drosophila melanogaster, 208 Drosophila Toll receptor, 221 Drug, active, 435 Drug, Pro-, 435 Drugs, 419, 433, 435, 436 Duty cycle, 128, 378 fast, 67 rectangular waveform, 374, 375, 378, 385 Dye fluorescence, quenching, 195 Dye–ligand affinity chromatography, 86 Dye–Trp proximity, 191 Dynamic peak range, 455
E e− - trapped ion interaction, 175 e− -beam-cloud overlap, 174 E/N, 391, 407, 408 ECD FT-ICR mass spectrum, 136, 137 ECD mass spectrum, 136 ECD of protein and peptide, 135–138 ECD, principles, 135 ECD, see Electron capture detection Effective collisions, 383 Effusive oven, 299, 342 EI, see Ionization, electron Electric circuit, switchable, 442, 443 Electric field, 205 gradient, 388 strength, 390, 404, 408 dipole, 442 higher-order multipole, 442 quadrupole, 442 Electrical breakdown, 391 Electrodes, coated, 439, 444 hyperbolic angle of, 263 Electrodynamic ion trap, 3, 4, 13, 14, 16, 25 Electromagnetic trap, linear, 357 Electron affinity, 24, 406 Electron association reactions, 411
519
Subject Index Electron attachment rate constant, 407 Electron beam, pulsed, 439, 445 Electron capture, 387, 406 cross-section, 135 detection, 484 dissociation, ECD, 21, 64, 89, 123, 130, 246 dissociative, 400 rate constant, 407, 408 thermal, 387 Electron detachment reactions, 411 Electron detachment, thermal, 387, 406, 409 Electron diffraction, 170, 171 instrument, 172 measurement, 175 pattern, 175 Electron energy, 447 distribution, 407 Electron gun, 172, 297 Electron impact ionization, EI, 36, 342, 503 Electron multiplier, 173, 452 Electron multiplying charge-coupled device, EM-CCD, 257 Electron photodetachment, 241, 254 Electron scattering, inelastic, 176 total, 175 Electron transfer, ET, 12, 20, 24, 60, coupled with PTR, 67 field-induced, 195 photo-induced, 191 plus CID, EtcaD, 64 Electron transfer dissociation, ETD, 6, 8, 15, 17–19, 21, 22, 24, 60, 64, 66, 67, 70, 72, 142 comparison with ETcaD, 65 non-dissociative, ETnoD, 64 multiple, 66 reaction, bio-ion/ion, 21 without dissociation, ET, 24 Electronic action spectroscopy, 252 spectrum, 252 Electronic excitation, 176 Electro-optic modulator, EOM, 302 Electro-optical chopper, EOC, 300, 302 Electrospray ionization, 3, 4, 7–12, 14, 15, 17, 24, 25, 36, 46, 47, 49, 52, 54, 62, 84, 91–93, 103, 154, 160, 207, 210, 215, 221, 253 source, 170, 257 orthogonal, 157 Electrostatic field, 140, 194–196, 392, 403 interaction, 194 lens, 128 e-mail, tranmission of results from unmanned site by, 503 EM-CCD, Newton, Andor Technologies, 257 End lens, 70, 126–128
End-cap electrode(s), 7, 157, 173, 190, 191, 262, 269–273, 276, 303, 337, 341, 345, 347, 350, 367, 368, 372 End-cap electrodes trap, 345, 346 AC voltage applied across, 300, 301, 303, 434 stretched out, 442 Endoproteinase Lys-C digestion, 24 Endrin, 469 Energy-resolved mass spectrometry, ERMS, 155, 157–159, 161, 164 Energy-variation study, 54 Enthalpy, 388, 394, 396 changes, standard, 397, 399 reaction, 398 Entropy, 388, 394, 396 changes, standard, 397, 399 reaction, 398 Enzymic activity, 430 Epicatechin (−), 155, 156, 162–164 Equilibrium, 393 Equilibrium constant, 397 Equilibrium, phase-partitioning, 493 Equine myoglobin, 218 Escherichia coli, 16, 140, 144, 220, 241 ESI, see Electrospray ionization ESI-FT-ICR, 140 ESI-LIT-TOF instrument, 51 ESI-QIT instrument, 51 ETD, see Electron transfer dissociation ETD/CID MS/MS, 64 Ethylbenzene, 492, 495, 497, 498, 500 Ethylbromide, 402, 403 ETnoD, multiple, 66 European Union Directive 98/83, 492 Exact hard-sphere scattering, EHSS, 222 Excimer laser, 252 Excitation probability, 360 non-resonant, 367, 369 resonant, 367, 369, 373, 374 Exciton Corporation, 255 Exogenous compounds, 433 Explosives, detection of, 388, 399 External calibration, 14 Extract(s), contaminated, analysis of, 439 Extractive electrospray ionization, EESI, 224
F Faraday cup, 172, 173, 207 Faraday plate, 390, 398 Fatty acid tail(s), 426, 430 Fatty acyl chains, 423 FC-43, see Perfluorotributylamine Fe3+, 86 FeCO2−, 70
520 FELIX, 246, 248, 249 Fiber optic, 420 Fick’s equations, 493 Field adjusting phase, 375–377 Field asymmetric waveform IMS, FAIMS, 206, 209, 228 Field test(s), 491, 496, 502, 505 Filament assembly, 448 Filtered noise field, 459, 460 Finite-element based program, 345 Finnigan 3D QIT, 67 Corporation, 466 LTQ, 420, 423 LTQ mass spectrometer, 9, 69, 71 Finnigan MAT, 440 First-Doppler effect, 332, 343 First-order Doppler shift, 342, 358 Fixed-charge derivatization, 91 Flavonoid, 154–157, 162–164 Flavonoid glycoside isomers, 49, 50 Fluoranthene, 9, 476 Fluoranthene anion, 68–70 Fluorescence, 294, 300, 334, 341, 348, 353, 418 decay, 190 detection sensitivity, 188 emission, 242, 256, 342, 349, 350, 353 spectrum, 186 excitation spectrum, 255, 256 hole, 266, 270, 271, 274, 275, 278 image from 40Ca+, 300, 307, 308 imaging system, 300 intensity, 187–189, 281, 348 lifetime, 186 measurement, 191, 193 lifetime, temperature dependence, 193, 195, 198 measurement, 191, 334 resonance energy transfer, FRET, 255, 283 spectroscopy, 240, 242, 254, 255 Fluorescent lifetime, 191, 192 Fluorophore, 418 FNF, see Filtered noise field Focusing device, 128 Forbidden optical transition, 332 Forward mass scan, 376, 377 Fourier transform, 174, 191, 460 ion cyclotron resonance, FT-ICR, 9, 15, 19, 42, 44, 45, 51, 121–144, 207, 242, 244–246, 255, 282, 293, 95 analysis, 126 instrument, 51 mass spectrometry, 388, 389, 458 principles of, 122–126 resolving power, 129 sensitivity, 130
Subject Index Fragmentation reaction, selective gas-phase, 107 threshold, 465 site-specific, 93 Franck–Condon factor, 24 Free energy, 388 Free-electron laser, FEL, 244, 246, 247, 251, 282 Free-jet expansion, 396 Frequency metrology, 328, 341, 357–359 Frequency of ion motion, 5 Frequency power spectrum, 269, 270 Frequency reference, 354 Frequency spectrum, 126, 348, 349, 460, 462 Frequency stability, 358 Frequency standard, 358 Frequency synthesizer, programmable, 301, 302, 305 Frictional force, 312, 313 FT-ICR, see Fourier transform ion cyclotron resonance Fundamental secular frequency, ωr,0, 262, 263, 268, 298, 337, 339, 343, 349, 350 Fundamental secular frequency, ωz,0, 262, 263, 268, 271, 298, 337, 339, 343, 349, 350 Fungicide, 446
G Ga3+, 86 Gas chromatography/mass spectrometry, GC/MS, 49, 439, 476, 478, 486, 496 Gas chromatography/tandem mass spectrometry, GC/MS/MS, 439, 440, 442, 454, 455, 457, 470, 486 Gas-phase acidity, 39, 40, 49 basicity, 39, 40 ion, 3 ion chemistry, 41 ions of opposite polarities, 3 reactions, 394 Gaussian peak, near-, 392 Gelatin, 154 General relativity theory, 330 Geometric parameter, η, 298 German National Institute, 345 Glucokinase, 133 Glucose polymer, 227 Glutamine, 129 Glycerol backbone, 423, 425 Glycerophosphocholine lipids, 22 Glycopeptides, 227 Glycoprotein, 134 Glycosylation, 21 Gravitational redshift, 359 Gravity wave, 330
521
Subject Index Greenwich Mean Time, GMT, 329 Grid(s), 389 Ground-positioning system, GPS, 330, 354
H H/D exchange analysis, 53 mass spectra, 48, 49 reaction, 37, 39 historical perspective, 36 deuterating agents, 38 doubly-protonated species, 47 flip-flop mechanism, 40 instrumentation, 42 ion trapping, 42 model compounds, 47 model peptides, 51 motivation for, 41 practical aspects, 40 proposed mechanisms, 38 protein, 51, 53 theory of, 37 H2, 310, 311 Halobacterium salinarum, 142 Haloethanes, 492, 497 Halomethanes, 492, 497 Hard-sphere model, 207, 264, 273 Harmonic potential, one-dimensional, 302 HASTE CID, see Collision-induced dissociation techniques, HASTE CID Hazardous substances, monitoring of, 388 HCT, see High Capacity Trap HD, 310, 311 Head-space, 500 Heart disease, molecular differentiation, 140 Heavy gases, presence of, 372 Heidelberg, 345 Helium, 368, 372, 375, 383, 440, 447 Heme, 430, 432 Hemoglobin, 219 Hemoglobin (Hb) tetramer, 215 Hepatitis B virus capsid protein, 220 Hepatitis C patient, cryoglobulins, 140 Heptabromodiphenylether, 480 Heptachlor epoxide, 469 Hewlett-Packard-Austin, 353 Hexapole ion trap, 340 Hexapole LIT, 14 Hg+, 346, 352–355, 357 High amplitude low frequency, HALF, 16 High amplitude short time excitation, HASTE, 98 High capacity trap, Bruker, 444 High performance liquid chromatography, HPLC, 154, 223, 435 /tandem mass spectrometry, 436 High resolution mass analysis, 14
High sensitivity, 15 High-energy synchrotron radiolysis, 101 Higher-order field, 262–264, 268, 271, 315, 316 Higher-order terms, 314, 315 HighQ Pulsed CID, see Collision-induced dissociation techniques, HighQ Pulsed CID Histidine, 24, 127 Histone PTM state, 70 Hitachi 3DQ mass spectrometer, 42 Hitachi M-8000 ion trap mass spectrometer, 6 Hole, in electrode, 260, 262–266, 269, 270–275, 279, 334 Homochirality, 228 Hot electron capture dissociation, HECD, 137, 140 HPLC, see High performance liquid chromatography HPLC/MS/MS, see High performance liquid chromatography/tandem mass spectrometry Human cerebrospinal fluid, 140 Hb variants, identification of, 228 HeLa cell, 144 nuclear protein, tryptic digest, 21 serum, 85 α-casein, 143 Hybrid FT-ICR instrument, 123, 131, 132, 138, 140 Hybrid instruments, 15 linear ion trap FT-ICR, 138, 139 LIT/FT-ICR instrument, 14 Q-TOF instrument, 213, 214 tandem mass spectrometers, 13 LIT /FT-ICR, 13 Orbitrap, 13 quadrupole/TOF, 13 triple quadrupole/LIT, 9 Hydrate ions, 396 Hydration reactions, 397 Hydrogen atom, labile, 45, 48, 49 non-labile, 49 Hydrogen bonding, intramolecular, 48 Hydrogen/deuterium (H/D) exchange, 36–55, 101, 221, 223 reaction, 36 Hydronium ion, 395 Hydroxyl radical probe, 101 Hyperbolic rods, 338 Hyperboloidal ion trap, 334, 345 Hyperfine transition, 332
I I−, 70 IA, 64, 71 IA coupled with CID, 70
522 Icosahedral capsid, 220 ICR cell, 122–144, 242, 243, 246 Image creation, 417, 421 Image current, 126 Images, chemically-selective, 418 Imaging, 225 mass spectrometry, 417, 426, 430, 435 spatial resolution, 225 system, 319, 320 IM-mass spectrometry, reviews, 209 Immunoaffinity chromatography, 86 IMS, see Imaging mass spectrometry IMS, see Ion mobility spectrometry IMS/MS, see Ion mobility spectrometry/mass spectrometry IMS-Q-TOF mass spectrometer, 208 In+, 355 In vitro chemical derivatization, 94 In vivo metabolic labeling, 94 Incident angle, 420 Informing power, 15 Infrared chromogenic cross-linker, IRCX, 109 Infrared multi-photon dissociation, IRMPD, 41, 109, 123, 133, 243, 244–247, 251–253, 282 In-source collision-induced dissociation, ISCID, 105 Instrument duty cycle, 12 Intermediate electrode, 374 Internal atomic oscillator, 330 Internal calibration, 14 Internal enegy, 369, 465 deposition, 368, 369 Internal standard, 433, 434 International Atomic Time, TAI, 329 Intersystem crossing rate, 198 Intramolecular interaction energy, 197 Intramolecular vibrational redistribution, IVR, 245, 249–251, 253 Ion activation, 5, 370, 439, 461 data-dependent, 64 infrared photon, 64 Ion attachment, IA, 60 Ion charge control (ICC) value, 281, 282 Ion clocks, current research, 352 Ion cloud, 174, 175, 188, 242, 243, 255, 258, 260, 264, 279, 280, 335, 337 overlap, 10, 264, 280, 348 size estimation, 278, 345 imaging, 342 increased density, 189 linear, 334 manipulation, 170, 342 spatial distribution, 278, 335, 348, 350, 358 trapped, 173, 349, 351
Subject Index Ion cyclotron motion, 123 Ion cyclotron resonance cell, 124, 242 mass spectrometer, 37 Ion detection, 11 efficiency, summary, 274 Ion ejection, 272 efficiency, summary, 274 Ion ensemble, spatial distribution of, 279 Ion fluorescence, 170, 307, 309, 310 Ion fragmentation, 7, 11, 212 Ion genealogy, 5 Ion guide, 13 RF-only, 208 Ion injection, 7, 10, 11 efficiency, 11 Ion internal energy, 61, 158 Ion isolation, 5, 10, 439, 455 Ion kinetic energy, 12, 369, 372, 446, 447 Ion manipulation, 72 Ion micro-motion, 345, 353, 359 Ion mobility spectra, 391 Type 1, 393, 395 Type 2, 393, 400 Type 3, 394, 403 Ion mobility spectrometer, hand-held, 396 Ion mobility spectrometry, IMS, 205, 387, 388, 417, 419 Ion mobility spectrometry/mass spectrometry, 394, 397, 399, 402, 403, 411 Ion mobility, IM, 205, 207 Ion mobility-mass spectrometry, 205–230 traveling wave, 205–230 applications, 205–230 Ion motion detection, 302 Ion motion frequency(ies), 348 Ion motion in QIT, dynamics of, 261 Ion motion, theoretical treatment, 262 Ion number density, 12, 16 Ion optical clock, 355 Ion parking, 16, 17, 25, 66 Ion photo-excitation, 242 Ion processing, multi-stage, 63 Ion production region, 310 Ion reaction vessel, 60 Ion selection, 370 Ion shutter, 389, 390, 392 Ion source, external, 446 moveable, 397 multiple, 62 Ion spin exchange, 328 Ion splat events, axial distributions of, 273 Ion storage time, variation of, 36 Ion swarm, 390 Ion temperature, 316, 319, 343, 354 Ion tomography study, 278
Subject Index Ion trajectory calculation, 263–266 Ion trajectory simulation, 189 Ion trajectory sImulation software package, ITSIM, 264, 276 Ion trajectory, Fourier analysis of, 263, 266, 268, 271 stable, 261, 262, 268 unstable, 262 Ion transmission mode, 12 Ion transmission time, 12 Ion transmission, efficiency, 7 Ion trap detector, 440 Ion trap dimensions, 5 Ion trap geometry(ies), 328 Ion trap housing, 258 Ion trap imperfection, 315 Ion trap loading, 317, 319, 321 Ion trap mass spectrometer, 60, 170, 187, 190, 239–284, 492 Ion trap parameters, choice of, 323 Ion trap technology, 170 Ion trap, 2D LC/MS, 447 Ion trap, micro, 295 Ion trap, miniature, design, 343 Ion trap, non-linear, 439 Ion trap, Paul-Straubel type, 347, 349 Ion trap, quadrupole, 455 Ion trap, quasi-miniature, 339 Ion trap, segmented, 338 Ion trapped simultaneously, 299, 307, 323 Ion trapping, 328 charge-sign independent, 63, 64 efficiency, 11, 349, 446 parameter, 261, 297, 320 technique, 328 Ion traps, millimeter-scale, 345, 352, 357 multi-pole, 335, 336, 339 Ion, doubly-charged, 223 forced motion of, 300-303 metal cluster, 169, 170, 173 multiply-charged, 241 sequence-informative, 70 Ion/ion chemical reaction, 60 Ion/ion chemistry, 7, 12, 15, 61, 72 Ion/ion ETD reaction, 9, 14 Ion/ion interaction, 348 Ion/ion proton transfer reactions, 15 Ion/ion reaction, 3, 4, 6, 8–11, 15, 17–20, 25, 61–63, 66, 68 cation-switching, 20 efficiency, 5 kinetics, charge-squared dependence of, 16 sequential, 14 tools, 4 vessel, 14, 64 Ion/molecule chemistry, 389
523 Ion/molecule interaction potential, 396 Ion/molecule processes, 392 Ion/molecule reaction, 62, 207, 387, 393, 395, 396, 445–467 association, 411 time, 446 Ion/molecule research, 440 Ion/neutral association, 392 Ionization efficiency, 12 Ionization energy, 447 Ionization time, 372 duration of, 441, 466 fixed, 466 Ionization, ambient, 225 Ionization, chemical, 36–40, 63, 69, 439, 447, 448, 453, 466 external, 450 hybrid, 448, 451, 454, 486 internal, 451 liquid, 439, 448–450, 481, 486 negative, 451 positive, 451 pulsed positive ion negative ion, see PPINICI reagent, 448, 481 gas pressure, 448 ions, 449 selective-ejection, 448, 449 self-ejection, 448 source, 70, 71 /tandem mass spectrometry, 483, 484 Ionization, desorption electrospray, 418 Ionization, electron, 299, 310, 444, 447, 453, 466 Ionization, internal, 444, 445 Ionization, self-chemical, 445 Ion-manipulation methodology, 60, 64 Ion–molecule complex, 37, 40, 42, 44 Ions, collisional cooling of, 262 Ions, laser-cooled, 345 Ions, mobility separation of, 212 Ions, negative, 388, 393, 406, 411 Ions, positive, 388, 411 Ions, precursor, 456, 457 Ions, simultaneous trapping of both polarities, 9 Ions, spectroscopy of, 239–284 Ion-trap dimension, r 0, 298 z0, 298 Ion-trap electrodes, 318 Ion-trap frequency, 320, 321, 323 relative shift, 318 Ion-trap-oscillator ensemble, 328 Ion-trapping device, 227 gas-phase, 73 IP-MALDI, see Matrix-assisted laser desorption, intermediate pressure
524 IRMPD action spectrum, 247–249 IRMPD FT-ICR mass spectrum, 134 IRMPD MS/MS, 140 IRMPD of protein and peptide, 133–135 IRMPD, principles, 133 Iron-containing ions, 6 ISCID mass spectral scan, 106 Isoaspartic acid residue, 22 Isobaric interferences, 429, 432 Isobaric ion identification, 417, 427 Isobaric ions, 428, 436 Isobaric separation, 429 Isobaric species, 427 Isobars, 427, 428 Isolation resolution, 457 Isolation window, 428, 433, 434 Isolation, low mass, 456 Isolation, notch, 457, 460 Isolation, two-step, 456, 457 Isoleucine, 137, 140 Isomer differentiation, 47, 49 Isomer diffraction pattern, calculated, 183 Isomer space filling structure, 181 Isophthalic acid, 48 Isotope effect, 311 Isotope-coded affinity tag, ICAT, 94, 96, 97 Isotopic cluster ions, 478 Isotopic patterns, 503 Isotopic peaks, 441 Isotopomer peak, 458 Isotopomers, 378 Italy, laws, 492 ITD, Finnigan Corporation, 370 ITD, see Ion trap detector ITD-700, 466 ITD-800, 466 ITMS, see Ion trap mass spectrometry ITQ, see Thermo Scientific ITQ, 451 iTRAQ approach, 97, 98 iTRAQ reporter ion, 99 iTRAQ-labeled peptide, 100 ITS-40, 440, 467 IVR killer mode, 251
J Jet Propulsion Laboratory, JPL, 329, 337, 339, 340, 352, 353, 355 JPL geodetic receiver, 354 Jumping, 377
K Kinetic constants, 394 Kinetic data, 387, 391, 394 Kinetic energy, 382, 383, 391, 431, 464
Subject Index Kinetic shift, 243, 245, 253 KinFit, 45, 47 Knudsen oven, 178
L Laboratoire national de métrologie et d’essais, Système de références temps espace, LNE-SYRTE, 331, 353 Lamb-Dicke parameter, 333, 339 regime, 333, 335 Laser beam, 242, 295, 297, 299, 323, 334, 347, 419 Laser capture microdissociation, 435 Laser cooling, 294, 301–304, 311–313, 318–321, 328, 332, 333, 335, 339, 341, 342, 356, 357 process, 358 Laser desorption/chemical ionization, 419 Laser diode, 328 Laser dye, 252 Laser hole, 266, 270, 271, 274, 275, 278, 280, 339 Laser modulation, 303, 305 Laser power, 281, 282, 422, 423 Laser spot size, 421 Laser tuning, 360 Laser, CO2, 243-246 Laser, fixed-wavelength, 240, 242, 243 Laser, Nd:YAG, 243, 252, 257 Laser, nitrogen, 243 Laser, titanium:sapphire, 252, 257, 281, 357 Laser, tunable IR, 244, 245, 247, 256 Laser, vacuum UV fluorine, 244 Laser-cooled mercury ions, string of, 339 Laser-induced fluorescence, LIF, 255, 257 Laser-induced reaction spectroscopy, 250 LC CID MS/MS analysis, 141 LC/MS, see Liquid chromatography/mass spectrometry LC/MS/MS, see liquid chromatography/tandem mass spectrometry LC/MSn, see Liquid chromatography/tandem mass spectrometry LC-ESI-MS, 156 LCM, see Laser capture microdissociation LC-MS/MS analysis, 85, 86, 96, 100, 140 LCQ, 444 Advantage ion trap mass spectrometer, 157 LD/CI, see Laser desorption/chemical ionization Leak valve method, 44 Lectin affinity chromatography, 86 Leucine, 137, 140 Leucine encephalin, protonated molecule of, 379–381
Subject Index Lifetime measurement, 193, 194 time-resolved, 189 Linear extended ion trap, LITE, 339, 354 Linear ion trap standard, LITS 1–4, 354 Linear ion trap standard, LITS project, 352 Linear ion trap, LIT, 9–11, 15, 24, 25, 42, 62, 63, 69, 86, 92, 98–100, 255, 294–297, 300–302, 310, 311, 313, 334, 337, 359, 424, 425, 442 chamber, 52 quadrupole array, 9 sketch of, 296 Linear polarizer, 302 Linear trap ensemble, 340 Linear two-dimensional (2D) quadrupole ion trap, 9, 337, 339 Linear two-ion system, 293, 294, 300, 303, 305, 307, 308, 315 Linearity, 491, 498, 505 Lipids, 419, 422, 424 Liquid chromatography, LC, 84, 208 /mass spectrometry, LC-MS, 49, 422, 442 /mobility separation, 228 -tandem mass spectrometry, LC-MS/MS, 64, 422, 442 Liquid nitrogen, 175 LIT, see Linear ion trap LIT-CID MS/MS, 141, 142 LIT-FT-ICR instrument, 140, 143 LIT-TOF system, 53 LMCO, see Low-mass cut-off Local atomic oscillator, 330 LOD, see Detection limits Logic atomic clock, 356 Low-critical energy process, 371 Low-mass cut-off, LMCO, 5, 64, 101, 104, 241, 372, 439, 465 LTQ, 420, 431, 432, 442 Orbitrap XL, 14 Luteolin, 155, 156, 162, 163 Luteolin-4′-O-glucoside, 50, 155, 156 Luteolin-7-O-glucoside, 50, 155, 156 Lysine, 24, 127, 129 Lysophosphatidylcholine, 430 Lysozyme, 54, 141
M Machinable macor, 346 Macromolecule ions, 19 metal-containing, 21 Macromolecule mixture analysis, 15 Macro-motion, 348, 352, 353 Magnesium, 319 Magnetic field, Earth, 317 residual, 317 Magnetron motion, 124
525 MALDI, 15, 42, 46, 84, 93, 126, 139, 156, 207, 210, 215, 225, 241, 243, 417–419, 422, 424, 429, 430, 433, 435, 436 MALDI-FT-ICR, 140 MALDI-QIT instrument, 51 Malondialdehyde, 88 Mapping, 377 Marseille, 329, 347, 359 Mass accuracy, 14, 15, 468 Mass analysis, 11, 13 data-dependent, 65 high-sensitivity, 67 Mass analyzer, 14 Mass discrimination, 392 Mass filter, 13 Mass isolation window, 10 Th, 157, 159 Mass range, 419 Mass ratio μ, 296, 303, 314, 322 Mass resolution, 310, 324, 468 higher, 439, 442 Mass spectrometric imaging, full-scan, 417, 429 tandem, 328, 417, 429 Mass spectrometry region, 310 Mass spectrometry, identifying ions by, 392 Mass spectrometry/mass spectrometry, see Tandem mass spectrometry Mass spectrum, ‘quadrupole like’, 468 Mass spectrum, simulated, 274–277 Mass spectrum, Synapt, 214, 219 Mass-resolving power, 14, 15 Mass-selected ions, optical spectroscopy of, 240 Mass-selective axial ejection, MSAE, 9, 11 instability scan, 371, 375 Mass-selective external ion accumulation, 138 Mass-selective instability, 259, 264, 272, 273, 276, 440, 441, 456 Mathieu equation, 348, 374 Mathieu parameters, 374 Matlab v. 7.0, The Mathworks, Inc., 228, 266 Matrix cluster ions, 429 Matrix effect(s), 12, 491, 499 Matrix ions, 424 Matrix solution, 421 Matrix-assisted laser desorption, atmospheric pressure, AP-MALDI, 419 intermediate pressure, 417, 430 intermediate vacuum, 420, 423, 431, 432 ionization, see MALDI Max Planck Institute, 359 m-Difluorobenzene, 37 Medial genticulate body, 425 Melittin, 67, 68 Membrane inlet mass spectrometry, MIMS, 491, 493, 494, 500–502, 505 Mercury, 327, 332, 334, 352, 356 Mercury ion, 329, 332, 334, 337, 339, 352 Messenger spectroscopy, 250
526 Metabolites, 418, 419, 422, 424, 435, 436 Metal cluster aggregation sources, 170–172 collisional relaxation, 175 ion source, 174, 175 Metal transfer, 20 Metal-ion affinity chromatography, IMAC, 86 Metal-ion insertion, 25 Metal-ion transfer, 15, 19, 20 gas phase, 20 Metal–oxide affinity chromatography, MOAC, 86 Metastable state, 342 Methane, 403 Methanococcus jannaschii, 144 Methanol, 481 Methionine residue, 91 Method detection limit, MDL, 485, 486 Methyl bromide, 401–403, 411 Mg+, 295, 299, 311 MgD+, 305, 310, 311 MgH+, 305, 310, 311, 324 Microcrystals, 435 Micro-extraction, 484 Micromotion, 316, 344 amplitude, 334, 335 minimization of, 344 Microwave frequency domain, 329, 333, 352 Microwave frequency standard, 334 Microwave interrogation, 339 Microwave-assisted D-cleavages, 24 Military preparedness, 411 Minimum energy conformation, 248, 251 Mirror image, 256 Mobile proton condition, 87, 90 Mobilities, differences in, 390 Mobilitiy measurement(s), 387, 389 Mobility cell, 227 Mobility coefficient, 396 Mobility data, 213, 226 Mobility gas, 218 Mobility separation, 224, 228 Mobility spectrum, 387 Mobilogram, 213, 214 Model proteins, top-down study, 18 Modes, electronic, 388 rotational, 388 translational, 388 vibrational, 388 Modification, S-type, 108 unique, 107 Modulation frequency, 305 Modulation voltage, 300, 301, 307 Molar Gibbs energy, ΔG, 38 Molecular beam, 295 Molecular dynamics, 161, 191, 219, 251 simulation, 191, 195, 197, 207 Molecular ion cluster, 379
Subject Index Molecular ion, photodissociation, 311, 323 Molecular ions, 310, 323 Molecular mechanics force field, 251 Molecular scattering curve, 178, 179 Molecular scattering data, 180 Molecular scattering intensity, 177, 178, 181, 182 Molecular weight distribution, 227 Molybdenum, 337, 345, 347 Monitoring, 388 prolonged, 491 Monoclonal antibodies, glycosylation of, 226 Monoisotopic mass, 129 Monoisotopomer, 378, 379 MS/MS, see Tandem mass spectrometry MSn, see Tandem mass spectrometry, multiple stages of Multiphoton dissociation, 250, 251 Multiple collisions, 369 Multiple decomposition channels, 369 Multiple photon process, 250 Multiple resonant frequencies, 480 Multiple stages of mass selectivity, MSn, 60 Multiply-charged anions, 7 Multiply-charged cations, 6 Multiply-charged ion, 3, 4, 18, 19, 61, 127 Multiply-charged precursor ions, 17 Multiply-charged reagent ions, 19 Multiply-protonated polypeptides, 6 Multipole storage-assisted dissociation, MSAD, 131 Multipoles, higher-order, 443 Multi-sector mass spectrometers, 368 Mutual ion storage, 6, 10 Mutual ion storage mode, 9, 13, 14 Mycobacterium tuberculosis, 144 Myoglobin, 133, 215, 218
N NADH dehydrogenase 1 beta sub-complex 3, 99 n-Alkyl bromide, 400–402 Nanoelectrospray ionization (nESI) source, 187, 190 Nano-liquid chromatography, 130 Nano-spray static tip, 62 Naringenin, 155, 156, 160, 162, 163 Naringenin-7-O-neohesperidoside, 155, 156, 160, 162, 163 NASA Deep Space network, 354 National Burean of Standards, NBS, 329 National Institute of Standards and Technology, NIST, 251, 329, 337, 338, 346, 353, 354, 356, 357, 396, 398 National Institutes of Health, 55 National Physical Laboratory, NPL, 329, 346, 359 National Research Council, NRC, Canada, 359
527
Subject Index Natural Sciences and Engineering Research Council, Canada, 284 N-Benzyliminodiacetoylhydroxysuccinimid, BID, 104 NCI, see Ionization, chemical, negative ND3, 37, 38, 40, 43, 51, 54 NDBA, 483 NDEA, 483 NDMA, 483 NDPA, 483 Negative chemical ionization source, 15 Negative ETD reagent ions, 14 Negative ions, 15 Neohesperidin, deprotonated, 47 Nephelometry, 154 Neurotensin, cross-linked, pELYENKPRRPYIL, 107, 108 Neutral loss, NL, 429 Newton’s equations of motion, 264 NH3, 329 Nickel acetate, 20 NIST Chemistry WebBook, 38 Nitrobenzene, 6 Nitro-compounds, 472 Nitrogen, 403, 406 Nitrophenols, 472 Nitrosamines, 483 NMEA, 483 Nobel Prize for Chemistry (2002), 127 Nobel Prize for Physics (1989), 328 Nomenclature, backbone product ions, 131 Non-exchanged isotopic contribution, 45 Non-linear field component, 260, 261, 263 Non-linear ion trap, see Ion trap, non-linear Non-linear resonance, 263, 268, 348, 350 Non-mobile proton condition, 87 Non-zwitterion, 247–249 Normalized collision energy, NCE, 463 Notch window, 460 Nozzle-skimmer dissociation, 131 NPIP, 483 NPYR, 483 N-terminal derivatization, 88 Nuclear magnetic resonance, NMR, 221 spectrometry, 154, 228 Nucleic acid, 54 mixture analysis, 16 Nucleoside, 54 Numerical simulation, 298
O O2, 319 Oak Ridge National Laboratory, 492 Octopole ion guide, 9, 15 o-Difluorobenzene, 37 Office of Water, 468
Off-resonance excitation, 41, 132 Oleic acid, 426 Oligodeoxynucleotide anions, 20 multiply-charged, 20 Oligonucleotide, 15, 54, 223 One-liter trap, 340, 341 Oppositely-charged ion populations, 4 Optical cavity-laser, 355 Optical frequency comb technique, 356, 357 Optical frequency domain, 333, 341 Optical frequency ion clock, 358 Optical frequency standards, clocks for, 341 Optical parametric oscillator, OPO, 247, 252 Optical spectrum, 240 Orbitrap, 12 Orbitrap mass spectrometer, 100 Organohalogen compounds, 497, 503, 505 Organomercurial agarose bead, 86 Organophosphate chemical warfare simulant, 228 Organophosphorus compounds, 403 Orthogonal reflectron TOF mass analyzer, 13 Oscillation frequency, 299, 303, 305, 308, 309, 314–316, 318, 319 Oversampling, 422
P P&T/GC/MS, see Purge-and-trap/gas chromatography/mass spectrometry Paclitaxel, 419 Pair correlation function, Fourier transform of, 171 Parallel ion parking, 16, 17, 25 Partial proton transfer reactions, 17 Paul trap, 5, 341, 349, 351; see also Quadrupole ion trap ion trap, 240, 242, 245, 246, 255, 257, 296, 346 PBDE-100, see 2,2′,4,4′, 6-Pentabromodiphenylether PBDEs, see Polybrominated diphenyl ethers PC, see Phosphatidylcholine PCBs, see Polychlorinated biphenyls PCI, see Ionization, chemical, positive PD, see Photodissociation PE, see Phosphatidylethanolamine Peak tailing, chromatographic, 445 Penning trap, 245, 296 Pentachlorophenol, 472 Pentapeptide, protonated, 51 Peptide backbone bond, multiple breakage, 67 Peptide backbone, 19 Peptide cation, multiply-charged, 71 Peptide identification, 96
528 Peptide ions, 372, 382 differentially-labeled, 97 multiply-protonated, 24 Peptide mass fingerprinting, 139, 140 Peptide mixture, simple, 19 labeled, 97 Peptide sequencing, 140 Peptide subset, chemical derivatization of, 86 Peptide, amino acid sequence, 84 analysis, 121 cross-linked, 102, 106 cysteine-containing, 86, 92, 101 dead-end modified, Type 0, 102 doubly-protonated, 71 highly-charged precursor, 67 histidine-containing, 86 identification of cross-linked, 105 intermolecular cross-linked, Type 2, 102–105, 107 intramolecular backbone bonds, 70 intramolecular cross-linked, Type 1, 102, 104 methionine-containing, 91 methionine sulfoxide-containing, 87 N-acetylgalactosamine-containing, 87 phosphorylated, 86, 223 photodissociation, 109 Peptide–polyphenol, 153–164 gas-phase affinity scale, 155, 161, 164 supramolecular assembly, 153–164 tannin, 153 Peptides, 419, 422, 424 multiply-deprotonated, 21 three-dimensional structures of, 207 Perfluoro-1,3-dimethyl-cyclohexane, PDCH, 6, 67 Perfluorotributylamine, PFTBA, 452, 454, 457 Periaueductual gray, 425 Period, injection, 454 isolation, 454 Permanent magnet, 126 Permittivity of vacuum, 302, 337 Pervaporation, 493 Pe-scan, 444 Pesticide(s), 477 Phase method, 320, 321, 323 Phase shift, 272 Phase-detection method, 305, 309 Phenanthrene, 452, 453, 455 Phenylboronic acid, 101 Phenylisothiocyanate, 89 Phosphatidylcholine, 423, 425–428, 431 Phosphatidylethanolamine, 428–430 Phosphatidylinositols, PI, 424 Phosphatidylserine, 428–430 Phosphocholine, 425 head group, 431
Subject Index Phospholipid ion, 429 Phospholipid(s), 422, 425, 426, 432, 433 Phosphopeptide ion, deprotonated, 19 Phosphopeptide ions, 19 CID, 22 doubly-protonated, 19 ETD, 22 methyl-esterified, 21 Phosphoric acid, losses of, 21, 142, 143 Phosphorus hexafluoride, 70 Phosphorylation, 21, 137 Photodissociation, PD, 240, 241, 252, 257, 282, 283, 369 Photoelectron spectroscopy, 182 Photo-induced dissociation, PID, 240 Photoionization beam, UV, 318, 319, 342 Photoionization, resonance-enhanced, 299 Photon detection, 319 PHPMS, see Pulsed high pressure mass spectrometry Phthalates, 472, 473 Phthalic acid, 48 isomers, 48 Physikalisch-Technische Bundesanstalt, PTB, 345, 359, 360 Planck’s constant, 341 Plasma chromatography, 206, 388 Plasma ECD, 138 mass spectrum, 138 Polarized compounds, 444 Polarized light microscopy, 435 Pole number, 336 Polybrominated diphenyl ethers, 478, 480, 481 Polychlorinated biphenyls, 439, 483–485 Polycyclic aromatic hydrocarbons, 392 Polyethylene glycol, PEG, 227 Polynuclear aromatic compounds, PNAs, 452, 454, 473 Polypeptide, 15 cation, multiply-charged, 21 ion, 9, 20 multiply-protonated, 21 Polyphenol, structure of, 153, 155 Polyproline lifetime measurements, 191 Polyproline peptide, 186, 190, 191 dye-derivatized, 189, 191 Polyvinylpolypyrrolidone, PVPP, 154 Porapak, 498 Porcine elastase, 18 Position specific mass spectrum, 421 Positive ion mode, 426 Post-ion/ion reaction (PTR) MS3 spectrum, 68 Post-mortem human brain tissue, 434 Post-translational modification, see PTM Potabilization, 497, 502 Potential well depth, 195, 299, 319, 335, 336, 339, 342, 345, 347, 350
Subject Index PPINICI, 448, 451 PQD-MS/MS, 98–100 Precision mass measurement, 309, 315 Precursor ion, 367, 368, 372, 373, 419, 427, 464, 465, 472 charge-state manipulation, 15 doubly-charged, 383 isolation, 61, 131 Precursor ions, higher-charged, 67 Pre-exponential factor, 405, 406, 411 Pre-ion/ion reaction (PTR) MS/MS spectrum, 68 Pre-scan, 422 Pressure, effect of increasing, 430 Prion diseases, 220 Prion protein, 220 Product ion, 367, 368, 383, 385 diagnostic, 91 manipulation, 18 mass spectra at selected working points, 384 in silicio, 18 mass spectra, simplification, 15, 17 mass spectrum, 10, 18, 21, 23, 66, 69, 84, 90, 91, 96, 108, 227, 372 multiply-charged, 69 selected ejection of, 373 transition, 427 a-type, 88, 89, 137 b-type, 67, 87–89, 98, 100, 109, 132, 133, 137 c-type, 69, 136, 137 y-type, 67, 87–89, 98, 100, 109, 132, 133, 137 z-type, 69, 136, 137 Product ions, 428 singly-charged, 383 Projet d’Horloge Atomique par Refroidissement d’Atomes en Orbite, PHARAO, 332 Proline residues, 16 Proline rich protein, PRP, 154, 155 primary structures, 156 synthetic, 154 synthetic B714, 155–157, 159, 161–164 IB8c, 155, 156, 163 IB934, 155, 156, 160, 162, 163 Protein A, 134 Protein analysis, 16 Protein and peptide depletion, 85 separation, 85 Protein backbone cleavage, 71 Protein conformation, 224 Protein database search, 21 Protein digestion, enzymatic, 59 Protein expression, 93 Protein folding, 101, 209 dynamics, 101 Protein identification, 14, 85
529 Protein interaction reporter, PIR, 105 Protein ions, fragmentation of, 16 Protein mis-folding diseases, 220 Protein mixture, 17 analysis of, 15 Protein quantitation, 93, 95 analysis, label-free, 94 stable isotope label, 94 Protein sequence analysis, 61 Protein sequence characterization, 72 Protein Sprouty2, 143 Protein stability information, 220 Protein structural characterization, 83 Protein structure, 101 Synapt, 219 Protein topology, low-resolution structure of, 101 Protein, analysis, 121 conformations of, 53 differential quantitative analysis, 93 highly-charged precursor, 67 multiply-protonated, 52 multiply-deprotonated, 52 Protein–polyphenols, interaction of, 154 Protein-protein interaction, 101, 102, 105, 106 Proteins, 422, 424 lubricating salivary, 154, 164 Proteome, human plasma, 208 Proteome, mouse brain, 86 Proteome, urinary, 208 Proteomic analysis, amniotic fluid, 140 Proteomics approach, LC-ESI-MS/MS, 229 Proteomics research, MS-based, 84, 223 Proteomics, application of FT-ICR, 139 Proteomics, bottom-up approach, 70, 84, 85, 101, 123, 139 Proteomics, major goals, 84 Proteomics, shotgun approach, 84, 140 Proteomics, top-down, 18, 60, 123, 139, 143 Proton affinity, 392, 449 Proton hydrate, 395 Proton mobility, 106 Proton transfer, 11, 12, 15, 24, 39, 64, 105, 395 multiple, 12, 19 reaction, PTR, 6, 16–18, 60, 449 sequential, 19 Proton-bound dimer(s), 403, 404, 406 Proxy marker, 60 PS, see Phosphatidylserine Pseudo-first order, 392, 401 Pseudo-potential trapping well, 5, 6, 339 Pseudopotential well, cross-section, 335, 336 PTM, 14, 59, 64, 72, 136, 142 PTM information, 21, 26 PTM motif, 70 PTR, 61, 66, 67, 70 Pulse width, 392
530 Pulsed axial activation, 369 Pulsed double ionization sources, 12 Pulsed dual ion source, 13, 14 Pulsed dual polarity ionization source, 12 Pulsed high pressure mass spectrometry, 389, 396, 398, 402, 406, 409 Pulsed laser, 187, 189 Pulsed Q collision-induced dissociation, PQD, 98, 99 Pulsed triple ionization source, 12 Pulsed-valve method, 44 Purge-and-trap/gas chromatography/mass spectrometry, 491, 492, 494, 496–498, 500–502, 505 Pyridine–pyridine, 397 Pyridine–water, 397
Q qcut-off, 371 QIT, see Quadrupole ion trap QIT Esquire 3000+, Bruker Daltonics, 257, 259, 261, 263, 264, 266, 270, 272, 275, 277, 278, 281, 282 QQQ, see Triple-stage quadrupole qr, 298, 368 QSTAR XL, Applied Biosystems/MDS Sciex, 13 Q-TOF hybrid instrument, 208, 223, 224 QTRAP, 64, 66 Q-Trap 2000, Applied Biosystems/MDS Sciex, 9 qu, 261 Quadratic phase relation, 459 Quadrupole array, 11, 12 Quadrupole bender, 171, 173 Quadrupole field, quasi-pure, 345 Quadrupole ion trap, 36, 42, 45, 46, 54, 61–64, 67, 70, 87, 88, 90, 96, 97, 101, 107, 190, 207, 208, 211, 240, 244, 253, 255, 258, 260, 262, 328, 339, 367, 369–371 electrodes, cross-section of, 265 instrument, 51, 104, 140 home-built, 186–188 hyperboloidal geometry of, 328 mass spectrometer, 60, 84, 155, 173 modification for spectroscopy, 257 modified hyperboloidal angle, 264, 268, 281 non-ideal, 261 Quadrupole mass filter, 215, 293, 295, 297, 334, 339, 467 Quadrupole mass spectrometer, 398 Quadrupole time-of-flight, QTOF, 97 Quadrupole-FT-ICR, 138 Quantification, 417, 432 Quantum entanglement, 356 Quantum information, 328
Subject Index Quantum jump, 342, 343 number, 360 Quartz crystal, beating, 329, 352 Quasi-stable, 372 QuEcHERS, 476 Quenching measurement, 191, 193 Quenching rate, 191, 194–196 model, 193–195, 197 temperature dependence of, 192–194 Quercetin, 155, 156, 162, 163 Quercetin-3-O-rhamnoside, 156 Quercetin-3-O-rutinoside, 155, 156, 162 Quercitrin, 50 QUISTOR, 328 qz, 173, 174, 191, 259, 261, 268–272, 278, 280–282, 350, 368 qz-axis, 370 qz-value, 446, 465
R Radar, development of, 329 Radial ion ejection, 11 Radial trap frequency, 300 Radiation pressure force, 301 Radiative lifetimes, 169 Radio frequency (RF) cavity, 332 ion trap, 169, 171, 238, 346 Radiofrequency domain, 332 Ramsey fringe, 332 Raster-step size, 422 Rat brain, 423, 424 tissue section, 425 Rate constant, 393, 394 ion/molecule reaction, 398 Rate equation analysis, 192 Rats, control, 432 Rayleigh length, 188 Reactant ion peak, RIP, 395 Reaction kinetics, 36 Reaction rate constant, measurement of, 387, 403, 404, 411 Reaction time, 392, 401 window, 16 Reaction vessel, 12 Rectangular wave voltage, periodic, 374 Rectilinear ion trap, RIT, 42, 44 Reduced mobility, 397, 406 coefficient, 391 K0, 207 Reference measurement, 321 Relative mass resolution, 293, 295 Remotely-controlled instruments, 494 Repeatability, 505 Reporter group ion, 98 Reproducibility, 491, 498 Residence time, 389
Subject Index Residual magnetic field, 317 Resolution, mass, 441, 467 Resonance ejection, 6, 132 scan, 440, 467 Resonance excitation, 157, 264, 294 /ejection, 382 Resonance frequency, 294, 313 ω+, 304, 305, 307, 317, 322 ω−, 304, 317, 322 Restriction of Hazardous Substances, RoHS, 478, 481, 482 Reversed-phase chromatography, 96, 101 Reversed-phase liquid chromatography, 156, 192 RF barrier, 9 RF circuit, 259, 260 RF drive voltage amplitude, 334, 335 RF electric field, 334 RF frequency, 298, 307, 314, 322, 341–343, 345–347 RF gain curve, 259 RF ion trap, 188, 294 RF linear trap, 332 RF modulation, 462 RF photon correlation, 342 RF potential, 175, 258, 259, 301, 316, 338, 345, 346, 350 RF power supply, 355 RF ramping, 191 RF unbalance, 9 Rhem–Weller observation, 197 Rhenate, ReO3− and attachment mechanism, 71, 72 Rhodamine 101, 272, 274, 275, 277 Rhodamine 590, 255, 256, 281 Rhodamine 640, 188, 189 Rhodamine 6G, 255 Ring electrode, 7, 191, 258–265, 270, 271, 275, 339, 345, 348, 350, 371, 374 stacked, 208 Ring trap, 347 Ring-down time, 241 Robert A. Welch Foundation, 55 Round rods, 338
S Saccharomyces cerevisiae, 96, 142 Safe Water Drinking Act, SWDA, 468 Sample gas inlet, 390 Sample preparation, 421 Saturn 2000 3D, 449, 467 Saturn 4D, 440 Saturn, Varian, 370 Saturn-I, 440, 467 Saturn-II, 467 Saturn-III, 467
531 Saturn-IV, 467 Scan function, 371, 372, 374 chemical ionization, 449 triple resonance, 442, 467 Scan speed, 443 higher, 439, 442 Scan table, 374, 382 Scan, data-dependent, 464 stepped normalized, 464 Scattering rate, 320 Sciatic nerve, 432 SCSI-MS instrument, sketch of, 294 Second, time unit, 329 Secondary ion mass spectrometry, SIMS, 418 Second-Doppler effect, 332, 334, 339, 354, 359 Secular frequency(ies), 16, 317, 368, 457, 462, 463 Secular frequency, see Fundamental secular frequency summary of calculated values, 267 SELECT, 91, 92, 100 Selected reaction monitoring, SRM, 427 Selective ion monitoring, SIM, 469 Selective ion storage, SIS, 467 Self-CI, see Ionization, self-chemical Sequence coverage, complete, 21 Sequence information, 21 complementary, 21 Serine octomer, conformational structure of, 228 Serum albumin, 134 Servo-loop, 343, 353 scheme, 330 Sewage treatment plant, 503, 505 Shewanella oneidensis, 144 Shuttle trap, 339 Sickle-cell anemia, 222 Side-chain losses, 24 Sigma Aldrich Corporation, 255 Signal-to-noise ratio, S/N, 383, 447 Silver nitrate, 20 SIMION model parameters, 265 SIMION v. 8, Scientific Instrument Services Inc, 261, 263–268, 273, 278, 345 Simulation trajectories and histograms, 195 Dye–Arg+, 195, 196 Dye–Trp, 195, 196 Trp–Arg+, 195, 196 Single ion oscillation frequency, 302 Single ion preparation, 341 Single ion, sympathetically-cooled, 291, 299, 300 Single trapped ion, 170, 333, 342, 345–347, 355, 357, 360 laser spectroscopy of, 170, 342, 345, 346 Single-ion mass spectrometry, 291, 292, 311, 318
532 Singly-charged anions, 6 Singly-charged ions, large, 17 Skimmer lens, 105 Slides, glass, 431 Slides, indium–tin oxide coated glass, 422 Slides, non-conductive plain glass, 422 Small molecule analysis, 419 Small molecule identification, 430 S-methyl 5,5′-thiodipentanoylhydroxysuccinimide, 106, 108 structure, 107 SN2 displacement, 400 Sodium, 425 Sodium acetate, 425 Sodium adduct ion, 431, 433 Sodium dodecyl sulfate polyacrylamide gel electrophoresis, SDS-PAGE, 101, 154 Sodium ion adduct, 423 Sodium phosphocholine, 427 Solid-phase isotope-labeling strategy, 96 Solvents, chlorinated, 492 Sonic spray ionization, SSI, 11, 62 SORI-CID, 122, 130, 132, 133, 135, 246 principles of, 132, 133 Source region, 105, 387, 389 corona discharge, 389 radioactive nickel foil, 389 ultraviolet discharge lamps, 389 Source, switchable, 447 Space charge, 467 control, 417, 422 interaction, 189 limit, 240 limited density, 174 Space-charging, 423, 424 Spatial resolution, 422 Spatially-resolved measurement, 293 Spectrograph, Shamrock 303i, Andor Technologies, 257 Spectroscopy, action (or consequence), 240, 241 Sphingomyelin, SPM, 425, 432, 433 Spinal cord, 432 Splenic tissue, cryosectioned, 435 Sprague-Dawley rat, 425 Spring constant, 302 Sr+, 355 SSI, 12 Stability diagram, 26, 268, 270, 348, 351, 353, 369–371, 373, 375, 382, 383, 385 boundaries of, 350, 352, 378, 382 computed and experimentally-determined, 380, 381 cross-section of, 351 theoretical, 374 Stability region, 298, 378 first, 378
Subject Index Stability, instrumental, 505 Stability, reliability, 505 Stable isotope labeling by amino acids in cell culture, SILAC, 94 Stable isotope labeling strategies, summary, 95 Stable trajectory region, 380 Stacked-ring ion guide, SRIG, 210 Stark shift, 359 Stearic acid, 426 Stilbene, 36 Stored waveform inverse Fourier transform, SWIFT, 131, 174–176, 191, 458–460 Streptavidin affinity chromatography, 86 Strong cation exchange (SCX) chromatography, 86 Strontium ion, 346 Structural characterization, 425 Styrene, 495, 497, 500 Sub-Doppler laser cooling, 341 Substatia nigra, 425 Sulfonium ion derivatization, 92 Sulfur dioxide, 6 Sulfur hexafluoride, 407 Supersonic jet, 245 Supplemental fields, dipole, 442, 462, 464 Supplemental fields, quadrupole, 442 Supplemental waveform, 66 Supplementary radiofrequency voltage, 347, 367 Supplementary RF signal, 16 Surface charge, 339 Surface waters, pollution of, 503 Surface-induced dissociation, SID, 130 Sustained off-resonance irradiation CID, see SORI-CID Swarm experiments, 407 Sweep frequencies, 369 Switching circuits, 374 Sympathetic cooling, 292, 300, 356, 360 Sympathetically-cooled single ion mass spectrometry, SCSI-MS, 292, 293, 295, 296, 299, 300, 303, 305, 309, 311, 312, 321, 323, 324 Synapt high definition mass spectrometer, HDMS, 210 Synapt instrument, 212, 215, 218, 221, 223–230 Synapt, modes of operation, 213 schematic diagram, 211 Syrian Hamster protein, 214 Syrian Hamster PrP protein, 220
T Taggants, 399 Tailored waveform, 16 Tandem mass spectrometric analysis, 96, 97
Subject Index Tandem mass spectrometric strategy, 105 Tandem mass spectrometry, 59, 87, 89, 91, 92, 104, 105, 122, 130, 367, 372, 417, 419, 424–428, 434–436 analysis, 97, 153–164, 214, 224, 226, 230 multiple stages of, 9, 14, 61, 62, 72, 83, 84, 105–107, 133, 212, 226, 367, 425, 426, 430–432, 436 functionality, 25 interrogation scheme, 64 scan type, 61 scan, 100 Tandem-in-space, 454 Tandem-in-time, 5, 454 Tantalum, 345, 346 Tekmar velocity XPT purge and trap, 496 Terahertz (THz) frequency domain, 329 Terephthalic acid, 48 Tetrachloroethylene, 492, 495, 501, 504 Tetrameric transthyretin (TTR) complex, 222 Tetramethylrhodamine, 191 Thermal equilibrium, 388 Thermalization, 389 Thermo Electron Corporation, 420 Thermo Finnigan LCQ mass spectrometer, 42, 43 Thermo Scientific, 440 Thermo Scientific ITQ, 447 Thermodyamic data, 387, 391, 394, 397, 398 Three-dimensional (3D) quadrupole ion trap, 5 Tickle activation, 347, 348, 372 Tickle frequency, 348, 349 Tickle voltage, 347, 348, 372 Time metrology, 328, 341, 359 Time-domain signal, 126, 207 Time-of-flight mass spectrometer, 293, 295, 419 Time-of-flight, TOF, 9, 211 Time-of-flight/time-of-flight, TOF/TOF, 97 Time-resolved measurement, 293 TiO2, 86 Tissue analysis, 417, 422 Tissue sample, 421, 429 Tissue section(s), 418, 430, 431, 433, 435 Tissue specimens, 424 Tissue studies, 423 Tissue surface, 419 Tissue, intact, 426 ovarian, 419 TMPP-Ac derivative, 90 TMPP-Ac-OSu, 88 TMPP-AcSC6F5 bromide, 88, 89 TOF, 12 TOF mass analyzer, 14, 171, 172, 208 TOF mass spectrum, 174 TOF, orthogonal acceleration, 225 TOF, see Time-of-flight mass spectrometer Toluene, 492, 495, 498, 501–504
533 Torus, 345 Total ion chromatogram, TIC, 472, 473, 476–478, 483, 484 Total ion count, TIC, 423 Total organic carbon, TOC, 483 Toxicological drug study, 435 Transcriptional editing process, 59 Transferrin, 141 Transition metal complex cations, 20 Transition metal ion insertion, 20 Transition, fluorescence of, 330 Transmission mode ETD, 23 Trap ring aperture, 188 diameter, 188 Trapped ion cloud, 135 Trapped ion electron diffraction, TIED, 169–171, 173 Trapped ion fluorescence, 186 spectroscopy, 187, 241 Trapped ion instrument, 60 Trapped ion laser excitation, 187 Trapped ion mass spectrometer, 72, 240–245, 247, 254, 282, 283 Trapped ions, 333 activated by IR laser, 54 dynamics, 169–199, 348 structure, 169–199 Trapping by proxy, 5 Trapping field imperfection, 316 Trapping field, hexapolar component of, 263 Trapping frequency calibration, 456 Trapping oscillation, 124 Trapping parameters, 349, 369 Trapping, selective, 456 Traps on micro-chips, 328 Traveling wave IMS, TWIMS, 206, 209, 212, 213, 215–225, 228 Traveling wave ion guide, TWIG, 209–212 Traveling wave, T-Wave, 209–212 Tributylamine, 220 Trichloroethylene, 492, 495, 498, 502–505 Triethylene, 220 Trihalomethanes, 492 Triple-stage quadrupole, 367–369, 427 /LIT mass spectrometer, 11 TriWave, 210, 211, 215 trp RNA binding protein, TRAP, 219 Trp-11 neurotensin, 20 Trp-cage protein charge states, 192 Tryptic lectin glycopeptide, 23 Tryptic peptide, 208 Tryptic protein digest, 219 Tryptophan, Trp, 186, 191, 219 Tumor detection, 419 Turning quadrupole, 6–8 T-wave, 211, 215, 216; see also Traveling wave IMS, TWIMS
534
Subject Index
Two-dimensional differential gel electrophoresis, 2D DIGE, 93 Two-ion crystal, 305 system, 300, 305, 319, 321, 323
vMALDI, see Matrix-assisted laser desorption, intermediate vacuum Volatile organic compounds, VOCs, 492–494, 499, 505
U
W
U.S. Environmental Protection Agency, USEPA, Method, 492 U.S. Naval Observatory, 329 U.S. Navy Observatory, USNO, 354 Ubiquitin, 54, 69, 126, 133 Ultra low expansion, ULE, spacer, 356, 357 Ultra-violet photon dissociation, UVPD, 41 Undersampling, 422 Unipolar mode excitation, 348 Universal constant, 330 Unstable, 368 Upper m/z limit, 14 Urinary metabolites, 208 USEPA Method 505, 484, 485 USEPA Method 521, 439, 470, 472, 481, 483, 484 USEPA Method 524.2, 470 USEPA Method 525.2, 469 USEPA Method 527, 470 USEPA Method 528, 470 USEPA Method 529, 470 USEPA Method 603, 498 USEPA Method 8260, 470 USEPA Method 8260B, 495–497, 505 USEPA Method 8270, 439, 468–470, 473, 475 USEPA Method SW-846, 470 USEPA Methods, 439, 467 USEPA, see U.S. Environmental Protection Agency UV photodissociation, 93
Wannier expression, 391 Wastewater(s), 497, 499, 502, 503 Water, 396, 405 drinking, 483, 492 industrial, 501 reagent, 486 surface, 486 Waters Corporation, 210, 213 Waveform, broadband, 456 Waveform, notch, 456, 457, 460–462 Wine, 153 Wine astringency, 153, 154, 164 Wine tasting, 155 Wine, organoleptic property, 153 Wire mesh, micrometric, 347 Working points, 371
V Van der Waals complex, 243 Van’t Hoff plot, 396, 399, 400 Varian 4000 ion trap, 442 Varian 4000MS, 449–451, 454 Varian 500MS, 444 Varian quadrupole ion trap mass spectrometer, 494 Varian Saturn, 440 Varian Star 3400X Saturn 2000 GC/MS, 496 Varian turbo DDS, 464 Velocity of light, 330 Vial shield, 450 Vibrational relaxation, 176 Vibrational temperature, 171 Vinblastine, 225 Vinyl chloride, 502 Virus capsids, 220 Virus tails, 220 Viscous damping force, 303, 345
X Xenon cations, 21 X-ray, 173, 221 scattering, 219 Xylene(s), 492, 500
Y Yb+, 355 Yeast enolase, tryptic digest, 91, 92 Ytterbium, 332, 334
Z z0, 173, 296 Zeeman effect, 354, 357, 359 Zidovudine, 228 Zoom scan mode, 443 ZrO2, 86 Z-spray source., 210, 211 Zwitterion, 247, 249 α-Helical PrP, 220 α-Helical PrP, β-sheet-rich structures, 220 α-Methylstyrene, 500 β-Casein phosphopeptide, 137 β-Galactosidase, 141 β-Lactoglobulin, 54 βr, 370, 372, 373, 380, 382 βu, 261, 262 βx, 350, 352, 353 βz, 268, 270–272, 276, 350–353, 370, 382, 383, 385 μ-Metal shield, 172