@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Science, Faculty of"@en, "Chemistry, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Zhang, Wenzhu"@en ; dcterms:issued "2011-03-03T06:58:10Z"@en, "1991"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """Absolute photoabsorption differential oscillator strengths (cross sections) for N Is and 0 Is inner shell excitation and ionization of N0₂ have been derived from the presently obtained high resolution electron energy loss spectra using dipole (e,e) spectroscopy. The N0₂ inner shell differential oscillator strength spectra are in good agreement with multichannel quantum defect theory calculations in both excitation energies and (differential) oscillator strengths. The N0₂ spectra were interpreted with the aid of the calculations. A resonance present in both the N0₂ N Is and 0 Is ionization continua was identified as excitation to an unbound molecular orbital. A consideration of the present spectra for N0₂ and earlier spectra for other molecules showed that additional prominent structures observed in previously reported N0₂ inner shell photoabsorption spectra obtained using synchrotron radiation were due to the presence of impurities. Absolute photoabsorption differential oscillator strengths for the valence shells of CF₄, CF₃CI, CF₂Cl₂ and CFCI₃ have been measured in the equivalent photon energy range up to 200 eV using dipole (e,e) spectroscopy. The present results are in good agreement with earlier reported optical measurements and electron impact measurements. The photoionization efficiencies and also the photoion branching ratios have been determined for CF₄, CF₃CI, CF₂C1₂ and CFCI₃ from time of flight mass spectra using dipole (e,e+ion) coincidence spectroscopy at equivalent photon energies ranging up to 80 eV. Absolute partial differential oscillator strength spectra for the molecular and dissociative photoions have been derived. The natures of the dipole induced breakdown pathways of CF₄, CF₃CI and CF₂C1₂ were investigated by combining the present differential oscillator strength measurements for molecular and dissociative photoionization with photoelectron data obtained in the present thesis work plus previously reported photoelectron branching ratios. On the basis of the present work, a revised set of absolute electronic state partial photoionization differential oscillator strengths for CF₄ are presented. Absolute electronic state partial photoionization differential oscillator strength spectra for CF₃CI, CF₂C1₂ and CFCI₃ in the photon energy range 41—160 eV have been derived by combining total differential oscillator strength spectra obtained in the present thesis work with photoelectron branching ratios obtained from photoelectron spectra measured using synchrotron radiation. Absolute photoabsorption differential oscillator strengths and for F Is, C Is, and Cl 2p,2s inner shell excitation and ionization of freon molecules CF₄, CF₃C1, CF₂C1₂, CFCI₃ and CCl₄ have been derived from the presently obtained high resolution electron energy loss spectra plus previously reported. Tentative assignments of the spectra were obtained using the MO picture and the potential barrier model."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/32006?expand=metadata"@en ; skos:note "O S C I L L A T O R S T R E N G T H S F O R P H O T O A B S O R P T I O N A N D P H O T O I O N I Z A T I O N P R O C E S S E S OF F R E O N A N D N 0 2 M O L E C U L E S B y Wenzhu Zhang B . Sc. (Physics) Tsinghua University, Bei j ing, C h i n a , 1982 A T H E S I S S U B M I T T E D I N T H E R E Q U I R E M E N T S D O C T O R O F P A R T I A L F U L F I L L M E N T O F F O R T H E D E G R E E O F P H I L O S O P H Y i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S C H E M I S T R Y We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A September 1991 © Wenzhu Zhang, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract Absolute photoabsorption differential oscillator strengths (cross sections) for N Is and 0 Is inner shell excitation and ionization of N 0 2 have been derived from the presently obtained high resolution electron energy loss spectra using dipole (e,e) spectroscopy. The N 0 2 inner shell differential oscillator strength spectra are in good agreement with mult i -channel quantum defect theory calculations i n both excitation energies and (differential) osciUator strengths. The N 0 2 spectra were interpreted with the aid of the calculations. A resonance present i n both the N 0 2 N Is and 0 Is ionization continua was identified as excitation to an unbound molecular orbital . A consideration of the present spectra for N 0 2 and earlier spectra for other molecules showed that additional prominent structures observed i n previously reported N 0 2 inner shell photoabsorption spectra obtained using synchrotron radiation were due to the presence of impurities. Absolute photoabsorption differential oscillator strengths for the valence sheUs of C F 4 , C F 3 C I , C F 2 C 1 2 and CFCI3 have been measured i n the equivalent photon energy range up to 200 e V using dipole (e,e) spectroscopy. The present results are i n good agreement w i t h earher reported optical measurements and electron impact measurements. The photoionization efficiencies and also the photoion branching ratios have been determined for C F 4 , C F 3 C I , C F 2 C 1 2 and CFCI3 from time of flight mass spectra using dipole (e,e+ion) coincidence spectroscopy at equivalent photon energies ranging up to 80 eV. Absolute part ia l differential oscillator strength spectra for the molecular and dissociative photoions have been derived. The natures of the dipole induced breakdown pathways of C F 4 , C F 3 C I and C F 2 C 1 2 were investigated by combining the present differential oscillator strength 11 measurements for molecular and dissociative photoionization wi th photoelectron data obtained i n the present thesis work plus previously reported photoelectron branching ratios. O n the basis of the present work, a revised set of absolute electronic state partial photoionization differential oscillator strengths for CF4 are presented. Absolute electronic state part ial photoionization differential osciUator strength spec-t ra for CF3CI, C F 2 C 1 2 and CFCI3 in the photon energy range 41—160 e V have been derived by combining total differential oscillator strength spectra obtained i n the present thesis work wi th photoelectron branching ratios obtained from photoelectron spectra measured using synchrotron radiation. Absolute photoabsorption differential osciUator strengths and for F Is , C Is, and C l 2p,2s inner shell excitation and ionization of freon molecules C F 4 , C F 3 C 1 , C F 2 C 1 2 , CFCI3 and CCI4 have been derived from the presently obtained high resolution electron energy loss spectra plus previously reported. Tentative assignments of the spectra were obtained using the M O picture and the potential barrier model. i i i Table of Contents Abstract ii List of Tables viii List of Figures x List of Abbreviations xiii Acknowledgements xiv 1 Introduction 1 2 Photoabsorption and Electron Energy Loss Spectra 5 2.1 Photoabsorption Cross Section and Differential Oscillator Strength . . . . 6 2.2 The Bethe-Born Theory of Fast Electron Impact 9 2.3 Spectral Analysis 17 2.3.1 Occupied and Unoccupied Molecular Orbitals 17 2.3.2 Transitions to Rydberg Orbitals 19 2.3.3 Transitions to V i r t u a l Valence Orbitals 21 2.3.4 Transitions to Ionization Continua 21 2.3.4.1 Delayed Onsets of Ionization Continua 22 2.3.4.2 X A N E S and E X A F S 22 2.3.4.3 Atomic and Molecular Photoionization at High Photo-electron Energies 23 iv 2.3.5 Photoelectron Spectroscopy and Dipole (e,e-fion) Spectroscopy . . 23 2.3.6 Potential Barrier Effects . . 25 2.3.7 The Equivalent Core (Z + 1) Analogy 26 2.3.8 The Absolute Differential Oscillator Strength Scale 26 3 Experimental Methods 28 3.1 The Dipole (e,e+ion) Spectrometer 28 3.2 The High Resolution Dipole (e,e) Spectrometer 32 3.3 Energy Cal ibrat ion 36 3.4 Sample Handl ing 37 4 Inner Shell Electron Energy Loss Spectra of N 0 2 at High Resolution: Comparison with Multichannel Quantum Defect Calculations of Dipole Oscillator Strengths and Transition Energies 39 4.1 Calculations 39 4.2 Results and Discussion 41 5 Absolute Differential Oscillator Strengths for the Photoabsorption and the Ionic Photofragmentation of C F 4 , CF 3 C1, CF 2 C1 2 and CFC1 3 59 5.1 Electronic Structures 59 5.2 Photoabsorption Differential Oscillator Strengths for the Valence Shells . 59 5.2.1 The C F 4 Photoabsorption Differential Oscillator Strengths . . . . 80 5.2.2 The CF 3 C1 Photoabsorption Differential Oscillator Strengths . . . 83 5.2.3 The CF2CI2 Photoabsorption Differential Oscillator Strengths . . 86 5.2.4 The CFC13 Photoabsorption Differential Oscillator Strengths . . . 90 v 5.2.5 Comparison of the Photoabsorption Differential Oscillator Strengths of C F 4 , CF3CI, C F 2 C 1 2 and C F C 1 3 93 5.3 The CF4 Photoabsorption Differential OsciUator Strengths for the C Is and F Is Inner Shells and the Valence SheU Extrapolat ion 94 5.4 Molecular and Dissociative Photoionization of CF4, CF3CI, C F 2 C 1 2 and CFCI3 97 5.5 Absolute Electronic State Part ia l Photoionization Differential OsciUator Strengths for C F 4 124 5.6 The Dipole Induced Breakdown 128 5.6.1 The Dipole Induced Breakdown of C F 4 129 5.6.2 The Dipole Induced Breakdown of C F 3 C 1 132 5.6.3 The Dipole Induced Breakdown of C F 2 C 1 2 136 6 Photoelectron Spectroscopy and the Electronic State Partial Differen-tial Oscillator Strengths of the Freon Molecules CF 3 C1, CF 2 C1 2 and CFCI3 Using Synchrotron Radiation from 41 to 160 eV 141 6.1 Photoelectron Spectra 142 6.2 Photoelectron Branching Ratios and Part ia l Photoionization Differential OsciUator Strengths 145 7 Absolute Dipole Differential Oscillator Strengths for Inner Shell Spec-tra from High Resolution Electron Energy Loss Studies of the Freon Molecules C F 4 , CF 3 C1, CF 2 C1 2 , CFC1 3 and CC1 4 164 7.1 Absolute Differential OsciUator Strengths 164 7.2 Electronic Configurations and Spectral Assignments 167 7.3 C Is Spectra 170 vi 7.4 F Is Spectra 180 7.5 C l 2s and 2p Spectra 183 8 Conclusions 190 Bibliography 191 vi i L i s t of T a b l e s 3.1 Reference energies of spectral calibration 37 3.2 Sources and purity of samples 38 4.3 Dipole-allowed transitions for N 0 2 42 4.4 Experimental and calculated data for N Is excitation of N 0 2 45 4.5 Experimental and calculated data for 0 Is excitation of N 0 2 55 4.6 Term values for N 0 2 and its Z + l analogue 0 3 57 5.7 Valence electronic configurations for C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 . . 60 5.8 Electronic ion states for C F 4 61 5.9 Electronic ion states for C F 3 C 1 62 5.10 Electronic ion states for C F 2 C 1 2 63 5.11 Electronic ion states for C F C 1 3 64 5.12 Differential oscillator strengths for C F 4 65 5.13 Differential oscillator strengths for C F 3 C 1 68 5.14 Differential oscillator strengths for C F 2 C 1 2 71 5.15 Differential oscillator strengths for C F C 1 3 74 5.16 Coefficients of the valence shell extrapolation formulas 79 5.17 Photoion branching ratios for C F 4 107 5.18 Photoion branching ratios for C F 3 C 1 109 5.19 Photoion branching ratios for C F 2 C 1 2 I l l 5.20 Photoion branching ratios for CFCI3 113 v i i i 5.21 Ion appearance potentials for C F 4 120 5.22 Ion appearance potentials for CF3CI 121 5.23 Ion appearance potentials for C F 2 C l 2 122 5.24 Ion appearance potentials for CFCI3 123 5.25 Electronic state partial differential oscillator strengths for C F 4 126 6.26 Ionization energies for C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 144 6.27 Photoelectron branching ratios for C F 3 C 1 146 6.28 Photoelectron branching ratios for C F 2 C 1 2 147 6.29 Photoelectron branching ratios for C F C 1 3 148 6.30 Electronic state partial differential oscillator strengths for C F 3 C 1 154 6.31 Electronic state partial differential oscillator strengths for CF2CI2 . . . . 155 6.32 Electronic state part ial differential oscillator strengths for CFCI3 156 7.33 Inner shell atomic oscillator strengths for C , F and C l atoms 166 7.34 Integrated sub-shell oscillator strengths per atom for C F 4 , C F 3 C 1 , C F 2 C 1 2 , CFCI3 and C C 1 4 166 7.35 Electronic configurations for C F 4 , C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and C C 1 4 . . . 168 7.36 Dipole-allowed transitions for C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 169 7.37 Experimental data for the inner shell excitations of C F 3 C 1 174 7.38 Experimental data for the inner shell excitations of C F 2 C 1 2 175 7.39 Experimental data for the inner shell excitations of C F C 1 3 176 7.40 Experimental data for the inner shell excitations of C F 4 and C C 1 4 . . . . 177 ix List of Figures 2.1 Schematic of scattering geometry 10 3.2 Schematic of the dipole (e,e+ion) spectrometer 29 3.3 Schematic of the high resolution dipole (e,e) spectrometer 33 4.4 Calculated quantum defects for N 0 2 40 4.5 I S E E L S spectra for N 0 2 N Is excitation i n the discrete and continuum regions 43 4.6 I S E E L S spectra for N 0 2 N Is excitation in the pre-ionization edge region 44 4.7 N 0 2 sample purity investigation 47 4.8 Comparison of I S E E L S spectra wi th earher optical measurements for N 0 2 49 4.9 I S E E L S spectra for N 0 2 0 Is excitation in the discrete and continuum regions 53 4.10 I S E E L S spectra for N 0 2 0 Is excitation in the discrete region 54 5.11 Differential oscillator strengths for valence shell photoabsorption of CF4 . 81 5.12 Differential oscillator strengths for valence shell photoabsorption of CF3CI 84 5.13 Differential oscillator strengths for valence shell photoabsorption of C F 2 C 1 2 87 5.14 Differential oscillator strengths for valence shell photoabsorption of CFCI3 91 5.15 Differential osciUator strengths for valence shell and inner shell photoab-sorption of C F 4 95 5.16 T O F mass spectrum of C F 4 at an equivalent photon energy of 80 eV . . 98 5.17 T O F mass spectrum of C F 3 C I at an equivalent photon energy of 45 eV . 99 x 5.18 T O F mass spectrum of CF2CI2 at an equivalent photon energy of 50 eV . 100 5.19 T O F mass spectrum of CFCI3 at an equivalent photon energy of 49 e V . 101 5.20 Photoion branching ratios for C F 4 103 5.21 Photoion branching ratios for C F 3 C 1 104 5.22 Photoion branching ratios for CF2CI2 105 5.23 Photoion branching ratios for CFCI3 106 5.24 Differential oscillator strengths for dissociative photoionization of C F 4 . . 116 5.25 Differential oscillator strengths for molecular and dissociative photoioniza-t ion of C F 3 C I 117 5.26 Differential oscillator strengths for dissociative photoionization of C F 2 C 1 2 118 5.27 Differential oscillator strengths for dissociative photoionization of CFCI3 119 5.28 Electronic state part ia l photoionization differential oscillator strengths for C F 4 127 5.29 Differential oscillator strengths for the dipole induced breakdown scheme o f C F 4 130 5.30 Dipole induced breakdown scheme for C F 4 131 5.31 Differential oscillator strengths for the dipole induced breakdown scheme of CF3CI 133 5.32 Dipole induced breakdown scheme for C F 3 C 1 134 5.33 Differential oscillator strengths for the dipole induced breakdown scheme o f C F 2 C l 2 139 5.34 Dipole induced breakdown scheme for CF2CI2 140 6.35 Photoelectron spectra of the freon molecules 143 6.36 Photoelectron branching ratios for C F 3 C 1 149 xi 6.37 Photoelectron branching ratios for CF2CI2 150 6.38 Photoelectron branching ratios for CFCI3 151 6.39 Electronic state partial differential osciUator strengths for CF3CI 157 6.40 Electronic state part ial differential osciUator strengths for C F 2 C 1 2 . . . . 158 6.41 Electronic state part ial differential osciUator strengths for C F C 1 3 159 7.42 C Is differential oscillator strength spectra for C F 4 , C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and C C I 4 i n the discrete region 172 7.43 C Is differential oscillator strength spectra for C F 4 , CF3CI, C F 2 C 1 2 , C F C 1 3 and C C I 4 i n the discrete and continuum regions 173 7.44 F Is differential osciUator strength spectra for freon molecules in the dis-crete region 181 7.45 F Is differential osciUator strength spectra for freon molecules i n the dis-crete and continuum regions 182 7.46 C l 2p differential oscillator strength spectra for C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and C C 1 4 i n the discrete region 185 7.47 C l 2p,2s and C Is differential osciUator strength spectra for C F 3 C 1 , CF2CI2, CFCI3 and C C 1 4 186 7.48 C l 2p and 2s oscillator strength distributions per C l atom for CF3CI, C F 2 C 1 2 , CFCI3 and C C 1 4 molecules and for the C l atom 187 x i i List of Abbreviations E E L S electron energy loss spectroscopy E X A F S extended X - r a y absorption fine structure F W H M ful l width at half maximum IP ionization potential I S E E L S inner shell electron energy loss spectroscopy M C Q D multichannel quantum defect M O molecular orbital M S - X c * multiple scattering Xa method OS oscillator strength P E S photoelectron spectroscopy T term value T O F time of flight T R K Thomas-Reich e -Kuhn V U V vacuum ultraviolet X A N E S X-ray absorption near edge structure X P S X - r a y photoelectron spectroscopy x i i i Acknowledgements I should like to express my sincere thanks to Dr. C E . Brion for his interest, assistance, encouragement and supervision throughout the course of my study. I appreciate having had the opportunity to work with him and other members in his research group. Special thanks are due to Dr. G. Cooper, who gave me a lot of help in my research, and also measured the freon photoelectron spectra, to Dr. H.K. Sze, who helped me in the early stages of my work, to Dr. X.M. Tong and Prof. J.M. Li, who performed the multichannel quantum defect calculation on the N0 2 molecule, to Dr. T. Ibuki, who helped to record the freon photoabsorption and photofragmentation data, to Prof. A.P. Hitchcock, who provided me with the background and Gaussian fitting programs, to Dr. J.H. Scofield, who supplied numerical data of calculated atomic photoionization cross sections, and to Dr. Y.H. Hong and G. Burton, who gave helpful comments and suggestions on the writing of my thesis. For invaluable technical assistance, discussions of things I could not grasp on my own, for advice, for guidance and for a few good references, I am very grateful to: W.F. Chan, S. Clark, M. Coschizza, P. Duffy, B. Greene, X. Guo, B. Hollebone, N. Lermer, B. Todd, E. Zarate, and Y. Zheng. I should also like to gratefully acknowledge a University of British Columbia Graduate Fellowship. The research work was also supported by operating grants from The National Sciences and Engineering Research Council of Canada. Finally, I should like to thank my parents. This thesis is dedicated to them. xiv Chapter 1 Introduction Absolute photoabsorption and photoionization differential oscillator strength (cross sec-tion) spectroscopies for molecular valence shell and inner shell excitation i n the V U V and soft X - r a y regions are of fundamental and applied interest. A quantitative knowl-edge of photoabsorption and photoionization processes is very important i n understand-ing the interaction of molecules wi th electromagnetic radiation. Detailed information on the transition energies and (differential) oscillator strengths is urgently needed i n a large number of scientific contexts, including studies i n aeronomy [1], astrophysics [2], planetary sciences [3] and radiation chemistry, physics and biology [4]. However, un-t i l comparatively recently, only l imited data were available from optical measurements using discharge lamps or X-ray tubes [5]. In the past decade, an increasing amount of quantitative spectral information over a wide energy range has been obtained with the increasing availability of tunable synchrotron radiation [6,7]. Alternatively, electron en-ergy loss spectroscopy ( E E L S ) can also be used to measure optical differential oscillator strengths by ut i l iz ing the vir tual photon field created by a fast scattered electron at neg-ligible momentum transfer. Under such conditions E E L S techniques are often referred to as dipole electron impact experiments. A recent review of such techniques and measure-ments has been published by Gallagher et al. [8]. Electron energy loss techniques have been used to obtain most of the experimental results reported in the present work. In addition photoelectron branching ratios from synchrotron radiation measurements have 1 Chapter 1. 2 been used to determine some partial differential oscillator strengths. Electron impact experiments are now widely recognized as providing physical infor-mation complementary to that obtained i n photon impact experiments. In fact one of the earhest demonstrations of quantization was the classic Franck-Hertz experiment [9] i n which electrons were used to probe atomic systems. Dipole electron impact techniques have been demonstrated i n the past decade to provide a very suitable alternative to V U V and soft X - r a y photoabsorption experiments for the measurement of optical quantities such as transition energies and dipole differential osciUator strengths or cross-sections for photoabsorption and photoionization processes [8]. The possibility of studying optical processes by electron impact was i n fact pointed out as early as 1930 by Bethe [10]. The use of electron energy loss spectroscopy for optical differential osciUator strength measurements was pioneered by the early experiments of Lassettre [11], Geiger [12] and V a n der W i e l [13,14,15]. In particular the direct techniques developed by V a n der W i e l and co-workers [13,14,15] and more recent development here at The University of Br i t i sh Columbia have provided versatile methods for the measurement of absolute differential osciUator strengths for photoabsorption and photoionization processes for valence sheU (for examples, see references [16,17,18]) and inner shell (for example, see reference [19]) processes. Some useful reviews on the electron energy loss spectroscopy studies and absolute optical differential oscillator strength measurements are to be found i n refer-ences [8,20,21,22,23,24,25,26]. In the present work optical (differential) osciUator strengths and transition energies for photoabsorption and photoionization have been obtained for a variety of valence and inner shell processes i n N 0 2 (chapter 4) and the freon molecules C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 (chapters 5 and 7) using the E E L S based techniques of dipole (e,e) and dipole (e,e+ion) spectroscopies. In addition partial optical differential oscillator strengths for valence Chapter 1. 3 shell photoionization (electronic states) have been derived for C F 3 C 1 , CF2CI2 and CFCI3 (chapter 6) from synchrotron radiation P E S measurements and the dipole (e,e) results. The NO2 molecule and the freons were selected because of their fundamental interest and also because of the importance of their interaction wi th energetic electromagnetic radiation i n a number of applications. For example N 0 2 is a toxic gas which is an important l ink i n the chain leading to the production of the atmospheric photochemical smog from air, sunlight, and automobile exhaust. Freons (in particular C F 2 C 1 2 and CFCI3) have been released to the atmosphere as a result of their widespread use i n industry and daily life, for example as refrigerants, foam blowing agents and aerosol propellants. The release of these molecules into the atmosphere has caused major concern as to their role [27,28,29,30,31,32,33] i n processes resulting i n depletion of the ozone layer which protects us and our planet from the damaging effects of short wavelength solar radiation. Quantitative spectroscopic information for these molecules i n the V U V and soft X - r a y regions of the electromagnetic spectrum are thus needed due to the presence of such energetic solar radiation i n the upper levels of the earth's atmosphere [34]. Prior to the present work very litt le quantitative spectroscopic information was available for N 0 2 and the freons i n the V U V and soft X-ray regions. The results of the work described i n this thesis are to be found i n the following publications: • W . Zhang, K . H . Sze, C E . B r i o n , X . M . Tong and J . M . L i , Chem. Phys. 140 (1990) 265. • W . Zhang, G . Cooper, T . Ibuki and C E . Br ion , Chem. Phys. 137 (1989) 391. • W . Zhang, G . Cooper, T . Ibuki and C E . Br ion , Chem. Phys. 151 (1991) 343. • W . Zhang, G . Cooper, T . Ibuki and C E . Br ion , Chem. Phys. 151 (1991) 357. Chapter 1. 4 • W . Zhang, G . Cooper, T . Ibuki and C E . Br ion , Chem. Phys. 153 (1991) 491. • G . Cooper, W . Zhang, C E . Br ion and K . H . Tan, Chem. Phys. 145 (1990) 117. Chapter 2 Photoabsorption and Electron Energy Loss Spectra The spectroscopic properties of molecular systems can be studied by using either photons or electrons as probes. In a photoabsorption experiment, when a photon of energy E = hv is resonantly absorbed by a target molecule, the process may be represented as Process : hv + AB —• AB* (2.1) Energy of photon : E where AB is the molecule i n its ini t ia l state and AB* in its final (bound or continuum) state w i t h an energy higher than the in i t ia l state by E. In an electron energy loss experiment wi th the same molecule, if an incident electron e transfers the same amount of energy E to the target molecule, the same final molecular state can be accessed. In this process Process : e + AB —> AB* + e (2.2) Energy of electron : E0 E0 — E, i n which E0 and (E0 — E) are the electron energies before and after the colhsion respec-tively. The energy loss E of the incident electron is analogous to the photon energy E i n the photoabsorption experiment. Moreover, a quantitative relation exists between the target transition intensities in the two experiments when the electron impact experiment is performed under the condition that the momentum transfer to the target is neghgibly 5 Chapter 2. 6 small . This parallel between photoabsorption and electron impact experiments is the basis of the major part of this thesis work. 2.1 Photoabsorption Cross Section and Differential Oscillator Strength The probabil i ty that a photon of energy E — hu w i l l be absorbed i n passing through a target is specified by a cross section o~(E). o~(E) has the dimensions of area and is often given a geometrical interpretation. Imagine that a circle of area o~(E) is centered on each target molecule i n the plane perpendicular to the incident photon beam. The circular area cr(E) has the property that any photon of energy E entering o~{E) is absorbed by the target [35]. Consider an incident photon flux N (N = number of photons per unit area per unit time) impinging on the target and passing through it wi th a path length dL. If no circles are hidden behind others within the distance dL, when viewed from the direction of the incident photon beam, the change in the photon flux is then dN = -Na(E)m dL (2.3) where ra is the target density (i.e. the number of target molecules per unit volume). The corresponding attenuation of the incident hght intensity ( I = Nhv) is accordingly given by dl = -Icr{E)m dL. (2.4) The condition corresponding to the requirement for the vahdity of the above linear rela-tionships is a(E)m dL « 1,, - (2.5) i n which the cr(E)mdL is the total area of circles centered on molecules contained i n a slab w i t h thickness dL and unit area surface, on which the photon beam is incident. Chapter 2. 7 This correspondence can be understood since, the smaller cr(E)mdL is compared wi th one (unit area), the smaller the chance that the circles are hidden behind others. A n equivalent of condition 2.5 is that the probabihty (—dN/N) of a given photon being absorbed i n passing a distance dL through the target is much smaller than one. In the language of electron scattering, this is also the condition for a single coUision i n an electron impact experiment (see the following section). In the situation where condition 2.5 is not val id for a path length L, equation 2.4 may be integrated over the path length and the Lambert-Beer Law is obtained, that is I = J 0 e - ° { E ) m L (2.6) where IQ and I are the light intensities before and after passing through the target. The quantum mechanical expression for the cross section for a process i n which a molecule having Z electrons wi th coordinates { r , } , whose components i n the direction of the electric field of the electromagnetic radiation are {XJ}, undergoes a transition from an in i t ia l state u 0 to a bound, final state un, upon resonantly absorbing a photon of energy E, is (in atomic units) | ( 2 - 1 6 ) where M is the number of atoms associated wi th atomic number zp. Bethe showed [26] that equation 2.13 can be reduced to where en(K) is a matr ix element for the in i t ia l and final states, en(K) = / <(ri, •••,r*)E ^ f ' u Q { r u • • •, rz) <*ri • • • dfz. (2.18) J 3=1 Chapter 2. 11 The cross section expressed i n equation 2.17, for a process i n which a fast electron trans-fers a given amount of energy and momentum, consists of two distinct factors, namely 4(k'/k)(l/K4) and \\en(K)\\2. The quantity 4(k'/k)(l/K4) may be evaluated from the measurable quantities k, k' and 8, which concern the incoming and outgoing electrons only, and it is therefore a purely kinematic factor. The second quantity |e„(.fY)|2, as shown i n equation 2.18, is target dependent and it determines the conditional probabil-i ty of a transition i n the target from state u0 to state un upon receiving a momentum transfer K. The conditional probabihty is because there is no unique correspondence between momentum transfer and energy transfer. The quantity |en(./iT)|2 is referred to as a dynamic factor [26]. The differential cross section in equation 2.17 can be also written as £ f = ! i i k - < * > <>•»> where the generalized oscillator strength fn(K) = j^MK)\\2 (2-20) is introduced as a straightforward generalization of the optical oscillator strength defined i n equation 2.8 [26]. The relationship between the electron scattering process, i n which a small momen-t u m is transferred, wi th the photoabsorption process can be revealed by considering the following: The term exK'r> i n equation 2.18 can be expanded i n a power series of K ei&*i = eiK*t = ! + iKx. + ilE^L + . . . + ( i n f i l l + . . . (2.21) assuming that the momentum transfer vector is in the x direction. W i t h this expansion, and considering the orthogonahty of the wavefunctions u0 and « „ , equation 2.18 becomes ' en(K) = (iK)ex + (iK)2e2 + • • • + {iK)*et + ••• (2.22) Chapter 2. 12 where the tth order multipole matr ix element et (t = 1 is electric dipole, t = 2 is electric quadrupole, etc) is given by 1 / z et = If / < ( r i r \" , r z ) £ x $ u 0 ( 7 V \" , r z ) dr1---drz. (2.23) The relationship between the optical and the generalized oscillator strength is obtained by combining equations 2.8 and 2.22, and is given by fn(K) = 2E[e21-r(el-2e1e3)K2 + (el + 2e1e5-2e2e4)K4 + ---] = fn + AK2 + BK4 + •••. (2.24) Thus i n the l imit of zero momentum transfer, the generahzed oscillator strength becomes equal to the optical oscillator strength, that is l i m UK) = fn, (2.25) A 2 — > 0 and therefore equation 2.19 becomes, i n the l imit of K2 —> 0, do-n 2 k' 1 = E l K i f - ( 2 ' 2 6 ) This is the Bethe-Born equation which gives the quantitative relationship between the process in which an electron is scattered at neghgible momentum transfer and the pho-toabsorption process. This is the basis of dipole (e,e) and dipole (e,e+ion) spectroscopies. For an ionization process i n which the target molecule undergoes a transition from the in i t ia l state UQ to a final continuum state ug, the differential electron scattering cross section corresponding to equation 2.13 is d2a 1 k' i ' - 2 dttdE 4TT2 k \\J e , ^ ' r « £ ( f i , • • •,r z)Vu 0(f i r . . , j ^ ) d f i • • • drzdr (2.27) Chapter 2. 13 and the parallel of equation 2.19 is 0 if and only i f both E/E0 and 0 approach zero. Earlier electron impact experiments were based on the above considerations by measuring the generalized oscillator strength for a fixed transition as a function of K2 and extrapolating to K2 = 0 to obtain the optical oscillator strength. For example, Lassettre et al. [23,24,40] measured the generalized oscillator strength for a particular transition (fixed E) as a function of the scattering angle 9 at a fixed impact energy E 0 . Alternatively, Hertel and Ross [41,42] obtained the generahzed oscillator strength for a particular transition (fixed E) at zero scattering angle, while varying the impact energy E 0 . However, the extrapolating procedures used were tedious and involved considerable uncertainty i n the resulting optical oscillator strength i n some cases due to the long extrapolation needed under the experimental conditions employed. A more direct approach was employed by Geiger et al. [12] and by V a n der W i e l et al. [13,14,15]. To this end, a very small K2 i n equation 2.33 can be approximated i n terms of the first Chapter 2. 15 order of x2 = (E/2E0)2 and 62 as [20] 2E0(x2 + 62). (2.34) W i t h this expression for small K2 and integrating equations 2.26 and 2.32 over the small scattering angle, i t is found that where 9Q is the half angle of acceptance of the scattered electrons [20]. Therefore, under experimental conditions of high impact energy (smaU x) and small scattering acceptance angle about zero degrees (small 60), the momentum transfer K is small , and therefore the electron scattering signal (which is proportional to the scattering cross section as discussed below) can be related to the optical oscillator strength through the Bethe-Born factor. This latter approach of direct measurements at negligible momentum transfer has been developed for a routine measurement procedure for oscillator strengths for various optical processes [8]. This direct method is preferable and has therefore been used i n the present work. The scattered electron signal is recorded as a function of energy loss E and this electron energy loss spectrum is then converted to a relative optical spectrum by a Bethe-Born factor. A n absolute optical differential oscillator strength scale is then established for the spectra by using a normalization method as described i n section 2.3.8. To convert an electron energy loss spectrum to a relative optical spectrum, the Bethe-B o r n factor i n equations 2.35 and 2.36 can be evaluated from the scattering kinematics and the spectrometer geometry as has been done with the dipole (e,e+ion) spectrometer described i n section 3.1. A simpler approach is to approximate equations 2.35 and 2.36 (2.35) and (2.36) Chapter 2. 16 as [20,43] dan = aE bf, n (2.37) and d*L = aE~b^L dE dE (2.38) where a and b are constants depending on the values of the impact energy of the incident electrons and the acceptance angle of the scattering electrons. In an actual experimental situation, i t is found that 2 < b < 3 [44]. The above equations are useful for Bethe-B o r n conversion over a l imited energy range, and have been used to obtain approximate optical differential oscillator strength spectra i n section- 5.3 and chapter 7 using the high resolution dipole (e,e) spectrometer which is described i n section 3.2. In the presently described electron energy loss experiments the gas pressure i n the collision chamber is adjusted so that the length of chamber is much smaller than the mean free path of the molecules to ensure that only single collision events occur when the incident electrons pass through the gas sample. This can be checked by estimating the value of the mean free path i n the present experiments, and also by seeing that peaks corresponding to multiple scattering processes are absent i n the spectra obtained i n the lower electron energy loss region (i.e. in valence shell spectra). Under this condition, which i n fact corresponds to condition 2.5 (since ami — 1 [45], where a and I are the cross section and the mean free path for the forward scattering, and m is the target density), a linear relationship exists between the electron scattering signal and the corresponding electron scattering cross section, similar to that for light intensity attenuation given i n equation 2.4. Therefore, the intensity of the electron energy loss spectrum is related to the photoabsorption differential oscillator strength as indicated i n equations 2.35 and 2.36 or i n the approximate equations 2.37 and 2.38. For this reason, the scattered electron Chapter 2. 17 signal i n a dipole electron scattering experiment drops off ~ E b times as quickly as does the corresponding photoabsorption intensity. 2.3 Spectral Analysis The profile and features of a photoabsorption differential oscillator strength spectrum are determined by the molecular wavefunctions before and after the transition as shown i n equations 2.8 and 2.11. To assign these features to particular transitions accurately, reliable theoretical calculations of both transition energies and oscillator strengths are needed. In chapter 4, such a high quality calculation for N 0 2 based on multichannel quantum defect theory [46] has been used to provide the spectral assignments. How-ever, such calculations are often not available, especially for larger polyatomic molecules, due to computational difficulties and the cost of computation. Therefore it is neces-sary to have an alternative method to analyze spectra with the aid of simple qualitative and semi-quantitative principles without involving extensive calculations. The qualita-tive principles concern general characteristic of various molecular orbitals, and of cor-responding transition processes. Regarding the semi-quantitative analysis, term values and quantum defects have proved to be useful concepts for providing tentative spectral assignments i n electron energy loss and photoabsorption studies [47,48,49]. In addition, relatively simple calculations such as G A U S S I A N 76 [50] are sometimes also helpful for spectral assignment. 2.3.1 Occupied and Unoccupied Molecular Orbitals In the molecular orbital ( M O ) theory, the state of an electron i n the molecule is described by a one-electron wavefunction, which is also referred to as a molecular orbital . The molecular orbitals can be conveniently classified into occupied inner shell (core) orbitals, Chapter 2. 18 occupied valence orbitals, unoccupied (virtual) valence orbitals and unoccupied Rydberg orbitals [47]. A n inner shell orbital closely surrounds a particular atomic site and orbital character is therefore dominantly atomic-like and non-bonding. The ionization energy of an inner shell electron (or binding energy of inner shell orbital) is close to the corre-sponding value for the isolated atom, differing only by a small chemical shift which is due to molecular effects. In contrast, occupied and unoccupied valence shell orbitals extend over the molecular framework, and they can be bonding, anti-bonding, or non-bonding. These molecular orbitals are therefore responsible for the formation and chemical prop-erties of the molecule. The ionization energies of electrons i n occupied valence orbitals are generally i n the energy range ~ 10-45 eV. As for Rydberg orbitals, they are large and diffuse, extending well beyond the spatial bounds of the ground state molecule. If an electron is excited to a Rydberg orbital , it does not distinguish the detailed spatial charge distribution of the molecular ion core and sees the ion core essentially as a point charge. A Rydberg orbital is therefore atomic-like and non-bonding. In addition to the bound states involving the above types of orbitals, the molecule may alteratively be i n an ionized (continuum) state in which case the electron is unbound. The present study provides an examination of the differential oscillator strength spec-t r u m resulting from excitations of electrons from the occupied valence shell or inner shell orbitals in the ground state molecule to the unoccupied Rydberg and vir tual valence or-bitals as well as for ionization to continuum states. The term value T is a useful quantity i n the spectral analysis. For a particular transition feature i n the spectrum, T is defined as [47] T = IP-E (2.39) where IP is the ionization potential of the electron i n the in i t ia l state and E is the transition energy between the in i t ia l and final states. T can therefore be thought of as the Chapter 2. 19 ionization energy of the excited, final state. The term values corresponding to the same final state, but different in i t ia l states, have been found to be related. In particular, since an electron i n a Rydberg state is mostly located far away from the molecular ion core, the hole location has l i t t le effect on its orbital energy. Therefore, Rydberg transitions associated wi th the same final state have similar term values. This is often referred to by saying that the Rydberg term value is transferable. In contrast, v i r tual valence orbitals are normally delocahzed over the framework of the molecule and their energies are sensitive to the hole location. Hence the vir tual valence term value is not expected to be transferable [51,52,53]. In addition to the term value of a feature, its transition intensity is another important consideration i n spectral analysis. The intensity of a transition induced by photons (or electrons under the conditions of neghgible momentum transfer) depends on the dipole moment matr ix element for the in i t ia l and final states as shown i n equation 2.8. Such transitions which obey dipole selection rules are called dipole allowed transitions. Given a dipole aUowed transition, another necessary condition for appreciable transition probability is that the two wavefunctions involved have significant spatial overlap. In the absence of a detailed oscillator strength calculation for a molecule a qualitative argument which is often used is that the intensity of a transition is expected to increase wi th the degree of spatial overlap. 2.3.2 Transitions to Rydberg Orbitals Since Rydberg orbitals in a molecule are atomic-like, the term values for Rydberg tran-sitions can be fitted into the Rydberg formula [47] T = R (2.40) (n - s,y Chapter 2. 20 where R is the Rydberg energy (13.605 e V ) , n is a quantum number, and Si is the d i -mensionless quantum defect of the s, p, d etc. type of Rydberg orbital labeled according to the angular momentum quantum number /. The quantum defect reflects the degree of penetration of the Rydberg orbital into the molecular ion core. The deeper the pene-trat ion, the larger the Si value (i.e. larger T). For molecules containing second and third row atoms, the typical range of quantum defects Si is 0.8-1.3 for s Rydberg series, 0.4-0.8 for p Rydberg series and 0-0.2 for d Rydberg series [47]. The typical magnitudes of term values for the lowest members of Rydberg series are i n the range 2.8-5.0 e V for the lowest ns member, 2.0-2.8 for the lowest np member and 1.5-1.8 for the lowest nd member [47]. Since both the inner sheU orbitals and Rydberg orbitals are essentially non-bonding, the excitation of an inner shell electron to a Rydberg state does not cause any appreciable change i n internuclear distance. A t room temperature the ground state molecules are populated mostly i n the lowest vibrational level, and the Franck-Condon region [54] of the transition w i l l mainly contain the lowest vibrational level of the Rydberg state. Therefore Rydberg features in inner shell spectra are generally expected to be sharper than those features associated with transitions to the vir tual valence orbitals which are mostly anti-bonding. Since Rydberg orbitals are large and have low probability density i n the molecular ion core region, where the inner shell orbital resides, the intensities of Rydberg transitions are usuaUy weak compared to transitions to v i r tua l valence orbitals. A s the quantum number n increases, the higher Rydberg orbitals have even less spatial overlap wi th the in i t ia l state orbital , and the transition intensities become smaller [55]. These properties of shape and intensity are useful for identifying Rydberg transition features i n inner shell spectra. Chapter 2. 21 2.3.3 Transitions to Virtual Valence Orbitals The orbitals which have spatial distributions most delocalized over the framework of molecule are valence orbitals and may be occupied or unoccupied by electrons. The vir-tual valence (unoccupied) states can either be located below (T > 0) or above (T < 0) the ionization hmit . The latter situation is due to potential barrier effects (see section 2.3.6 below). Since the in i t ia l orbital generally has better spatial overlap wi th v i r tua l valence orbitals than w i t h Rydberg orbitals, transitions to vir tual valence states are expected to have larger intensities than those to Rydberg states. Unlike Rydberg orbitals, which are essentially non-bonding, v ir tual valence orbitals are mostly antibonding i n character, and the excitation of an electron to these vir tual valence orbitals w i l l lead to a significant change i n internuclear distance. According to the Franck-Condon principle, the features for transition to these states are expected to be vibrationally broadened, particularly i n inner shell spectra. Thus the vir tual valence features i n inner shell spectra are usually more intense and broader than those features associated with Rydberg states. 2.3.4 Transitions to Ionization Continua For the hydrogen atom, the photoionization differential oscillator strength is a maximum at ionization threshold and monotonically decreases at higher energies [25,56]. Such a hydrogenic profile in the atomic photoionization differential oscillator strength is also observed for many-electron atoms when an electron i n a sufficiently deep inner shell is ionized [25,57]. Centrifugal effects and molecular field effects have been observed i n the photoionization differential oscillator strength spectra of molecules, and such effects lead to the various types of non-hydrogenic spectral behavior discussed below. Chapter 2. 22 2.3.4.1 D e l a y e d O n s e t s of Ioniza t ion C o n t i n u a A delayed onset i n the spectral distribution occurs when the photoionization intensity is depressed near the ionization threshold and then increases to a maximum at an energy above the threshold. This phenomenon, which may be explained i n terms of centrifugal effects [25,56], has been observed for some photoionization processes i n many-electron atoms and molecules, for example the 2p —> ed transition i n Ne (e represents a continuum state), the 3d —> ef transition in K r [25] and the C l 2p continua i n the freon molecules studied i n the present work (see figure 7.47). When an electron is ionized to a continuum state of higher angular momentum the centrifugal repulsion prevents the ionized electron from approaching close to the ion core and thereby reduces its wavefunction overlap with the in i t ia l state. Hence the photoabsorption is depressed near the ionization threshold by centrifugal effects which become more significant wi th increasing angular momentum. A n absorption maximum occurs when the photoelectron energy has increased sufficiently to overcome centrifugal effects. 2.3.4.2 X A N E S a n d E X A F S Inner shell photoionization spectra are also called \"above edge X - r a y absorption spec-t r a \" . The spectra can be divided into two regions. The spectral structures i n the high photoelectron energy region are referred to as extended X-ray absorption fine structure ( E X A F S ) and this has been interpreted as being due to scattering processes where the high kinetic energy photoelectron emitted by an atom is weakly scattered by only one neighboring atom i n a single scattering process [58]. In contrast, the X - r a y absorption near edge structure ( X A N E S ) located in the low photoelectron energy region is explained i n terms of multiple scattering of the low kinetic energy photoelectron by neighboring Chapter 2. 23 atoms i n the molecule [58]. The shape resonances (see section 2.3.6 below) i n the inner shell spectra are actually special cases of X A N E S [58]. A photoelectron energy of ~40 eV has been used, somewhat arbitrarily, to divide the spectra X A N E S and E X A F S [58]. 2.3.4.3 Atomic and Molecular Photoionization at High Photoelectron Energies A theoretical study of the N Is photoionization of N 2 molecule by Dehmer and D i l l [59] showed that i n the high energy continuum the total molecular ionization cross section is close to twice the atomic cross section i n magnitude. This is because the in i t ia l core state has an atomic-hke charge distribution and also because the escape of the energetic photoelectron is not significantly altered relative to the free atom case (i.e. the final orbital is also atomic-hke). Thus, to a first approximation, the molecular inner shell photoionization differential oscillator strength for high photoelectron energies is equal to the sum of the corresponding differential oscillator strengths for the constituent atoms. Molecular effects, including E X A F S discussed i n the previous section 2.3.4.2, are weak modulations on the sum of the atomic differential osciUator strengths. This relationship between molecular and atomic photoionization is the basis for the inner shell spectral normalization procedures used i n the present study (see section 2.3.8) and also by others [44,60]. 2.3.5 Photoelectron Spectroscopy and Dipole (e,e-f-ion) Spectroscopy W h e n a photon excites a molecule to a state above its first ionization threshold, ionization may occur where an electron is ejected, but this need not occur 100% of the time [56]. The ratio of the number of photo-ejected electrons to the number of photons absorbed is the photoionization efficiency [8]. Chapter 2. 24 In a photoionization process where a particular electron i n an orbital is ejected, the molecular ion is left i n a corresponding electronic ion state. Subsequently the molecular ion relaxes and may remain as a stable molecular ion or dissociate into fragments (charged and/or neutral). The photoelectron spectrum indicates the intensity of the ejected pho-toelectrons at a given photon energy as a function of electron kinetic energy. B y using Einstein's photoelectric equation, the scale can be converted from electron kinetic en-ergy into orbital binding energy. In this spectrum, to a first approximation, each peak corresponds to the ejection of an electron from a particular orbital or more precisely, to the production of the molecular ion in a particular electronic state. The probability for a molecule to produce a molecular ion i n a particular electronic state upon absorbing a photon can be expressed by the electronic state branching ratio which is the ratio of the corresponding peak area to the area of aU peaks i n the photoelectron spectrum. Accu-rate branching ratios are only obtained i n an experiment if the variation of the analyzer transmission efficiency with electron kinetic energy is taken into account. The photofragmentation processes following the electron photo-ejection can be studied by photoionization mass spectrometry or the equivalent electron impact technique of dipole (e,e+ion) coincidence spectroscopy [8]. In the dipole (e,e+ion) experiment, the photoions are detected in coincidence with the energy loss electrons as a function of mass-to-charge ratio. Similar to the electronic state branching ratio, the photoion branching ratio is obtained as the ratio of the area of one ion peak to the area of al l ion peaks i n the mass spectrum. The part ia l photoionization differential oscillator strengths for production of the elec-tronic states and photoions can be obtained by taking the triple product of the total photoabsorption differential oscillator strengths and the photoionization efficiencies w i t h the branching ratios for producing the electronic states and the photoions, respectively. Chapter 2. 25 Furthermore, the dipole induced breakdown pathways can be investigated by appropriate combination of the two kinds of partial photoionization differential oscillator strengths (see chapter 5). 2.3.6 Potential Barrier Effects Original ly studied for many-electron atoms where potential barriers are formed due to the presence of centrifugal potentials [25], potential barrier effects have also been investigated for many molecules (for examples, see references [59,61,62,63,64]). Such effects have been found to occur i n the inner shell spectra of N 0 2 (chapter 4) and the freon molecules (chap-ter 7). The potential barrier effects are manifested i n inner shell photoabsorption and electron energy loss spectra by intense resonance features (shape resonances) accompa-nied by a suppressed ionization continuum and/or Rydberg structures. The mechanism of the effects has been studied earlier [61,62,65]. Briefly, if a double well potential exists i n a molecule, w i t h a potential barrier on the perimeter of the molecule, it is possible that the spatial distribution of a particular wavefunction is mostly wi th in the inner well or the outer well . A n inner well wavefunction, which can be at an energy either above or below the ionization energy of the molecule, has a large spatial overlap wi th the in i t ia l state wavefunction which resides mainly in the inner well and this results i n strongly enhanced spectral features [65]. In contrast, the outer well wavefunctions have small am-plitudes i n the inner well region, and this leads to very low intensities for transitions to the outer well states [65]. Rydberg states belong to the outer well manifold of states [65]. The causes of potential barriers have been studied theoreticaUy and experimentally. For example, centrifugal effects have been assigned as the cause of the shape resonances i n the inner shell spectra of the N 2 molecule [59]. The presence of electronegative hgands surrounding a central atom i n a molecule is also believed to cause a potential barrier. Chapter 2. 26 Such an explanation was used to account for the inner shell spectra of molecules such as S F 6 , S e F 6 , T e F 6 and C 1 F 3 [62,63,64]. 2.3.7 The Equivalent Core (Z + 1) Analogy Inner shell spectra can also be analyzed using the (Z + 1) equivalent core model [66,67]. Consider an inner shell electron, bound to a particular atomic site wi th atomic number Z , which is excited to an unoccupied vir tual valence or Rydberg orbital . Since, i n the final state, there is one less electron i n the core to shield the nucleus, the promoted electron w i l l see the atom as having an effective atomic number of (Z + 1) so that the term values for the inner shell transitions w i l l be approximated by those for valence excitations i n the ( Z + l ) core analogous species. For example, the term values for the N 0 2 N Is transitions are expected to be comparable with those for the valence transitions for the (Z + 1) core analogous species 0 3 as discussed i n chapter 4. 2.3.8 The Absolute Differential Oscillator Strength Scale Since i n general only relative intensities are obtained for both photoabsorption and Bethe-B o r n converted E E L S spectra, calibration (normalization) procedures are required i n order to estabhsh an absolute differential osciUator strength scale. Normalizat ion may be achieved using several different approaches: 1. Normalizat ion at a single photon energy to a pubhshed absolute measurement or calculation for the molecule. 2. Normalizat ion using the partial T R K sum rule. For example, i n chapter 5, the part ia l T R K sum rule (see section 2.1) has been used to normahze valence sheU Bethe-Born converted E E L S spectra so that the area under the normahzed valence Chapter 2. 27 shell differential oscillator strength spectrum, as a function of photon energy (eV), is equal to the number of valence shell electrons plus a small correction corresponding to Pauh excluded transitions [38]. Similarly, in chapter 7 the part ial T R K sum rule has been used to verify the differential oscillator strength scale estabhshed using atomic differential osciUator strengths (see method 3 below). 3. Normalizat ion of inner sheU spectra i n the high photoelectron energy region to atomic differential osciUator strengths (see section 2.3.4.3). The atomic differential osciUator strengths (cross sections) can be obtained either from calculations [57, 68,69] or from recommended semi-empirical values [70]. This approach has been used i n the normalization of inner shell Bethe-Born converted E E L S spectra i n section 5.3 and chapter 7. Chapter 3 Experimental Methods The results reported i n this thesis were obtained using dipole (e,e+ion) and dipole (e,e) spectroscopies. The dipole (e,e+ion) spectrometer was used to obtain the valence shell photoabsorption, photoionization and photofragmentation differential oscillator strength spectra and photoionization mass spectra for the freon molecules discussed i n chapters 5. The inner shell photoabsorption and photoionization differential oscillator strength spec-t ra for NO2 i n chapter 4 and for the freon molecules i n chapter 7 were derived from electron energy loss spectra measured using the high resolution dipole (e,e) spectrometer. The photoelectron branching ratios and partial electronic state photoionization differen-t ia l oscillator strengths for the freon molecules discussed i n chapter 6 were obtained from photoelectron spectroscopy measurements made at the Canadian Synchrotron Radiat ion Facil i ty at the University of Wisconsin by D r . G . Cooper. 3.1 The Dipole (e,e+ion) Spectrometer The dipole (e,e+ion) spectrometer was originally built and operated at the F O M insti-tute i n Amsterdam [13,14,15,71,72,73,74,75]. In 1980 this instrument was moved to the University of Br i t i sh Columbia where it has been modified [76,77]. Details of the con-struction of the apparatus and its operation have been described i n references [13,14,15, 71,72,73,74,75,76,77]. The form of the spectrometer as used for the work discussed i n this thesis is shown i n figure 3.2. 28 ( e , e + j o n ) S P E C T R O M E T E R ELECTRON GUN COLLISION CHAMBER ANGULAR SELECTION EINZEL LENSES DECELERATING LENS TIME OF FLIGHT MASS SPECTROMETER AND ION LENSES PRIMARY BEAM DUMP DELAYS ELECTRON ANALYSER PRINTER PLOTTER DISC DRIVES ENERGY LOSS PDP 11-03 ION T A C Figure 3.2: Schematic of the dipole (e,e+ion) spectrometer. Legend: TAC—time to amplitude convertor P D P 11-03—computer Chapter 3. 30 A narrow (1 m m diameter) beam of fast electrons (8 keV) is produced from a black and white television electron gun with an indirectly heated oxide cathode (Phil ips 6AW59) . The incident electrons interact wi th target molecules i n the collision chamber. In the scattered channel the electrons i n a small cone of 1.4 x 1 0 - 4 steradians about the zero scattering angle pass through an angular selection aperture and are transported and decelerated by the Einze l lenses and the decelerating lens as shown i n figure 3.2. The electrons then are energy analyzed by the hemispherical electron analyzer (~1 eV F W H M resolution) and are detected by a channel electron multiplier (Mul lard B 4 1 9 A L ) . The positive ions produced i n the collision chamber are extracted at 90 degrees to the incident electron beam into the time-of-flight ( T O F ) drift tube and are then detected by the ion multiplier (Johnston M M 1 - 1 S G ) . A homogeneous electric field (400 V c m - 1 ) across the colhsion chamber and an accelerating lens system ensure uniform collection of ions wi th up to ~ 2 0 eV excess kinetic energy of fragmentation, independent of the in i t ia l direction of dissociation [73,74,77]. The length of the drift tube and the final ion kinetic energy are such that mass spectra wi th a mass resolving power ( r a / A m ) of 50 can be obtained from a time-of-flight analysis. Magnetic shielding is achieved w i t h Helmholtz coils and high permeability mumetal shields. The spectrometer is evacuated wi th turbo molecular pumps to provide the clean vacuum environment desirable for quantitative ion and electron spectroscopy. Two modes of operation of the spectrometer are possible: 1. Photoabsorption differential oscillator strengths. According to the Bethe-Born theory discussed i n section 2.2, the small momentum transfer (K) condition is to be satisfied to ensure that the collected scattered elec-trons are associated wi th dipole transitions i n the target molecule. In the present Chapter 3. 31 studies for which the energy loss E < 200 eV, this condition is met by choos-ing the scattering kinematics of the spectrometer: high electron impact energy (Eo = 8 keV) and small half scattering acceptance angle (9Q = 6.7 x 1 0 - 3 radians) about zero degrees. The electron energy loss spectrum can be converted to a rela-tive optical spectrum by using equations 2.35 and 2.36 through a Bethe-Born factor which can be readily obtained from the E,EQ,80 values. The absolute differential oscillator strength scale for the optical spectrum can then be estabhshed using the part ial T R K sum-rule (section 2.12). 2. Time of flight mass spectra and ionic photofragmentation differential oscillator strengths. T O F mass spectra at a fixed energy loss (i.e. photon energy) can be obtained i n an electron-ion coincidence measurement wi th the aid of a time to amphtude converter ( T A C ) . The T A C is started by a pulse signal from an electron of a given energy loss and stopped by a pulse from the ion signal. The T A C generates an output pulse w i t h an amphtude proportional to the time between the start and stop pulses. This time is proportional to ^/m/e , a characteristic quantity for a specific ion which has mass ra and positive charge e. A T O F mass spectrum is then constituted from the pulse height distribution measured using a P D P - 1 1 / 0 3 computer v i a an analogue to digital converter and software routines. As has been discussed i n section 2.3.5, the ionic photofragmentation branching ratios at a fixed energy loss for a particular ion are obtained from the ratio of the area under the corresponding peak to the area under al l ion peaks. The photoionization efficiency is measured as the number of ions produced by each energy loss electron. The photofragmentation differential osciUator strength for a particular ion can then be obtained by taking the triple Chapter 3. 32 product of the absolute photoabsorption differential oscillator strength (derived from the non-coincident energy loss measurement described i n the measurement mode above), the photofragmentation branching ratio, and the photoionization efficiency. 3.2 The High Resolution Dipole (e,e) Spectrometer The high resolution dipole (e,e) spectrometer was built i n this laboratory i n the early 1980s. F i g . 3.3 shows a schematic diagram of the instrument. The design and con-structional details have been described i n detail in reference [78] and hence only a brief description w i l l be given here. Electrons are produced from a direct current heated thoriated tungsten filament located w i t h i n an externally adjustable mount i n an oscilloscope electron gun body (Cliftronics C E 5 A H ) . The voltage for the filament cathode (C) , grid (G) , anode (A) and the second element of the focussing Einzel lens F are all floated at the negative of the impact energy (typically i n the range 2-3.7 keV i n the present work) wi th respect to the grounded first and third elements of the focusing lens F . The electron gun provides a nar-row (1 m m diameter) electron beam. The beam is then retarded by the two element lens Li at the monochromator entrance to the required pass energy of the monochromator, which is a hemispherical electron energy analyzer. The monochromated beam then exits through a v i r tua l slit formed by the accelerating lens L 2 (ratio 1:20) and is brought into focus at the entrance (P 4 ) of the stainless steel reaction chamber after passing through a second accelerating lens L3 (ratio 1:5). The beam is then transported to the collision chamber ( C C ) by the Einzel lens L 4 to collide wi th the molecules under study. The exiting main beam and scattered electron beam then pass through a zoom (energy-add on) lens L 5 , retarding lenses L 6 (ratio 5:1) and L 7 (ratio 20:1) to a vir tual slit formed MONOCHROMATOR DETECTOR ELECTRON GUN p T T c \" T Til—HjT W n A N A L Y S E R U L 7 TURBO PUMP r 360 L/S 0 SCALE llXllllJllll}. ' I\" \" ' 10 20 cm TURBO PUMP 450 L/S 8 •8 Co TURBO PUMP 360 L/S WHIM/MA Figure 3.3: Schematic of the high resolution dipole (e,e) spectrometer. Legend: A anode G grid P i ~ P s apertures C cathode G A S gas inlet Q 1 - Q 9 deflectors C C collision chamber H V high voltage T tube D decoupling transformer L1-L7 lenses V valve F forcusing lens L / S liter per second co co Chapter 3. 34 at the analyzer entrance. The half angle of acceptance (0Q) of the scattered electrons is 3.0 x 1 0 - 3 radians about zero degrees. Following energy analysis, the scattered electrons are detected by a channeltron electron multiplier (Mul lard B 4 1 9 A L ) mounted behind the analyzer exit aperture. The channeltron is operated at high voltage of 3.5 keV and the signal is decoupled using a ferrite core transformer. Magnetic shielding of the various regions of the spectrometer is provided by hydrogen-annealed mumetal enclosures, exte-rior to the vacuum housing. Turbo molecular pumps are used to provide a clean vacuum environment. H i g h resolution, high sensitivity, and high stabihty are achieved by this spectrometer due to the following features [78]: 1. Separate differential pumping of the four vacuum chambers of the spectrometer has alleviated the problems of surface contamination, retuning and frequent cleaning that occur wi th single chamber instruments. This ensures long term stabihty and maintains high sensitivity and good resolution. In addition, thermal stability of the hot filament is not affected when the sample is introduced into the spectrometer. 2. The advanced electron optics allow transmission of large beam currents and also minimize the effects of scattering of the beam from slit edges and the surfaces of the analyzer into the detector. This design also permits operation at zero degree scattering angle since the main (unscattered) primary electron bean is strongly suppressed due to the high energy selectivity of the zoom lenses. 3. The large hemispherical electron energy analyzers (mean radius RQ = 19 cm = 7.5 in) chosen provide high transmission and high resolution for a relatively high pass energy. This also permits a high impact energy, which is necessary for optical differential oscillator strength measurements, while st i l l retaining reasonable lens Chapter 3. 35 ratios. The pr imary (unscattered) electron beam is used to tune up the spectrometer by steering the beam w i t h analyzer deflection voltages, lens voltages and the deflectors ( Q x to Qg). Each of the deflectors consist of two pairs of electrostatic plates. The colhmation and direction of the electron beam can be monitored wi th electrometers connected to the apertures (P1 to P 8 ) and to the cone of the channeltron. The small currents on the channeltron cone are measured with a floated vibrating reed electrometer (Cary, model 401). To obtain an energy loss spectrum, a voltage equal to the energy loss corresponding to the inelastic scattering is added on top of the voltages already applied to the analyzer system, from the lens L 5 onward. Thus by regaining their energy loss, the scattered electrons are transmitted through the analyzer system to the channeltron and an energy loss spectrum is obtained by using a suitable offset energy loss voltage and scanning the energy loss region of interest. The channeltron signals are processed by high gain pre-amplifier and amplifier-/discr iminator units ( P R A models 1762 and 1763 respectively). The spectrum is recorded by using a P D P 11/023 computer, and/or a Nicolet 1071 signal averager operated i n a multichannel scaling mode where the channel address is stepped synchronously wi th the voltage on L 5 and the analyzer. The spectrum is monitored on-line v ia a V T 1 0 5 graphics terminal . The energy resolution, which can be defined as the ful l width at half maximum height ( F W H M ) of the electron beam at the analyzer exit, depends on the pass energies selected for the monochromator (EM) and analyzer (EA)- The theoretical resolution for this spectrometer has been derived to be [78] AEFWHM = 0.003 y 7 ^ + E\\. Chapter 3. 36 In practice, the experimental resolution achieved is shghtly better than the above theo-retical value. The energy resolutions specified i n the spectra reported i n the present work were obtained by measuring the profile of the primary electron beam, and were within the range 0.030—0.3 e V F W H M . The choice of the resolution to be used is determined by the spectral region of interest as well as natural hnewidth considerations. For exam-ple, when structures i n the pre-ionization edge region of a molecule are to be examined, high resolution is chosen; when a spectrum over a long energy range, including both pre-ionization edge and continuum regions, is to be scanned, a lower energy resolution is used. The length of time need to collect an inner shell spectrum ranges from a few hours to several days, depending on the primary beam current and the intensities of the transitions involved. 3.3 E n e r g y C a l i b r a t i o n To calibrate the energy loss scale of the electron energy loss spectrum measured using the high resolution dipole (e,e) spectrometer, the sample and a suitable reference gas were introduced simultaneously into the spectrometer, and the spectrum of the mixture was measured to avoid any problems of chemical-dependent energy shifts and contact potentials. The absolute energy scale was then estabhshed by calibrating a prominent spectral feature due to the sample wi th respect to a known feature of the reference gas. Sodhi and B r i o n have carefully measured the energy of a number of selected inner shell atomic and molecular transitions for calibrating energy loss spectra [51]. The particular calibration energies used i n this work are hsted i n table 3.1. The cahbration corrections have been found to be very small (<0.03 eV) due to the efficient differential pumping of the different regions of the spectrometer (see section 3.2). The energies of the various spectral features were visually determined with the aid of computer software. Chapter 3. 37 The energy loss scale for the dipole (e,e+ion) spectrometer has been calibrated using one of the following methods: 1. Using a prominent feature of known energy from previously published high resolu-t ion spectrum of the sample molecule. 2. Using the reported appearance potential for production of a particular ion i n the mass spectrum of the sample molecule. 3.4 S a m p l e H a n d l i n g The samples studied were introduced to the reaction chambers of the respective spec-trometers through gas inlet systems including Granvi l le-Phi l l ips series 203 stainless steel leak valves. Appropriate gas regulators were used for the respective gaseous samples depending on the type of gas i n the gas cylinder. In case of CFCI3 the l iquid sample was degassed by freeze, pump and thaw cycles before the sample was allowed to evaporate into the spectrometer v ia the leak valve. Cylinders of commercially available NO2 were found to contain various amounts of N O impurity. The N O was removed by cooling and pumping the cyhnder prior to use. No dimers (i.e. N 2 O 4 ) were expected to be present [79] at the low pressures employed in the presently reported experiment. Table 3.1: Reference energies for inner shell spectra 0 Inner shell transition Transition energy b (eV) S F 6 S 2 P l / 2 —» t 2 g 184.54(5) N 2 N Is —> n*(v = 1) 401.10(2) C O C I s — > T V ' ( V = 0) 287.40(2) \"Reference energies are taken f r o m ref. [51]. ^Uncertainties are shown in brackets, e.g. 184.54(5) means 184.54 ± 0.05 e V . Chapter 3. 38 A l l the samples studied i n the work presented i n this thesis were obtained commer-cially and their stated purities are listed i n table 3.2. Table 3.2: Source and purity of samples Sample Source Pur i ty(%) N 0 2 Matheson 99.5 C F 4 Matheson 99.7 C F 3 C 1 Matheson 99.0 C F 2 C 1 2 Matheson 99.0 CFCI3 P C R 99.0 Chapter 4 Inner Shell Electron Energy Loss Spectra of N 0 2 at High Resolution: Comparison with Multichannel Quantum Defect Calculations of Dipole Oscillator Strengths and Transition Energies The N 0 2 inner shell E E L S spectra reported i n this chapter were measured using the high resolution dipole (e,e) spectrometer described in section 3.2. 4.1 Calculations The multichannel quantum defect ( M C Q D ) theory calculations were carried out by Tong and L i [80] and the details of the calculations are similar to those reported earlier for SO2 [53]. F i g . 4.4 shows the calculated quantum defects for a i , &i, b2 and a 2 symmetries. The calculated term values (IP-E) are obtained from the (state-dependent) quantum defects. The calculated (electronic) oscillator strengths / ' s for discrete transitions are represented (see figs. 4.5a and 4.9a below) as / / A , where the width A ( F W H M ) is estimated from the experimental spectra (~0.9 e V for the 6a! , 2fel5 and 7a1 final states, 3 e V for 5b2 and ~0 .5 e V for the lower Rydberg states). The representation of oscillator strength i n the discrete transition region has been discussed in section 2.1. Oscillator strengths for the higher Rydberg states and i n the ionization continuum are represented as differential oscillator strengths, df/dE. 39 Chapter 4. 40 < Z> O 2.0 1.0 0.0 10 0 4 3-1 r m = N 0 2 — pbi — fibi -1 ; i »3t>1 1 , 45 P\" Pt>2 |/ 'f\"3b2 | / f-ib 2 4 p d b 2 1 , •10 10 0 TERM VALUE (eV) 10 IL — f ° 2 4 J ca2 1 , Figure 4.4: Calculated quantum defects for excitations of NO2. Chapter 4. 41 4.2 Results and Discussion The NO2 molecule is of C2v symmetry, and the ground state, independent particle, elec-tron configuration, including the unoccupied (virtual) valence orbitals, may be written as [81,82,83] ( 1 6 2 ) 2 ( l a Q 2 ( 2 ^ ^ ^ O Is N Is valence orbitals ( 2 6 1 ) ° ( 7 a 1 ) 0 ( 5 6 2 ) ° ; 2 A X . unoccupied (virtual) valence orbitals A s a result of the unpaired electron i n the half-filled 6ai valence orbital , any transition from the core orbitals to any other final orbital results i n a pair of doublet final excited states, depending on the coupling of electron spins [82]. Similarly, core ionization leads to the 3 A i and x A i states of N O j [84,85]. The sphtting caused by interaction between the two 0 Is orbitals (lfe2 and lcii) is expected to be negligibly small due to their essentially atomic-like nature. The dipole-allowed transitions for N 0 2 are shown i n table 4.3. F i g . 4.5 shows the measured low-resolution (0.14 e V F W H M ) I S E E L S and M C Q D calculated N Is excitation spectra of N 0 2 i n the discrete (pre-ionization edge) and ion-izat ion continuum regions. The absolute differential oscillator strength scale was estab-lished using procedures described below. The N 0 2 sample was purified as explained i n section 3.4 above (see also further discussion below), but a very small amount of residual N O is s t i l l present. F i g . 4.6 shows the high-resolution (0.090 e V F W H M ) N Is I S E E L S spectrum of the pre-ionization edge region for an N 0 2 sample of higher purity, together w i t h the M C Q D calculation. The M C Q D calculated oscillator strengths are presented i n the discrete and continuum transition regions, as discussed i n section 4.1. The measured energies and oscillator strengths together with the calculated oscillator strengths and Chapter 4. 42 Table 4.3: Dipole-allowed transitions i n N 0 2 from the 2AX ground state for C21, symmetry F i n a l configuration Dipole-allowed core hole occupied final state orbital vir tual M O ax \\2AX,22AX a ax 12B1,22B1 a ai b2 l2B2,22B2a a 2 12B2,22B2 a2 b2 l2B1,22Bl h ai l2B1,22B1 a h bi l2Au22Ai b2 ai l2B2,22B2a b2 b2 \\2Ai,22Ax a i continuum 3AU1A1 continuum 3 A 2 , M 2 h continuum *Bi?Bx h continuum 3B2* B2 \" O n l y one doublet state wil l result f r o m transitions to final states in which the 6 a i orbital is doubly occupied or empty. term values are shown i n table 4.4. Also shown are the assignments based on a consider-ation of the M C Q D calculations. The energies of the 3 A i and 1 A i N Is ionization edges are assigned according to X P S measurements [84,85,86] using the triplet-singlet sphtting of 0.70 e V reported by Davis et al. [84]. The energies of features 1 and 2 are close to the values reported for the two largest peaks in the N Is photoabsorption spectrum (see fig. 4.8b below) of N 0 2 from ref. [82], but peak 1 is much less intense relative to peak 2 i n the I S E E L S spectrum. The smaU peak, just below 400 eV, barely visible i n fig. 4.5b but present to a much larger extent i n the photoabsorption spectrum [82], is clearly due to the presence of a small amount of an N O impuri ty (see discussion below) which has Chapter 4. 43 ^ 1 5 H T > © 10-CM 5 H o ui cn f— Dl o o (a) MCQD Calculation 3 A i edge 2bi 6ai 7a 4 p a i N 0 2 N1s | 5 b 2 I5.0 5.0 -5.0 -15.0 TERM VALUE (eV) ( b ) 2bi ISEELS Experiment cn 5H OH AT edge 0=0\" E 0 = 3 7 0 0 e V A E = 0 . 1 4 e V i—r i 2 ftN1s edges\" l in in i 34 914 16 4 0 0 4I0 4 2 0 4 3 0 4 4 0 ENERGY LOSS (eV) 4 5 0 Figure 4.5: (a) M C Q D calculation and (b) low-resolution N Is ISEELS spectrum of NO2 in the discrete and continuum regions. Spectrum (b) as shown is not Bethe-Born corrected. Chapter 4. 44 > 0) CM I o X r -o u i Dl I— CO OH o o CO o rr U l ( a ) MCQD Calculation 1 0 - 2bi 5-6 a i 0 -Aiedge l P b l 7 Q l A 4 p ° 1 m 4 s a i n , LH K5D ' £ 0 \" T E R M V A L U E ( e V ) 0.0 10-( b ) ISEELS Experiment 2b 0=0* E 0 = 3 0 0 0 e V A E = 0 . 0 9 0 e V 5-0 x 3 N1s edges 1—I I I llllll II T 3 4 5 678101214 15 • • 9 \" ' ? 4 0 0 410 E N E R G Y L O S S ( e V ) Figure 4.6: (a) MCQD calculation and (b) high-resolution N lfi ISEELS spectrum of NO2 in the pre-ionization edge region. Spectrum (b) as shown is not Bethe-Born corrected. Chapter 4. 45 Table 4.4: Experimental and calculated data\" for N Is excitation of N 0 2 Experimental [this work M C Q D calculation feature energy term value oscillator 6 term oscillator assignment (eV) strength value 0 strength f inal orbital 3 A X 1Ald ( x l O \" 2 ) (eV) ( x l O \" 2 ) (to 3 A i l i m i t ) d 1 401.04 11.56 2.7 13.12 2.10 6a x 2 403.28 9.32 8.9 10.62 8.26 3 ~ 408 ~ 5 4.80 0.20 4sai 4 408.92 3.68 3.22 0.60 4pbx 5 409.48 3.12 (3.82) 3.05 0.38 4pax 6 410.08 2.52 (3.22) 2.40 0.02 4pb2 7 410.48 2.12 (2.82) 1.98 0.06 5sai 8 410.68 1.92 (2.62) 1.89 1.46 7ai(aJ) 9 410.84 1.76 (2.46) 1.70 0.04 ZdW 10 411.04 1.56 (2.26) 1.50 0.72 3d2ai 1.44 0.18 5pbi 11 411.28 1.32 ' (2.02) 1.39 0.14 5pai 12 411.52 1.08 (1.78) 0.96 0.02 4dbx 0.89 0.30 4do<2i 13 411.76 0.84 (1.54) 0.82 0.08 6p&i 0.80 0.06 6pax 0.78 0.10 4 i i 2 a i 0.60 0.02 5dW 14 411.96 0.64 (1.34) 0.57 0.12 5c?ofli 0.53 0.04 lpbx 0.52 0.02 1pax 0.51 0.02 5d2a,i 0.40 0.06 6c?oai 0.36 0.02 6d2ai 0.29 0.04 7dodi I P ( 3 A X ) 412.6 6 0 0 — oo 15 412.72 (0.58) I P ^ A i ) 413.30 6 0 16 416.16 -3.56 7.3 -2.45 9 562(6$) \"See fig. 4.5 a n d 4.6. ' ' A kinematic B e t h e - B o r n conversion of E2,5 has been applied. c W i t h respect to the 3Ai l imit . d N o t e assignment for 1 J 4 i t e r m value must be interpolated since M C Q D calculation is shown for 3Ai terms only. ^AiMi, IPs f r o m X P S [84,85,86]. Chapter 4. 46 an intense N Is —> n* band at 399.7 e V [51,87,88,89]. M u c h larger contributions from N O were found i n spectra produced from gas samples taken directly from commercially supphed gas bottles of N 0 2 without any further purification. The N O contribution i n the N 0 2 could be diminished to an almost negligible level (as shown i n figs. 4.6b and 4.7a) by repeated freezing, pumping and thawing cycles of the gas cylinder. The presence of impurities may be the reason for the significant differences between the spectral features observed i n the present I S E E L S N Is and 0 Is spectra of N 0 2 and those shown i n the previously published photoabsorption results [82]. Therefore a careful study has been made of spectra obtained from commercial cylinders of N 0 2 before and after fractionation. In addition, comparisons are made wi th the known I S E E L S spectra of possible impurities. The results of these investigations are shown i n fig. 4.7. The peak at 399.5 e V observed i n fig. 4.5b is clearly shown to be due to an N O impuri ty by the vibrational structure present i n a high-resolution (0.068 eV F W H M ) I S E E L S spectrum obtained using a sample from an unfractionated commercial N 0 2 cyhn-der (fig. 4.7b). This vibrational structure is identical i n profile and energy position to that observed i n an earlier reported high-resolution I S E E L S spectrum of N O [88], which is shown as an insert i n fig. 4.7b. A further possible impuri ty that could complicate N 0 2 N Is spectra is N 2 , since peak 1 ( N Is —> 6ax) of the N Is spectrum of N 0 2 (see fig. 4.5b and table 4.4) is at 401.04 eV, whereas the intense Is —» n*(v = 1) peak of N 2 is known to be at 401.10 eV [51]. This is clearly shown by the spectra i n fig. 4.7c. i n which comparable amounts of N 2 and N 0 2 were admitted simultaneously (see comparison of spectra 4.7c and 4.7d wi th 4.7a). The familiar pattern [21,51,89,90] of the vibrationally resolved N Is —• TT* transition of N 2 can clearly be seen, superimposed on the N 0 2 ( N Is —> 6ai) band at ~401 eV. F i g . 4.7d shows the spectrum when the relative contribution of N 2 is greatly increased. F i g . 4.7a shows the same spectral region for an N 0 2 sample Chapter 4. 47 in LU cn N 0 2 samples HR ISEELS (a) NO ISEELS -ref [88] (c) (d) ^ /N02 • 0=0* E 0 = 3 0 0 0 e V A E = 0 . 0 9 0 e V Purified Cylinder Sample N02 Commercial Cylinder NO + N02 Addition of N 2 to cylinder sample NO + N 2 + N02 •MO excess - NO + N 2 + N02 3 9 8 4 0 0 4 0 2 4 0 4 4 0 6 E N E R G Y L O S S ( e V ) 4 0 8 Figure 4.7: Investigation of sample purity in N0 2 cylinders'by high resolution ISEELS measurements in the N Is region: (a) purified N0 2 ; (b cyhnder N0 2 (unpurified) showing NO impurity and the high resolution NO spectrum of Tronic et al. [88]; (c) Addition of N 2 to cyhnder N0 2 ; (d) same as (c) but with excess N 2 . Chapter 4. 48 that has been purified by repeated freezing, pumping and thawing of the cylinder. The earlier pubhshed photoabsorption results [82] show a prominent peak at ~400 eV i n the N Is spectrum and also a number of other significant differences from the present N Is (figs. 4.5 and 4.6) and 0 Is (see figs. 4.9 and 4.10 below) I S E E L S spectra of N 0 2 . We have therefore digitized the previously pubhshed N Is and 0 Is photoabsorption spectra attributed to N 0 2 [82] and transposed them from a wavelength ( A ) to an energy scale (eV). The respective photoabsorption results (figs. 4.8b and 4.8f) are compared on the same energy scales with similar resolution N Is I S E E L S spectra of pure N 0 2 (this work), N 2 [89], and N O [87] i n figs. 4.8a-4.8d, and w i t h 0 Is I S E E L S spectra of N 0 2 (this work), 0 2 [89], N O [87], and H 2 0 [91] i n figs. 4.8e-4.8i. It should be noted that some differences i n relative intensities may occur since the photoabsorption spectral intensities were presumably derived [82] from a photoplate. It can be concluded from a consideration of the data i n figs. 4.7 and 4.8 that both the discrete and continuum regions of the inner-shell photoabsorption spectra 1 of ref. [82] contain appreciable contributions from impurities that probably include N O and N 2 i n the case of N Is , and 0 2 , N O , and H 2 0 i n the case of 0 Is. This implies that peaks attributed to double excitation i n the photoabsorption spectra [82] are in fact due to impuri ty gases. In view of these considerations, no further comparison between the present I S E E L S measurements and the experimental photoabsorption results of ref. [82] w i l l be made for either the N Is or O Is spectra. That the experimental N Is I S E E L S spectra obtained i n the present work are of the N O molecule and contain no significant contributions from impurities is further confirmed by the M C Q D calculations, which predict features that correspond extremely well wi th 1 In comparing the various spectra, it should be noted that the uncertainties in the energy (wavelength) scales of the photoabsorption spectra [82] were stated to be to ± 0.4 eV (N Is and ± 1 eV (0 Is). In the present ISEELS work the energy scales are considered to have an uncertainty of ± 0.02 eV. Chapter 4. 49 4 0 0 410 4 2 0 4 3 0 5 3 0 5 4 0 5 5 0 5 6 0 ENERGY LOSS (eV) F i g u r e 4.8: Comparison of N Is and 0 Is ISEELS spectra of N 0 2 (a) and (e) with photoabsorption measurements (b) and (f) from ref. [82]. Also shown ((c) and (d)) are N Is ISEELS spectra of N 2 [89] and N O [87], as well as O Is ISEELS spectra of H 2 0 [91], 0 2 [89] and N O [87] ((g) and (i), respectively). Chapter 4. 50 the present experiment (but not the photoabsorption spectra [82]) wi th regard to both transition energies and relative intensities (see table 4.4 and figs. 4.5 and 4.6). The M C Q D calculations are therefore used for the spectral assignments (see table 4.4 and figs. 4.5 and 4.6). There is a small difference between experimental and calculated energies for the N Is to 6ai and N Is to 2b\\ virtual-valence excitations, but agreement is quite good for the higher valence (7ai(a^) and 562(&2)) a n d Rydberg excitations. It should be noted that the 7a! peak is both predicted and observed i n the discrete portion of the spectrum. This is contrary to the conclusions of Schwarz et al. [82]. W h e n an inner-shell electron i n the open-shell molecule NO2 is excited to an orbital above the (singly occupied) 6a! orbital there wi l l be three unpaired electrons and the final states accessible according to dipole selection rules are i n two doublet series ( 1 2 B ! , and 2 2 B i , for example [82]). Following Slater's treatment of three unpaired electrons [92,93], the wavefunctions of the two doublet states can be represented as linear combinations of the Slater determinants for the spin degenerate states. The coefficient of the hnear combinations are determined by the interactions (exchange integrals) among the electrons and the relative intensities for the transitions to the two doublet states can be calculated accordingly. Two situations involving different couplings can be envisaged. In the first case, if one electron is far away from the other two electrons (i.e. the situation for higher Rydbergs or ionization) the two doublet states would result from weak coupling of the remote Rydberg (or ionized) electron wi th the two more strongly coupled electrons (in the core and 6a x orbitals, respectively) and these states would be associated with the 3 A i and x A i ionization l imits . In this situation the relative intensities of the doublet states would be [94] approximately 3 : 1, as observed in X P S measurements [84]. The second type of situation occurs when the three electrons are much closer together and this corresponds to excitation to the normally unoccupied vir tual valence orbitals, which are of course quite Chapter 4. 51 localized. In this situation al l interactions between the three electrons are important and a consideration of the various couphngs leads to relative intensities for transitions to the two doublet states leading to the 3 A X and x A i ion states of p : 1 (where 1 < p < 3, depending on the interactions) wi th the lower-lying doublet state being the more intense. The unsymmetrical profiles of peaks 2 and 16 i n the N Is spectra (figs. 4.5b and 4.6b) and peaks 2 and 10 i n the 0 Is spectra (figs. 4.9b and 4.10b) of N 0 2 are consistent wi th these considerations. Thus two doublet series are expected throughout the I S E E L S spectra. However, the linewidths and densities of states, together wi th the lower relative intensity expected for those states leading to the higher energy 1A1 l imi t , result i n only the lower energy, higher intensity doublet series being clearly identifiable i n the present work. The asymmetry of peaks 16 ( N Is, fig. 4.5) and 10 ( 0 Is , fig. 4.9) may alternatively be due to the inherent nature of continuum resonance line shapes, as discussed i n refs. [96,97]. The M C Q D calculation (table 4.4) predicts a relatively low intensity for the 4s(a!) Rydberg state, which is at best rather weak (feature 3) on the low-energy side of fea-ture 4 (see also the discussion below for the situation i n the 0 Is spectrum) i n the measured spectrum (figs. 4.5b and 4.6b), i n keeping w i t h the fact that on the basis of a purely atomic-hke selection rule a Is —> 4s transition would be formally dipole for-bidden. Such atomic-hke selection rules have generally been found to be a reasonable guide for interpreting intensities of core-to-Rydberg molecular transitions, given the es-sential atomic-like character of both the in i t ia l and final orbitals [20]. In general, the predicted overall distribution of Rydberg intensity compares quite favorably wi th the high-resolution experimental spectrum (fig. 4.6). In the continuum (fig. 4.5) the position and intensity of the N Is to 5&2(&2) resonance is fairly accurately predicted (table 4.4), as is the intensity of the underlying continuum relative to the discrete structure below the ionization edge (see fig. 4.5). A width of 3 eV has been assigned to the fe2 resonance, Chapter 4. 52 as indicated by the measured spectrum. The presently obtained N Is and 0 Is inner-shell excitation spectra of N 0 2 (see figs. 4.5b, 4.6b, 4.9b and 4.10b) have each been placed on an approximate absolute dipole (i.e. optical) differential oscillator strength scale, according to the normalization principles discussed i n sections 2.3.4.3 and 2.3.8. The normalization procedures used were as follows. F i rs t ly a straight-line extrapolation of the pre-ionization edge region below the respective 6a ! excited states was used to estimate a baseline for the respective inner-shell spectra. The lower resolution N Is (fig. 4.5b) and 0 Is (fig. 4.9b) spectra were then respectively normalized to the known atomic nitrogen Is and twice the atomic oxygen Is photoionization differential oscillator strengths [57,68,70], at 25 eV above the respective ionization edges. No kinematic Bethe-Born conversions (equations 2.35 and 2.36) were apphed to the spectra shown i n figs. 4.5b, 4.6b, 4.9b and 4.10b. W h i l e such conversions are of the order of ~ Eb (3 > b > 2) as approximated i n equations 2.37 and 2.37, their effect i n the N Is and 0 Is inner-shell region w i l l be small (only about 20% and 15% variation, respectively) over the energy loss (E) ranges of the spectra shown i n figs. 4.5b and 4.9b. The slightly higher resolution spectra (figs. 4.6b and 4.10b) were then normahzed to the respective lower resolution spectra (figs. 4.6b and 4.9b). Despite the approximations and uncertainties in the procedure outlined above, it can be seen (figs. 4.5, 4.6, 4.9 and 4.10) that quite good agreement exists between the differential oscillator strength spectra obtained for the normahzed experimental spectra and those derived from the M C Q D calculations. The experimental optical oscillator strengths for transitions to the 6a i , 2b\\ and 562 orbitals as shown i n table 4.4 have been obtained by integrating the respective peak areas i n figs. 4.5 and 4.6 and then applying an estimated Bethe-Born conversion factor of E2'5. Good quantitative agreement exists between the measured and calculated values. Chapter 4. 53 > 0) CM I o X I-o 2 Ui OH h-Ul OH O O o OH Ul 10H 5\" 0-5H 0 ( a ) MCQD Calculation AT edge 6 a i 2 b ' 7a, 4pb, 4pa! 4sa! \\ —P JlJrJkl N 0 2 01s :5b 2 1 5 . 0 5.0 - 5 . 0 -15.0 T E R M V A L U E ( e V ) ( b ) ISEELS Experiment 2bi A i e d 9 e 6 = 0' E 0 = 3 7 0 0 e V AE=0.14eV r—i 1 l l l 12 34 89 J L_ 10 5 2 0 5 3 0 5 4 0 5 5 0 5 6 0 5 7 0 E N E R G Y L O S S ( e V ) 5 8 0 Figure 4.9: (a) MCQD calculation and (b) low-resolution 0 Is ISEELS spectrum of N02 in the discrete and continuum regions. Spectrum (b) as shown is not Bethe-Born corrected. Chapter 4. V to o Cr i o n i > 0) CN I - 5H o Ld cr r— LO cr O 0 10H 5H OH (a) MCQD Calculation N 0 2 01s 2bi 6a-7a-4pbi 4sa1^pai J O Aiedge I J l k I0.0 5.0 TERM VALUE (eV) 0.0 ( b ) ISEELS Experiment 6a-2b-/ 0=0' E 0 = 3 7 0 0 e V AE=0.090eV 4sa id, , i 01s edges i r I 2 i i i II—r 45 678 9 __i I— Figure discrete 530 540 ENERGY LOSS (eV) 4.10: (a) MCQD calculation and (b) high-resolution O Is ISEELS spectrum of N 0 2 in region. Spectrum (b) as shown is not Bethe-Born corrected. Chapter 4. 55 Table 4.5: Experimental and calculated data\" for 0 Is excitation of N 0 2 Experimental this work feature energy (eV) term value oscillator b strength ( x l O - 2 ) M C Q D calculation term value 0 (eV) oscillator strength ( x l O \" 2 ) assignment final orbital (to 3AX l imit) 1* 1 2 3 4 5 6 7 8 530.32 532.36 536.34 537.76 538.20 538.82 539.18 539.6 I P ^ A x ) 541.3 e I P ^ A i ) 541.97 e 10 547.18 10.98 8.94 4.96 3.54 3.10 2.48 2.12 1.7 4.1 8.1 (4.21) (3.77) (3.15) (2.79) (2.37) 540.52 0.78 (1.45) -5.88 4.3 13.12 10.62 4.80 3.22 3.05 2.40 1.98 1.89 1.70 1.50 1.49 1.48 1.44 1.39 0.96 0.89 0.84 0.83 0.82 0.80 0.78 0.60 0.57 0.54 0.53 0.52 0.52 0.51 0.40 0.37 0.37 0.36 0.29 0 -2.45 3.68 5.84 0.54 0.52 0.40 0.06 0.20 3.06 0.08 1.06 0.06 0.06 0.08 0.06 0.02 0.40 0.04 0.04 0.02 0.02 0.24 0.02 0.14 0.02 0.02 0.02 0.02 0.08 0.06 0.02 0.02 0.04 0.04 6ax 26i(ftf) Asa,! 4pbx 4pax 4pb2 5sax 7ax(a*) 3dbx 3d2ax 3db2 3da2 5p6i 5pax 4dbx 4doax 4db2 4da2 6pbx 6pax 4d2ax 5dbx 5 d o a i 5db2 5da2 7PW 7pax 5d2ax 6^0^! 6db2 6da2 6d2ax IdQd-y oo \"See fig. 4.9 and 4.10. * A kinematic B e t h e - B o r n conversion of E2,5 has been applied . c W i t h respect to the 3AX l imi t . d N o t e assignment for 1A\\ t e r m value must be interpolated since M C Q D calculation is shown for 3AX terms only. ^AtMi, IPs f r o m X P S [84,85,86]. Chapter 4. 56 Figs. 4.9a and 4.9b show the measured I S E E L S 0 Is spectra of N 0 2 at low resolution (0.14 e V F W H M ) , together wi th the M C Q D calculation. F i g . 4.10 shows the high-resolution experimental 0 Is spectrum and the M C Q D calculation i n the below-edge region i n somewhat greater detail . The approximate differential oscillator strength scale for the experimental spectra as shown i n figs. 4.9 and 4.10 was estabhshed as described above. The measured energies and oscillator strengths together with calculated oscillator strengths and term values, are shown in table 4.5. Also shown are the assignments based on a consideration of the M C Q D calculations. The values of the experimental oscillator strengths for transitions to the 6a i , 2bx and 562 orbitals reported i n table 4.5 include apphcation of an estimated Bethe-Born conversion of E 2 5 to the peak areas i n the spectra. The energies of the 0 Is 3 A X and XAX ionization edges are assigned according to X P S measurements [84,85,86] using the singlet-triplet splitt ing of 0.67 e V reported by Davis et al. [84]. As has already been discussed wi th reference to fig. 4.8 the earlier reported O Is photoabsorption spectrum of N 0 2 contains major contributions from impurities, and thus no effective comparison wi th the present work can be made. W h i l e the M C Q D calculation is i n quite good quantitative agreement with respect to the transition energies and also the relative intensities observed i n the present I S E E L S work the differences with experiment are greater than i n the case of the N Is excitation (compare figs. 4.5 and 4.6 w i t h figs. 4.9 and 4.10). These differences, particularly noticeable for the 6oti and 2&i states probably arise from the fact that the present calculations use C2v symmetry for the 0 Is excited and ionized states of N 0 2 and thus do not take account of the broken symmetry caused by a localized Is hole i n one of the O atoms. It can be seen that the 4sGt! Rydberg state is clearly visible (feature 3) in figs. 4.9b and 4.10b i n contrast to the situation i n the N Is spectrum where the corresponding transition is extremely weak. The greater intensity of the 4sai peak in the 0 Is spectrum reflects a further relaxation Chapter 4. 57 Table 4.6: Term values for N Is and 0 Is excitation of N 0 2 Transit ion Measured term value (eV) a Est imated term value (eV) 3 A X l imi t X A X l imit from (Z -f 1) analogue 6ai 11.6 12.3 11.4 2&i 9.3 10.0 8.7 7ai 1.9 2.6 4 562 -3.6 -2.9 -6a x 11.0 11.7 -26i 8.9 9.6 -7ai 1.7 2.4 -5fc2 -5.9 -5.2 -N Is N Is N Is N Is 0 Is 0 Is O Is O Is \"This work, from N0 2 ISEELS; see tables 4.4 and 4.5 ' f rom data for valence-shell excitation of O 3 using geometry and exchange correction—see ref. [82]. of the atomic-like (s —> s) selection rule due to the broken symmetry. The below-edge lax ( ° i ) final state (peak 8) is quite prominent, as is the above-edge fe2 (&2) resonance. The M C Q D calculation reproduces these features as well as the rest of the spectrum, quite well both for transition energies and oscillator strengths. In accord wi th the equivalent core model discussed in section 2.3.7, the term values of the N Is core excitations i n N 0 2 would be expected to be similar to those for the corresponding valence-shell excitations of the ( Z + 1) analogue, O 3 , if geometry and exchange corrections are taken into account [82]. This situation for core-excited N 0 2 and valence-excited O 3 has been discussed in some detail by Schwarz et al. [82]. The term values of the O 3 valence excitation obtained from extensive C I calculations were adapted [82] to the N 0 2 geometry and allowance was made for exchange effects (since core-excited N 0 2 is open-shell). Table 4.6 shows the adapted 0 3 valence-shell term values [82] i n comparison wi th the presently measured N Is term values for several transitions. A quite reasonable correspondence exists, given the various approximations Chapter 4. 58 involved [82] i n obtaining the estimated term values of O 3 valence excitation. In a similar fashion the term values for the 0 Is excitations of N O (see table 4.6) could be used to obtain estimates of the valence-shell excitation energies of the species N O F . A consideration of the data i n tables 4.4-4.6 shows that the term values for given core-to-virtual-valence transitions show systematic differences, w i t h the 0 Is values being consistently lower than those for N Is. In contrast, and as expected (see discussed i n section 2.3.1), the term values for Rydberg transitions to the same final orbitals are almost identical for N Is and 0 Is excitation. In conclusion, the present work demonstrates the use of high quahty M C Q D calcula-tions of transition energies and osciUator strengths to provide definitive interpretations of high resolution inner-shell electronic excitation spectra for the NO2 molecule. Chapter 5 Absolute Differential Oscillator Strengths for the Photoabsorption and the Ionic Photofragmentation of C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 The relative photoabsorption spectra and time of flight ( T O F ) mass spectra used to derive absolute differential oscillator strengths (total and partial) for C F 4 , C F 3 C 1 , C F 2 C 1 2 and CFCI3 were obtained using the dipole (e,e+ion) spectrometer described i n section 3.1. 5.1 Electronic Structures The point group symmetries for C F 4 , C F 3 C 1 and C F C 1 3 , and C F 2 C 1 2 are T 4s Rydberg transition [104,119]. Similarly, Rydberg transitions (5e —> 5s and/or 5e —» 3d), (4e —> 4s and/or l a 2 —> 4p) and 4ax —» 4s are the assignments for the features at 11.5 eV, 13.5 eV and 16.5 e V Chapter 5. 84 80r > CD N I o O OH E-H ( a ) CP .1.0 60 40 OH O EH < i—i o CO o <: I—H H 0. 0 ^ 15 r OH W & H Q & O i—i E-H PH OH O CO PQ < O E-H O K OH 10 0 l o o o c >-o z w o E w 0.5 z o H 4p) and 2ai —» 3s [119]. A broad feature (fig. 5.12a) centered at ~40 eV is observed i n the present photoab-sorption measurement for C F 3 C 1 and is also seen i n the work of Lee et al. [101]. A similar but even more prominent broad feature at ~40 e V is also present i n the photoabsorption differential oscillator strength spectrum of CF4 (fig. 5.11a). Similar but weaker features are also seen i n the photoabsorption differential oscillator strength spectra of CF2CI2 (fig. 5.13a) and CFCI3 (fig. 5.14a) at photon energies ~40 eV and ~ 3 7 e V respectively. The amphtudes of these structures decrease as the number of F atoms decreases. In this regard it is noteworthy that these structures occur close to the region of the onsets of ionization of the inner valence (F 2s) orbitals (see chapter 6 and ref. [120]) and it is possible that they are associated wi th the cross sections for these processes. However it should be noted that M S - X a calculations and P E S partial cross section measurements of CF3CI indicate a max imum i n some of the outer valence part ial photoionization chan-nels i n a similar energy range [121]. The authors [121] pointed out that these predicted and observed outer valence structures are not due to shape resonances but are probably associated wi th scattering (diffraction) of the photoelectrons by the neighboring atoms i n the molecule (see also refs. [122,123,124,125]). Above a photon energy of 24 eV the present results agree wi th those reported by Lee et al. [101] to better than 10%. However in the 16-24 eV region the cross sections reported by Lee et al. [101] are ~20 % higher than the present results (figs. 5.12a and 5.12c). Similar discrepancies i n this same energy region (16-24 eV) have also been observed i n the comparisons between our results (figs. 5.11 and 5.13) and those of Lee et al. [101] and W u et al. [110] for the molecules CF4 and CF2CI2. A s discussed i n section 5.2.1, the differences between the present results and those reported by Lee et al. [101] and W u et Chapter 5. 86 al. [110] i n the energy region 16-24 eV are most likely due to a systematic error i n the use of the Sn film i n the latter work. The data of Jochims et al. [107] were single point normalized to the photoabsorption measurement reported by Rebbert and Ausloos [108] whereas the data of K i n g and McConkey [104] were put on an absolute scale at a single point to an average of the photoabsorption measurements reported by Jochims et al. [107] and the electron impact measurement reported by Huebner et al. [126]. The previously reported single point normalized results [104,107] are of the same overall shape but are considerably higher than the present results below 23 eV (fig. 5.12c). Above 23 e V the data of K i n g and McConkey [104] approach the present data with increasing photon energy. Our data are also lower than the photoabsorption measurements of Gilbert et al. [106] below 17 eV, whereas at higher energy their data diverge rather drastically from all other measurements (fig. 5.12c). The absolute photoabsorption cross sections of C F 3 C 1 from 124-270 e V have been measured by Cole and Dexter [109] using synchrotron radiation. Agreement w i t h the present work is good below 160 eV but the measurements diverge at the higher energies and exhibit a 20% difference at 190 e V (fig. 5.12b). 5.2.3 The CF 2 C1 2 Photoabsorption Differential Oscillator Strengths The presently reported measurements on C F 2 C 1 2 are compared wi th earlier photoab-sorption measurements [109,110] obtained using synchrotron radiation hght sources i n figs. 5.12a and 5.13b. In fig. 5.13c the present photoabsorption data are also compared i n the low energy region (8.5-36 eV) with previously reported results obtained using syn-chrotron radiation [107,110] and the helium Hopfield continuum [106] as light sources. The previous photoabsorption data derived from small momentum transfer electron i m -pact measurements [104] have been digitized from the reported diagram and are also shown i n fig. 5.13c. Chapter 5. 87 i > N I o 80 K H O K E-H oo o E-1 < hJ r—I u o < r—I OH W E n r—t Q 40 0. 0 K 10 O E-H DH PH O CO CQ < o O 0L (a) V > u z w o i l ! w 0.5K z o < N z 2 0.0 • this work o ref. [108] 20 40 60 PHOTON ENERGY (eV) • P h Abs , dipole (e,e) ° P h Abs [110] 20 40 60 80 100 \\ ^ \" j ^ ^ Polynomial fit • P h Abs , dipole (e,e) o P h Abs [109] 120 80 40 • Ph Abs, dipole (e,e) 0 Ph Abs [110] 1 Ph Abs [107] 'A * Ph Abs [106] • Ph Abs [111] a Ph Abs from EELS [104] 80 120 160 0 ^ 200 10 20 30 40 P H O T O N E N E R G Y (eV) Figure 5.13: A b s o l u t e photoabsorption differential oscillator strengths for the valence shell of CF2CI2. a) 8.5-100 e V (insert shows ionization efficiency), b) 70-200 e V . c) 8.5-36 e V (expanded scale). Chapter 5. 88 The different energy resolution characteristics of the present dipole (e,e) techniques and optical methods (i.e. the large differences i n energy resolution and the fact that the energy resolution changes wi th photon energy i n optical experiments—see section 5.2) comphcates comparison of the various data sets at lower energies, particularly i n the region of discrete excitation. A similar difficulty exists i n comparing the present results at 1 e V F W H M resolution in the discrete excitation region w i t h the data [104] derived from intermediate impact energy E E L S at ~0.05 eV resolution. In particular, meaningful comparisons of such data wi th the present work are not possible below 12.5 e V and the higher resolution data from references [104,106,107] are omitted from fig. 5.13c i n this region. The shape of the presently reported photoabsorption spectrum i n the energy region below 25 e V (fig. 5.13c) is similar to the shapes of the spectra reported i n refs. [110, 104,107]. However considerable variations exist i n the absolute magnitudes of the various data sets. In the structured low energy region below 15 eV Rydberg transitions have been assigned to the various features. In particular, the peak at 13 e V just discernible i n the present low resolution work has been attributed to the Rydberg transitions 462 —> 3d and/or 3a 2 —> 4s [49,104,119]. The Rydberg transitions 3a 2 —> 3d and/or 66x —• 5p and (2a 2+5&i) —> 4s are the assignments for the structures i n the 15-17 e V region [104,119] (unresolved i n the present work). The maximum at 19.5 e V has been assigned to the 7ai —* 4s Rydberg transition [49,104,119], while the broad shoulder at ~24.5 eV has been suggested by R o b i n [49] to be due to a Rydberg transition originating from the carbon 2s orbitals. A term value of ~ 3 e V can be derived for this transition from the ionization potential of the C 2s orbitals (2ai + l& 2 ) of C F 2 C 1 2 presented i n chapter 6. The magnitude of this term value is consistent with the assignment (2a 1 ? 162) —> 3s or (2a x , 162) —> 3p, since the typical magnitudes of Rydberg term values are i n the range 2.8-5.0 e V for the lowest s orbital and 2.0-2.8 for the lowest p orbital [47]. It should be noted that the Chapter 5. 89 Rydberg transition assigned to a particular structure depends largely on the ionization energies taken from P E S measurements (i.e. different ionization energies give rise to different term values (see sections 2.3.1 and 2.3.2) for the same structure and therefore the assignment of the particular structure can be different i n uti l iz ing ionization energies from different P E S measurements). The rise i n photoabsorption differential oscillator strength at 200 e V i n fig. 5.13b is caused by excitation of C l 2p inner shell electrons [119]. Above a photon energy of 24 eV the present results are i n excellent quantitative agreement w i t h those reported by W u et al. [110]. However i n the 16-24 eV region the cross-sections reported by W u et al. [110] are ~ 1 0 % higher than the present work (figs. 5.12a and 5.13c ). Similar discrepancies i n this energy region have been observed i n earlier comparisons between our results (figs. 5.11 and 5.12) and those of Lee et al. [101] for the molecules C F 4 and C F 3 C 1 . As pointed out i n section 5.2.1, the discrepancies i n the 16-24 e V energy region are most likely due to a systematic error i n the use of a Sn film i n the work of W u et al. [110] and Lee et al [101]. The data of K i n g and McConkey [104] and of Jochims et al. [107], which were single point normalized to the photoabsorption measurements reported by Person et al. [ I l l ] and Rebbert and Ausloos [108] respectively, are of similar shape but proportionally higher than the present results below 24 eV (fig. 5.13c). In the 24-30 eV energy region the data of K i n g and M c C o n k e y [104] approach the present data w i t h increasing photon energy, however, above 30 eV their data fall below ours. The agreement between the present results and those of Gilbert et al. [106] is good i n the region 12-16 e V , but above 16 e V their data diverge rather drastically from all other measurements (fig. 5.13c). The absolute photoabsorption cross sections of C F 2 C 1 2 from 124 e V to 270 e V have been measured by Cole and Dexter [109] using synchrotron radiation. Agreement wi th the present work is good below 130 eV but the measurements diverge at the higher energies and exhibit a 20% difference at 190 e V (fig. 5.13b). Chapter 5. 90 A broad feature of low intensity (fig. 5.12a) centered at ~ 4 0 eV is observed. Similar but more prominent broad features were also seen i n the photoabsorption differential oscillator strength spectra of CF4 (fig. 5.11a) and CF3CI (fig. 5.12a). They were i n -terpreted (section 5.2.2) as being associated wi th inner valence ionization or possibly scattering (diffraction) of the outgoing (outer valence) photoelectrons by the neighboring atoms i n the molecules [121,122,123,124,125]. The presently observed feature (fig. 5.12a) probably also has similar origins. F i n a l assignment must await detailed P E S experi-ments and theoretical calculations. A even weaker feature at ~ 3 7 e V is also present i n the photoabsorption differential oscillator strength spectra of CFCI3 (fig. 5.14a) 5.2.4 The CFC1 3 Photoabsorption Differential Oscillator Strengths The presently reported measurements are compared wi th earlier photoabsorption mea-surements [109] obtained using synchrotron radiation as the light source'in fig. 5.14b. In fig. 5.14c the present photoabsorption data are also compared i n the low photon en-ergy region wi th previously reported results obtained using synchrotron radiation [107] and the hehum Hopfield continuum [106] as hght sources. The photoabsorption data (~0.05 e V F W H M resolution) derived from small momentum transfer electron impact measurements [104] were reported both i n the form of integrated oscillator strengths (in a table) and differential oscillator strengths (in a diagram). In fig. 5.14c i n order to make meaningful comparisons with the present results, which were obtained with lower resolution (1 e V F W H M ) , below 12.5 eV the results from ref. [104] are shown as av-erage differential oscillator strengths (i.e. the integrated oscillator strengths divided by the corresponding energy intervals). This has been done i n order to smooth away sharp structures i n the discrete region. Above 12.5 eV the results are shown as differential oscillator strengths digitized from the reported diagram [104]. Chapter 5. 91 i > I . O E-H O JZi W OH E-H in OH O < r — I o w o < I—I E-H 55 H OH OH OH Q 55 O E-H OH OH O Ul QQ < O E-H O OH (a) 80 40 1 > ; \\ £ 1 . 0 u 2 W • • • • • • • u z o • • o d IIVZINOI • • V o d IIVZINOI • • • • • this work ° ref. [108] CFC1 20 40 PHOTON ENERGY (eV) Ph Abs, dipole (e,e) •°»»* • 0 20 40 60 80 100 10 \\ ( b ) •—- Polynomial fit • Ph Abs, dipole (e,e) o Ph Abs [109] 0 L 160 120 • Ph Abs, dipole (e,e) o Ph Abs [107] & ° P h Abs [106] ° r # <*P D ph Abs from EELS [104] 80 120 160 200 P H O T O N E N E R G Y (eV) Figure 5.14: Absolute photoabsorption differential oscillator strengths for the valence shell of CFCI3. a) 6-100 eV (insert shows ionization efficiency), b) 65-200 eV. c) 6-38 eV (expanded scale). Chapter 5. 92 The different energy resolution characteristics of the present dipole (e,e) techniques and optical methods (i.e. the large differences i n energy resolution and the fact that the energy resolution changes wi th photon energy i n optical experiments—see section 5.2) complicates comparison of the various data sets at lower energies, particularly i n the region of discrete excitation. In particular meaningful comparisons of such data w i t h the present work are not possible below 12.5 e V for CFCI3, and therefore the higher resolu-t ion data from refs. [106,107] are omitted from fig. 5.14c i n this region. The presently reported photoabsorption spectrum is similar i n shape to those reported i n refs. [104,107] i n the energy region below 24 eV (fig. 5.14c). The absolute magnitudes of the photoab-sorption cross-sections, however, exhibit considerable variations between the various data sets. Rydberg transitions have been assigned to the features i n the low energy region below 24 eV. In particular, the peak at 10 eV just discernible in the present low resolu-tion work has been assigned to the Rydberg transition 4e —> 4s [104,119]. Similarly the broad structure at ~ 1 5 eV has been attributed to the Rydberg transitions 2e —> 4s and 4ai —> 4s [104,119]. The maximum at ~17.5 e V and the shoulder at ~22 eV have been recently assigned to the Rydberg transitions 3ax —» 4s and l e —> 4s [119]. A feature only just discernible at ~ 3 7 e V is similar to progressively less intense features observed i n C F 4 (fig. 5.11a), CF3CI (fig. 5.12a) and C F 2 C 1 2 (fig. 5.13a). A s discussed earlier (section 5.2.2) such features are perhaps due to the onsets of inner valence (i.e. F 2s) ionization processes or possibly scattering (diffraction) of the outgoing (outer valence) photoelectrons by the neighboring atoms i n the molecule [121,122,123,124,125]. A more definite assignment of this feature must await more detailed P E S part ial cross section experiments and theo-retical calculations. The rise of differential oscillator strength at ~200 e V (fig. 5.14b) is due to excitation of the C l 2p inner shell [119]. Chapter 5. 93 The electron impact measurements of K i n g and McConkey [104] and the photoab-sorption measurements of Jochims et al. [107] are considerably higher than the present work below 24 e V . Above 24 eV the data of K i n g and McConkey [104] approach the present results w i t h increasing photon energy. The agreement between the present re-sults and those of Gilbert et al. [106] is good i n the region 13-16 e V , but above 16 eV their data diverge rather drastically from all other measurements. The absolute pho-toabsorption cross sections of C F C 1 3 from 124 eV to 270 eV have been measured by Cole and Dexter [109] using synchrotron radiation and at 124 eV these results agree with the present work wi th in 10%. However the measurements diverge at the higher energies and exhibit a 25% difference at 190 eV (fig. 5.14b). 5.2.5 Comparison of the Photoabsorption Differential Oscillator Strengths of C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 Although the total integrated oscillator strengths for the four freons al l have similar val-ues (i.e. ~ 33, see section 5.2), the shapes of the photoabsorption differential oscillator strength spectra vary considerably i n going through the series CF4, CF3CI, C F 2 C 1 2 and C F C 1 3 , w i t h the oscillator strength becoming progressively more concentrated i n the low photon energy region as the number of C l atoms increases. W i t h increasing number of C l atoms i n the freon molecules, the photoabsorption differential oscillator strength decreases more rapidly i n the 20-40 eV region, while above ~40 e V the rate of decrease be-comes smaller. For example, at energies ~35 eV above the photoabsorption onsets of the respective spectra the differential oscillator strengths for C F 4 , C F 3 C 1 , C F 2 C 1 2 and CFCI3 are 0.29, 0.26, 0.20 and 0.14 e V - 1 respectively, whereas at ~100 e V above the onsets the values are 0.09, 0.07, 0.06, 0.06 e V - 1 respectively. These differences between the differen-t ia l oscillator strengths for C F 4 , C F 3 C 1 , C F 2 C 1 2 and CFCI3 occur because an increasing Chapter 5. 94 number of molecular orbitals wi th predominantly C l 3p atomic character contribute to the differential oscillator strengths. This effect is due to the Cooper min imum [25,127] i n the C l 3p atomic orbital photoionization cross-sections at ~35-40 e V [68]. 5.3 The C F 4 Photoabsorption Differential Oscillator Strengths for the C Is and F Is Inner Shells and the Valence Shell Extrapolation In this section the extrapolated C F 4 differential oscillator strengths for valence excitations obtained using formula 5.41 and also the differential oscillator strengths for C F 4 C Is and F Is excitations obtained from normalization of Bethe-Born converted E E L S spectra are compared wi th the reported results from direct X-ray absorption measurements. The absolute differential oscillator strengths of inner shell excitations for other freon molecules w i l l be reported i n chapter 7. A s discussed i n section 5.2 above, an analytical function was fitted to the valence shell experimental data for C F 4 from 70 to 200 eV and was used to estimate the differential os-cillator strength for valence shell photoabsorption from 200 e V to infinity. This estimated valence shell contribution is shown by the dashed hnes i n figs. 5.15a (valence shell), 5.15b (C Is region) and 5.15c (F Is region) i n the photon energy regions 200-250, 270-350 and 670-740 e V respectively. For the comparison to be made below, we have digitized literature results of C Is and F Is inner shell photoabsorption measurements [128,129] and they are also shown i n fig. 5.15b and 5.15c by the sohd and dash-dot hnes respec-tively. In the region below the C Is IP (i.e. the K edge), there is excellent agreement between our valence shell estimation of the absolute differential oscillator strengths us-ing the analytical function and the direct photoabsorption measurements obtained using X - r a y characteristic hnes at 248.0 eV (fig. 5.15a) and 277.4 e V (fig. 5.15b). In the present work, we have also converted the earlier reported relative inner shell Chapter 5. 95 i > CM I o o z LJJ CH \\— (/) CH O 60 40 2 0 -o IE CL 0 V A L E N C E Ph Abs, dipole (e,e) this work Polynomial fit Ph Abs [128] 60 40 20 0 d 20 o ui o p 10h z CH 50 C 1s 100 150 200 • Ph Abs from EELS [130] Polynomial fit Ph Abs (valence and C ls)[128] • Ph Abs (valence)[128) X Extrapolation (valence)[128] K-edge 1 = 0 o t— CL. CH O Ul CD < O 0 (b) 280 F 1s .,300 K-edge 320 340 (c) Ph Ab« from EELS [130] Polynomial fit (valence) Ph Abs (valence) and C ls[68] Ph Abs (valence + C Is + F ls)[128] Ph Abs (valence + C ls)[128] Extrapolation (valence)[128] Ph Abs (valence + C Is + F ls)[129] Ph Abs (valence + C ls)[129] 20 10 0 5 '^k^^r^j^^ 0 680 700 720 PHOTON ENERGY (eV) 740 F i g u r e 5.15: Absolute photoabsorption differential oscillator strengths for (a) the Valence-shell, (b) the C Is region and (c) the F l s region of C F 4 . In each spectrum the dashed line represents the fitted (valence shell, 70-200 eV) and extrapolated valence shell contribution. See text for details. Chapter 5. 96 electron energy loss spectra ( ISEELS) of C F 4 [130] to absolute photoabsorption differen-t ia l oscillator strengths by using known absolute atomic photoionization differential os-cillator strengths, according to the normalization principles discussed i n sections 2.3.4.3 and 2.3.8. The procedures used were as follows. First ly, the estimated background and valence shell continuum contributions were subtracted from the raw I S E E L S data [130] by straight hne extrapolation of the pre-ionization edge spectral intensity. (A modified subtraction procedure is used i n chapter 7.) The resulting spectrum due to inner shell excitation alone was then converted to an approximate relative optical spectrum by using equations 2.35 and 2.36 where the estimated kinematic (Bethe-Born) factor was taken to be proportional to E2-5. The absolute photoabsorption differential oscillator strengths were then obtained by normalizing the relative optical spectrum at a reasonably high photoelectron energy (i.e. ~25-30 eV above the Is edge) to the sum of corresponding atomic Is cross-sections for the respective constituent atoms [57,70,68]. The carbon Is atomic cross-section at 30 e V above the ionization potential was used to place the C Is spectrum of C F 4 on the absolute differential oscillator strength scale (fig. 5.15b). Simi-larly, four times the fluorine Is atomic photoionization cross-section at 25 e V above the F Is ionization potential was taken for the normalization of the C F 4 F Is spectrum. The so obtained C Is inner shell absolute photoabsorption differential oscillator strength spectrum of C F 4 was then added to the valence continuum contribution determined by the present analytical function to give the total photoabsorption differential oscillator strengths i n the C Is region as shown i n fig. 5.15b. Similar procedures were used for the F Is spectrum shown i n fig. 5.15c. Our results are compared to those obtained by direct photoabsorption measurements [128,129]. In the C Is region, the presently converted I S E E L S results are higher than the optical measurements [128], especially i n the discrete region. This could be due to the uncertainty involved in the background subtraction Chapter 5. 97 procedure. It should be noted that the differential oscillator strengths for C F 4 C Is is re-derived i n chapter 7 using a modified background subtraction procedure. However, even after the re-derivation, the differential oscillator strength as obtained i n chapter 7 is st i l l higher than the optical results i n the discrete transition region. This discrepancy may part ly be due to a systematic error i n the present normahzation method, since for highly fluorinated molecules such as C F 4 [44] the presence of potential barriers may depress the photoabsorption cross-section corresponding to the central atom i n the region of the con-t inuum immediately above the IP. As for the F Is spectra, the converted I S E E L S spectra are i n good agreement with the optical measurements [128,129] i n the continuum region with in statistical error. The differences between the photon measurements and the con-verted I S E E L S spectra i n the discrete regions of the C Is and also the F Is spectra might be due to hne saturation (bandwidth) effects which can seriously perturb the intensities of discrete transitions i n photoabsorption measurements [26,131]. Since electron impact excitation is non-resonant, such bandwidth effects do not occur i n electron energy loss spectra [8,26]. The photoabsorption spectrum of ref. [129] shows an energy shift which is probably due to energy calibration errors. 5.4 Molecular and Dissociative Photoionization of C F 4 , C F 3 C 1 , C F 2 C 1 2 and CFC1 3 T i m e of flight mass spectra have been measured using the dipole (e,e+ion) spectrometer described i n section 3.1 i n the equivalent photon energy ranges 15.5-80 eV, 12.5-80 eV, 11.5-70 e V and 11.5-49 e V respectively for C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 . Typica l T O F mass spectra are shown i n figs. 5.16, 5.17, 5.18 and 5.19, obtained respectively at 80 e V for C F 4 , at 45 e V for C F 3 C 1 , at 50 e V for C F 2 C 1 2 , and at 49 eV for C F C 1 3 . The positive ions detected i n the T O F mass spectra of C F 4 were CF3 , C F 2 , C F + , C + , F + Chapter 5. 98 2 3 4 5 TIME OF FLIGHT (/^sec) Figure 5.16: Time of flight mass spectrum of CF 4 at 80 eV equivalent photon energy. Chapter 5. 99 2 3 4 T I M E O F F L I G H T ( / x s e c ) Figure 5.17: Time of flight mass spectrum of CF 3C1 at 45 eV equivalent photon energy. Chapter 5. 100 10-in *E D >. i_ O • — _Q v_ O >-\\— CO 0+ C F 2 C I 2 • N + 0 + F + hy=50eV n CFCI + I ] CF 2 CI + CCI2 CFClJ n n 1 1 1 r Z 3 4 TIME OF FLIGHT (/usee) Figure 5.18: Time of flight mass spectrum of CF2C12 at 50 eV equivalent photon energy. Chapter 5. 101 % 10-0 CFCI hi/=49eV 5 C l + C F c+ F, N + 0 + F + N2 C C I + I 1 n i C F C I + . n C F C l £ CCl£ m I ft | ccij W T T T I 2 3 4 TIME OF FLIGHT (/xsec) Figure 5.19: Time of flight mass spectrum of CFC1 3 at 49 eV equivalent photon energy. Chapter 5. 102 and the doubly charged ion C F 2 + . The CF4 molecular ion was not found, i n accord wi th previous work [132]. For C F 3 C 1 , the molecular ion C F 3 C 1 + and the fragment (dissocia-tive) ions C F 2 C 1 + , C F + , CFC1+, C F + , C C 1 + , C l + , C F + , F+, C+ as well as the doubly charged ions C F 2 C 1 2 + and C F C 1 2 + were detected. The positive fragment (dissociative) ions CFC1+, C F 2 C 1 + , CC1+, CFC1+, C F + , CC1+, C l + , C F + , F+, C+ and the doubly charged ion C F 2 C 1 2 + were detected for C F 2 C 1 2 . The molecular ion C F 2 C l J was not ob-served i n contrast to the situation for CF3CI. In the T O F spectra of CFCI3, the positive fragment ions CC1+, CFC1+, CC1+, CFC1+, C C 1 + , C l + , C F + , F+ and C+ were detected. Unhke the situation for C F 3 C 1 , the molecular ion (CFCI3 ) was not detected. Also no doubly charged ions were observed for CFCI3 i n contrast to the other freons. Photoion branching ratios determined by integrating the mass peaks i n the T O F spectra are re-ported diagrammatically i n figs. 5.20, 5.21, 5.22 and 5.23 for C F 4 , C F 3 C 1 , C F 2 C 1 2 and CFCI3 respectively, and also numerically i n respective tables 5.17, 5.18, 5.19 and 5.20. It can be seen that the dominant ion produced from C F 4 by ionizing radiation i n the region 15.5-80 e V is CF3 (see table 5.17 and fig. 5.20) and the molecular ion C F 3 C 1 + is only observed w i t h any appreciable intensity below 30 eV (see table 5.18 and fig. 5.21). In comparing the branching ratios of various photoion fragments produced by pho-toionization of CF3CI, C F 2 C 1 2 and C F C 1 3 , it is found that the excited molecular ions favor fragmentation processes involving loss of one or more C l atoms over loss of an F atom. For example the molecules C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 have a larger yield of the CF3 , C F 2 C 1 + and CFCLj\" ions (per C - C l bond) resulting from loss of a C l atom than of the C F 2 C 1 + , CFChj\" and CCht ions (per C - F bond) from lose of an F atom respectively. These phenomena are al l consistent wi th the fact that the bond strength of a C - C l bond (3.58 eV) is weaker than that of a C - F bond (4.84 eV) . It is also observed that the intensities of the ion fragments C + and F + are highest for the photoionization of C F 4 Chapter 5. 103 100 o < cr o < cr CD O O I— o X Q_ 50 0 10 0 10 0 10 0 10 0 0 w v.. —i CF, + III It ****** ** ** M XAB C D E F CF, + C F + C + F + tt 10 20 30 40 50 60 70 80 PHOTON ENERGY (eV) Figure 5.20: Branching ratios for dissociative photoionization of CF4. The vertical arrows repre-sent expected thermodynamic appearance potentials (see table 5.21) and the hatched hnes indicate the vertical ionization energies [99,98] for production of the electronic states of C F ^ . Chapter 5. 104 1 0 0 4 0 g 2 0 o 0 E 100 o 2 5 0 < CD -L 0 0 . 5 0 1 0 CF3CI ! ABCOEF Q H ..... I J ~l 1 r T r CF3CI + -1—•—1 1 r CF2CI+ -1 1 1 1— -1 1 1 r . C F C I \" i 1 1 1 f 1 1 C F 2 + —-*i 1 1 1 1 1 1 0 2 0 4 0 6 0 8 0 X A 6 EF G H u 1 J CCI+ • • • • • 1 1 1 1 1 ' ' C l + . 1 -t ; 1 1 1 1 1 III / CF+_ 1 \" 1 1 1 1 ' . : F+_ III! 1 1 i \" 1 II 11 1 1 1 ; . : c + -CF2CI+T 2 0 ' 1 4 0 6 0 8 0 2 0 2 0 1 0 0 PHOTON ENERGY (eV) Figure 5.21: Branching ratios for dissociative photoionization of CF3C1. The vertical arrows represent expected thermodynamic appearance potentials (see table 5.22) and the hatched lines indicate the ver-tical ionization energies presented in chapter 6 and reported in ref. [100] for production of the electronic states of CF 3C1 + . Chapter 5. 105 O I— < Cd O z o z < CD Z O O (— o IE Q_ 20 10 0 100 50 0 0.2 0 5 0 10 0 CF2CI2 CFCI2+ y * . . . \"nidi (III II C C I A C E G J K (L.M) N O . . . - I I I 1 1 1 - \\ CF2CI + 'v. s \\ Xfcj D F (H,T) jiijii ij II II A C E G J K (L.U) N 5 1 ' 1 1 1 1 1 ' C l + 11 1 \"1 1 1 1 1 1 CF+ _ i i / 1 1 1 1 1 1 • • • * CCI2+ 1 1 1 1 1 1 f. ' ...\" CFCI+ • — i i i i i i 1 . i i . . . . . . . . . • • * F + \" i i \" i i i i ii i n ' c + ~ -1 1 1 1 1 1 - CF2+ II. ' i \" i i i i i CF2CI++-4 2 0 20 10 0 10 5 0 2 0 2 0 20 40 ' 60 ' 8 0 20 40 P H O T O N E N E R G Y ( e V ) 60 Figure 5.22: Branching ratios for dissociative photoionization of CF2C12. The vertical arrows represent expected thermodynamic appearance potentials (see table 5.23) and the vertical lines indicate the vertical ionization energies presented in chapter 6 and reported in ref. [100] for production of the electronic states of CF2Clj\". Chapter 5. 106 0 1 0 0 o I— < DC 5 0 o o X o 2 5 cc DQ CFCI mi % t it XABC D EF 0 H I 0! r 2 0 -1 0 0 - | 1 1 r 1 0 C C h + T 1 1 1 r— r C F C I 2 + - i 1 1 1 1 r CCIo + ~i i C F C I + ~r~^ 1 1 1 1 1 r 2 0 3 0 4 0 t l f I f I I XABC 0 EF Q fit I C C I + 1 1' l i C I + _ 1 I / 1 1 1 \" \" \" * C F + -1 1 1 1 1 1 I * • ' F + \" t 1 i i i i i i 1 1 : c + -2 0 PHOTON ENERGY (eV) 3 0 1 0 5 0 2 0 0 10 5 0 0 2 1 0 4 0 5 0 Figure 5.23: Branching ratios for dissociative photoionization of CFC13. The vertical arrows represent expected thermodynamic appearance potentials (see table 5.24) and the hatched lines indicate the ver-tical ionization energies presented in chapter 6 and reported in ref. [100] for production of the electronic states of CFCI J . Chapter 5. 107 Table 5.17: Photoion branching ratios of CF4 Photon energy (eV) Branching ratio (%) CF+ CF+ 15.5 100.00 16.0 100.00 16.5 100.00 17.0 100.00 17.5 100.00 18.0 100.00 18.5 100.00 19.0 100.00 19.5 100.00 20.0 100.00 20.5 99.86 0.14 21.0 99.74 0.26 21.5 98.52 1.48 22.0 95.86 4.14 22.5 92.08 7.92 23.0 89.98 10.02 23.5 89.19 10.81 24.0 89.17 10.83 24.5 89.91 10.09 25.0 90.90 9.10 25.5 91.39 8.61 26.0 92.63 7.37 26.5 93.54 6.46 27.0 93.99 6.01 27.5 94.83 5.17 28.0 94.44 4.56 28.5 95.58 4.42 29.0 95.83 4.17 CF+ F+ C + C F 2 , * continued on next page Chapter 5. 108 Table 5.17: (continued) Photon energy Branching ratio (%) (eV) CF+ CF+ CF+ F+ C+ 29.5 95.62 4.20 0.18 30.0 95.47 4.23 0.30 31.0 95.03 4.31 0.66 32.0 94.06 4.49 1.46 33.0 92.91 4.83 2.27 34.0 91.90 4.99 3.11 > 35.0 89.97 5.42 4.29 0.17 0.15 36.0 88.64 5.40 5.05 0.38 0.53 37.0 86.90 5.56 5.65 0.73 1.16 38.0 85.28 5.64 5.94 1.17 1.97 39.0 83.17 5.78 6.34 1.65 3.06 40.0 81.49 6.07 5.99 2.20 4.25 41.0 80.44 6.12 5.27 2.91 5.26 42.0 81.10 6.04 4.87 3.10 4.90 43.0 81.12 6.37 4.27 3.48 4.77 44.0 81.19 6.36 3.79 3.69 4.82 0.14 45.0 81.62 6.61 3.62 3.96 4.05 0.14 46.0 82.11 6.41 3.63 3.81 3.75 0.29 47.0 81.74 6.58 3.49 4.15 3.64 0.40 48.0 82.02 6.37 3.52 4.14 3.53 0.41 49.0 81.25 6.23 3.73 4.25 3.88 0.64 50.0 81.45 5.85 3.56 4.78 3.87 0.49 55.0 76.90 6.20 4.48 6.70 5.10 0.63 60.0 74.35 6.39 5.05 7.97 5.22 1.03 65.0 72.61 6.77 5.46 9.01 5.15 1.00 70.0 70.96 6.85 5.42 10.46 5.19 1.13 75.0 69.08 6.52 6.09 11.48 5.65 1.17 80.0 67.29 6.42 6.45 12.52 6.18 1.14 Chapter 5. 109 Table 5.18: Photoion branching ratio for C F 3 C 1 Photon Branching ratio (%) energy (eV) C F 3 C l + CF 2 C1+ CF+ CFC1+ CFt CC1+ C1+ C F + F+ C + C F 2 C 1 2 + 12.5 7.83 92.17 13.0 2.32 97.68 13.5 1.80 98.20 14.0 1.94 97.77 0.29 14.5 1.54 1.06 97.00 0.40 15.0 1.39 3.19 95.17 0.25 15.5 0.97 8.84 90.03 0.16 16.0 0.90 16.10 82.93 0.07 16.5 0.64 24.95 74.27 0.14 17.0 0.55 30.53 68.78 0.14 17.5 0.65 34.54 64.66 0.15 18.0 0.57 35.97 63.28 0.19 18.5 0.44 36.87 62.54 0.16 19.0 0.59 37.19 61.70 0.10 0.42 19.5 0.34 37.58 59.90 0.15 1.31 0.72 20.0 0.34 36.06 58.76 0.10 2.45 2.29 20.5 0.56 33.55 56.87 0.19 4.51 4.32 21.0 0.40 30.11 52.85 0.16 9.46 7.02 21.5 0.45 28.23 50.29 0.32 12.92 7.78 22.0 0.42 28.04 48.30 0.42 14.49 8.33 22.5 0.38 28.37 47.20 0.52 15.29 8.23 23.0 0.44 28.99 46.09 0.55 15.22 8.09 0.62 23.5 0.40 29.51 45.99 0.54 15.17 7.65 0.74 24.0 0.27 30.44 45.13 0.52 14.90 7.88 0.86 24.5 0.27 30.85 44.77 0.49 14.51 7.85 1.26 25.0 0.41 31.64 43.53 0.53 14.23 8.11 1.54 25.5 0.37 32.52 43.02 0.48 13.54 8.14 1.93 26.0 0.43 32.94 42.20 0.44 13.03 8.61 2.35 26.5 0.23 33.67 41.49 0.43 12.69 8.58 2.92 27.0 0.29 34.68 40.46 0.35 12.32 8.56 3.34 27.5 0.26 35.10 40.05 0.33 12.12 8.28 3.87 continued on next page Chapter 5. 110 Table 5.18: (continued) Photon energy (eV) Branching ratio (%) CF3C1+ C F 2 C 1 + CF$ CFC1+ C F j CC1+ C1+ CF+ F+ C+ C F 2 C 1 2 + 28.0 0.14 35.95 38.58 0.38 11.96 0.08 8.89 4.01 28.5 0.30 36.34 37.84 0.41 11.83 0.17 8.66 4.45 29.0 0.21 37.20 36.60 0.32 11.74 0.22 8.91 4.81 29.5 0.20 37.09 35.64 0.30 11.79 0.33 9.42 5.32 30.0 0.16 37.76 34.75 0.36 11.84 0.39 9.48 5.26 31.0 0.19 38.81 33.32 0.35 12.02 0.70 9.24 5.38 32.0 0.12 39.60 31.40 0.23 12.12 0.87 9.85 ' 5.82 33.0 0.06 40.77 29.49 0.31 12.13 0.97 10.37 5.83 0.08 34.0 0.05 40.62 27.55 0.28 12.97 1.35 10.63 5.90 0.20 0.46 35.0 0.06 40.93 25.77 0.31 13.47 1.43 11.15 5.94 0.22 0.74 36.0 0.07 40.70 24.32 0.40 13.81 1.55 11.83 5.84 0.36 1.13 37.0 0.00 39.36 23.42 0.33 14.43 1.58 12.70 5.91 0.63 1.63 38.0 0.00 38.98 21.98 0.35 14.79 1.76 13.17 5.87 0.81 2.10 0.17 39.0 0.00 38.21 21.38 0.38 15.39 1.60 13.82 5.63 0.96 2.38 0.25 40.0 0.00 37.02 21.10 0.34 15.86 1.45 14.75 5.45 1.10 2.62 0.31 41.0 0.00 36.61 20.65 0.35 15.93 1.33 15.66 5.30 1.22 2.64 0.30 42.0 0.00 36.14 20.78 0.36 16.23 1.22 15.44 5.44 1.41 2.66 0.32 43.0 0.00 36.17 20.66 0.31 16.32 1.17 15.82 5.53 1.40 2.27 0.36 44.0 0.00 35.98 20.88 0.43 16.17 1.03 15.83 5.56 1.48 2.28 0.35 45.0 0.00 35.37 21.06 0.37 16.28 0.91 16.47 5.77 1.40 2.01 0.36 46.0 0.00 35.62 21.45 0.35 16.18 0.88 15.95 5.81 1.45 1.86 0.44 47.0 0.00 35.18 21.15 0.34 15.92 0.96 16.28 6.15 1.54 2.04 0.44 48.0 0.00 34.96 21.44 0.31 15.61 0.95 16.21 6.30 1.68 2.11 0.43 49.0 0.00 34.65 21.25 0.29 15.56 0.99 16.68 6.54 1.67 1.98 0.39 50.0 0.00 34.70 21.10 0.34 15.15 0.97 16.66 6.81 1.84 2.03 0.40 55.0 0.00 31.48 22.11 0.32 15.88 0.87 16.46 7.76 2.22 2.25 0.64 60.0 0.00 30.34 20.63 0.13 15.93 0.93 17.28 8.44 3.01 2.73 0.48 65.0 0.00 29.43 20.15 0.23 15.59 0.97 18.17 8.48 3.41 3.06 0.53 70.0 0.00 29.19 20.02 0.41 15.70 0.94 17.60 8.44 3.79 3.24 0.68 75.0 0.00 28.14 19.27 0.29 15.96 0.97 18.33 8.43 4.58 3.60 0.44 80.0 0.00 27.74 19.07 0.20 14.86 1.04 18.82 8.57 5.37 3.72 0.61 Chapter 5. Ill Table 5.19: Photoion branching ratio for CF2CI2 Photon Branching ratio (%) energy (eV) C F C t f CF 2 C1+ C C L t CFC1+ CF+ CC1+ C1+ C F + F+ C+ C F 2 C 1 2 + 11.5 100.00 12.0 100.00 12.5 100.00 13.0 100.00 13.5 100.00 14.0 0.28 99.72 14.5 1.11 98.89 15.0 1.73 98.27 15.5 3.79 96.21 16.0 6.81 93.19 16.5 9.40 90.60 17.0 11.12 88.88 17.5 12.01 87.56 0.44 18.0 13.18 85.50 1.31 18.5 12.94 83.67 0.42 2.96 19.0 12.57 80.67 1.09 5.67 19.5 12.04 77.66 1.80 8.49 20.0 11.72 73.84 2.90 10.58 0.39 0.56 20.5 11.12 72.44 3.68 11.19 0.56 1.02 21.0 10.85 71.23 4.19 11.52 0.93 1.28 21.5 11.18 69.53 4.60 12.18 0.95 1.56 22.0 11.00 67.56 5.16 12.52 1.33 2.42 22.5 10.69 66.37 5.35 12.71 1.55 3.33 23.0 10.47 65.13 5.63 13.44 1.39 3.93 23.5 10.42 63.31 6.01 13.57 1.64 5.05 24.0 10.43 61.58 6.15 13.60 0.32 2.04 5.87 24.5 10.31 60.80 6.31 13.66 0.32 1.98 6.62 25.0 10.73 59.84 5.71 13.50 0.56 1.87 7.80 25.5 11.03 59.26 5.45 13.24 0.83 2.25 7.93 continued on next page Chapter 5. 112 Table 5.19: (continued) Photon Branching ratio (%) energy (eV) C F C 1 J CF 2 C1+ CC1+ CFC1+ CF+ CC1+ C1+ CF+ F+ C+ C F 2 C 1 2 + 26.0 11.58 58.29 5.17 13.20 1.13 2.54 8.09 26.5 12.23 57.49 5.10 13.06 1.15 2.88 8.09 27.0 12.31 56.57 4.90 13.23 1.50 3.19 8.29 27.5 12.78 55.43 4.93 12.94 1.72 3.65 8.55 28.0 13.30 54.70 4.64 12.90 1.93 4.16 8.37 28.5 14.05 53.23 4.85 12.63 2.05 4.64 8.55 29.0 14.34 52.07 4.72 13.23 2.43 4.81 8.41 29.5 14.76 51.02 4.57 13.24 2.50 5.35 8.56 30.0 15.29 50.03 4.67 13.15 2.54 5.50 8.81 31.0 16.50 46.94 5.01 13.58 2.99 6.16 8.56 0.26 32.0 16.88 44.91 4.89 13.96 3.21 6.82 8.63 0.69 33.0 17.98 42.47 5.00 14.46 3.45 7.59 8.26 0.79 34.0 18.41 39.49 5.66 14.72 3.28 8.65 8.48 1.31 35.0 18.74 37.26 5.78 15.01 3.30 9.83 8.54 1.54 36.0 18.67 35.29 5.75 15.10 3.36 10.93 8.46 0.60 1.83 37.0 18.46 33.15 6.41 15.20 3.21 12.21 8.73 0.57 2.07 38.0 18.66 31.59 6.51 15.16 2.98 13.09 8.73 0.69 2.40 0.19 39.0 18.80 30.77 6.55 15.36 3.06 13.83 8.60 0.61 2.23 0.18 40.0 18.13 29.61 7.26 15.40 2.64 14.65 9.04 0.78 2.11 0.38 42.0 17.83 28.64 7.37 15.25 2.62 15.14 10.08 0.90 1.75 0.42 44.0 17.05 27.67 7.44 15.40 2.78 15.87 10.80 0.89 1.64 0.46 46.0 16.46 27.70 0.09 7.31 15.00 2.81 16.20 11.35 0.94 1.64 0.49 48.0 16.16 27.62 0.00 7.22 15.07 3.02 16.37 11.47 1.05 1.43 0.59 50.0 15.73 27.82 0.10 7.12 15.18 3.29 16.39 11.15 1.17 1.55 0.51 55.0 14.85 26.80 0.08 6.27 14.80 3.58 18.04 11.33 1.39 2.26 0.60 60.0 13.39 26.70 0.11 6.53 14.74 3.62 19.75 11.00 1.65 2.17 0.33 65.0 13.20 25.54 0.15 6.18 14.69 3.56 20.60 10.70 2.13 2.80 0.47 70.0 13.10 25.13 0.17 6.14 14.13 3.75 20.59 10.65 2.67 2.88 0.79 Chapter 5. 113 Table 5.20: Photoion branching ratio for C F C 1 3 Photon Branching ratio (%) energy (eV) CC1+ C F C L J CC1+ CFC1+ CC1+ C1+ C F + F+ C+ 11.5 100.00 12.0 100.00 12.5 100.00 13.0 100.00 13.5 100.00 14.0 0.53 99.47 14.5 1.19 98.81 15.0 1.61 98.39 15.5 2.24 97.76 16.0 2.54 97.46 16.5 2.90 96.51 0.59 17.0 3.15 95.63 1.22 17.5 2.32 93.69 0.46 3.54 18.0 1.86 89.98 1.12 6.55 0.49 18.5 2.02 84.87 2.14 10.17 0.23 0.57 19.0 1.72 81.16 2.97 12.57 0.36 1.22 19.5 1.91 79.11 3.43 13.38 0.40 1.75 20.0 1.60 77.30 3.74 13.99 0.58 2.79 20.5 1.63 76.81 3.85 13.68 0.57 3.46 21.0 1.43 75.34 4.23 13.97 0.28 0.53 4.22 21.5 1.66 74.23 4.22 13.72 0.88 0.45 4.85 22.0 1.63 73.38 4.35 13.63 1.36 0.73 4.92 22.5 1.55 72.36 4.70 13.05 1.93 0.80 5.61 23.0 1.56 70.43 4.63 13.11 3.14 0.87 6.26 23.5 1.64 69.77 4.02 12.96 3.83 1.24 6.54 continued on next page Chapter 5. 114 Table 5.20: (continued) Photon Branching ratio (%) energy (eV) CC1+ C F C 1 J CC1+ CFC1+ CC1+ C1+ C F + F+ C+ 24.0 1.55 69.50 4.34 11.95 4.65 1.48 6.52 24.5 1.75 68.22 4.10 11.83 5.61 1.89 6.60 25.0 1.67 67.53 3.99 11.59 6.21 2.74 6.29 25.5 1.71 66.25 3.80 11.69 6.71 3.35 6.49 26.0 1.64 66.24 3.61 11.75 7.08 3.78 5.89 26.5 1.73 64.51 4.04 11.16 8.15 4.54 5.88 27.0 1.77 63.68 3.63 11.58 8.32 5.06 5.95 27.5 1.94 62.15 4.26 11.72 8.32 5.38 6.24 28.0 1.91 60.72 4.31 12.10 8.88 5.73 6.36 28.5 2.08 60.27 4.11 12.47 8.89 6.24 5.96 29.0 1.94 59.00 4.35 12.53 8.93 6.40 6.39 0.44 29.5 2.06 57.76 4.75 13.58 8.78 6.53 6.00 0.54 30.0 2.20 56.92 4.73 12.97 8.65 7.56 6.19 0.78 31.0 2.09 49.82 4.91 14.74 9.32 10.98 6.65 1.49 32.0 2.34 47.44 5.28 15.44 9.09 12.37 6.20 1,84 33.0 2.20 44.84 5.42 15.59 9.16 13.67 7.03 2.09 34.0 2.00 43.00 5.72 16.25 9.11 14.66 7.12 2.14 35.0 2.27 41.75 5.67 16.50 9.10 15.86 7.03 1.83 36.0 2.13 38.12 5.73 17.84 8.49 17.92 7.49 2.28 37.0 2.15 35.56 6.93 18.37 8.42 18.67 7.89 2.00 38.0 2.18 34.26 5.98 19.06 8.72 19.85 7.91 0.04 2.00 39.0 2.15 34.07 6.82 18.22 8.89 19.85 7.81 0.21 1.97 40.0 2.05 33.22 6.75 17.29 8.66 20.83 9.16 0.20 1.84 41.0 2.18 31.79 6.90 18.29 9.02 20.89 8.88 0.24 1.82 43.0 2.23 30.21 6.64 18.39 9.71 21.37 9.51 0.27 1.67 45.0 2.01 29.68 6.54 17.55 9.96 22.48 9.72 0.29 1.77 47.0 2.00 29.41 6.63 17.40 10.27 22.72 9.59 0.35 1.63 49.0 1.85 29.87 6.34 16.67 10.32 23.33 9.61 0.30 1.71 Chapter 5. 115 and decrease i n the order of C F 4 > C F 3 C 1 > C F 2 C 1 2 > C F C 1 3 . In the dipole (e, e+ion) experiments the ratio of the total coincident ion signal to the forward scattered energy loss signal at each energy loss gives the relative photoion-izat ion efficiency as defined i n section 2.3.5. Our data show 20 eV , 23.5 eV, 17.5 eV and 14.5 eV to be the lowest photon energies at which the photoionization efficiencies (rji) for C F 4 , C F 3 C 1 , CF2CI2 and C F C 1 3 reach approximately respective constant val-ues. M a k i n g the reasonable assumption that the photoionization efficiency is unity at high energy [8,20,114], we therefore obtain the result that 77, for C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 reach 1.0 respectively at ~20 eV, 23.5 eV, 17.5 e V and 14.5 eV. The pho-toionization efficiency curves for C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 are shown as inserts to respective figs. 5.11a, 5.12a, 5.13a and 5.14a, and values of rji are hsted i n respective tables 5.12, 5.13, 5.14 and 5.15. Compared to previously reported work [103], our C F 4 ion-ization efficiency curve is higher and has less structure (insert to fig. 5.11a), probably due to lower energy resolution. The previously reported photoionization efficiencies [108] of C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 determined using neon resonance lamp radiation at ~16.75 eV are i n generally good agreement wi th the present results (inserts to figs. 5.12a, 5.13a and 5.14a). Absolute differential osciUator strengths for production of the fragment (dissociative) ions are obtained by taking the triple product of the photoabsorption, the photoionization efficiency and the photoion branching ratio at each photon energy. The absolute partial differential oscillator strengths for production of the molecular and fragment ions are shown i n fig. 5.24 and table 5.12 for C F 4 , i n fig. 5.25 and table 5.13 for C F 3 C 1 , i n fig. 5.26 and table 5.14 for C F 2 C 1 2 , and i n fig. 5.27 and table 5.15 for C F C 1 3 . Tables 5.21, 5.22, 5.23 and 5.24 present the appearance potentials ( ± 1 eV) for the production of ion species respectively from C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 measured i n Chanter 5. 116 i > CM I o *— Y A Q r n I— o z Ul cn \\— in cn o o GO o 3 0 < N Z o o I— o X Q_ cn < Q_ 0 5 0 2 0 2 0 2 0 0 . 3 0 \\ 11 X A B C 0 E F CF 4 C F 3 . CF?+ 2 . C F + c + J i l l _ L F + C F ++ 2 H 6 0 3 0 f= 2 0 4 0 6 0 8 0 PHOTON ENERGY (eV) 0 5 0 2 0 2 0 2 0 0 . 3 0 o LU CO I CO CO o O < N O o I— o X Q_ < Q_ Figure 5.24: Absolute differential oscillator strengths for dissociative photoionization of CF4. The vertical arrows represent expected thermodynamic appearance potentials (see table 5.21) and the hatched lines indicate the vertical ionization energies [99,98] for production of the electronic states of C F j . Chapter 5. 117 > CD ° 0 . 3 C F 3 C 1 0 20 10 o z LiJ C£. I— to O r -< 5 0 to o X A D EF G H ! f Q Z o t— < M Z o o I — o X 3 0 2 0 10 0 0 . 3 0 5 0 \\ :\\ C F 3 C I + C F 2 C I + C F C I + V C F . + 2 0 4 0 -1 1— 6 0 § e X A D E F G H T J CCI + — v ' • • • • • J f *» • • • C l + \"1 1 1 1 1 Ill y • • . C F + _ • • • 1\" 1 1 1 1 III II 1 * • 1 1*\" 1 1 1 1 1 C + . . . 11 11 • • • . ~ 1 1\" 1 1 1 C F 2 C I + + • • • • • < Q_ 8 0 2 0 4 0 PHOTON ENERGY (eV) 6 0 8 0 0 . 5 0 3 0 2 0 0 . 5 0 0 . 5 0 0.1 0 Figure 5.25: Absolute differential oscillator strengths for molecular and dissociative photoionization of CF3CI. The vertical arrows represent expected thermodynamic appearance potentials (see table 5.22) and the hatched lines indicate the vertical ionization energies presented in chapter 6 and reported in ref. [100] for production of the electronic states of CFsCl\"1\". Chapter 5. 118 > X I— o 10 U J o QT w ^ 60 or o £ 40 % 20 o < 0 zO.04 OC UJ U_ U_ o h -< M Z O o I — o X or < CL 0 2 0 5 0 CFCI2+ • * C F 2 C I 2 — • • • • • • • . - • • • • • • • • i 1 1 CF2CI + \\ • • • • • • \"iiiiidi i l 1 1 l i CCI2+ AC EG J K (L.M) N O • i . • • • • 1 * CFCI + • * • * •»• • • • • • # ^ • • . . 1 i i C F 2 + m • • • ' • • . - . . « / , \\ \\ ii i n i II A C 1 G J K fi 6 CCI+ • * • . • 1 ~ 1 1 1 1 1 11V A' c i + -• • • • / ~\"1 1 1 1 1 1 ,<\\ CF + -11; • « . / • « \" • ( I I I 1 F+ M i l ..«. 1 1 • • • 1 1 • 1 l C+ JI 111 • * . * • 1 H 1 1 1 1 CF2CI++ • • • • • • • • • • 20 40 60 80 20 40 PHOTON ENERGY (eV) 60 0 4 0 5 0 0.5 0 0.5 0 0.1 0 80 Figure 5.26: Absolute differential oscillator strengths for dissociative photoionization of C F 2 C I 2 . The vertical arrows represent expected thermodynamic appearance potentials (see table 5.23) and the ver-tical lines indicate the vertical ionization energies presented in chapter 6 and reported in ref. [100] for production of the electronic states of C ^ C l J . Chapter 5. 119 > I (— o z LxJ cr t— if) cr' O . 2 0 8 0 o < 5 o LIJ t 2 £ 0 < o o 5 o X °- 0 cr < CL m & i 11 i c F c i 3 n i t n 1 rr,+ XABC 6 EF G fi T J V^^-l • /•••••••• c c i 3 + \"T 1 1 i l l * C F C ! 2 + 1 1 1 1 1 1 1,'., ••••• .-1 1 / | ' \" 1 | 1 1 1 1 C F + 1 t ; ' • - T ' \" 1 1 1 1 1 1 1 • • • 1 1 1 1 1 1 F + 1 i t 1 • • • 1 1 1 1 1 1 C F C I + ' r— 1 1 1 1 1 1 1 1 1 ! 1 1 1 c + 1 1 1 r - — , ! , , 3 0 4 0 O.i 0 0. 0 PHOTON ENERGY (eV) Figure 5.27: Absolute differential oscillator strengths for dissociative photoionization of CFCI3. The vertical arrows represent expected thermodynamic appearance potentials (see table 5.24) and the hatched lines indicate the vertical ionization energies presented in chapter 6 and reported in ref. [100] for pro-duction of the electronic states of CFCI3 . Chapter 5. 120 Table 5.21: Calculated and measured appearance potentials for production of charged species from C F 4 Process Appearance potential (eV) calcu- exp eriment al [ref.] l a ted a this work 6 [135] [136] [137] [138] [139] [103] [140] (1) C F + + F 14.74 16 15.9 15.4 16.0 16.2 15.35 15.56 15.52 (2) C F + + F 2 19.32 21 22.45 22.33 20.3 22.2 (3) C F J + 2 F 20.92 (4) C F + + F 2 + F 22.07 (5) C F + + 3 F 23.66 30 22.85 27.32 22.6 (6) C + + 2 F 2 28.23 (7) C + + F 2 + 2 F 29.82 (8) C++4F 31.42 35 29.5 31.5 (9) F + + F 2 + F + C 35.99 35 24.0 36 (10)F++3F+C 37.58 (11) CF1++? 44 44.3 (12) C F 2 + + ? 42.7 \"Using thermochemical data from ref. [132], assuming zero kinetic energy of fragmentation. 6 ± 1 eV. Chapter 5. 121 Table 5.22: Calculated and measured appearance potentials for the production of charged species from C F 3 C 1 Process Appearance potential (eV) Calculated\" Experimental [ref.] This work* [137] [141] [107] [142] [140] [143] [144] [145] [146] [147] (1) CF3C1+ 12.6 12.5 13 12.8 12.45 12.39 12.43 12.91 12.6 (2) CF2C1+ 14.5 15.0 15.5 14.25 16.15 (3) CF++C1 12.7 12.5 12.7 12.95 12.55 12.65 12.57 13.06 (4) CFC1+ 14.5 20.45 (5) CF^+FCl 16.5 (6) CF++F+C1 19.1 19 20 21.0 18.85 18.84 18.85 (7) CC1+ 28 (8) C1++CF3 16.8 (9) C1++CF2+F 20.4 19.5 21 21.0 (10) C1++CF+F2 24.1 (11) C1++CF+2F 25.7 (12) Cl++C+F+F2 29.7 (13) C1++C+3F 31.3 (14) CF++F+FC1 19.3 (15) CF++F2+C1 20.2 20.28 (16) CF++2F+C1 21.8 23.5 22.6 25.0 22.00 (17) F++CF2+C1 24.9 (18) F++CF+FC1 27.6 (19) F++CF+F+C1 30.1 31 (20) F++C+FC1+F 33.2 (21) F++C+Fa+Cl 34.1 34 (22) F++C+2F+C1 35.7 35.0 (23) C++F3C1 24.3 (24) C++F2-I-FC1 25.4 (25) C++F2+F+C1 28.0 (26) C++3F+C1 29.6 33 31 (27) CF2C12++?? 38 (28) CFC1 2 ++?? 40 \"Using thermochemical data from refs. [133,132,134] assuming zero * ± 1 eV. kinetic energy of fragmentation. Chapter 5. 122 Table 5.23: Calculated and measured appearance potentials for the production of charged species from C F 2 C i 2 Process Appearance Potential (eV) Calculated\" Experimental [ref.] \"This\" [107] [142] [143] [ I i i ] [146] [ U 7 ] -work 6 (1) C F 2 C 1 2 + 11.8 11.75 11.75 12.31 11.87 (2) C F C 1 2 + 14.0 14.15 13.81 13.30 (3) CF 2 C1+ 11.5 12.10 11.99 12.55 11.96 (4) C C 1 2 + 46 (5) CFC1+ 18.5 17.76 18.60 (6) C F 2 + + C 1 2 14.6 14.90 (7) CF 2 ++2C1 17.1 17.5 17.22 16.98 (8) CC1+ 24 21.60 (9) C 1 + + C F 2 + C 1 18.5 18.76 (10) C1++CF+FC1 21.2 20 (11) C1++CF+F+C1 23.8 (12) C1++C+FC1+F 26.8 (13) C l + + C - r - F 2 + C l 27.7 (14) C1++C+2F+C1 29.3 (15) CF++FC1+C1 17.3 (16) C F + + F + C 1 2 17.4 17.65 17.35 (17) CF++F+2C1 19.9 20 20.20 19.84 (18) F + + C F + C 1 2 25.7 (19) F++CF+2C1 28.2 (20) F++C+FC1+C1 31.2 (21) F + + C + F + C 1 2 31.3 (22) F + + C + F + 2 C 1 33.8 36 (23) C++2FC1 20.5 (24) C + + F 2 + C 1 2 23.5 (25) C++FC1+F+C1 25.1 (26) C++2F+C1 2 25.1 (27) C + + F 2 - r 2 C l 26.0 (28) C++2F+2C1 27.6 31 (29) C F 2 C 1 2 + + ? ? 38 \" U s i n g thermochemical da ta f r o m refs. [133,132,134] assuming zero kinetic energy of fragmentation. b ± 1 e V . Chapter 5. 123 Table 5.24: Calculated and measured appearance potentials for the production of charged species from CFCI3 Process Appearance potential (eV) calculated\" Experimental [ref.] \"this' [107] j l i i j [148] [147] [146] [141] work 6 (1) C F C L t 1L8 11.46 (2) C C L t 14 13.50 13.25 12.77 13.8 (3) C F C 1 J 11.5 11.65 11.57 11.97 (4) C C t f 17.5 17.0 17.12 (5) CFC1+ 16.5 16.0 16.02 17.41 16.95 (6) CC1+ 21 20.5 20.00 (7) C1++CF+C1 2 19.6 18.5 (8) C1++CF+2C1 22.0 (9) C1++C+FC1+C1 25.0 (10) C l + + C - r - F + C l 2 25.1 (11) C1++C+F+2C1 27.6 (12) CF++C1+C1 2 15.7 15.7 (13) CF++3C1 18.2 18 18.35 18.10 (14) F + + C C 1 3 21.6 (15) F + + C + C 1 + C 1 2 29.6 (16) F++C+3C1 32.0 38 (17) C + + F C 1 + C 1 2 20.8 (18) C++FC1+2C1 23.3 (19) C + + F + C 1 + C 1 2 23.4 (20) C++F+3C1 25.9 29 \"Using thermochemical data from refs. [133,132,134] assuming zero kinetic energy of fragmentation. b± 1 eV. Chapter 5. 124 the present work and appearance potentials for the various processes calculated from ther-modynamic data [132,133,134], assuming zero kinetic energy of fragmentation. Previously reported values are also shown i n tables 5.21, 5.22, 5.23 and 5.24 for comparison. The calculated thresholds are denoted by arrows on figs. 5.20, 5.24, 5.21, 5.25, 5.22, 5.26, 5.23 and 5.27, together w i t h the vertical ionization energies for production of the electronic ion states of respective molecules (see section 5.1). 4 5.5 Absolute Electronic State Partial Photoionization Differential Oscillator Strengths for C F 4 In this section the electronic state partial differential oscillator strengths for C F 4 have been re-derived based on the earlier reported results and the new information available from the present investigations. Those for other freon molecules w i l l be reported i n chapter 6. Electronic state part ial photoionization cross sections have been reported for C F 4 [105] for production of the X, A, B, C and D ion states up to 70 e V photon energy. Yates et al. [149] have reported P E S branching ratio measurements for C F 4 consistent wi th the cross-section data of Carlson et al. [105]. However, as discussed i n section 5.2.1 above, the reported absolute values of the electronic state partial cross-sections for the X, A, B, C and D states [105] may be incorrect due to the method used for the absolute measurement and/or the effects of higher order radiation (see section 5.2.1). It should be noted that the E ion state has an adiabatic ionization potential of ~34 e V [117], while the appearance potentials of C + and F + are ~35 eV i n the present work (see table 5.21). Therefore, we can tentatively set the sum of the differential oscillator strengths for production of the E and F ion states to be equal to the sum of those for producing C + and F + . W i t h these considerations i n m i n d , we have re-analyzed the P E S data for C F 4 reported by Carlson Chapter 5. 125 et al. [105] as follows: 1. In the energy region below the adiabatic ionization energy of the E state (~34 eV) , photoelectron branching ratios were calculated from the previously reported exper-imental part ial cross-sections [105]. These were then combined wi th the presently reported photoabsorption and photoionization efficiency results (table 5.12) to ob-tain improved estimates of the electronic state partial photoionization differential oscillator strengths of the X, A, B, C and D states of C F 4 . 2. A t higher photon energies up to 70 eV, the originally reported relative intensities for the X, A, B, C and D ion states [105] were used to part i t ion a differential oscillator strength equal to the present total photoabsorption (table 5.12) minus an appropri-ate allowance for the combined (E+F) state differential oscillator strengths. Since there are no direct experimental measurements, the combined (E+F) state differ-ential oscillator strengths have been equated to the sum of the measured differential oscillator strengths for the production of the C + and F + ions (this work) for the reasons described above. These revised electronic state part ial differential oscillator strengths are presented nu-merically i n table 5.25 and are shown together wi th the M S - X a calculations [150] and previous P E S data [105] i n fig. 5.28. The revised values show only small differences from the original P E S data [105] for the B, C and D states. Somewhat larger differences exist for the (<20 and 25-35 eV) and (<23 eV) state part ial cross-sections. In general, the (resolutionless) M S - X a calculations give a reasonable semi-quantitative description of the trends i n the measured partial cross-sections. The presently estimated differential oscillator strengths for the combined (E+F) states are comparable wi th the calculation i n terms of shape and magnitude. It is noteworthy that for the X state, the presently Chapter 5. 126 Table 5.25: Electronic state partial differential oscillator strengths for the photoionization of C F . Photon Electronic state differential oscillat or strength ( 1 0 _ 2 e V _ 1 ) energy (eV) X A B C b E+F 6 15.5 0.81 16.0 3.64 16.5 7.80 17.0 14.95 17.5 22.47 18.0 9.58 20.26 18.5 12.21 22.00 19.0 14.65 22.82 19.5 10.99 25.72 2.97 20.0 11.58 22.82 7.71 21.0 14.56 24.60 9.45 22.0 18.43 21.47 12.46 23.0 16.34 15.62 8.33 11.10 24.0 16.76 12.57 6.84 11.99 25.0 18.31 11.05 6.36 9.38 26.0 17.97 9.92 6.17 6.78 28.0 18.62 8.70 5.93 5.35 0.42 30.0 19.22 8.72 5.33 4.94 0.72 32.0 17.98 8.76 5.83 5.06 0.92 34.0 16.31 10.42 5.73 6.01 1.31 36.0 14.06 10.62 5.59 6.76 2.05 0.36 38.0 12.52 10.61 5.34 7.36 2.31 1.24 40.0 11.52 10.39 5.11 7.33 2.12 2.52 45.0 9.38 8.60 4.16 5.02 1.70 2.51 50.0 7.78 7.37 3.69 3.40 1.85 2.28 55.0 6.44 6.16 3.21 3.69 1.61 2.82 60.0 5.48 5.81 3.20 3.76 1.25 2.96 65.0 4.58 4.58 3.30 3.28 1.28 2.81 70.0 4.21 4.01 2.81 3.11 1.11 2.83 \"Based on branching ratios calculated from previously published PES measurements [105] combined with the present photoabsorption and photoionization efficiency data (table 5.12) and using values of the (E+F) state partial differential oscillator strengths estimated from the partial photoionization differential oscillator strength sum (C ++F +). See text for details. 'The electronic state differential oscillator strength sum (E+F) was set equal to the partial pho-toionization differential oscillator strengths (C ++F +). See text for details. Chapter 5. 127 O i— O UJ CO I CO CO o O < N O O t— o X Q_ 30 20 10 0 30 20 10 0 15 10 5 0 20 15 10 5 0 2 cn < CL 0 0 CF. t S t •C + + F + X J I I 1 — 1 l _ c D i i i i 1 , 1 E + F E 20,? i o 10 o 20^ to ct: O o to O 10 0 10 5 o g 15| 10 t| 5 -0 2 2 O < M 1 § O 0 - 4 15 65 0 o X Q_ CC < 25 35 45 55 PHOTON ENERGY (eV) Figure 5.28: Electronic state partial photoionization differential oscillator strengths of C F 4 . Open circles—previously reported partial photoionization cross—section ( electronic states) measurements obtained by PES [105] (see text). Solid circles—revised values obtained by combining the present photoabsorption and photoionization data with PES branching ratios derived from reference [105]. See section 5.5 for details. Solid lines—MS-Xa calculations [150]. Squares—summed ( C + + F + ) partial differential oscillator strengths. See text for details. Chapter 5. 128 revised P E S cross-sections show a resonance at ~17.5 e V which is predicted by the calculation but not exhibited i n the P E S data as originally presented [105]. 5.6 T h e D i p o l e I n d u c e d B r e a k d o w n In photoionization once the photon energy exceeds the upper l imit of the Franck-Condon region, the internal energy of the molecular ion is independent of the photon energy for a given electronic state and the remainder of the energy is carried by the photoelectron according to Einstein's photoelectric equation. O n the assumption that fragmentation ratios for dissociative ionization from each electronic state of the ion are constant when the photon energy is above the Franck-Condon region [77], the partial differential oscilla-tor strengths for the production of the singly charged molecular or any stable dissociative (fragment) ion should be a fixed hnear combination of electronic state part ia l photoion-ization differential oscillator strengths at al l photon energies. This general approach and possible exceptions such as autoionization, internal conversion to other electronic ion states and multiple ionization have been discussed i n ref. [77]. The dipole induced breakdown patterns of many small molecules have been investigated w i t h considerable success by this type of analysis, for example, see refs. [77,114]. The dipole induced break-down schemes for C F 4 , CF3CI and CF2CI2 discussed below have been obtained using the presently reported absolute differential osciUator strengths for molecular and dissocia-tive photoionization. The electronic state partial photoionization differential oscillator strengths for CF4 are as reported i n section 5.5, while those for CF3CI and CF2CI2 are obtained from the triple product of the presently measured total photoabsorption differ-ential oscillator strengths (section 5.2.2), the photoionization efficiencies (section 5.4) and the electronic state branching ratios obtained from synchrotron radiation P E S measure-ments reported i n section 6.2 and ref. [121,151] (see details below). The dipole induced Chapter 5. 129 breakdown of C F C 1 3 is not reported since insufficient P E S data is available. 5.6.1 The Dipole Induced Breakdown of C F 4 Since stable C F 4 is not observed on the time scale of the T O F mass spectrometer and because CFjj\" is the only ion produced below the state ionization potential (adiabatic IP = 21.7 eV [99])-see table 5.21, the X, A and B electronic states of singly ionized C F 4 must exclusively lead to production of C F 3 . A consideration of the Franck-Condon w i d t h of the C state and the appearance potential of CF^\" (~21 eV) indicates that CF£ can be formed from the C state. Similarly the smaller fragments can only be formed from the D and/or higher states. In the region below ~40 eV, C F + can only be formed from the D state. W i t h these considerations i n mind the following relationships between the differential oscillator strengths for formation of C F 3 , C F j \" , C F + , C + and F + , and those for production of the X, A, B, C, D, E and F ion states are found to provide a reasonably consistent rationalization of the breakdown: ^ ( C F + ) = i | ( X - M + B + 0.62C7) (5.42) § ( C F J ) = ; § ( 0 . 3 8 C ) (5.43) £ ( C + + F+) = 1L{E + F). (5.45) F i g . 5.29 shows the breakdown relationship as a function of photon energy. In scheme I, the originally presented P E S data [105] have been used. The agreement is quite good considering that the P E S data and the ion photofragment differential oscillator strengths are from independent measurements using different techniques w i t h independent means of estabhshing the absolute scales. However, as discussed i n section 5.2.1, we have some reservations concerning the absolute values reported i n the original P E S study [105] and, Chapter 5. 130 Figure 5.29: Absolute differential oscillator strengths for the proposed dipole induced breakdown scheme of C F 4 . Solid circles—present dipole (e, e+ion) experimental data, sohd lines—sums of electronic state partial differential oscillator strengths (scheme I using original PES data [105]; scheme II using revised PES data). See text for details. (a,d) Photofragmentation to CFJ\". (b,e) Photofragmentation to CFJ. (c,f) Photofragmentation to CF+. Chapter 5. 131 PES Dipole Ionization Ground State Molecule States (IP eV) 0 00 Fragmentation (%) (e, e+ion) Ion Products E F 40.3 43.3 1 100 7100 C F + 3 C F+2 C F + C+ + F + Figure 5.30: Proposed dipole induced breakdown scheme for the ionic photoionization of CF4. See text for details. Chapter 5. 132 therefore, have considered a set of revised P E S data as discussed i n section 5.5. The revised electronic state cross-sections are used i n scheme II of fig. 5.29. Scheme II shows a somewhat better overall fit to the ion data. However it should be remembered that localized variations of the overall breakdown pattern might be expected at lower energies due to the autoionization levels preceding the various ionization hmits . Nevertheless the general consistency of the fits for C F 3 , CFJ\" and C F + over the entire energy range up to 70 eV i n both schemes lends confidence to the essential correctness of the breakdown scheme as proposed above. F i g . 5.30 shows i n diagrammatic form the main features of the proposed dipole induced breakdown scheme for C F 4 . Further details of the break-down pattern of C F 4 must await detailed photoelectron-photoion coincidence studies as a function of photon energy. 5.6.2 The Dipole Induced Breakdown of CF 3 C1 Since the molecular ion C F 3 C 1 + and the ion C F 3 are the ions produced from C F 3 C 1 (see table 5.22) below the A state ionization potential [100], the X state must lead to production of C F 3 C 1 + and C F 3 . Considering the Franck-Condon region of the A state [100] and the appearance potentials of C F 2 C 1 + and C F C 1 + , it appears that the ions C F 2 C 1 + and C F C 1 + are the fragments produced from the dipole induced breakdown of the A state. B y taking account of the present energy resolution (1 e V F W H M ) , the Franck-Condon width and the ionization energy of the D state [100] and the appearance potential of C F j , it is suggested that the C F j ion is formed from the D electronic state of the molecular ion. Similarly, the ions C l + , C F + and C C 1 + can be formed from the F, G and H states (ref. [100] and chapter 6) respectively and the ions F + and C + from the / state (chapter 6). W i t h these considerations i n mind the following relationships between the part ial differential osciUator strengths for formation of C F 3 C 1 + , C F 2 C 1 + , C F 3 , Chapter 5. 133 > a, Figure 5.31: Absolute differential oscillator strengths for the proposed dipole induced breakdown scheme of CF3CI. Solid circles—present dipole (e,e+ion) experimental data. Solid hnes—sums of the electronic state partial differential oscillator strengths obtained using the presently determined photoab-sorption differential oscillator strengths and PES branching ratio data in the 21-41 eV [121] and 41-80 eV (chapter 6) regions. See text section 5.6.2 for details. Crosses—(d//djT)[0.5(/+/)] obtained the same way as the solid hnes using PES data presented in chapter 6. Chapters. 134 Ground State Molecule PES Dipole Ionization Figure 5.32: Proposed dipole induced breakdown scheme for the ionic photoionization of CF3CI. See text for details. Chapter 5. 135 CFC1+, C F + , C C 1 + , C l + , C F + , F+ and C + , and those for production of the X, A, B, C, D, E, F, CT, H and (I+J) electronic states are found to provide a reasonably consistent rationahzation of the dipole induced breakdown of C F 3 C 1 with in the energy range of the present data: £ ( C F 3 C 1 + ) = ^(OMX) (5.46) ^LCF 2 C 1 + ) = ^ ( 0 . 9 5 i + B + 0.6C) (5.47) dE diLi d E { C ¥ t ) = i E ( 0 ' 9 9 * + ° A d + ° A t > ) ( 5 - 4 8 ) £ ( C F C 1 + ) = £ ( 0 . 0 5 i ) (5.49) ^ (CF+) = £ ( 0 . 6 7 J + E) (5.50) IE = 7 E ^ (5.51) S(C1+> = 1E{^ (5-52» i < C F + > = I E ^ (5.53) £ ( F + ) = £ [ 0 . 5 ( 7 + i)[ (5.54) £ ( C + ) = £ [ ( 0 . 5 ( 7 + j)]. (5.55) F i g . 5.31 shows these breakdown relationships as a function of photon energy. The elec-tronic state part ia l differential osciUator strengths used i n fig. 5.31 were obtained from the triple product of the presently measured total photoabsorption differential oscillator strengths (section 5.2.2), the photoionization efficiencies (section 5.4) and the electronic state branching ratios obtained from P E S measurements i n the 21-41 eV [121] and 4 1 -80 e V (chapter 6) regions. The measurement i n chapter 6 includes the inner-valence photoelectron bands. Since only the sum of the branching ratios for the I and J states could be reported (chapter 6), the partial differential oscillator strengths for the produc-t ion of the I and J states are represented i n the form of a sum i n the above breakdown Chapter 5. 136 relationships. A s can be seen from fig. 5.31 the relationships are reasonably successful i n reproducing both the shapes and magnitudes of the photoion part ial differential oscilla-tor strengths. F i g . 5.32 shows the presently proposed overall dipole induced breakdown scheme for C F 3 C 1 . A more detailed investigation of the breakdown patterns of C F 3 C 1 must await photoelectron-photoion coincidence studies as a function of photon energy. 5.6.3 T h e D i p o l e I n d u c e d B r e a k d o w n of C F 2 C 1 2 Since the molecular ion C F 2 C l J is not observed on the time scale of the T O F mass molecular ion. Similarly the fragment ions C C l J , C F j , C C 1 + , C l + , F + and C + can be formed from the (N+O), G, K, J, N and (L+M) electronic states (ref. [100] and chapter 6) respectively, and the ions C F C 1 + and C F + from the (H+I) states [100]. W i t h these consideration i n m i n d the following relationships between the part ia l differential oscillator strengths for the formation of C F C l J , C F 2 C 1 + , CCTj\", CFC1+, C F + , C C 1 + , C l + , C F + , F + and C + , and those for the production of the (X+A), (B+C), D, (E+F+G), (H+I), J, K, (L+M) and (N+O) electronic states are found to provide a reasonably consistent rationalization of the dipole induced breakdown of C F 2 C 1 2 wi th in the energy range of the present data: spectrometer and because C F 2 C 1 + is the only ion detected (see table 5.23) below the D state ionization potential of C F 2 C 1 2 [100], the X, A , B and C electronic states of the molecular ion must exclusively lead to production of the C F 2 C 1 + ion. Considering the Franck-Condon region of the D state [100] and the appearance potential of C F C l ^ , it appears that C F C l J is the fragment ion produced from the D electronic state of the df_ dE £. dE (5.56) (5.57) Chapter 5. 137 ^ ( C C l + ) = ^ [ 0 . 0 1 5 ( i V + O)] J ^ ( C F C l + ) = -^[0.38(77 + 7~)] %(CF+) = %[0.38(E + F + G)} ^ ( F + ) = ^ [ 0 . 2 ( A > + O)] ^ ( C l + ) = ^[J + 0.647? + 0.485(iV + 6)} 7 F ( C F + ) = T F [ 0 - 6 2 ( ^ + /) + 0.2(iV + 6)] (5.58) (5.59) (5.64) (5.61) (5.63) (5.62) (5.60) i^(C+) = M(L + M) + 0.1(N + d)]. (5.65) F i g . 5.33 shows these breakdown relationships as a function of photon energy. The elec-tronic state part ial differential oscillator strengths used i n fig. 5.33 were obtained from the products of the presently measured total photoabsorption differential osciUator strengths (section 5.2.3) and the electronic state branching ratios obtained from P E S measure-ments i n the 27-41 e V [151] and 41-70 e V ( chapter 6) regions. The measurements i n chapter 6 includes the inner-valence photoelectron bands. Due to the unresolvabihty of certain features i n the photoelectron spectra reported i n chapter 6 and i n ref. [151], the branching ratios for the production of the relevant states were reported as sums, i.e. (X+A), (B+C), (E+F+G), (77+7), (L+M) and (N+O), therefore the corresponding part ia l differential oscillator strengths are presented i n similar form i n the above break-down relationships. In spite of the uncertainties involved i n such a simple rationale the correspondence of the proposed scheme wi th the partial photoionization differential os-cillator strengths (fig. 5.33) is reasonably good i n terms of both shape and magnitude. F i g . 5.34 shows the presently proposed dipole induced breakdown scheme for CF2CI2. A Chapter 5. 138 more detailed analysis of the dipole induced breakdown scheme of C F 2 C 1 2 may be ob-tained from photoelectron-photoion coincidence studies as a function of photon energy. Chapter 5. 139 V K W (— 0-w g 50r y*0.15(E+F+G c F . c r U 25-W O £0.03r u gjo.oo-I 4 o * 2^ o 5 <* a 10-z o g * o a. oL < < CU 47(E+F+G) o 0.015(N+0) C C L CFC1 0.38(H+T) C F 2 + / , 0.38(E+F+G) o 5 0 5 0 10.5 po 1 20 40 60 c c r 0.36K 0 . 6 4 K V v o J+0 .64K+0 .485(N+0^ + o 0.62(H-fI)+0.2(N+0) C F 0.62(H+I) o 0.2(N+0) F + ••••• o (L+M)+0.1(N+0) Q * , (L+M) 20 40 60 P H O T O N E N E R G Y ( e V ) Figure 5.33: Absolute differential oscillator strengths for the proposed dipole induced breakdown scheme of C F 2 C 1 2 . Solid circles—present dipole (e,e-)-ion) experimental data. Sohd lines—sums of electronic state partial differential oscillator strengths using the presently determined photoabsorption differential oscillator strengths and P E S branching ratio data in the 27-41 eV [151] and 41-70 eV (chapter 6) regions. See section 5.6.3 for details. Chapter 5. 140 Ground State Molecule CFaClg States (IPeV) PES Dipole Ionization X+A 12.26.12.53 13.11.13.45 D 14.36 E+F+S 15.9,16.30.16.9 H+l 19.3 J 20.4 K 22.4 L+M 27.2 N+5 38.6.41.4 Figure 5.34: Proposed dipole induced breakdown scheme for the ionic photoionization of CF 2C1 2 . See text for details. Chapter 6 Photoelectron Spectroscopy and the Electronic State Partial Differential Oscillator Strengths of the Freon Molecules CF 3 C1, CF 2 C1 2 and CFC1 3 Using Synchrotron Radiation from 41 to 160 eV The presently reported P E S spectra for C F 3 C 1 , C F 2 C 1 2 and CFCI3 were measured by D r . G . Cooper at the Canadian Synchrotron Radiat ion Facility ( C S R F ) located at the A L A D D I N facility at The University of Wisconsin [152]. The measurement method and apparatus have been described previously [121,153,154,155,156]. Briefly, a 1200 l i n e / m m holographic grating i n a Grasshopper Monochromator is used to monochromate syn-chrotron radiation from the A l a d d i n storage ring in Stoughton, Wisconsin. The grating has useful output at photon energies above 40 eV. The sample is photoionized i n a free gas jet and the photoelectrons are energy analyzed at the pseudo magic angle using a Leybold Heraeus LHS-11 hemispherical photoelectron spectrometer. The spectrometer is isolated from the optical elements of the beam hne by two stages of differential pumping. The overall resolution of the monochromator/spectrometer apparatus in the present ex-periments was i n the range 0.2-0.3 eV F W H M , depending upon the photon energy (since A A oc AE/E2 is constant, where E is photon energy) and the pass energy (typically 12.5 or 25 eV) used i n the electron analyzer. The transmission efficiency of the photoelectron spectrometer has been found to be effectively constant over the range of photoelectron kinetic energy used i n the present work. 141 Chapter 6. 142 6.1 Photoelectron Spectra Photoelectron spectra of CFCI3 and CF2CI2 obtained with synchrotron radiation at 80 eV energy are shown i n figs. 6.35a and 6.35b respectively, while the spectrum of CFCI3 at hv = 90 e V is shown i n fig. 6.35c. The assignments of the photoelectron bands up to 30 eV ionization energy indicated i n the figures follow those given by Cvitas et al. [100], while the assignments of the higher energy bands are the same as those of Potts et al. [157]. Green's function calculations within the two particle-hole Tam-Dankoff approximation (2ph T D A ) method performed by Cambi et al. [158] indicate that the one electron description of ionization is not adequate to describe ionization of the inner-valence orbitals of freon molecules (the orbitals wi th predominantly C l 3s and F 2s character), and that each inner-valence orbital wi l l give rise to a number of many-body photoelectron states. Therefore the molecular orbital labels for the bands > 25 eV ionization energy i n fig. 6.35 are an approximation and give only the major one electron configuration involved i n the ion states. Unlike Potts et al. [157] we do not observe clear separations between the F 2s lb^1 and l a ] \" 1 states in CF2CI2 or the F 2s le~* and la ] \" 1 states i n C F 3 C 1 . The reason for this is unclear since the electron energy resolution of the presently reported measurements should be adequate to observe the splittings seen by the authors of ref. [157]. The estimated vertical ionization energies for the inner valence states obtained i n the present work are given in table 6.26. The features i n the CFCI3 and CF2CI2 spectra (see fig. 6.35) at apparent ionization energies of 46.8 and 48.7 e V (CFC1 3 ) and 47.1 and 48.8 e V ( C F 2 C 1 2 ) are i n fact due to C l 2j»3^2 and 2p\\~j2 states respectively produced by third order radiation from the monochromator. The kinetic energies of these peaks match those expected on the basis of the reported ionization energies of the C l 2p~x states of CFCI3 (207.20 eV (2^3/2), Chapter 6. 143 10 5-CO UJ 0 10 L U 5 -> h-< as 5\" 0 3e CFCI: (2e,4a,) (46,50,) \\ ' (lo2,5e)N>;; i , r h„ = l20eV la, ' ' I L j hv=80 eV 3a. . . : -:: .: • A le2a' v!°i A . Cl 2p-'{3rd Order) 3/2 '/2 rr i 1 r (5a,,la2,3b,) .*' (2b,,4a,) (202,6a,) .5 ( 4 b 2 , 4 b , ) • .\\2b, : » ^ 3a, T h!/ = 80eV CF2Ch Cl 2p-' (3 r d Order) (2a,,lb2) lb, la, 3/2 '* r r i r 3e - I 1 hv=90eV 5ai la24e' m • f CR.CI 2e j 5e • t '4a\\U le lai 2a, T 10 2 0 3 0 4 0 5 0 BINDING ENERGY (eV) Figure 6.35: Photoelectron spectra of: a) CFC13 at hu = 80 eV (Insert shows hu = 120 eV); b) CF2C12 at hu = 80 eV; c) CF3CI at hu =90 eV. The assignments shown on the spectra follow those given in refs. [100] and [157] (see section 6.1 for details). The peaks above 45 eV in a) and b) are spin-orbit doublets from Cl 2p ionization by third order radiation (see text). 1 Chapter 6. 144 Table 6.26: Vert ical ionization energies for the inner valence regions of C F 3 C 1 , CF2CI2 and CFCI3 : Molecule Ionization ener gy (eV) this work\" ref. [157] C F 3 C I 2a\\ le l a i 2ai le l a x 26.9 40.0 42.5 26.3 40.0b 42.8 6 C F 2 C 1 2 (2oi,l6 2) I61 lax (201,163) l&i l a i 27.0 38.6 41.4 — 39 41 CFCI3 le 2ai lax le 2ai l a i 25.3 27.6 40.0 25.5 27.6 39.5 A ± 1.0 e y. 6 T h e ionization energies for the l e and l a i orbitals of CF3CI given i n the text of ref. [157] are 42.8 a n d 44 e V respectively, however the spec t rum shown in fig. 2c of ref. [157] is not consistent wi th these values. T h e numbers given i n the table therefore have been estimated f r o m fig. 2c of Potts et al. [157]. Chapter 6. 145 208.81 eV (2p-/ 2) [159]) and C F 2 C 1 2 (207.47 eV (2p-/ 2), 209.10 e V {2p~f2) [159]) for hv = 240 eV. The assignment as spin-orbit doublets is supported by the relative intensities and spacing of the two peaks i n each case. Further support is gained from the fact that no such spin-orbit doublets appear within the same spectral range of the (first order) spectra at other photon energies (for example see the insert to fig. 6.35a which shows the spectrum of CFCI3 at hv = 120 eV) . F i g . 6.35c shows the spectrum of CF3CI at hv = 90 eV. 6.2 Photoelectron Branching Ratios and Partial Photoionization Differential Oscillator Strengths The photoelectron spectra were analyzed by least squares fitting the data to Gaussian peak shapes plus a hnear background function. For some molecular orbitals individual photoelectron branching ratios could not be obtained since the photoelectron bands were too heavily overlapping each other at the resolution used in the present experiments. In these cases branching ratios for groupings of peaks are presented below. Photoelec-tron branching ratios of C F 3 C 1 , C F 2 C 1 2 and CFCI3 are presented i n tables 6.27-6.29 and i n figs. 6.36-6.38 respectively. Where the separate contributions from overlapping peaks could not be adequately determined from the curve f i tt ing procedure, a combined branching ratio is presented. In fig. 6.36 are also plotted the branching ratios reported by Bozek et al. [153] for C F 3 C 1 from 41 to 70 eV. Above 70 e V the branching ratios of the F 2s l e and lai orbitals begin to become significant so that comparison of the presently reported results wi th those of ref. [153] is no longer appropriate since Bozek et al. [153] d id not make measurements i n the inner valence region. Novak et al. [151] did not directly report their measured photoelectron branching ratios for C F 3 C 1 and C F 2 C 1 2 , instead they quoted part ial photoionization cross-sections which they derived from the Chapter 6. 146 Table 6.27: Photoelectron branching ratios of C F 3 C 1 Photon Branching Ratio (%) Energy (eV) 5e 5a x l a 2 4e 3e 4a x 2e 3a x 2a x ( l e + 1 41 4.4 10.4 14.1 31.2 21.1 4.0 13.2 1.6 45 4.0 10.7 12.0 27.1 21.0 4.6 15.7 4.7 0.3 50 3.9 10.1 9.7 26.9 19.3 5.6 17.1 7.2 0.4 55 4.2 10.1 8.9 24.1 18.2 5.8 17.2 9.5 1.9 60 5.4 11.3 6.9 25.3 18.0 6.5 17.0 8.5 1.0 65 5.9 9.1 { 52.4 } 8.3 15.8 7.3 1.4 70 6.5 8.8 8.2 24.1 21.8 6.0 16.7 6.4 1.4 75 7.4 8.3 7.6 23.3 21.0 5.4 16.6 5.7 0.8 3.8 80 6.5 7.9 6.3 21.9 19.0 5.5 16.0 5.6 0.8 10.5 85 7.0 7.9 4.6 26.0 16.0 5.6 15.0 5.0 0.4 12.5 90 7.5 5.2 9.4 17.0 22.0 5.1 14.4 4.5 0.4 14.6 95 7.5 6.7 8.3 18.5 18.7 5.1 13.6 4.6 0.5 16.6 100 7.9 6.2 9.7 18.5 19.9 5.0 14.1 4.4 0.6 13.5 105 7.9 6.6 7.5 21.1 15.9 4.0 12.7 4.8 0.4 19.0 110 9.1 8.9 6.9 21.1 15.1 5.6 12.5 5.2 0.5 15.2 115 8.6 7.4 8.9 19.2 16.9 4.4 12.4 5.1 1.0 16.2 120 8.8 8.3 8.8 21.0 17.1 4.5 13.2 4.8 0.9 12.7 130 9.4 8.5 8.6 20.8 16.8 4.8 12.1 4.8 0.4 13.8 140 9.5 8.0 8.5 18.0 15.1 4.5 11.5 4.4 1.0 19.8 150 10.5 8.0 8.1 16.6 15.2 5.3 11.8 5.6 1.1 17.9 160 10.1 7.9 8.1 16.5 17.2 4.6 13.0 6.1 1.4 15.0 Chapter 6. 147 Table 6.28: Photoelectron branching ratios of CF2CI2 Photon Branching Ratio (%) Energy (eV) (4fr2+4fri) (2a1+6ai) Zb2 (5ai+la2+3&i) (26!+4ai) 2b2 3ai (2ai+l&2) (l&i+lai) 41 3.9 5.4 9.4 46.6 13.8 12.9 8.0 45 3.3 6.5 9.1 44.1 14.4 14.2 8.6 50 4.2 7.5 8.5 43.8 14.1 12.9 8.9 55 5.5 8.2 9.1 42.0 15.4 11.1 8.5 60 7.5 7.2 8.9 38.8 14.9 11.6 9.3 2.0 65 7.4 8.0 8.4 39.4 16.0 10.0 8.1 2.7 70 7.0 6.9 6.6 33.0 13.8 8.2 9.9 1.6 13.1 75 8.5 7.0 6.0 34.5 15.2 8.2 6.3 1.5 12.9 80 8.7 6.6 6.5 34.8 14.7 9.2 6.3 2.1 11.2 85 8.7 6.9 6.2 32.7 14.4 7.9 6.2 2.2 14.7 90 10.0 8.1 5.2 32.5 15.2 6.5 6.7 3.8 12.1 95 10.6 9.3 6.2 31.5 14.9 7.2 6.0 3.5 10.8 100 11.1 9.6 7.0 32.6 13.5 8.6 5.8 1.4 10.6 105 11.7 9.5 7.4 32.2 13.0 8.1 5.8 2.5 9.7 110 12.3 10.6 6.7 31.9 13.1 7.5 5.4 2.1 10.2 120 12.2 11.1 7.4 30.6 11.1 7.6 5.1 2.6 12.5 130 11.7 10.2 8.4 30.1 11.1 8.5 5.4 4.5 10.2 140 12.8 11.1 8.3 29.2 11.4 7.2 5.5 3.0 11.6 150 12.2 12.9 7.9 27.5 12.1 6.4 6.2 2.8 12.0 160 13.2 9.1 8.8 26.5 12.2 6.5 6.4 2.8 14.4 Chapter 6. 148 Table 6.29: Photoelectron branching ratios of CFCI3 Photon Branching Ratio (%) Energy (eV) ( l a 2 + 5e + 4e + 5a x ) 3e (2e + 4ai) 3ai ( le + 2ai) l a i 41 20.7 29.0 33.8 10.6 5.8 45 20.8 27.9 35.2 10.2 6.0 50 22.8 27.7 31.5 10.7 7.4 55 28.3 27.3 29.8 10.0 4.6 60 28.4 25.5 30.0 10.1 6.0 65 31.0 26.4 30.1 8.7 3.8 70 28.3 24.0 29.7 6.4 6.5 5.2 75 28.9 21.8 28.3 9.1 7.3 4.6 80 32.4 21.4 27.8 7.5 4.4 6.4 90 32.6 18.1 25.2 7.5 7.7 9.0 100 36.2 18.2 23.2 7.6 6.8 8.1 110 39.2 17.9 22.5 7.4 6.4 6.7 115 41.6 17.8 22.4 7.0 5.7 5.5 120 40.6 18.3 21.2 6.7 6.6 6.7 130 41.4 17.4 20.4 7.0 6.4 7.6 140 42.1 17.2 20.7 7.3 7.4 5.4 150 41.8 17.0 21.8 6.7 4.7 7.9 Chapter 6. 149 6^ O rr CD X o < cc CD o rr r— O LU _ J UJ O I— o X CL 4a, • Rf2 8 ° D 8 • . . . . • . CF3CI b e i 1 1 1 1 1 5a, . WcP •••• 0 • . • • a • • a r 1 1 1 1 r— 2e 1 1 1 1 1 1 ° °° ° l a 2 0 0 0 • • » • ' • • . . • . ' • • • * • • 1 1 1 1 1 ' 3a, 0 0 t • 0 0 ^ 0 8 • • • • - ' 4e * \" • . • 3 • • • • • • 0 a • • • -I • 1 r i • 2a, -. . • • • • • • « . « • « « • • - • • • le*la, • • > \" i 0 0 ) W o . . . • oe -. . . • ) -) 60 ' I0O 140 60 100 140 0 0 20 10 0 10 0 10 0 20 0 PHOTON ENERGY (eV) Figure 6.36: Photoelectron branching ratios for CF3C1. Sohd circles-this work; open circles-ref. [151]; open squares-ref. [153]. Chapter 6. 150 20 ^ 10 o cn CD 0 10 X o 0 -z. < o cn \\-o UJ 0 40 LLI 30 O O 20 C L 10 CF2CI2 4b2+4b, -1 r i r -202+ 6ai T r 1 1 1 R 3b2 - i 1 1 1 1 r 5a,+ la2*3b| t o 60 -i 1 — 100 —i r 140 2b+4a, - r 1 r— 1 1 r 2b2 ° o • t ; 20 10 0 10 -I 1 1 r-3a i % • . • • • • • • • • • 2a,+ lb2 10 10 0 10 -l r lb,+ la, 60 loo\" \"wo r 0 20 - 10 0 PHOTON ENERGY (eV) Figure 6.37: Photoelectron branching ratios for CF2C12. Solid circles-this work; open circles-ref. [151]. Chapter 6. 151 2 401 Q=30| CD X 20 O CFCh • • • la2+5e+4e*5a cr co 10 Z 0 ° 30 cn 1 h-o LLJ 201 U J g 101 o X * - 0 -i r 60 3e • • • • • • • • • • • 2e*4a, i 1 1 1—• 3a, i i T r le*2ai la, 30 20 10 0 10 0 10 0 10 100 140 60 100 PHOTON ENERGY (eV) 140 Figure 6.38: Photoelectron branching ratios for CFC1 3. Chapter 6. 152 product of their branching ratios and previously published photoabsorption cross-sections for CF3CI [101] and C F 2 C 1 2 [110]. We have therefore recalculated photoelectron branch-ing ratios from their published tables of cross-sections [151] i n order to facilitate valid comparisons w i t h the present work. The so obtained photoelectron branching ratios of Novak et al. [151] are plotted i n figs. 6.36 and 6.37. Note that unlike i n the present work and that of ref. [153], Novak et al. [151] did not obtain separate branching ratios for the 5 a i , l a 2 and 4e orbitals of C F 3 C 1 and instead presented combined values for these bands. Thus i n these cases their data [151] is not represented in fig. 6.36. F i g . 6.36 shows that i n general there is good agreement between the presently reported photoelectron branching ratios and the two previously pubhshed data sets [151,153] for CF3CI from 41 to 70 eV. The agreement seen i n fig. 6.37 between the present data for C F 2 C 1 2 and that of ref. [151] from 41 to 70 eV is equally good. Some significant and systematic differences between the presently reported results and those of Bozek et al. [153] do exist however, for the 5 a l 5 l a 2 , 4e and 3e orbitals of C F 3 C 1 . Whi le the combined branching ratio sum (5ai+la2-)-4e-r-3e) of ref. [153] is very similar to that of the present work (not plotted i n fig. 6.36), their branching ratios for the 5a x and 4e orbitals are smaller than those presented here, while those for the l a 2 and 3e are larger. These differences are due largely, if not entirely, to aspects of the respective curve fitt ing procedures used. It appears that Bozek et al. [153] fitted the overlapping band system comprised of ionization from the 5 a 1 ; l a 2 , 4e and 3e orbitals with peaks of equal width , which is not necessarily vahd. Therefore i n the present work we have allowed the fitting program to determine individual widths for each peak i n the spectrum. The major difference i n the two procedures is that the l a j 1 band is determined to be substantially narrower and the 4 e _ 1 band wider i n the present work than i n ref. [153]. The He II spectrum of C F 3 C 1 reported by Cvitas et al. [100] clearly shows a narrower Chapter 6. 153 profile of the la2 band and a wider 4 e _ 1 band, supporting the presently employed curve f i t t ing strategy. We therefore believe the presently reported branching ratios for the 5a i , l a 2 , 4e and 3e orbitals of C F 3 C 1 to be more accurate than those of Bozek et al. [153]. The photoelectron branching ratio sum (5a 1+la2+4e+3e) reported by Novak et al. [151] agrees very well w i t h the equivalent summation performed on the presently reported data and that of ref. [153] (not shown in fig. 6.36). Using total photoabsorption differential oscillator strengths for C F 3 C 1 , CF2CI2 and CFCI3 presented i n sections 5.2.2-5.2.4 along with the presently reported photoelectron branching ratios, we have derived part ial photoionization differential oscillator strengths for each of the electronic ion states of the three molecules. These are given numerically i n tables 6.30-6.32 and are shown graphicaUy in figs. 6.39-6.41. Bozek et al. [153] and No-vak et al. [151] used previously published photoabsorption cross-sections for C F 3 C 1 [101] and CF2CI2 [HO] i n order to derive partial photoionization differential oscillator strengths from their electronic state branching ratios. Since their branching ratio data is very sim-ilar to the presently reported values except where noted above, any differences i n the shapes of the electronic state part ial photoionization differential oscillator strengths be-tween the values reported i n refs. [151,153] and the present values shown i n figs. 6.39 and 6.40 from 41 to 70 eV are caused solely by differences i n the total photoabsorp-tion differential oscillator strengths presented i n sections 5.2.2, 5.2.3 and reported i n refs. [101,110]. These differences are discussed i n sections 5.2.2 and 5.2.3. However, Caulet t i et al. [160] used a method of internal cahbration i n order to directly measure part ia l photoionization cross-sections (differential oscillator strengths) at 40.81 eV for the outermost 5e orbitals of C F 3 C 1 and the (4& 2+4&i), (2a 2 +6ai) and 3b2 orbitals of CF2CI2. We therefore compare the present data with that of ref. [160] i n figs. 6.39 and 6.40. The part ia l photoionization differential osciUator strength for the 5e orbitals of C F 3 C 1 at Chapter 6. 154 Table 6.30: Electronic state part ial photoionization differential oscillator strengths for C F 3 C 1 Photon Differential oscillator Strength ( l O ^ e V - 1 ) \" Energy (eV) 5e 5a x l a 2 4e 3e 4 a : 2e 3 a ! 2a x ( l e + l < 41 1.24 2.95 3.99 8.83 5.97 1.12 3.73 0.44 45 0.95 2.53 2.85 6.42 4.96 1.08 3.71 1.10 0.07 50 0.78 2.04 1.95 5.41 3.88 1.12 3.44 1.45 0.08 55 0.76 1.83 1.60 4.35 3.27 1.05 3.10 1.72 0.35 60 0.89 1.86 1.14 4.16 2.96 1.07 2.79 1.39 0.17 65 0.87 1.34 { 7.72 } 1.22 2.32 1.07 0.20 70 0.86 1.17 1.10 3.22 2.90 0.80 2.23 0.86 0.19 75 0.89 0.99 0.91 . 2.79 2.53 0.65 2.00 0.68 0.10 0.46 80 0.72 0.88 0.71 2.44 2.11 0.61 1.78 0.63 0.09 1.17 85 0.72 0.81 0.48 2.69 1.65 0.58 1.55 0.52 0.04 1.29 90 0.67 0.47 0.85 1.54 1.99 0.46 1.30 0.41 0.03 1.32 95 0.62 0.56 0.69 1.54 1.56 0.42 1.13 0.38 0.04 1.38 100 0.64 0.50 0.78 1.49 1.60 0.40 1.13 0.36 0.05 1.09 105 0.57 0.48 0.54 1.52 1.15 0.29 0.92 0.34 0.03 1.37 110 0.60 0.59 0.46 1.39 1.00 0.37 0.83 0.34 0.04 1.01 115 0.53 0.46 0.55 1.19 1.04 0.27 0.77 0.31 0.06 1.00 120 0.50 0.47 0.50 1.19 0.97 0.25 0.75 0.27 0.05 0.72 130 0.48 0.43 0.44 1.06 0.86 0.24 0.61 0.25 0.02 0.70 140 0.41 0.34 0.37 0.78 0.65 0.19 0.50 0.19 0.04 0.86 150 0.41 0.31 0.31 0.64 0.59 0.20 0.45 0.22 0.04 0.69 160 0.36 0.28 0.29 0.59 0.62 0.16 0.47 0.22 0.05 0.54 V(Mb) = 1.0975 x 102(d//a\\E)(eV-1). Chapter 6. 155 Table 6.31: Electronic state part ial photoionization differential oscillator strengths for CF2CI2 ^ Photon Differential oscillator strength (10~2eV *) ° Energy (eV) (4b2+46i) (2ai+6ai) Zb2 (5ai+la2+3&i) (2b1+4a1) 2b2 3ai (2ai+lb2) (l&i+lai) 41 0.84 1.15 2.02 9.99 2.96 2.76 1.70 45 0.63 1.24 1.74 8.48 2.76 2.72 1.64 50 0.71 1.26 1.43 7.33 2.37 2.16 1.49 55 0.81 1.22 1.36 6.27 2.30 1.67 1.27 60 1.03 0.99 1.23 5.36 2.05 1.60 1.29 0.28 65 0.95 1.04 1.09 5.11 2.07 1.30 1.05 0.35 70 0.79 0.77 0.74 3.69 1.54 0.92 1.11 0.18 2.02 75 0.89 0.74 0.63 3.63 1.59 0.86 0.66 0.16 1.36 80 0.86 0.65 0.65 3.46 1.46 0.91 0.63 0.20 1.11 85 0.77 0.61 0.55 2.92 1.29 0.71 0.56 0.20 1.31 90 0.83 0.68 0.44 2.70 1.26 0.54 0.56 0.32 1.00 95 0.82 0.71 0.48 2.43 1.15 0.55 0.46 0.27 0.83 100 0.77 0.66 0.48 2.25 0.93 0.60 0.40 0.10 0.73 105 0.77 0.62 0.49 2.12 0.86 0.53 0.38 0.16 0.64 110 0.76 0.66 0.43 1.98 0.81 0.46 0.34 0.13 0.64 120 0.66 0.60 0.40 1.65 0.60 0.41 0.28 0.14 0.67 130 0.55 0.48 0.39 1.41 0.52 0.40 0.25 0.21 0.48 140 0.52 0.45 0.34 1.19 0.47 0.29 0.22 0.12 0.47 150 0.44 0.46 0.28 0.99 0.43 0.23 0.22 0.10 0.43 160 0.43 0.29 0.29 0.86 0.40 0.21 0.21 0.09 0.47 V(Mb) = 1.0975 x 1 0 2 ( e V - 1 ) . Chapter 6. 156 Table 6.32: Electronic state part ial photoionization differential oscillator strengths for C F C 1 3 Photon Differential oscillator strength (10 2 e V x ) a Energy (eV) ( l a 2 + 5e + 4e + 5a x) 3e (2e + 4ai) 3 CM I r— o z U l CY. t— 00 cr: O 2-2-oo 0 o UJ p CK Q 0 < M Z o o I o CL 0 0.5 < CL 0.5H 0 CFCI la2+5e+4e+5a. — i — 3e T 1 2e + 4a 3at > • • -i 1 . • • • le + 2a, • • • T • • * T r la • • • • 60 80 100 120 140 PHOTON ENERGY (eV) -2 •0 -2 -0 -4 •2 -0 -0 •0.5 0 -0.5 -0 Figure 6.41: Electronic state partial photoionization differential oscillator strengths (cross-sections) of CFC13. Chapter 6. 160 40.81 e V reported by Caulet t i et al. [160] is much lower (~ 50%) than that measured i n the present work (see fig. 6.39), but the results for C F 2 C 1 2 (see fig. 6.40) are i n good agreement. The fact that Caulet t i et al. [160] used an internal standard and i n addition d id not make their measurements at the magic angle, but attempted to correct for this using approximate theoretical values of the photoelectron asymmetry parameter 0, may account for the large discrepancy for the 5e orbitals of CF3CI. M S - X a calculations of the outer valence photoelectron branching ratios and par-t ia l photoionization differential oscillator strengths of CF3CI were reported by Bozek et al. [153]. These proved to be i n generally good agreement wi th their experimental values. The presently reported branching ratios and partial photoionization differential oscillator strengths, especially those for the 5a! , l a 2 and 4e orbitals, further improve the agreement between theory and experiment from 41 to 160 eV. The theoretical results [153] predict weak resonances at high photon energies (>100 eV) i n the part ial photoionization cross-sections of al l the valence orbitals of C F 3 C 1 caused by scattering of the photoelectrons. These resonances should also be visible in the photoelectron branching ratios. W h i l e the presently reported branching ratios and differential oscillator strengths for C F 3 C 1 shown i n figs. 6.36 and 6.39 suggest the possibility of such weak resonances for various orbitals (e.g. the 3ai and 2e), the data do not unequivocally corroborate their existence. The general features of the photoelectron branching ratios and electronic state par-t ia l photoionization differential oscillator strengths for C F 3 C 1 , C F 2 C 1 2 and CFCI3 shown i n figs. 6.36-6.41 may be understood qualitatively i n terms of the atomic orbital char-acters of the molecular orbitals from which the photoelectron bands are derived. Of course, this discussion impl ic i t ly invokes the independent particle and the frozen orbital approximations, and ignores molecular effects such as shape resonances. Such a model was formulated by Gelius [161] in a quantitative way and should increase i n validity as Chapter 6. 161 the photoelectron energy increases. It can be used quantitatively i n X - r a y photoelectron spectroscopy but is only qualitatively useful at the photoelectron energies involved i n the present work. Therefore, we may understand shapes of the molecular orbital branching ratios and part ia l photoionization differential oscillator strengths shown i n figs. 6.36-6.41 by reference to the C 2p, C 2s, F 2p, F 2s, C l 3p and C l 3s atomic orbital differential oscillator strength curves [68]. For example, the \"lone pair\" molecular orbitals that have predominantly C l 3p atomic character (the 5e i n C F 3 C 1 , the 46 2 , 46 1 ; Qa,\\ and 3b2 i n CF2CI2 and the l a 2 , 5e, 4e and ba\\ in CFCI3) increase in branching ratio from 50 to 160 eV. This is because there is a Cooper min imum [25,127] i n the C l 3p atomic or-b i ta l photoionization cross-sections at 35-40 eV [68], so that above this energy they first increase i n intensity then decrease very slowly wi th increasing photon energy. Similarly the branching ratio and part ial differential oscillator strength characteristics of the molec-ular orbitals of essentially fluorine lone pair character (the la2, 4e and 3e i n C F 3 C 1 and the 3 b l and l a 2 i n C F 2 C 1 2 ) , are very much hke those of F 2p atomic orbitals [68]. The molecular orbitals having C - F and C - C l bonding character exhibit differential oscillator strength and branching ratio curves which are generally consistent w i t h mixed C / F and C / C l character. The inner valence molecular orbitals of C F 3 C 1 , C F 2 C 1 2 and CFCI3 (the 3ax orbitals and al l lower ly ing levels) may be expected to show partial photoionization differential oscillator strengths more consistent with atomic-hke behaviour than the outer valence orbitals since they exist closer to the atomic centres and thus w i l l be less influenced by molecular bonding and symmetry effects. Electron correlation w i l l , however, be more important for ionization of the inner valence orbitals [158], leading to several many-body ion states for ionization of a given molecular orbital . The 3ax orbitals of al l three molecules have predominantly C 23 character and exhibit very similar photoionization Chapter 6. 162 differential oscillator strengths from 65 to 160 eV. These differential oscillator strengths are also i n semi-quantitative agreement wi th Hartree-Slater theoretical calculations of the C 2s atomic orbital differential osciUator strengths [68] (within 20%). In addition, however, the 3a ! differential oscillator strengths of C F 3 C 1 show different behaviour from those of CF2CI2 and CFCI3 below 60 e V . For the latter two molecules the 3a x differential osciUator strengths continue to rise to lower energy, whereas those for C F 3 C 1 show a max imum at 55 e V photon energy. These types of differences can arise near threshold due to the influence of the molecular field on low energy (slow moving) photoelectrons; i n fact this maximum is predicted by the M S - X o - calculations for C F 3 C 1 [153]. The inner valence orbitals of C l 3s character (the 2a x for C F 3 C 1 , the 2a x and lb2 for C F 2 C 1 2 and the l e and 2ai for C F C 1 3 ) have very low photoionization differential oscillator strengths from 41 to 160 eV, wi th magnitudes approximately i n proportion to the number of C l atoms i n the molecule. Comparison wi th the Hartree-Slater calculations [68] for C l 3s atomic orbitals reveals that the presently reported differential osciUator strengths for these molecular orbitals (per chlorine atom) are substantiaUy lower ( 40-75%) than those expected on the basis of the Gehus model [161]. The inner valence orbitals of predominantly F 2s character (the l e and lax for C F 3 C 1 , the l b i and l a x for C F 2 C 1 2 and the l a i for CFCI3) show similar behaviour, i.e. the differential oscillator strengths are approximately proportional to the number of F atoms i n the molecules, but are significantly (10-40%) less than predicted using the Gelius model [161] and the atomic F 2s differential oscillator strengths given i n ref. [68]. One possible reason for this is that electron correlation effects cause the inner valence C l 3s and F 2s orbitals to lead to a large number of many-body final ion states [158], many of which may not be readily visible i n the photoelectron spectra above what is assumed to be the (non-spectral) background. Such intensity wiU not be included in the data analysis procedure, and this w i l l result Chapter 6. 163 i n branching ratios and partial photoionization differential oscillator strengths for these orbitals which are too low. The present work clearly demonstrates that the part ia l photoionization differential oscillator strengths (cross-sections) of inner valence orbitals of C 2s and especially F 2s character are quite substantial i n the freon molecules and certainly cannot be ignored when attempting to use photoelectron branching ratios and total photoabsorption dif-ferential oscillator strengths to calculate partial photoionization differential oscillator strengths for valence level electronic ion states. In the absence of direct experimental measurements of the branching ratios or differential oscillator strengths of the inner va-lence orbitals, it may be possible to use estimates based on atomic data. Al though the inner valence part ia l photoionization differential oscillator strengths determined i n the present work are substantially lower than would have been estimated from atomic data, it is uncertain whether this is a real effect or a consequence of the many-body nature of the inner valence electronic ion states, leading to underestimates of their true intensities. Photoelectron spectra wi th much higher signal/noise ratio and possibly higher resolution and/or photoelectron/photoion coincidence measurements w i l l be required to answer this question w i t h greater certainty. Chapter 7 Absolute Dipole Differential Oscillator Strengths for Inner Shell Spectra from High Resolution Electron Energy Loss Studies of the Freon Molecules C F 4 , CF3CI, CF 2 C1 2 , CFCI3 and CC1 4 The CF3CI, C F 2 C 1 2 and CFCI3 inner shell E E L S spectra used to derive absolute differen-t ia l oscillator strengths were measured using the high resolution dipole(e,e) spectrometer described i n section 3.2. The C F 4 inner shell E E L S spectra are from refs. [130,162] and those for CCI4 are from ref. [163]. 7.1 Absolute Differential Oscillator Strengths Absolute photoabsorption differential osciUator strength (cross section) spectra have been obtained from the presently measured inner shell electron energy loss spectra. The proce-dures used i n the present chapter for obtaining absolute inner sheU differential oscillator strength spectra were as foUows (Similar types of investigation have previously been re-ported by Mclaren et al. [44] and also in section 5.3 above). First ly, the underlying continuum due to valence and less tightly bound inner shells plus any non-spectral back-ground was fitted to a function of the form a(E — b)c + d in the energy region just before the inner sheU edge of interest (E is the energy loss and a, 6, c and d are constants). This fitted function was then extrapolated to higher energy and subtracted from the partic-ular inner shell electron energy loss spectrum. The resulting spectrum which should be due to the particular inner sheU excitation alone was then converted to an approximate 164 Chapter 7. 165 relative optical spectrum using a kinematic Bethe-Born factor estimated to be propor-tional to E2'5 (see section 2.2 and equations 2.37 and 2.38). The absolute scale was then obtained by normalizing the relative optical spectrum at a sufficiently high photoelectron energy (typically ~25-35 e V above the IP) to the sum of the corresponding atomic sub-shell optical differential oscillator strengths for the appropriate constituent atoms at that photoelectron energy. The principles underlying this normalization method have been discussed i n sections 2.3.4.3 and 2.3.8. For this procedure semi-empirical total atomic cross sections [57] were used to obtain relevant subshell atomic cross sections (differen-t ia l oscillator strengths) by subtracting estimated contributions from shells wi th lower excitation energies from the total . The accuracy of the background subtraction procedures used to obtained the sub-shell molecular differential oscillator strength spectrum was considerably improved by further constraining the fitting of the background region such that the resulting differen-t ia l oscillator strength spectrum in the higher (photoelectron) energy continuum region matched the shape of the appropriate summed (i.e. C Is, F Is , C l 2p and C l (2p+2s)) atomic oscillator strength distributions (see, for example, figure 7 below which illustrates the results of such constrained procedures for C l 2p,2s spectra). A s discussed i n sec-t ion 2.3.4.3, molecular effects, such as E X A F S , are of low amplitude relative to the direct ionization continuum and w i l l thus have httle adverse effect on these procedures as long as the energy range used for matching molecular and atomic spectra is sufficiently large. A s discussed i n section 2.3.4.3 at sufficiently high photoelectron energy, the molecu-lar photoionization differential oscillator strength contributed by each atom in a molecule can be considered, to a first approximation, to be equal to the corresponding atomic pho-toionization differential oscillator strength. Therefore the integrated oscillator strengths for the molecule and for the constituent atom above a sufficiently high photoelectron Chapter 7. 166 Table 7.33: Estimations of integrated atomic oscillator strengths (OS) for inner shell excitations i n C , F , and C l atoms E0 - IPa OS above E0b OS transferred 0 OS below E0d C Is 38.7 1.393 0.0913 0.5157 F Is 48.0 1.5 0.275 0.225 C l 2p 60.2 e 4.8 -0.9026 2.1026 aE0 is the photon energy below which the integrated oscillator strength for the given inner shell excitation is to be estimated. Eo — IP is thus the photoelectron energy. 6The integrated oscillator strength from Eo to 1.5 x 106 eV (effectively infinity) as estimated from calculation [69]. See text for details. coscillator strength transferred to other shells due to the Pauli excluded transitions from the present sub-shell to the already occupied orbitals [112]. dUsing the partial TRK sum-rule: the sub-shell oscillator strength below Eo = ( number of electrons in the inner shell) — (OS above EQ) — (OS transferred). ewith reference to the Cl 2pi/2 IP-Table 7.34: Integrated sub-shell oscillator strengths per atom for C F 4 , C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and CCI4 below E0 Exci ted shell E0 - IPa A t o m 6 C F 4 CF3CI C F 2 C 1 2 CFCI3 C C 1 4 C Is 38.7 0.52 0.57 0.57 0.57 0.54 0.53 F Is 48.0 0.23 0.23 0.23 0.23 0.22 C l 2p 60.2C 2.10 1.99 1.92 1.90 1.84 aEo — IP is the photoelectron energy below which oscillator strength for a given sub-shell is to be integrated. 6frpm table 7.33. cwith reference to the Cl 2p>y2 IP. Chapter 7. 167 energy should be approximately equal. Therefore, assuming that the oscillator strength transfer due to Pauh excluded transitions [38] to already occupied orbitals is similar for both the molecule and summed atomic situation, the integrated molecular oscillator strength per atom is expected (on the basis of the part ial T R K sum-rule considerations, see sections 2.3.4.3 and 2.3.8) to be comparable to the corresponding atomic oscillator strength integral up to the same photoelectron energy. The integrated sub-shell oscilla-tor strengths for C (Is) , F (Is) and C l (2p) atoms have been estimated using the partial T R K sum-rule considerations and are shown in table 7.33. The integrated oscillator strength above the photon energy E0 was obtained by integrating the calculated pho-toionization differential oscillator strength [68,69] from E0 to infinity (actually to the l imit of the relativistic Hartree-Slater calculation at 1.5 x 10 6 eV [69]. The contribution to the oscillator strength above 1.5 x 10 6 e V is estimated to be less than 2 x 1 0 - 6 ) . The oscillator strength transfer due to P a u l i excluded transitions is taken from Hartree-Slater calculations [112]. The adequacy of the presently employed background subtraction and normalization procedures is evident from table 7.34 from which it can be seen that the presently obtained integrated molecular oscillator strength per atom is very close to the corresponding integrated atomic oscillator strength. 7.2 Electronic Configurations and Spectral Assignments The C F 3 C 1 and C F C 1 3 molecules are of C 3„ symmetry, C F 2 C 1 2 has C 2 u symmetry and C F 4 and CCI4 have T j symmetry. The ground state electron configurations may be written as i n table 7.35. The valence shell electronic configurations of C F 3 C 1 , C F 2 C 1 2 , CFCI3 and C F 4 are as reported i n the earher P E S studies i n references [99,100,117,157] and i n this thesis work (chapter 6). The inner shell electronic configuration for C F 4 is as reported i n X P S measurements [86]. Gaussian 76 calculations were carried out for C F 4 , Chapter 7. 168 Table 7.35: Electronic configurations for the C F 4 , C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and C C 1 4 molecules\" Molecule C l Is F Is C Is C l 2s C l 2p C F 4 6 lt62lal 2al C F 3 C 1 ul 2al\\e4 3al 4a? ba22eA C F 2 C 1 2 lbjlal 2allbl 3a\\ 2bj4al ba23bl4bl2bllal<6al CFCI3 lalle4 2a\\ 3al 2eHal 5al3e4la224e46al5e4 C C 1 4 lt62lal 2al 2t623al 3t624allt\\le44tl Molecule Valence orbitals 0 V i r t u a l valence orbitals (unoccupied) C F 4 CF3CI C F 2 C 1 2 CFCI3 C C 1 4 3al2t*4al3tlle44tlltl 6al3e47al8al4e49al5e46e4la210al7e4 7al3bl8al5b29al6b210al4b2llal2a25b2 7b212al3a26b28b2 7a28al6e49all0al7e48e4llal9e410e42a 5al546a26tpe4742tl 5a°5t 2 (C-F)* 11a? 8e°12a? (C-Cl)* (C-F)* 13a?96° 76?14a? v v 'N v ' (C-Cl)* (C-F)* 12a? l le° 13a? (C-Cl)* (C-F)* 7a°8i 2 ( c - c i ) * \"See section 7.2 for details. fcfrom ref. [86]. cfrom chapter 6 and refs. [99,100,117,157]. Chapter 7. 169 Table 7.36: Dipole-allowed transitions between orbitals for C3v, C2v and Tj symmetries Init ial state F i n a l state C3v ( C F 3 C 1 and CFCI3) 0 l —> a i , e a 2 —> e e — • a i , e C2v ( C F 2 C 1 2 ) a i —> ax,6i,62 a 2 —> 61,62 61 —> ai,6i fe2 —> aub2 Td ( C F 4 and CC1 4 ) ax — • t2 t2 — • < 2 , i i , e , a i e —> i 2 , i i C\\| I o cn c CD L_ C O o _o ' o C/J o D 4 c CD o CD Q 0 c s — i 1 1 1 r -299 * 300 301 / ^ L \\ I I I I I A5-67 8 9 C F 4 AE=70meV C F 3 C I AE=67meV \\ A 2 / V s - A II II III II I B 3 5 8 10 C1s edge 1 I I II I 2 3 4 5 7 C i s edge 4 -0 -8 -C F 2 C I 2 AE=65meV N I I I I I M l I K • 1 2 3 4 56 7 9 C1s edge # » : \\ C F C I 3 AE=67meV I I I I I i i £ / \\ 2 3 4 5 . 6 7 C i s edge \\ ecu V AE=250meV 1 2 C1s edge ~ l I [ I I I I j I I I I [ I I T\" 290 295 300 Photon Energy (eV) 4 0 4 0 4 0 Figure 7.42: Differential osciUator strength spectra for C Is excitation of CF 4 , CF3C1, CF2C12, CFC13 and CCI4 in the discrete region. The CF 4 and CCI4 spectra were derived from EELS measurements in refs. [162,163] respectively. See section 7.3 and tables 7.37-7.40 for details. Chapter 7. 173 12 10 8 T> ^ CD 4 CM I O 2 ^-^ 0 _C CD -i 1 1 1 1 1 r r — — i 1 1 r 00 o 4-= 2 cn o O O 4 CD 0 -2-0 C1 s C F 4 A E = 5 0 0 m e V 9 C F 3 C I 1 A E = 3 0 0 m e V 11 C F 2 C I 2 1 A E = 3 0 0 m e V C F C I A E = 2 9 0 m e V 10 i 3 ecu i A E = 3 5 0 m e V -1 1 1 r 0 10 20 - 1 1 1 r 30 40 Photoelectron Energy (eV) Figure 7.43: Differential oscillator strength spectra for C Is excitation of CF 4 , CF3C1, CF2C12, CFC13 and CCI4 in the discrete and continuum regions. The CF 4 and CC14 spectra were derived from EELS measurements in refs. [130,163] respectively. See section 7.3 and tables 7.37-7.40 for details. Chapter 7. 174 Table 7.37: Experimental data for C Is, F Is and C l 2s, 2p excitations of C F 3 C 1 Feature\" Photon Oscillator Term value Possible energy (eV) strength6 (xlO\"2) (eV) assignment (final orbital) C Is 1 294.16 4.8 6.15 l l O i 2 296.66 16.4 3.65 8e 3 297.3 3.0 12ai 4 298.07 2.24 4p 5 298.82 1.49 5s 6 298.93 1.37 bp 7 299.46 0.85 6s IP 8 .(300.31)c 303.12 0 -2.81 shape resonance 9 ~318.0 —17.7 XANES F Is 1 690.51 N 4.53 llax 2 692.60 I 15.5 2.44 8e+12ai 3 694.5 0.5 ? IP (695.04)«v 0 4 720 -25 XANES Cl Is D 2823.5 6.7 l l a i E 2827.4 2.8 12ai+8e F 2827.3 2.4 4p IP 2830.2 0 Cl 2s 9 IP 271.5 (278.84)c 7.3 0 virtual valence Cl 2p3/ 2 Cl 2pi/2 C12p3/2 C12p1 / 3 Cl 2p 1 201.13 1.6 6.70 l l O ! 1' 202.71 1.1 6.73 lid! 2 204.31 0.6 3.52 5.13 8e+12oi 3,2' 205.74 2.09 3.7 4p 8e+12ai 4 206.26 1.57 5s 5 206.56 1.27 5p 3' 207.32 2.12 4p 2^3/2 IP 4' (207.83)d 0 208.0 1.44 5s 5' 208.26 1.18 bp 2P1/2 IP (209.44)d 0 6 214.8 -7 -5.4 N 7 8 219.6 228.0 -11.8 -20.2 -10.2 I -18.6 J XANES \"For features, see figs. 1 and 2 for C Is, figs. 3 and 4 for F Is, figs. 5 and 6 for Cl 2s and 2p and ref. [166] for Cl Is. I^ntegrated oscillator strengths. See sections 7.3-7.5 for details. cfrom XPS measurement [165]. dfrom XPS measurement [164]. Chapter 7. 175 Table 7.38: Experimental data for C Is, F Is and C l 2s, 2p excitations of CF2CI2 F e a t u r e 0 P h o t o n O s c i l l a t o r T e r m v a l u e P o s s i b l e e n e r g y s t r e n g t h ' ' ( e V ) a s s i g n m e n t ( e V ) ( X l O \" 2 ) ( f i n a l o r b i t a l ) C I s 1 2 9 2 . 7 3 3 . 7 6 .22 1 3 o i 2 2 9 3 . 6 2 6.2 5 .31 9i)2 3 2 9 5 . 0 5 3 . 8 8 7 & i 4 2 9 5 . 6 0 3 . 3 3 1 4 a i 5 2 9 6 . 4 8 2 . 4 6 4 p 6 2 9 6 . 8 7 2 . 0 6 4 p ' 7 2 9 7 . 4 6 1 . 4 7 5 s 8 2 9 7 . 6 1 .33 5 p 9 2 9 8 . 1 0 0 . 8 3 6 s I P ( 2 9 8 . 9 3 ) c 0 1 0 3 0 1 . 9 3 - 3 s h a p e r e s o n a n c e 11 3 1 5 . 6 - 1 6 . 7 X A N E S F I s 1 6 8 9 . 4 0 - \\ 5 . 26 1 3 a i + 9 6 2 2 6 9 1 . 9 0 £ 9 . 6 2 .78 7 6 1 + 1 4 a 1 3 6 9 3 . 3 7 1 1 .31 5 p I P ( 6 9 4 . 6 8 ) c J 0 4 6 9 7 . 5 4 - 2 . 8 6 s h a p e r e s o n a n c e 5 7 1 7 . 9 2 - 2 3 . 2 4 X A N E S C l I s D 2 8 2 3 . 0 6 .6 13ax+9b2 d E 2 8 2 6 . 4 3 .2 1 4 a 1 + 7fc l d F 2 8 2 7 . 2 2.4 4 p I P ( 2 8 2 9 . 6 ) 0 C l 2 s 11 I P 2 7 1 . 3 ( 2 7 8 . 6 3 ) c 7.3 0 v i r t u a l v a l e n c e C l 2 p 3 / 2 C l 2 p x / 2 C l 2 p 3 / 2 C l 2p1/2 C l 2 p 1 2 0 0 . 7 3 2.1 6 .74 1 3 a i 2 2 0 1 . 3 9 0 . 7 6 . 0 8 962 1 ' 2 0 2 . 3 2 1.2 6 . 7 8 1 3 a ! 2 ' 2 0 3 . 0 1 0.4 6 . 0 9 9(>2 3 2 0 4 . 0 4 1 .7 3 . 4 3 7 6 i 4 2 0 4 . 4 0 3 . 0 7 1 4 a i 5 2 0 5 . 1 7 2 .3 4 p 6 2 0 5 . 3 8 2 . 0 9 V 7 , 3 ' 2 0 5 . 8 0 1 . 6 7 3 . 3 5 s 7 & i 4 ' 2 0 6 . 0 2 3 . 0 8 1 4 a ! 5 ' 2 0 6 . 7 6 2 . 3 4 4 p 6 ' 2 0 7 . 0 2 2 . 0 8 4 p ' 2 p 3 / 2 I P (207.47)<= 0 7 ' 2 0 7 . 7 8 1 .32 5 s 2 p i / 2 I P ( 2 0 9 . 1 0 ) e 0 8 2 1 6 . 3 - 8 . 8 - 7 . 2 ~ \\ 9 2 2 2 . 8 - 1 5 . 3 - 1 3 . 7 S X A N E S 1 0 2 3 0 . 6 - 2 3 . 1 - 2 1 . 5 J °For features, see figs. 1 and 2 for C Is, figs. 3 and 4 for F Is, figs. 5 and 6 for Cl 2s and 2p and ref. [166] for Cl Is. integrated oscillator strengths. See sections 7.3-7.5 for details. cfrom XPS measurement [165]. d9&2 and 7&i have been mistakenly denoted as 10&2 and 661 in ref. [166]. efrom XPS measurement [164]. Chapter 7. 176 Table 7.39: Experimental data for C Is , F Is and C l 2s, 2p excitations of CFCI3 Feature\" Photon Oscillator Term value Possible energy strength* (eV) assignment (eV) (xl0~2) (final orbital) C Is 1 291.42 3.3 6.12 12ai 2 292.24 9.7 5.30 lle(l) 3 292.83 4.71 lle(2) 4 294.22 3.32 13ai 5 294.87 2.67 Ap 6 296.10 1.44 5s 7 296.71 0.83 6s IP 8 (297.54)c 300.7 0 -3.16 shape resonance 9 305.51 -7.97 10 318.60 -21.1 XANES F Is 1 688.7 \\ 5.63 12ax-|-lle 2 690.42 V 3.8 3.91 13ai IP (694.33)c 1 0 2 697.80 -3.47 shape resonance 3 720.0 -25.67 XANES Cl Is D 2822.8 6.5 12ai E 2825.7 3.6 lle+13ai F 2827.1 2.2 4p IP (2829.3) 0 Cl 2s 8 IP 270.9 (278.24)c 7.3 0 virtual valence Cl 2p3/2 Cl 2p 1 / 2 Cl 2p3/2 Cl 2p1 / 2 Cl 2p 1 200.65 3.6 6.55 12ax 2 201.79 0.5 5.41 lle(l) 1' 202.33 1.4 6.48 12ai 3,2' 203.15 0.9 4.05 5.66 lle(2) lle(l) 4 203.96 2.9 3.24 13ai 3' 205.11 3.70 lle(2) 4' 205.62 3.19 13ai 2P3/2 IP (207.20)d 0 2P1/2 IP (208.81)d 0 5 216.6 9.4 -7.8 -s 6 222.7 15.5 -13.9 V XANES 7 233.7 26.5 -24.9 J \"For features, see figs, 1 and 2 for C Is, figs. 3 and 4 for F Is, figs. 5 and 6 for Cl 2s and 2p and ref. [166] for Cl Is. ''Integrated oscillator strengths. See text sections 7.3-7.5 for details. cfrom XPS measurement [165]. dfrom XPS measurement [164]. Chapter 7. 177 Table 7.40: Experimental data for C Is, F Is and C l 2s, 2p excitations of C F 4 and CCI4 Feature\" Photon Oscillator Term value Possible energy6 strength0 (eV) assignment (eV) (XlO\" 2) (final orbital) C F 4 C Is 1 297.45 ^ 4.35 \\ 2 297.77 4.03 ) 5t2 plus 3 298.54 3.26 ( outer well states 4 298.81 2.99 J 5 299.45 2.35 3p 6 299.66 > 29 2.14 ? 7 299.87 1.93 1 8 300.31 1.49 4s 9 300.60 \\ 1.20 4p 10 301.03 ) 0.77 5s IP (301.8)d -/ 0 F Is 1 ~690.5 ~ 5 5ai 2 692.9 1 s 25.4 2.6 5t2(l) 2 ~ 694 1.2 5t2(2) IP (695.52)d> ) 0 CCLi C Is 1 290.9 9.9 5.5 8t2 2 294.5 1.8 5s IP (296.38)d 0 3 ~303 ~7 shape resonance Cl 2s 4 271.8 6.2 virtual valence IP (278.04)d 0 C12p 3 / 2 C12p 1 / 2 C12p 3 / 2 C12pi / 2 Cl 2p 2p3/2IP (207.04) e 0 2p 1 / 2IP (208.73)6 0 1 -217.7 -10.7 -9 >| 2 -224.7 -17.7 -16 i XANES 3 -235.2 -28.2 -26.5 J \"See figs. 1 and 2 for C Is features, figs. 3 and 4 for F Is features, and figs. 5 and 6 for Cl 2s and 2p features. experimental results for CF 4 are from [162] and those for CC14 are from [163] Integrated oscillator strengths. See text sections 7.3-7.5 for details. dfrom XPS measurement efrom XPS measurement 165]. 164]. Chapter 7. 178 have been assigned assuming that the term values for a particular type (i.e. ( C - C l ) * and (C-F)*) of antibonding orbital have similar values i n the different molecules. The oscil-lator strength concentrated wi th in a broad band (feature 1) i n the CCI4 spectrum which corresponds to the dipole allowed transition to the 8t2 C - C l antibonding orbital [163] is redistributed (see fig. 7.42 and tables 7.37-7.40) upon successive fluorination among the transitions to 12ai ( ( C - C l ) * , feature 1) l i e ( ( C - C l ) * , feature 2 and 3, see discussion be-low) and 13ax ( (C-F)* , feature 4) for C F C 1 3 (see table 7.39); to 13a! ( ( C - C l ) * , feature 1), 9b2 ( ( C - C l ) * , feature 2), 7h ( (C-F)* , feature 3) and 14a x ( (C-F)* , feature 4) for C F 2 C 1 2 (see table 7.38); to l l a x ( (C-Cl ) * , feature 1), 8e ( (C-F)* , feature 2) and 12a x ( (C-F)* , feature 3) for CF3CI (see table 7.37); and finally concentrates in a single broad band again i n C F 4 (see table 7.40), corresponding to the transition to the 5t2 ( C - F ) * antibond-ing orbital [150]. Under the present assignment, the term values for transitions from the C Is orbitals to ( C - C l ) * antibonding orbitals are found to be between 5.4-6.2 e V while those to ( C - F ) * antibonding orbitals are between 3.3-3.9 eV. In the C F C 1 3 spectrum the feature 2 and the shoulder shown as feature 3 are ~0.6 e V apart and they are probably the two Jahn-Teller components of the transition to the l i e v i r tua l orbital . A Jahn-Teller split t ing of ~0.8 eV has been observed i n P E S studies [167]. It is interesting to note from fig. 7.43 that a minimum followed by an above IP m a x i m u m is observed near the ionization edge i n the C Is spectra of C F 3 C 1 , C F 2 C 1 2 , CFCI3 and CCI4. In proceeding from C C 1 4 to C F 3 C 1 , the m i n i m u m shifts towards the IP and finally moves above the IP i n CF3CI. In the case of C F 4 , there is no localized near-edge max i m um above IP. A potential barrier due to the presence of the halogen ligands i n the freon molecules is expected to affect the absorption spectra as discussed i n section 2.3.6. CCI4 and C F 4 belong to the same symmetry point group T j w i t h the same central atom (C) and they each have similar ax and t2 manifolds of unoccupied valence Chapter 7. 179 orbitals (see table 7.35). Theoretical calculations on the L i F g system [168] wi th varying L i - F distance have shown that the longer the L i - F distance, the stronger the inner well strength, i.e. the lower the energy of state supported by the potential , w i t h respect to the ionization l imit (i.e. a larger term value). Consistent observations on spectra for the molecular series S i B r 4 , S i C l 4 and S i F 4 [169], B B r 3 , B C 1 3 and B F 3 [170], and P B r 3 , P C 1 3 and P F 3 [171] have also shown that the corresponding v ir tua l valence feature and/or shape resonance feature shifts to higher energy with respect to the IP of the central atom as the hgand changes from B r to C l to F . In the present situation, the fact that the term value of the prominent C Is —> t2 transition i n C C 1 4 (5.4 eV) [163] is larger than that i n C F 4 (~3.4 eV) [130,162] illustrates that the inner well potential of C C 1 4 is stronger and this is consistent wi th the C - C l bond length (1.77 A ) for C C 1 4 being greater than the C - F bond length (1.32 A ) for C F 4 [172]. Moreover, the presence and the absence of the above IP near edge spectral maximum i n the C C 1 4 and C F 4 spectra respectively is consistent wi th the C F 4 inner well strength being so small that an above IP shape resonance is not seen. We assign the maxima features above the IP to transitions to shape resonance states (see fig. 7.43 and tables 7.37-7.40). These shape resonances may be associated w i t h v i r tua l orbitals involving C l 3d components. Such an interpretation is consistent w i t h the absence of such resonances i n the case of C F 4 . The weak structures superimposed on the broad band at ~298 eV i n C F 4 are probably due to transitions to outer well states which are not Rydberg i n character. The remainder of the pre-ionization edge features i n al l molecules are interpreted as transitions to Rydberg states. The Rydberg transition features are usually sharp i n shape as discussed i n section 2.3.2. The narrow shapes and similar term values of features 8 for C F 4 , 5 for C F 3 C 1 , 7 for C F 2 C 1 2 , 6 for C F C 1 3 and 2 for C C 1 4 (fig. 7.42) suggest that they are associated w i t h Rydberg orbitals of the same type (s). Other Rydberg transitions are Chapter 7. 180 similarly identified. The fact that feature 9 is well resolved from feature 8 for C F 4 suggests that the corresponding Rydberg transitions (p) are associated w i t h the shoulders 6 for C F 3 C 1 and 8 for C F 2 C 1 2 . S imilarly features 10 for C F 4 , 7 for C F 3 C 1 , 9 for C F 2 C 1 2 and 7 for C F C 1 3 are a l l of the same type (s), and also features 5, 6 and 7 for C F 4 , 4 for C F 3 C 1 , 5 and 6 for C F 2 C 1 2 , and 5 for CFCI3 are of the same type (p). It is interesting to note that the very weak 6s Rydberg transition (feature 2) in the C C 1 4 spectrum increases i n intensity wi th successive fluorination so that it is much more intense for C F 3 C 1 than for CCI4. W h e n the series proceeds to C F 4 , the Rydberg feature merges w i t h the single band. 7.4 F Is S p e c t r a Figs. 7.44 and 7.45 show the presently obtained high resolution (short range) and low resolution (long range) F Is differential oscillator strength spectra for C F 3 C 1 , C F 2 C 1 2 and CFCI3 along with that for C F 4 presented i n section 5.3. The spectra i n fig. 7.45 are presented wi th photoelectron energy as the x axis. Energies, term values, integrated oscillator strengths up to IPs and possible assignments for transitions i n the F Is spectra are presented i n tables 7.37-7.40. It has recently been demonstrated from X-ray absorption experiments for CF3CI, C F 2 C 1 2 and C F C 1 3 that transitions from C l Is to C - F antibonding orbitals have appre-ciable intensities [166] i n addition to those to ( C - C l ) * antibonding orbitals, even though the C l Is orbital is localized i n a very small spatial region at the C l atomic sites i n the molecule. In keeping wi th this observation [166] transitions to both C - F and C - C l an-t ibonding orbitals are used to interpret the F Is spectra obtained i n the present work. The strong band (feature 2) below the F Is IP in the C F C 1 3 spectrum (fig. 7.44) can therefore be attributed to the transition to the 13ai (C-F) * antibonding orbital . Since Chapter 7. 181 I > 5 CD CM 1 4 CP c 2 CD Z O = 0 \"u O 1 - i 1 1 1 1 1 i 1 1 i r A s • * * * * I \\ ; F1s edge CF 4 :* AE=500meV S i • I I 1 /** 2 3 » A, C F 3 C I \\,AE=260meV 3 F 1 s edge C F 2 C I 2 A E = 1 4 0 m e V 3 F 1 s edge CFCI3 AE=265meV 1 £ 2 F 1 s edge 1 — r -690 1 1 1 1 1 r 686 690 694 698 Photon Energy (eV) F i g u r e 7.44: Differential oscillator strength spectra for F Is excitation of C F 4 , CF3CI, CF2CI2 and CFCI3 in the in discrete region. The CF4 spectrum was obtained from section 5.3. See section 7.4 and tables 7.37-7.40 for details. Chapter 7. 182 6- A > cd CM I o 4 -ii i « cn c CD •+-> 00 o ' u CO O o 1 -1 -c CD CD 0-1 • O H i t t tiimTr s C F * A E = 5 0 0 m e V C F 3 C I A E = 3 1 0 m e V C F 2 C I 2 A E = 3 0 0 m e V ' CFCI3 4 5 A E = 3 0 0 m e V T 0 10 20 30 ~40~ 50 Photoelectron Energy (eV) 0.5 0 Figure 7.45: Differential oscillator strength spectra for F Is excitation of CF4, CF3CI, CF2CI2 and CFCI3 in the discrete and continuum regions. The CF4 spectrum was obtained from section 5.3. See section 7.4 and tables 7.37-7.40 for details. Chapter 7. 183 the overall band is asymmetric, the expected weaker transitions to the ( C - C l ) * anti-bonding orbitals (12a i , l l e ) here therefore been assigned to the lower energy shower (feature 1). S imilar ly transitions to the (13ai+96 2 , (C-Cl)*) and (7&i+14ai,(C-F)*) are assigned to features 1 and 2 i n the F Is spectrum of C F 2 C 1 2 . Likewise, the l l a i ( (C-Cl)*) and (8e+12ai) ((C-F)*) final orbitals have been assigned to the two features 1 and 2 i n the C F 3 C 1 spectrum. The broad C F 4 pre-ionization edge band has been attributed to the transitions to the r 2 and ar states [150]. The shoulder 1 i n the C F 4 spectrum may correspond to the transition to the 5ax orbital . The partially resolved features 2 and 3 may be due to transitions to the 5r 2 orbital where the degeneracy is removed by the Jahn-Teller effect. In the long range F Is spectra (fig. 7.45) a min imum followed by a maximum is observed near the edge i n the spectra of C F C 1 3 , C F 2 C 1 2 and possibly C F 3 C 1 as i n the C Is spectra. The min imum shifts from below the IP for C F C 1 3 to above the IP for C F 2 C 1 2 and C F 3 C 1 . The above IP spectral maxima, hke those i n the C Is spectra are probably due to transitions to vir tual orbitals involving C l 3d participation (shape resonances), since the structure is not present i n the C F 4 spectrum. The fact that these above edge resonances are relatively more intense i n the C Is spectra than i n the F Is spectra (compare figs. 7.43 and 7.45) is consistent with spatial overlap considerations for the respective in i t ia l and final states if the later have C l 3d character. 7.5 Cl 2s and 2p Spectra The presently obtained high resolution C l 2p differential oscillator strength spectra for CF3CI, C F 2 C 1 2 and C F C 1 3 are shown i n fig. 7.46. Gaussian peaks were fitted to the pre-ionization edge spectra i n the low energy region. Three peaks for C F 3 C 1 and five peaks Chapter 7. 184 for C F 2 C 1 2 and CFCI3 were fitted as shown i n fig. 7.46. Long energy range, low resolution C l 2p, C l 2s and C Is spectra of C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and C C 1 4 are shown i n fig. 7.47 on a common photoelectron energy scale (respective to C l 2pi/2). The CCI4 differential oscillator strength spectrum was obtained by converting the previously reported E E L S spectrum [163] using the method outhned i n section 7.1. It should be noted that the CCI4 E E L S spectrum had to be digitized from the figure i n ref. [163] and therefore the resulting \"noise level\" on the spectrum i n fig. 7.47 is greater than for the other spectra shown. In fig. 7.48 the molecular spectra are shown as differential osciUator strength per C l atom i n the molecule i n order to facihtate comparison. The C l atomic spectrum (also shown i n fig. 7.48) for C l 2p and C l 2s excitations [57] was used to normahze the C l 2p molecular spectra (see section 7.1 for details). Energies, term values, integrated osciUator strengths and possible assignments for various transitions i n the C l 2s, 2p spectra of C F 3 C 1 , C F 2 C 1 2 , C F C 1 3 and C C 1 4 are presented i n tables 7.37-7.40. Also shown i n tables 7.37-7.39 are results from C l Is X-ray absorption spectra for C F 3 C 1 , C F 2 C 1 2 and CFCI3 recently reported by Perera et al. [166]. Two series of structures related to the C l 2p 3 / 2 and C l 2p1/2 ionization potentials respectively are observed i n the C l 2p spectra of C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 (fig. 7.46). The intensities of the series related to the C l 2p 3 / 2 IP are greater than those for C l 2p\\/2 as expected. The present assignment of the lower energy structures is based on the transitions to the normaUy unoccupied vir tual valence molecular orbitals. O n l y two peaks (1 and 1') are observed at low energy in the C F 3 C 1 spectrum (fig. 7.46) and they assigned to transitions to the l l a i C - C l antibonding orbital . The less intense peaks 2 and 2' can then be attributed to transitions to the C - F antibonding orbitals, 8e and 12a!. In the C F 2 C 1 2 spectrum two pairs of peaks (1 and 1', 2 and 2') have been fitted to the Chapter 7. 185 - i 1 r - i 1 1 1 1 r Cl 2p edges 3 4 5 3/2 i—n 1 2 c , i 2 a ^ ; A E = 6 0 m e V 200 -1 1 1 1— 2 0 5 1 r 2 1 0 Photon Energy (eV) Figure 7.46: Differential oscillator strength spectra for Cl 2p excitation of CF3CI, CF 2 C1 2 and CFC1 3 in the discrete region. See section 7.5 for details. Chapter 7. 186 10 8 I > CD CM I o CP c CD CF 3 CI AE=300meV 9 Cl 2s i C 1s C F 2 C I 2 A E = 3 0 0 m e V C 1s CFCI 3 A E = 2 9 0 m e V I B I 5 6 7 8 Cl 2s C C k A E = 3 5 0 m e V 1 1 1 1 2 3 4 Cl 2s C 1s T 1 1 1 r -1 1 1 r 12 10 8 6 4 2 0 18 16 14 12 10 8 6 4 2 0 0 20 40 60 80 100 Photoelectron Energy (eV) Figure 7.47: Differential oscillator strength spectra for Cl 2p, Cl 2s and C Is excitations of CF 3 CI, C F 2 C I 2 , C F C I 3 and CCU in the discrete and continuum regions. The C C I 4 spectrum was derived from EELS measurements in refs [163]. See section 7.5 and tables 7.37-7.40 for details. Chapter 7. 187 Figure 7.48: Differential oscillator strength spectra per Cl atom for Cl 2p, Cl 2s and C Is excitations of CF3C1, CF2C12, CFCI3 and CCLj (measured, this work and ref. [163]) and for Cl 2p and Cl 2s excitations of the Cl atom (ref. [57]). See section 7.5 for details. Chapter 7. 188 spectrum at low energy and we interpret these as being due to transitions to the 13a! and 9&2 C - C l antibonding orbitals. Notice that the fitted peak 2, related to the C l 2p3/2 hole state, is more intense than peak 2' associated wi th the C l 2px/2 hole. Peaks (3 and 3', 4 and 4') are assigned as transitions to the C - F antibonding orbitals 7bi and 14aj. In the CFCI3 spectrum peaks 1 and 1' can be attributed to transitions to the 12a x C-C l antibonding orbital . Considering the term values and of peaks 4 and 4', it is likely that they are due to transitions to the 13a! C - F antibonding orbital . The two pairs of peaks (2 and 2', 3 and 3') can then assigned to transitions to the two components of the l i e antibonding orbital whose degeneracy is removed due to the broken symmetry of CFCI3 when an electron i n the C l inner shell is promoted. Since 2' and 3 share the same fitted Gaussian peak i n the spectrum, the fact that peak 2 has lower intensity is not inconsistent wi th the above observation that the series related to C l 2p3/2 exhibits larger intensities. In the present assignments 7.37-7.39 the term values for transitions to C - F antibond-ing orbitals are i n the region 4-7 eV , while those for C - C l antibonding orbitals are i n the region 3-3.5 eV , similar to the observation on the C Is spectra (section 7.3). Rydberg structures are heavily overlapped and are superimposed on the rising ionization edge. Tables 7.37-7.40 list the tentative assignments. The long energy range spectra i n figs. 7.47 and 7.48 show delayed onsets (see sec-tion 2.3.4.1) which resemble the C l atomic situation [25]. In the spectra per Cl atom shown i n fig. 7.48, al l four molecular spectra are very similar i n overall shape at pho-toelectron energies above ~25 eV. It should be noted that i n this higher energy region al l molecular spectra exhibit rather similar low amplitude E X A F S modulations on the photoionization continuum (see sections 2.3.4.2 and 2.3.4.3) wi th respect to the situa-t ion i n the smooth atomic C l spectrum [57] indicated by the solid hne i n fig. 7.48. The Chapter 7. 189 similar modulations i n this region are due to the fact that E X A F S is the result of sin-gle scattering processes i n which the high energy photoelectron is scattered by only one neighboring atom as discussed i n section 2.3.4.2. In the present situation the neighboring electron scatterers are C , F and C l atoms, each of which is at a similar distance (i.e. C l - C (~1.75 A ) , C l - F (~2.53 A ) and C l - C l (~2.88 A ) ) from the ionized C l atom i n each of the different molecules [172]. Similar types of E X A F S modulation pattern have been seen i n the C l 2p spectra of C H 2 C 1 2 , C H C 1 3 and C C 1 4 [163]. From fig. 7.47 it can be seen that there are also additional structures i n the spectra for C F 3 C 1 (features 6 and 8), C F 2 C 1 2 (features 8, 9 and 10) and C F C 1 3 (features 5,6 and 7) and C C 1 4 (features 1, 2 and 3) i n the near edge region below ~ 2 5 e V photoelectron energy. The term values and magnitudes of these X A N E S structures are different i n the various molecules (see fig. 7.48), since X A N E S is the the effect of multiple scattering of the outgoing photoelectron by the neighboring atoms as discussed i n section 2.3.4.2. The C l 2s spectra are broadened due to the short lifetime of the excited states wi th respect to autoionizing decay to the underlying C l 2p continuum. The energies and possible assignments of these features are shown in tables 7.37-7.40. C h a p t e r 8 C o n c l u s i o n s This work has presented dipole (e,e) and dipole (e,e+ion) experimental results including the total and part ia l (photoion) absolute optical differential osciUator strength spectra of C F 4 , CF3CI, C F 2 C 1 2 and CFCI3 in the valence sheU excitation region; the high res-olution absolute optical differential oscillator strength spectra for C l 2p, C l 2s, C Is and F Is inner sheU excitation of C F 3 C 1 , C F 2 C 1 2 , and C F C 1 3 and for N Is and 0 Is inner shell excitation of N 0 2 . This work has also reported the absolute photoionization part ia l (electronic state) differential osciUator strength spectra for C F 3 C 1 , C F 2 C 1 2 and CFCI3 and proposed dipole induced breakdown pathways for C F 4 , C F 3 C 1 and C F 2 C 1 2 by combining the results from the dipole (e,e) and dipole (e,e+ion) measurements and previously pubhshed synchrotron radiation measurements. The good agreement between the differential osciUator strength spectra obtained i n the present work using the dipole (e,e) technique and previously measured spectra using synchrotron radiation where available, plus the good agreement between the dipole (e,e) experimental and the M C Q D theoretical results has demonstrated that the E E L S tech-nique is a useful alternative to experiments ut i l iz ing synchrotron radiation for studying photoabsorption and photoionization processes. The present measurements have con-siderably extended the available oscillator strength data base for photoabsorption and photoionization processes i n N 0 2 and the freon molecules C F 4 , C F 3 C 1 , C F 2 C 1 2 and C F C 1 3 i n the vacuum U V and soft X-ray regions. 190 Bibliography J . Heicklein. Atmospheric Chemistry (Academic, New York, 1978). Jr. L . Spitzer. Physical Processes in the Interstellar Medium (Wiley, New York, 1978). J . W . Chamberlain. Theory of Planetary Atmospheres (Academic, New York, 1978). M . Inokuti (ed.). Proceedings of the Workshop on Electronic and Ionic Collision Cross Sections Needed in the Modeling of Radiation Interaction with Matter 6— 8 December 1983 (Argonne National Lab . Report No . A N L 84-28, Argonne, IL , 1984). U . Becker and D . A . Shirley. Physica Scripta T 3 1 (1990) 56. H . Win ick and S. Doniach. Synchrotron Radiation Research (Plenum, New York, 1980). E . K o c h (ed.). Handbook on Synchrotron Radiation V o l . I, (North Hol land, 1983). J . W . Gallagher, C E . Br ion , J . A . R . Samson, and P . W . Langhoff. J. Phys. Chem. Ref. Data 17 (1988) 9. J . Frank and G . Hertz. Verhandl. Deut. Physik. Ges. 16 (1914) 457 and 512. H . Bethe. Ann. Physik. 5 (1930) 325. E . N . Lassettre and S .A . Francis. J. Chem. Phys. 40 (1964) 1208. J . Geiger. Z. Phys. 181 (1964) 413. M . J . V a n der W i e l . Physica 49 (1970) 411. M . J . V a n der W i e l and G . Wiebes. Physica 54 (1971) 411. M . J . V a n der W i e l and G . Wiebes. Physica 53 (1971) 225. G . Cooper, T . Ibuki , Y . Iida, and C E . Br ion . Chem. Phys. 125 (1988) 307. W . Zhang, G . Cooper, T . Ibuki , and C E . Br ion . Chem. Phys. 137 (1989) 391. E . B . Zarate, G . Cooper, and C E . Br ion . Chem. Phys. 148 (1990) 277. 191 Bibliography 192 [19] E . B . Zarate, G . Cooper, and C E . Br ion . Chem. Phys. 1 4 8 (1990) 289. [20] C E . B r i o n and A . Hamnett . Advan. Chem. Phys. 4 5 (1981) 1. [21] C E . B r i o n . Comments At. Mol. Phys. 1 6 (1985) 249. [22] C . Backx and M . J . V a n der W i e l . i n : Proceedings of the IV International Confer-ence on Vacuum Ultraviolet Radiation Physics, edited by E . - E . K o c h , R . Haensel and C . K u n z Hamberg, July 1974. [23] E . N . Lassettre and A . Skerbele. iw.Methods of Experimental Physics, 3B , edited by D . Wi l l iams (Academic Press, New York, 1974). [24] E . N . Lassettre. imChemical Spectroscopy in the Vacuum Ultraviolet, edited by C . Sandorfy and M . B . R o b i n (Reidel, Boston, 1974) p. 43. [25] U . Fano and J . W . Cooper. Rev. Mod. Phys. 4 0 (1968) 441. [26] M . Inokuti . Rev. Mod. Phys. 4 3 (1971) 297. [27] J . W . Chamberlain and D . M . Hunter. Theory of Planetary Atmospheres, 2nd E d . (Academic Press, New York , 1987) pp. 122-140. [28] R .S . Stolarski. Sci. Am. 2 5 8 (1988) 30. [29] H. I . Schiff. Nature 3 0 5 (1983) 471. [30] D . J . Wuebbles. J. Geophys. Res. 8 8 C (1983) 1433. [31] G . Brasseur, A . De Rudder, and C . Tricot. J. Atmos. Chem. 3 (1985) 261. [32] Y . Susumu. Kogai 2 3 (1988) 1. [33] M . J . M o h n and F .S . Rowland. Rev. Geophys. Space Phys. 1 3 (1975) 1. [34] H . Massey and A . E . Potten. Royal Institute of Chemistry, Lecture Series, No 1. Atmospheric Chemistry (Roy. Inst. Chem. , London, 1961) pp. 1-27, fig. 1. [35] R . Eisberg and R. Resnick. Quantum Physics of Atoms, Molecules, Solids, Nuclei, and Particles (John Wi ley & Sons, New York, 1974) pp. 53-54. [36] P . W . M i l o n n i and J . H . Eberly. Lasers (John Wi ley & Sons, New York,1988) p. 222, p. 226. [37] H . A . Bethe and E . E . Salpeter. Quantum Mechanics of One- and Two-Electron Atoms (Springer-Verlag, Ber l in , 1957) p. 256. Bibliography 193 [38] J . A . Wheeler and J .A.Bearden. Phys. Rev. 46 (1934) 755. [39] L . D . Landau and E . M . Lifshitz . Quantum Mechanics. Non-relativistic Theory translated by J . B . Sykes and J.S. Bel l (Pergamon, London, 1965), 2nd ed. [40] R . A . Bonham and M . F ink . High Energy Electron Scattering, A C S Monograph 16.9 (Van Nostrand, New York, 1974). [41] I .V. Hertel and K . J . Ross. J. Phys. B 1 (1968) 697. [42] I .V. Hertel and K . J . Ross. J. Phys. B 2 (1969) 285. [43] A . P . Hitchcock, G . R . J . Wi l l i ams , C E . Br ion , and P . W . Langhoff. Chem. Phys. 88 (1984) 65. [44] R . M c L a r e n , S . A . C . Clark , I. Ishii, and A . P . Hitchcock. Phys. Rev. A 36 (1987) 1683. [45] R . P . Feynman, R . B . Leighton, and M . Sands. The Feynman Lectures in Physics V o l . I (Addison-Wesley, Reading, Massachusetts, 1963) p. 43-4. [46] M . J . Seaton. Phys. Soc. London 88 (1966) 801. [47] M . B . R o b i n . Higher Excited States of Polyatomic Molecules V o l . I (Academic Press., New York , 1974). [48] M . B . Robin . Higher Excited States of Polyatomic Molecules V o l . II (Academic Press., New York , 1975). [49] M . B . Robin . Higher Excited States of Polyatomic Molecules V o l . I l l (Academic Press., New York , 1985). [50] J.S. Binkley, R . Whiteside, P . C Hariharan, , R. Seeger, W . J . Hehre, M . D . Newton, and J . A . Pople. G A U S S I A N 7 6 , Program no. 368, Quantum Chemistry Program Exchange, Indiana University, Bloomington, I N , U . S . A . [51] R . N . S . Sodhi and C E . Br ion . J. Electron Spectrosc. Relat. Phenom. 34 (1984) 363. [52] R . N . S . Sodhi and C E . Br ion . J. Electron Spectrosc. Relat. Phenom. 37 (1985) 97;125;145. [53] K . H . Sze, C E . Br ion , X . M . Tong, and J . M . L i . Chem. Phys. 115 (1987) 433. [54] G . Herzberg. Molecular Spectra and Molecular Structure, V o l . I (D. V a n Nostrand Company, Inc, Princeton, 1950). Bibliography 194 [55] C . Fabre and S. Haroche. i n : Rydberg State of atoms and molecules, edited by R . F . Stebbings and F . B . Dunning, (Cambridge University Press, London, 1983). [56] J . Berkowitz. Photoabsorption, Photoionization and Photoelectron Spectroscopy (Academic Press, New York, 1979). [57] E . B . Saloman, J . H . Hubbel l , and J . H . Scofield. At. Data and Nucl. Data Tables 38 (1988) 1. [58] A . Bianconi . i n : EXAFS and Near Edge Structure, Springer Series i n Chemical Physics, edited by A . Bianconi , L . Incouia and S. Striprich, V o l . 27 (Springer, Berhn, 1983) p. 118. [59] J . L . Dehmer and D . D i l l . J. Chem. Phys. 65 (1976) 5327. [60] G . Cooper, T . Ibuki , and C E . Br ion . Chem. Phys. 140 (1990) 147. [61] V . I . Nefedov. J. Struct. Chem. 9 (1968) 217. [62] J . L . Dehmer. J. Chem. Phys. 56 (1972) 4496. [63] K . H . Sze and C E . Br ion . Chem. Phys. 137 (1989) 353. [64] K . H . Sze and C E . Br ion . Chem. Phys. 140 (1990) 439. [65] J . L . Dehmer. i n : Photophysics and Photochemistry in the Vacuum Ultraviolet, edited by S.P. M c G l y n n , G . L . Findley and R . H . Hueber (Reidel, Dordrecht, 1985) p34. [66] G . R . Wight , C E . Br ion , and M . J . V a n der W i e l . J. Electron Spectrosc. Relat. Phenom. 1 (1972/73) 457. [67] G . R . Wight and C E . Br ion . J. Electron Spectrosc. Relat. Phenom. 3 (1974) 191. [68] J . J . Yeh and I. L i n d a u . Atomic Data and Nucl. Data Tables 32 (1985) 1. [69] J . H . Scofield. Theoretical Photoionization Cross Sections from 1 to 1500 keV, Lawrence Livermore Laboratory Report N o . UCRL-51326 (1973). [70] B . L . Henke, P. Lee, T . J . Tanaka, R . L . Shimabukuro, and B . K . Fujikawa. Atomic Data and Nucl. Data Tables 27 (1982) 1. [71] M . J . V a n der W i e l , T h . M . El-Sherbini , and L . Vriens. Physica 42 (1969) 411. [72] C . Backx, T h . M . El-Sherbini , and M . J . Van der W i e l . Chem. Phys. Lett. 20 (1973) 100. Bibliography 195 [73] C . Backx and M . J . V a n der W i e l . J. Phys. B 8 (1975) 3020. [74] C . Backx, G . R . Wight , R . R . Tol , and M . J . V a n der W i e l . J. Phys. B 8 (1975) 3007. [75] C . Backx, R . R . To l , G . R . Wight , and M . J . V a n der W i e l . J. Phys. B 8 (1975) 2050. [76] F . Carnovale, A . P . Hitchcock, J . P . D . Cook, and C E . B r i o n . Chem. Phys. 6 6 (1982) 249. [77] F . Carnovale and C E . Br ion . Chem. Phys. 74 (1983) 253, and references therein. [78] S. Daviel , C E . Br ion , and A . P . Hitchcock. Rev. Sci. Instr. 55 (1984) 182. [79] T . Yamazaki and K . K i m u r a . Chem. Phys. Letters 43 (1976) 502. [80] W . Zhang, K . H . Sze, C E . Br ion , X . M . Tong, and J . M . L i . Chem. Phys. 140 (1990) 265. [81] O . Edqvist , E . L i n d h o l m , L . E . Selin, L . Asbrink, C E . K u y a t t , S.R. Mielczarek, J . A . Simpson, and I. Fischer-Hjalmars. Physica Scripta 1 (1970) 172. [82] W . H . E . Schwarz, T . C Chang, and J .P . Connerade. Chem. Phys. Letters 49 (1977) 207. [83] C R . Brundle, D . Neumann, W . C Price, D . Evans, A . W . Potts, and D . G . Streets. J. Chem. Phys. 53 (1970) 705. [84] D . W . Davis , R . L . M a r t i n , M . S . Banna, and D . A . Shirley. J. Chem. Phys. 59 (1973) 4235. [85] P. F i n n , R . K . Pearson, J . M . Hollander, and W . L . Jolly. Inorg. Chem. 10 (1971) 378. [86] K . Siegbahn, C . Nordhng, G . Johansson, J . Hedman, P . F . Heden, K . Hamrin , U . Gelius and T . Bergmark, L . O . Werme, R. Manne, and Y . Baer. ESCA Applied to Free Molecules (North-Holland, Amsterdam, 1969). [87] G . R . Wight and C E B r i o n . J. Electron Spectrosc. Relat. Phenom. 4 (1974) 313. [88] G . C K i n g , M . Tronc and F . H . Read. J. Phys. B 13 (1980) 999. [89] A . P . Hitchcock and C E . Br ion . J. Electron Spectrosc. Relat. Phenom. 18 (1980) 1. Bibliography 196 [90] G . C . K i n g , F . H . Read, and M . Tronc. Chem. Phys. Letters 52 (1977) 50. [91] G . R . Wight and C . E . Br ion . J. Electron Spectrosc. Relat. Phenom. 4 (1974) 25. [92] J . C . Slater. Phys. Rev. 38 (1931) 1109. [93] L . Paul ing and E . B . Wi l son . Introduction to Quantum Mechanics ( M c G r a w - H i l l , New York , 1935). [94] P . A . Cox. Struct. Bonding 24 (1975) 59. [95] C . E . B r i o n and K . H . Tan. Chem. Phys. 34 (1978) 141. [96] D . A . Outka and J . St6hr. J. Chem. Phys. 88 (1988) 3539. [97] D . Arvanit is , H . Rabus, L . Wenzel, and K . Baberschke. Z. Physik D 11 (1989) 219. [98] P.S. Connel l . Energy Res. Abstr. 12 No . 9891 (1987) [99] C R . Brundle , M . B . Rob in , and H . Basch. J. Chem. Phys 53 (1970) 2196. [100] T . Cvi tas , H . Giisten, and L . Klas inc . J. Chem. Phys. 67 (1977) 2687. [101] L . C . Lee, E . Phi l l ips , and D . L . Judge. J. Chem. Phys. 67 (1977) 1237. [102] L . C . Lee, X . Wang, and M . Suto. J. Chem. Phys. 85 (1986) 6294. [103] G . R . Cook and B . K . Ching . J. Chem. Phys. 43 (1965) 1794. [104] G . C . K i n g and J . W . McConkey. J. Phys. B 11 (1978) 1861. [105] T . A . Carlson, A . Fahlman, W . A . Svensson, M . O . Krause, T . A . Whi t leyr , F . A . G r i m m and M . N . Piancastelli , and J . W . Taylor. J. Chem. Phys. 81 (1984) 3828. [106] R . Gi lber t , P. Sauvageau, and C . Sandorfy. J. Chem. Phys. 60 (1974) 4820. [107] H . W . Jochims, W . Lohr , and H . Baumgartel . Ber. Bunsenges. Phys. Chem. 80 (1976) 130. [108] R . E . Rebbert and P. Ausloos. J. Res. Natl. Bur. Std. 75A (1971) 481. [109] B . E . Cole and R . N . Dexter. J. Quant. Spectrosc. Radiat. Transfer 19 (1978) 303. [110] C . Y . R . W u , L . C . Lee, and D . L . Judge. J. Chem. Phys. 71 (1979) 5221. Bibliography 197 [111] J . C . Person, D . E . Fowler, and P .P . Nicole. Argonne National Laboratory Report ANL-75-60 Part I. 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 M . Inokuti . Private communication (1985). M . A . D i l l o n and M . Inokuti . J. Chem. Phys. 82 (1985) 4415. Y . I ida, F . Carnovale, S. Daviel , and C E . B r i o n . Chem. Phys. 105 (1986) 211. T . Ibuki , G . Cooper, and C E . Br ion . Chem. Phys. 129 (1989) 295. W . R . Harshbarger, M . B . R o b i n , and E . N . Lassettre. J. Electron Spectrosc. Relat. Phenom. 1 (1972/1973) 319. I. Novak, A . W . Potts, F . Quinn, G . V . M a r r , B . Dobson, L H . Hil l ier , and J . B . West. J. Phys. B 18 (1985) 1581. C E . B r i o n , Y . Iida, F . Carnovale, and J .P . Thomson. Chem. Phys. 98 (1985) 327. W . Zhang, T . Ibuki , and C E . Br ion . Chem. Phys. , manuscript i n preparation. A . W . Potts , I. Novak, F . Q u i n n , G . V . M a r r , B . R . Dobson, and L H . Hill ier and J . B . West. J. Phys. B 18 (1985) 3177. G . M . Bancroft , J . D . Bozek, J . N . Cutler , and K . H . Tan. J. Electron Spectrosc. Relat. Phenom. 47 (1988) 187. J.S. Tse. J. Chem. Phys. 89 (1988) 920. B . M . Addison-Jones, K . H . Tan, B . W . Yates, J . N . Culter , and G . M . Bancroft and J.S. Tse. J. Electron Spectrosc. Relat. Phenom. 48 (1989) 155. C R . Natoh. i n : EXAFS and Near Edge Structure, Springer Series i n Chemical Physics, edited by A . Bianconi , L . Incouia and S. Striprich, V o l . 27 (Springer, Berhn, 1983) p. 43. A . Bianconi , M . Del l 'Ar icc ia , A . Gargano, and C R . Nato l i . i n : EXAFS and Near Edge Structure, Springer Series i n Chemical Physics, edited by A . Bianconi , L . Incouia and S. Striprich, V o l . 27 (Springer, Ber l in , 1983) p. 57. R . H . Huebner, D . L . Bushnell , R . J . Celotta, S.R. Mielzarek, and C E . K u y a t t . Argonne National Laboratory Report ANL-75-60 Part I. J . W . Cooper. Phys. Rev. 128 (1962) 681. V . N . Sivkov, V . N . A k i m o v , A . S . Vinogradov, and T . M . Z imkina . Opt. Spectrosc. (USSR) 60 (1986) 194. Bibliography 198 [129] R . E . L a V i l l a . J. Chem. Phys. 58 (1973) 3841. [130] G . R . Wight and C E . Br ion . J. Electron Spectrosc. Relat. Phenom. 4 (1974) 327. [131] R . D . Hudson. Rev. Geophys. and Space Phys. 9 (1971) 305. [132] H . M . Rosenstock, K . D r a x l , B . W . Steiner, and J .T . Herron. J. Phys. Chem. Ref. Data 6 Suppl . 1 (1977). [133] S . G . Lias, J . E . Bartmess, J . F . Liebman, J . L . Holmes, and R . D . Levin and W . G . M a l l a r d . J. Phys. Chem. Ref. Data 17 No. 1 (1988). [134] D . D . Wagman, W . H . Evans, V . B . Parker, I. Halow, S . M . Bailey, and R . H . Schumm. NBS Tech. Note 270-3 (U.S. Government Pr int ing office, Washington, 1968). [135] M . M . Bibby, B . J . Toubehs, and G . Carter. Electron. Letters 1 (1965) 50. [136] R . I . Reed and W . Snedden. Trans. Faraday. Soc. 54 (1958) 301. [137] V . H . Dibeler, R . M . Reese, and F . L . Mohler . J. Res. NBS 57 (1956) 113. [138] C . Lifshitz and F . A . Long. J. Phys. Chem. 69 (1965) 3731. [139] T . A . Walter, C . Lifshitz , W . A . Chupka, and J . Berkowitz. J. Chem. Phys. 51 (1969) 3531. [140] C J . Noutary. J. Res. Natl. Bur. Std. 7 2 A (1968) 479. [141] J . W . Warren and J . D . Craggs. Mass Spectrometry (The Institute of Petroleum, London, 1952) p. 36. [142] J . M . Ajel lo , W . T . Huntress, Jr., and P. Rayermann. J. Chem. Phys. 64 (1976) 4746. [143] K . Watanabe, T . Nakayama, and J . M o t t l . J. Quant. Spectrosc. Radiative Transfer 2 (1962) 369. [144] L . M . Leyland, J .R . Majer , and J . C . Robb. Trans. Faraday Soc. 66 (1970) 898. [145] J . B . Farmer, I .H.S. Henderson, F . P . Lossing, and D . G . H . Marsden. J. Chem. Phys. 24 (1956) 348. [146] F . C . Y . Wang and G . E . Leroi . Ann. Isr. Phys. Soc. 6 (1984) 210. [147] H . Schenk, H . Oertel, and H . Baumgartel . Ber. Bunsenges. Phys. Chem. 83 (1979) 683. Bibliography 199 148] R . K . C u r r a n . J. Chem. Phys. 34 (1961) 2007. 149] B . W . Yates, K . H . Tan , G . M . Bancroft, L . L . Coatsworth, and J.S. Tse. J. Chem. Phys. 83 (1985) 4906. 150] J . A . Stephens, D . D i l l , and J . L . Dehmer. J. Chem. Phys. 84 (1986) 3638. 151] I. Novak, J . M . Benson, and A . W . Potts. J. Electron Spectrosc. Relat. Phenom. 41 (1986) 175. 152] G . Cooper, W . Zhang, and C . E . Br ion . Chem. Phys. 145 (1990) 117. 1531 J . D . Bozek, G . M . Bancroft, J . N . Cutler , K . H . Tan, and B . W . Yates. Chem. Phys. 132 (1989) 257. 154] K . H . Tan\", . G . M . Bancroft, L . L . Coatsworth, and B . W . Yates. Can. J. Phys. 60 (1982) 131. 1551 B . W . Yates, K . H . Tan , L . L . Coatsworth, and G . M . Bancroft. Phys. Rev. A 31 (1985) 1529. 1561 K . H . Tan, P . C . Cheng, G . M . Bancroft, and J . W m . M c G o w a n . Can. J. Spectrosc. 29 (1984) 1381. 1571 A . W . Potts , I. Novak, F . Quinn , G . V . M a r r , B . Dobson, L H . Hil l ier , and J . B . West. J. Phys. B 18 (1985) 3177. 1581 R. C a m b i , G . Ciu l lo , A . Sgamelotti, F . Tarantelh, R . Fantoni, A . Giard in i -Guidoni , M . Ros i , and R. T i r ibe lh . Chem. Phys. Lett. 90 (1982) 445. 1591 E . J . A i t k e n , M . K . B a h l , K . D . Bomben, J . K . Gimzewski , and G.S . Nolan and T . D . Thomas. J. Am. Chem. Soc. 102 (1980) 4873. 1601 C . Caulet t i , C . Furlani , C . P u l i t i , V . I . Nefedov, and V . G . Yarzhenski. J. Electron Spectrosc. Relat. Phenom. 31 (1983) 275. 1611 U . Gelius. J. Electron Spectrosc. Relat. Phenom. 5 (1974) 985. 1621 M . Tronc, G . C . K i n g , and F . H . Read. J. Phys. B 12 (1979) 137. 163] A . P . Hitchcock and C . E . Br ion . J. Electron Spectrosc. Relat. Phenom. 14 (1978) 417. [164] E . J . A i t k e n , M . K . B a h l , K . D . J . K . Gimzewski , G.S . Nolan , and T . D . Thomas. J. Am. Chem. Soc. 102 (1980) 4873. Bibliography 200 [165] S . A . Holmes. M . S . Thesis (Oregon State University, Oct 1974). [166] R . C . C . Perera, P . L . Cowan, D . W . Lindle , R . E . L a V i l l a , T . Jach, and R . D . Deslat-tes. Phys. Rev. A 43 (1991) 3609. [167] J . H . D . E l a n d . Photoelectron Spectroscopy 2nd E d . (Butterworths, London, 1984) p. 176. [168] A . A . Pavlychev, A . S . Vinogradov, and I .V. Kondrat 'eva. Sov. Phys. Solid State 26 (1984) 2214. [169] A . A . Pavlychev, A . S . Vinogradov, T . M . Zimkina , D . E . Onopko, and S . A . Ti tov. Opt. Spectrosc. (USSR) 47 (1979) 40. [170] E.I . Ishiguro, S. Iwata, Y . Suzuki , A . M i k u n i , and T . Sasaki. J. Phys. B 15 (1982) 1841. [171] E.I . Ishiguro, S. Iwata, Y . Suzuki , A . M i k u n i , H . Kanamor i , and T . Sasaki. J. Phys. B 20 (1987) 4725. [172] A . D . Mi tche l l and L . C . Cross, (eds.) Tables of Interatomic Distances and Con-figuration in Molecules and Ions Special publication No. 11 (Chemical Society, London, 1958). "@en ; edm:hasType "Thesis/Dissertation"@en ; edm:isShownAt "10.14288/1.0060156"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Chemistry"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Oscillator strengths for photoabsorption and photoionization processes of feron and NO₂ molecules"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/32006"@en .