Mass Spectrometry Volume 9

A Review of the Recent Literature Published between July 1984 and June 1986

By M. E. Rose

The Royal Society of Chemistry

Copyright © 1987 The Royal Society of Chemistry
All rights reserved.
ISBN: 978-0-85186-338-2

Contents

Chapter 1 Ionization Processes and Ion Dynamics By C. Lifshitz,
Chapter 2 Structures and Reactions of Gas–phase Organic Ions By M. A. Baldwin,
Chapter 3 The Chemistry of Gas–phase Ion Clusters By A. J. Stace,
Chapter 4 Developments and Trends in Instrumentation By T. R. Kemp,
Chapter 5 Applications of Computers and Microprocessors in Mass Spectrometry By J. R. Chapman,
Chapter 6 Reactions of Organic Negative Ions in the Gas Phase By J. H. Bowie,
Chapter 7 Analysis of Mixtures by Mass Spectrometry Part I: Developments and New Applications of Gas Chromatography/Mass Spectrometry By R. P. Evershed,
Chapter 8 Analysis of Mixtures by Mass Spectrometry Part II: Liquid Chromatography/Mass Spectrometry and Supercritical Fluid Chromatography/Mass Spectrometry By M. E. Rose,
Chapter 9 Mass Spectrometry Applied to Natural Products: Steroids By I. Howe,
Chapter 10 Drug Metabolism, Pharmacokinetics, and Toxicity By D. J. Harvey,
Chapter 11 Metal–containing and Inorganic Compounds Investigated by Mass Spectrometry By J. Charalambous,
Chapter 12 The Current State of Quantitative Metal Analysis by Mass Spectrometry By D. E. Pratt, J. Eagles, and R. Self,
Subject Index, 431,
Author Index, 439,

CHAPTER 1

Ionization Processes and Ion Dynamics

BY C. LIFSHITZ

1 Introduction

Previous reviews of ionization processes and ion dynamics by I. Powis set a certain format, to which the present reviewer will try to adhere. Thus, the topics to be covered will be the same, except for the discussion of van der Waals molecules and reactions of ion clusters, which are dealt with in a separate chapter in this volume. Considerable progress has been made in the last two years in our understanding of ion dynamics – dynamics of ionization processes as well as dynamics of uni- and bimolecular ion reactions. My task has been to present work related to fundamental aspects of the behaviour of relatively simple molecular ions. More complex systems will be dealt with in coming chapters. While the molecules themselves are rather simple, the physics and chemistry involved in ionization processes and ion dynamics is quite complex. I have therefore chosen to try to explain, in a rather simplified manner and in some detail, terms which mass spectrometrists do not normally encounter, at the risk of repeating some of the definitions given previously.

2 Ionization Processes

2.1 Molecular Photoionization.- Molecular photoionization can be studied by photoelectron spectroscopy (PES) which is a well established technique and gives valuable information on the energetics of the photoionization processes and therefore on molecular electronic structure. Where is the dividing line between the spectroscopic and dynamical aspects of molecular photoionization? This has been discussed in several recent review articles. Dynamical information is gleaned from measurements of relative cross-sections for individual processes, as a function of the incident photon energy, and from cross-sections as a function of the angle between the polarization vector of the light and the direction of the ejected photoelectron. The availability of intense tunable polarized radiation provided by synchrotron sources has made all of these experiments possible and one is talking about triply differential photoelectron measurements when the photoelectron intensity is measured as a function of the wavelength of the incident light, of the photoelectron kinetic energy and of the ejection angle. Furthermore, synchrotron radiation has enabled the extension of the energy scale to hitherto unavailable ranges and – as a result – the study of phenomena (such as shape resonances and configuration interactions, see coming paragraphs) which are more prevalent at higher energies.

The study of cross-sections and angular distribution parameters, 3, as a function of photon energy and photoelectron energy has revealed interesting structures, which have been discussed and explained previously in this series. Three major categories are distinguishable: shape resonances, autoionization resonances and Cooper minima.

Resonances normally appear as an intensity maximum superimposed on the photoionization continuum cross-section. Most shape resonances studied to date have been observed for ionization of core (K or L shell) electrons. The resonance observed is due to the temporary excitation of the core electron to an unoccupied orbital of the molecule. The electron-ion pair remains temporarily bound in spite of the energy exceeding the ionization limit, due to the existence of a centrifugal barrier particularly for d,f, etc. electrons of large 1. A potential well is formed by the combination of long-range attraction and short-range repulsion. If the well is deep enough to support a (quasi-) bound energy level of the electron, there will be a strong enhancement in the photoionization cross-section. In addition to enhancement of cross-sections, shape resonances influence the angular distributions of photoelectrons and lead to non-Franck-Condon effects in vibrationally resolved spectra. They are specifically a molecular phenomenon induced by the angular anisotropy of the molecular potential. In view of the short lifetimes of the respective quasibound states, shape resonances result in broad structures which may extend over several eV. Shape resonances have been observed for ionization of valence electrons in addition to ionization of core electrons, but are much less understood in the former case. Previous experimental studies of shape resonances in several molecules have been extended and several additional molecules added. The molecules studied include N2, CO2, CCl4, SiCl4, GeCl4, and BF3. As noted earlier, there has been a strong interplay between experiment and theory and theoretical calculations have been carried out for N2, NO, cyanogen as well as other linear molecules. Calculated features, particularly for CO2, are sharper and more intense than experimental ones and new calculations are needed to elucidate the discrepancies.

Cooper minima are minima in the photoelectron cross-section caused by the disappearance of a transition to a major channel in the continuum. The latter is caused by a change in sign, as a function of photon energy, of the matrix element, which characterizes the transition. Minima in both the cross-sections and 3 values were found for each of the first five orbitals of SiCl4 and experimental and theoretical work was compared with earlier work on CCl4. A change in sign in the matrix element requires a node in the orbital. The Cooper minimum in CCl4 is associated with the radial node anticipated for the 3p subshell in chlorine. In the case of the 2p orbital of fluorine, no node is expected and thus no Cooper minima are seen in angle-resolved photoelectron cross-sections of CF4. Theoretical analyses of Cooper minima in angular distributions and cross-sections were carried out for HCl and HI.

Autoionization is a multichannel process. Photoionization takes place via absorption into a discrete excited state followed by decay into an underlying continuum. Autoionization is the most important process in molecular photon absorption between 10 and 20 eV. Contrary to shape resonances, autoionization resonances are usually sharp. Also, electronic autoionization is a two-electron resonance and electron correlation must be included in its theoretical treatment. In most cases, autoionizing states consist of an excited Rydberg electron bound to an excited ion (also called “core”). A molecular ion core can store the energy needed to ionize a Rydberg electron in any of its three modes – electronic, vibrational or rotational.

Spin-orbit autoionization between the ionization thresholds for the II and 2Π1/2 states of HI+ has been studied theoretically by applying an ab initio approach to the multichannel quantum defect theory (MQDT) and experimentallyby angle-resolved photoelectron spectroscopy applying the constant ionic state (CIS) method. The CIS method has also been employed to study autoionization in N2O2. In the CIS method the ionizing photon energy is advanced in synchronization with the analysis of the kinetic energy of the photoelectron so that the data are always collected under conditions that correspond to a given final ionic state. Angle-resolved photoelectron spectroscopy has detected atomic autoionization of the Br atom following very fast dissociation of core-excited HBr.

A window resonance results from a strong destructive interference from direct photoionization that causes a sharp depression in the cross-section rather than a maximum in intensity. Window resonances were studied for N2O2 by angle-resolved photoelectron spectroscopy and for S2 by molecular beam photoionization. New vibrational assignments for the autoionization bands of O2 were made on the basis of isotope shifts. Rotational lines in autoionizing Rydberg states were resolved for NO. The latter study combined rotational cooling of the neutral molecules through supersonic expansion with high-resolution VUV ionization achieved by non-linear optical effects – so-called “four-wave mixing”. Vibrational autoionization was studied in PF3. In this type of autoionization, vibrational energy of the molecular ion core is converted to electronic energy of the ionized Rydberg electron. The current theory of vibrational autoionization predicts a propensity rule Δv= –1, where Δv is the change in vibrational quantum number in the autoionization process. In contrast to this rule, vibrational autoionization in PF3 is occurring with Δv ≤ –13. This and other examples suggest that a more general theory of vibrational autoionization is required. Theoretical studies have been carried out’ concerning electronic autoionization in H2. A theoretical study of cdmplex resonances near ionization thresholds has been applied to the N2 photoionization spectrum. A complex resonance consists of a central peak surrounded by a broad distribution of satellites and has been interpreted as due to the simultaneous autoionization of a dense series of Rydberg levels converging to a low-lying threshold and of a single low-n level (the “interloper”) pertaining to a series with a much higher threshold.

Several other systems for which autoionization phenomena were studied include silane, studied by threshold photoelectron spectroscopy (TPES) and photoionization mass spectrometry (PIMS), the OH radical and 1,1-dichlorodifluoroethene, both studied by PIMS. New and interesting results were found for the thoroughly studied case of vibrational autoionization in molecular hydrogen. Lifetimes of radiative excited levels of H2 were measured. Fluorescence survives when ionization is energetically possible and most levels decay via three competing channels: fluorescence, predissociation and autoionization.

Following photoionization resonances, another topic of great current interest involves correlation effects in the ionization of molecules and the breakdown of the molecular orbital picture. These phenomena may have far – reaching implications concerning ion fragmentations and our understanding of mass spectrometry in general. In the excitation energy range up to ~20 eV there is usually a one-to-one correspondence between the number of occupied molecular orbitals in the molecule and the number of bands (i.e. “ionization energies”) appearing in its photoelectron spectrum (except for well known and well understood splittings such as the ones due to spin-orbit or Jahn-Teller interactions or other effects due to ionization of open-shell molecules). This is no longer the case for ionization of inner-valence orbitals by excitation energies above ~20 eV or in some cases even at lower energies. The PES of these inner-valence MO’s demonstrate a great number of lines (satellites). This phenomenon arises primarily from multiple-electron excitations and the satellites are therefore shifted from the main beam. For some inner-valence orbitals, electron correlation effects are so large that single-electron ejection occurs in a minority of events. This can be understood in the following pictorial way: When an electron is ionized from a molecular orbital a “hole” is produced. This is called, in the terminology of Cederbaum et al., a single-hole (1h) configuration. If the ionized electron comes from an inner-valence orbital, there are many excited electron configurations adjacent in energy, for example states in which two holes are formed and an electron occupies an orbital which was empty in the original molecule. Such states are called two-hole, one-particle (2h-1p) configurations, where holes and particles refer to orbitals occupied and unoccupied in the molecular ground – state configuration, respectively. If the states are of the right symmetry, the ionic wavefunction is a superposition of the various lh, 2h-1p etc. configurations. This comes about because the electrons are “correlated” and the “configurations interact” and hence the name “electron correlation effects” or “configuration interaction”. This implies that if one can associate a well defined physical meaning to the concept of an inner-valence orbital this is not the case for the corresponding hole. The satellite lines in the spectrum borrow the oscillator strength from the main line through their mixing coefficients. If a main line can no longer be distinguished one is talking about the breakdown of the molecular orbital picture. Multielectron excitations clearly play a very important role in photoionization dynamics.

According to Koopmans’ theorem the ionization energy is equal in value and opposite in sign to the binding energy of the electron in the molecular orbital. Due to the breakdown of the molecular orbital picture, and since some ionic states are more prone, for symmetry reasons, to configuration interactions than others, the ordering of ionization energies is often modified. This is particularly evident in some molecules such as COS, cyanoethylenes, transition-metal complexes’, and nitrosomethane.

From the theoretical point of view, a Green’s function method based on a so-called “two-particle-hole extended Tamm-Dancoff Approximation” (2ph ex-TDA) has been quite valuable for calculating complete valence-shell ionization spectra of small- and medium-sized molecules, and the breakdown of the MO picture of ionization as a general phenomenon in the inner-valence region is now well established. In some cases the quantitative agreement with experiments is not so good due to basis set limitations and to the missing of three-hole – two-particle excited configurations.

The two major experimental methods for studying configuration interaction states are by tunable synchrotron radiation and dipole (e,2e) spectroscopy. In dipole (e,2e) spectroscopy, absolute oscillator strengths are obtained for photoionization processes, by fast electron impact. The incident electron which ionizes the molecule ejects an electron from the molecule (eej) and is scattered (escatt). Coincident detection of eej and escatt at a given energy loss simulates tunable energy PES. Among the molecules studied by these methods are CO2, H2S, HBr and H2O. Theoretical calculations of the configuration interaction states of CO2+ using an ab initio SCF-CI method showed that 3h-2p excited configurations are necessary to explain the line positions and intensities. Valence-shell binding–energy spectra of HBr at values of the energy loss (photon energy) of 20, 30, 40 and 45 eV are reproduced in Figure 1. The [??]2Π and [??]2Σ+ states of HBr+ give single peaks and thus are well described by a single-particle picture. In contrast, the [??]2Σ state of HBr+ is seen to split into a number of final ion states (MET, multiple electron transitions), clearly demonstrating the breakdown of the molecular orbital picture above 20 eV.

A powerful experimental method developed in recent years is Fluorescence Excitation Spectroscopy (FES). For partial photoionization cross-sections associated with ions produced through the ejection of inner electrons, only FES has provided threshold measurements. FES was employed to measure partial photoionization cross sections associated with the A2Π and B2Σ+ states of CO+; it has demonstrated the electronic autoionization of five Rydberg series converging to the B2Σ+ state into the CO+ A2Π state rather than to CO+ X. This was concluded on the basis of comparison of the CO+ (A-X) fluorescence cross-section dependence upon photon energy with that of the total photoionization cross section over the same energy range. Using molecular nitrogen as an example, it was demonstrated that FES can be used to measure partial photoionization cross-sections in external electric fields. The threshold for production of the N2+ (B2Σu) state was found to shift linearly with the square root of the applied field. Recently, the technique of dispersed fluorescence was coupled with synchrotron radiation excitation for the first time. The vibrational branching ratio for (v’=1)/(v’=0) in N2+2B+Σu was observed to undergo resonant enhancement at an excitation energy, hv = 29 eV, which was attributed to a shape resonance. Polarization fluorescence measurements give valuable information on photoionization dynamics. Fluorescence polarization from molecular photoions reflects the degree of alignment of the molecular ion in the laboratory-fixed frame, which is in turn determined by the relative dipole strengths for degenerate photoionization channels which have different symmetries in the molecule-fixed frame, i.e. different magnetic sublevels are not equally populated. This field has been pioneered by Zare and coworkers and has recently been employed for comparison of 12CO2+ and 13CO2+ alignment following photoionization of carbon dioxide. The observed [??] state alignment for 12CO2+ is smaller than for 13CO2+. In 12CO2+ the [??] state is perturbed by the [??] state, the alignment of which is of opposite sign to that of the [??] state, while in 13CO2+ the [??] state is unperturbed.

(Continues…)Excerpted from Mass Spectrometry Volume 9 by M. E. Rose. Copyright © 1987 The Royal Society of Chemistry. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.