)VISSI_J RETURNING MATERIALS: Place in book drop to “saunas remove this checkout from ——. your record. FINES will v be charged if book is returned after the date stamped below. PHOTOIONIZATION IIASS SWO STUDY OF SELECTED ORGANIC COMPOUNDS By Mahdi MOini A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department Of Chemistry 1985 “I V L' ABSTRACT PHOTOIONIZATIQI “ASS spsc'raommc STUDY OF SELECTED ORGANIC COMPOUNDS By Mahdi Moini Photoionization mass spectrometry (PIMS) is a technique for investigating the interaction Of ionizing radiation with gases. Threshold energies for simple and dissociative ionization can be measured, and the data combined with other thermochemical information to provide enthalpies Of formation Of the product. The ionization and fragmentation Of 0-, m— and p-dibromobenzene upon excitation in the 8.5-13 eV range have been studied, and [threshold energies for formation of CeHcBrs’, CoHsBr’ and Com’ from each precursor have been measured. The ionization energies of 0-, m- and Mrs are distinct, but the appearance energies of CoHaBr’ agree within experimental error for all three isomers and those Of 0034’ are also the same (higher in value, but within experimental error to that for Br loss), independent of the substituent distribution on the parent neutral. The results show that substituent isomerization occurs in CeHaBrt" prior to dissociation. Heats of formation for the ions have been calculated, and a estimate is given for that Of neutral o-benzyne: Emswefla) 2 370 t 20 kJ mol". Ionization and fragmentation Of p-dioxane (CoHeOn) and p-dioxane- Do have been followed by photoionization mass spectrometry, with particular emphasis on their dependence on photon energy and sample pressure. Appearance energies 'for the major ions have been determined Mahdi M0 ini and combined with other thermochemical data to Obtain heats Of formation of the cations. The mass spectra are strongly pressure- dependent, and show that some ions are formed by. both unimolecular and bimolecular steps. Both CcHeO:* and CchOa’ are Observed and their appearance energies have been measured. Ion-molecule reaction mechanisms which lead to these and other fragment ions are discussed. The role of hydrogen bonding in the. ion-molecule reactions has also been studied. The appearance energies for the protonated - parent and twelve cationic fragments of lZ-crown-4 have been measured and their heats Of formation have been calculated. The ionization energy Of 12-crown-4 is 8.86 t 0.1 eV. The threshold energies for fragmentation indicate that most Of the decompositions occur with insignificant kinetic shift. Possible mechanisms for the fragmentation and probable structures for some of the daughter ions are suggested. Farrokh Azadeh My Mother My Father ii ACKNOWLEDGMENTS First and foremost I thank my wife, Farrokh, and my mother and father for their support and encouragement throughout my scientific career. I thank Professor G. E. Leroi for his guidance and his extra help on personal matters. I also thank Professor J. Allison, who served as my second reader, and the other members Of my Guidance Committee, Professor R. II. Schwendeman and Professor E. G. LeGoff. I thank Marty Rabb who was a tremendous help in solving electronic problems, and also for his friendship during my stay at Michigan State University. I am grateful to Dr. T. Atkinson for keeping the PIMS computer "alive". Help from members Of the electronics, machine and glass shops is also gratefully acknowledged. I appreciate the support Of Ms. Emma Buycks who typed this dissertation, and of other members of the Chemistry Department who in one way or another helped me during my graduate study. During the two portions Of my stay at MSU I met different groups of people. From them I learned a lot, experienced widely ranging points Of view, and also found many friends. I also enjoyed and learned a great deal from WKAR-TV (PBS), which is like shade in darkness. Financial support from the Chemistry Department is gratefully acknowledged. iii. TAILS OI’CONTBITS Page' LIST OF TABLES .......................... vii LIST OF FIGURES ......................... viii CHAPTER I. INTRODUCTION ..................... 1 CHAPTER II. AN OVERVIEW OF PM ................. 4 Autoionization in Ate-s ............... 9 Autoionization in Molecules ............. l3 Dissociation and Predissociation. . ......... 20 Quasiequilibriul Theory ............... 24 Kinetic Shift Measure-eats .............. 28 Ionization Threshold Laws .............. ~ 30 I. Internal Causes . . . . . . . . . . ...... 33 a. Autoionization ............... 33 b. Dissociation and Predissociation ...... 33 c. Internal Energy Distribution ........ 34 d. Franck-Condon Effect ............ 34 II. External Causes ................ 38‘ a. Telperature ................ 38 b. Pressure .................. 42 c. Instrulental Effects ........... . 42 Heats Of Formation fOr Gas Phase Cations ....... 43 Couple-eatery Experimental Techniques ........ 48 a. Photoelectron Spectroscopy ......... 48 b. Photoelectron-Photoion Coincidence Spectroscopy ......... 51 CHAPTER III. EXPERIMENTAL APPARATUS AND PROCEDURES A. INTRODUCTION .................. 54 B. INSTRUMENT. .................. 54 1 Light Source ................ 54 2. Monochromator ............... 57 3. Interaction Region and Salple Inlet Systea ............ 58 4. Photon Transducer ............. 59' 5 Ion Optics and.Mass Spectra-eter ...... 60 iv CHAPTER IV. CHAPTER V. CHAPTER VI. 6. Ion Transducer. . . . . .......... 7. Vacuum SYItO-I. .............. 8. Interlock Systea. . . . . . . .,. . . . . . C. EXPERIMENTAL PROCEDURE ............. Saaples . . . . . . . SystemLPump—Down. . . Leap Spectra. . . . . . Mass Spectra. . . . Experimental Set-Up . . . Data Correction .- ............. O O O O O O O O O O O O O O O O O ODOIwar-I PHOTOIONIZATION AND FRAGMENTATION OF 0-, m- p-DIBROMOBENZENE ............ . . . A O IWDWTION O O O O O O O O O O O O O O O O O O 8. RESULTS .................... C. DISCUSSION. . ................. PHOTOIONIZATION MASS SPECTRCMETRIC STUDY OF 1,4-DIOXANE: THERMOCHEMISTRY AND ION- MOLECULE REACTIONS .................. INTRODUCTION. ........... . . . ...... RESULTS AND DISCUSSION. ............... A. Protonated p-Dioxane. . . . ........ B. Cali-:02+ .................. C. Unimolecular Dissociation of p—Dioxane at Low Pressure ......... D. Other Products Of Ion-Molecule Reactions ................. E. Ion With m/z Higher Than 89 . . . . . . . . Ion With a/z = 133. . . . . . ....... Ion With a/z = 116 and 114. . . . . . . . . F. Ion With a/z Less Than 88 ......... Cations With m/z = 73, 71 and 59. ..... Cations With a/z = 45, 44 and 43. ..... PHOTOIONIZATION MASS SPECTRIMETRIC STUDY OF 12-CROWN-4 ....... . ......... INTRODUCTION. . . . . ..... . . . . . . . . . . . RESULTS ................... . . . . DISCUSSION ......... . ........... . . i 3823838383 2 322’. 89 89 92 101 104 106 109 110 110 110 112 114 115 118 Page A. PIE Measurements and Theraochemistry. . . . 118 3. Structures and Energies of Fragments. . . . 118 CeHis0e*(a/z = 176) . ...... . . . . . 121 CeH1303’(m/z = 133) + CzHaO ........ 123 CeH1302*(m/z = 177) + 02H302 ........ 123 CeHsOa*(a/z = 101) + CaHiiO ........ 123 CsHst*(m/z = 89) + CeH'rOz ........ . 124 CeHst*(m/z = 88) + 0411302 ......... 125 CeHsO* (l/z = 73) + CzHaO + 0211902 ..... 125 CzH502*(I/z = 61) + (3sz + caho . . . . . 126 Cz&02*(m/z = 59) + 02113.0 + CeHsO + H2. . . 126 CsHsO+ (m/z = 58) + Colist + 01120. ..... 127 CszO+ (m/z = 45) + CeEOC + CzHaO ..... 127 CzHeO* (m/z = 44) + C4502 + CzHeO ..... 127 C2HaO’ (a/z = 43) + 04150: + CszO ..... 128 LIST OF REFERENCES ......... ' ............... 131 vi Table IV-l. Iv—z e Iv-3 O VI-l. VI-Z. LIST OF TABLES Threshold Energies and Thermochemical Data Relevant to the Photoionization and Fragmentation Of p-, m- and o-Dibromobenzene . Comparison of First Ionization Energies Of the Dibromobenzene Isomers Obtained by PIMS and PBS . Vertical Ionization Energies (in eV) and Suggested Sy-etry Assigiments Of Dibromobenzene Cation Electronic States . . . . . . . . . Threshold Energies and Heats Of Formation for Ions Produced from p-Dioxane at Low Saaple Pressure................. Ionization and Appearance Energies (2988) Of 12- Crown-4 and its Fragment Cations . . . Theraochemical Data Relevant to the Ionization and Dissociation Of 12-Crown-4- . . . vii Page 86 87 88 113 119 129 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure II-l II-2 II-3 II-4 II-5 II-6 II-7 II-8 I I-9 LIST 05' F1" Autoionization Of Tho-Electron Excited States in Heliua (from reference 24) . . . . . . . Autoionization-Inner Electron Transitions (froa reference 3). Autoionization Resulting from Excitation of an Electron from the (inner) 3s Orbital of Argon to up Levels (from reference 25). . Autoionization Converging to the Higher Energy Spin-Orbital Component Of the Ground State Configuration of the Art Ion (from reference 24) . Two alternative nodes of atoaic De- Excitation Which Follow. Creation of an Inner Shell Vacancy. (From K. Siegbahn, et a1.; ESCA Almquist and Wiksells, Uppsala, 1967) . . . . Autoionization (Thick-Line) and Direct Ionization (Dashed Line) in the Oxygen Molecule (from reference 35). . . . . . . . . . . . Schematic Diagraa Of Potential Curves Of Electronic States and Vibronic Transitions in a Diatomic Molecule (froa reference 27). . THO Electronic States Of a Molecule With Their Continuous Tera Spectra for Which Predissociation is Possible (froa Reference 38). . A Schematic Illustration Of Several Possible Types Of Photodissociation Of a Polyatomic Ion; (a) Electronic Predissociation; (b) Excitation to a Dissociative Vibrational Level Of the Ground Electronic State; (c) Vibrational Predissociation; (d) Direct Photodissociation on a Repulsive Surface (from Reference 39). viii Page 10 11 12 14 15 18 19 22 23 Figure Figure Figure Figure Figure Figure Figure Figure Figure II-10 II-11 II-12 II-13 II-14 II-15 II-16 II-17 II-18 Potential Energy Surface Showing a Random Walk in Phase Space leading to Dissociation Of a Polyatomic Ion or Molecule (froa reference 47). Fragmentation Of the Methanol Cation Showing Both Statistical Decomposition [CH20H* + H] and Non- Statistical Dissociation Through an Isolated Excited Electronic State [Clis+ + OH] (from reference 48). . Hypothetical Threshold Cross-Section Behavior for Frequently Encountered Excitation Processes (a) Direct Ionization by Electrons, (b) Photoionization, (c) Photoexcitation (from Reference 15) . . Experiaental Cross-Sect ion near Threshold for (a) Electron Inpact Ionization Of Rel iul, (b) Photoionization Of Nitric Oxide, (c) Double Ionization of helium by Electrons (from Reference 15). . Potential Curves for a Diatomic Molecule AB, Showing the Effect Of Predissociation on the Electron or Photon Iapact Spectrua (from Reference 67) . - - . - . . . . Internal Thermal Energy Distribution for Benzene at 25°C (from Reference 80). . . Franck-condon Factors and Hypothetical Direct Photoionization Yield Curves Expected for Snell, Moderate and Large Bond Length Changes in a Vertical Ionization (from Reference 48). . Photionization Cross-Section Of H20 (from Reference 15) . . Photoionization Cross-Section of CH: (froa Reference 15). . ix Page 27 29 31 32 35 36 39 40 41 Figure Figure ‘ Figure Figure Figure Figure Figure Figure Figure II-19 II-20 II-21 Ill-22 III-1 IV-1 IV-2 IV-3 I V—4 Time-Resolved Experimental (circles and squares) and Calculated (smooth lines) Daughter (CeHs*)PIE Curves in the Photofragmentation of Bromobenzene (from Reference 82) . . Schematic Diagraa Showing Kinetic Shift and Reverse Activation Energy in the Experimental Observation Of the fragmentation: M + hv 4 F” + N (from Reference 86). . . . . Principles Of UV Photoelectron Spectra. (a) Photoelectron Spectrua, (b) Electronic Configurations Of a Molecule and its Ionized States, (c) Energy Level Diagraa (from Reference 93). Vibrational Structure Of a Phtoelectron Hand and Vertical and Adiabatic Ionization Energies (from Reference 93). . BA-Baffle; EN-Entrance Slit; Eli-Exit Slit; GR-Grating; IS-Ion Source; IT- Ion Transducer; LP-Lamp; Pl-First Differential Pumping Port; P2-Second Differential Pumping Port; P3- Monochromator Pumping Port; P4-Samp1e Chamber Pumping Port; P5-Ouadrupole Chaflaer Pumping Port; PT-Photon Transducer; OP-Ouadrupole; OS- Ouadrupole Support; LE-Ion Lense. . The PIE Curves for the Ortho-, Mats- and Para-D ibromobenzene Parent Ions. . . Fine Structure on the PIE Curve of P- Dibromobenzene. . . . . . . . The PIE Curves Of the Clain-IBr+ Daughter Ions Generated by Photofragmentation Of Each Of the Isomeric Dibromobenzenes. . . . The PIE Curve Of the Cello+ Daughter Ions Generated by Photofragmentation of Each Of the Isomeric Dibromobenzenes. . . . . . . . Page 44 47 50 52 55 73 74 75 76 Figure V-l Figure V-2 Figure V-3 Figure V—4 Figure V-5 Figure V-6 Figure V-7 Figure V-8 Figure V-9 Figure VI-l Figure VI-2 Mass Spectrum of p-Dioxane Under Full Illumination Of the Hydrogen Many- Line Pseudo Continuum. . . Photoionization Efficiency Curves For Parent Ion (A) and Protonated Parent Ion (0).. . . . . Ion Yield vs. p—Dioxane Sample Pressure Under Full Illumination Of the Hydrogen Lamp. 177(0), 133 (A), 116 (n), 73 (0). 59 (9). . . . . Ion Yield vs. p-Dioxane Sample Pressure Under Full Illumination Of the Hydrogen Lap. 89 (O), 57 (A), 45 (D). 44 (o). 43 (V) . . . . Ion Yield vs. p-Dioxane Sample Pressure Under Full Illumination of the Hydrogen Lamp. 88 (e): 87 (A), 58 (e). . . . Ion Yield vs. p-Dioxane Sample Pressure Under Lyman-c (1215.6 A) Illumination. 177 (O), 133 (A), 116 (D), 73 (0)0 44 (.79 43 (4)- Ion Yield vs. p-Dioxane Sample Pressure Under Lyman-a (1215.6 A) Illumination. 89 (O), 88 (A), 87 (I). 45 (0) Photoionization Efficiency Curves for Parent and Daughter Ions of 1,4- Dioxane. 048302*(Al). 043102*( 0 )a Cs&0*(0), Csflso (53): 02350‘(O). CzHeO’ (a) , CzH30* (A3). Mass Spectrum Of l , 4-D ioxane , Obtained with Photon Energ Of 10. 20 eV (1215A), at the Specified Sample Pressures (in torr). Photoionization Efficiency Curves For Protonated and Daughter Ions Of 12- Crow-4. . . . . . . . . Photoionization Efficiency Curves For Daughter Ions Of 12-Crown-4. xi Page 91’ 93 95 96 97 98 99 102 105 116 117 OHAPTERI INTRODUCTION The precise determination Of heats Of formation for cations in the gas phase has assumed considerable importance in recent years. The most precise ionic heats Of formation have been Obtained from photoionization experiments.1 Photoionization mass spectrometry (PIMSP is primarily a spectroscopic technique in which the yield of ions from a sample gas is measured as a function of ion mass and photon energy. Thus, relative photoionization and fragmentation cross sections are measured, and they provide a wealth of information about the ions. Quantities such as ionization and appearance energies Of various ionization and fragmentation processes as well as vibrational and electronic energy level spacings Of molecular ionic states Often can be directly determined from PINS experiments. Modern photoionization measurementsM can provide information on these subjects with a precision Of about one kJ. PINS, coupled with a few other techniques,“ can Often fully characterize molecular ionization and fragmentation processes. The gas phase chemistry of ions has been an active area Of research ever since the advent Of mass spectrometryr The study of gas phase ions is Of interest because it can provide information about the structure of the neutral precursor molecule from the distribution Of fragment ions Observed in mass spectrometry, because of its importance in the study Of interstellar clouds of gas and dust, and also because it makes possible comparisons between the reactions of gas phase ions and those Of ions in solution. Two major differences between these two phases are the stability and de—axcitatiOn Of ions. In solution, ions are stabilized by salvation and any excited states are rapidly de-excited by collisions. In contrast, at low pressures gas phase ions are isolated and collision free. This thesis describes the photoionization mass spectrometric investigation Of ortho-, meta- and para-dibromobenzana, Of unlabelled and perdauterated 1,4-dioxana, and Of the macrocych crown ether 12- ~ crown-4. The interaction Of ionizing radiation with gases leads first to ions with a distribution Of internal energies. In competition with other deactivation channels these ions may isomerize and/or dissociate if their internal energy exceeds the thresholds for such reactions. Isomarization among dihalobenzanas has been Observed both in solution“ and in rare gas matrices." Reported here are the results Of a PIMS study Of 0-, m-, and p-dibromobanzene in the gas phase in which both ionization and dissociation are Observed. From the measured appearance energies and sunlisry thermochemical data, upper limits for the heats Of formation Of the parent ions and for CsHsBr‘ and CeHa’ have been Obtained which are more reliable than previously-reported values. The enthalpy of 3 formation Of neutral o-banzyne (Celia), a likely intermediate in many organic reactions,’ is also estimated. In the case of p-dioxane, threshold ionization or appearance energies have been measured for the following ions: 0.11.0.5 013.03% 0.11.0.3 0.11.03 0.11.03 0.11.03 0.3.03 0.3.03 0.0.0.4, 0.9.0.5 CaDeO‘ and CaDsO‘. In order to follow ion molecule reactions in this system, the mass spectra have been studied as functions of the sample gas pressure, the ion source rapellar' voltage, and the photoionization energy. Pardautarated p-dioxana was studied in order to more firmly distinguish PIEs of "adjacent" ions and to measure possible isotope effects on threshold energies. For 12-crown-4, the PIEs Of thirteen maior fragments have been measured and its fragmentation pattern has been compared with that of the proton-bound“ dimer Of p—dioxane. OHAPTERII AN OVERVIEW OF PINS The absorption of a monochromatic beam Of light by a homogeneous absorbing system is well described by the familiar combined Bear-Lambert law. One form of the law commonly employed in photochemical studies is I --.. = lo-cCL (II-1) Io where 1. represents the light intensity (or number Of photons) Of strictly monochromatic light incident per unit Of time at the front Of a column of a single absorbing species Of concentration C moles per liter; I is the light intensity per unit time transmitted through the column Of material, it is the column length in centimeters; and t (liter/mola-cm), the molar extinction coefficient, is a constant for a given pure absorbing species at a given wavelength and is a measure of the probability that the photon-molecule interaction will lead tO absorption Of the photon. The simplest Of the photoabsorption systems are those involving absorption Of light by atoms. The essential feature of the absorption Of radiation by atoms or molecules is that these processes are quantized; that is, transitions occur between disCrete, wall-defined energy levels. The simplest line spectrum is that Of the hydrogen atom. This spectrum consists, in the visible and near ultraviolet, Of series of lines whose separations and intensities decrease in a regular manner toward shorter wavelengths. Balmer was first to describe precisely the location and distribution Of these lines by means Of the simple formulau n1 2 A = G (II-2) Ina-4 where n: is an integer equal to 3,4,5..., G is a constant equal to 3645.6A and A is the wavelength in Angstrom units. Today this equation is normally written as 5 = R(1/23 - 1/n12) (II-3) where 5 is the wave number of the line characterized by integer n: and R is the Rydberg constant, equal to 109,737.32 cm".u Other series can also be Obtained by changing m in equation (II-3) from 2 to m = 1,3,4... and allowing m to take the value 2,3...; 4,5...; 5,6... respectively. Therefore, in general for the hydrogen atom one can write 3 = R(1/m2 - l/niz.) (II—4) where n: > m and both are integers. For the ground state Of the hydrogen atom n: = 1 and m can take any integer number from 2 to on. When n: approaches a high limit which is called the series limit, 3 = R/nz2 . (II-5) In principle, an infinite number Of lines lie at the series limit. Later, Bohr formulated the fundamental relation between the terms Of an atom and its structure. In general for spectra of hydrogen-like ions Bohr’s general formula is 3 = Rzz(l/n22 - l/Dlz) (II-6) where Z is the atomic number and R is the Rydberg constant. R = quaflch’ in which n is the reduced mass, e the charge of an electron, c the velocity Of light and h is Planck’s constant. By solving the Schrodinger equation for the hydrogen-like atoms one derives the same equation as the Bohr equation. Extending from the limit Of the discrote energy levels there is a continuous region of possible energy values, also called the continuum. In absorption the series limit corresponds to the separation Of an electron from the atom with zero velocity; that is, where the electron and proton are at infinite separation but at ”rest". If the transition takes place from the ground state Of the neutral atom to the ground states Of the ionized atom, the energy of the series limit gives directly the ionization energy. This for the H—atom is equal to 13.636 eV. Beyond this limit there is a region Of continuous absorption. This absorption corresponds to photoionization of the atoms and is continuous because any excess energy can be removed as kinetic energy Of the electron and positive ion. This process may be represented by hv = 18 + k (II-7) where IE is ionization energy of the atom and k is the sum Of the kinetic energies of the electron and the positive ion. The ion is always several thousand times heavier than the electron, so that, in order to conserve momentum, nearly all the kinetic energy must be taken by the electron. It is therefore usual to neglect the kinetic energy Of the positive ion in most discussions. The energy dependence Of the intensity of the absorption lines as they merge into the continuum is Of great interest. This behavior may be conveniently discussed in terms Of the (unitlass) oscillator strength, f, originally a concept Of the classical theory Of radiation. The relation between the absorption cross section Cs (cm') for a discrete transition and the oscillator strength for that discrete transition to an excited state n is given by”'“ r. = (no/see ) 1:. a» (II-8) where m is the mass of the electron, v is the frequency Of light, a is the electron charge, and c is the velocity of light; the integral is carried out over the entire absorption line. The cross section t in the photoionization continuum is related to df/dE, the density Of oscillator strength, by the equation“ s = (nah/mc)df/dE (II-9) c = (1.098x10'15a2eV)df/dE; df/do is of order“ 2m/3h||2 or s( 0) ~ 41'e2/3hc|lz ~ 10‘” ar". an order Of magnitude estimate Of do) for the continuum will be ((0) = 413/3(10'2x10'1°) ~ 10310-13 (:3 = 10 BI). Indeed the values Of absorption cross sections for molecules in the - continuum are in the range of one to many tens Of megabarns. Much Of the above discussion on hydrogen-like atoms also applies to all atoms and molecules. That is, at energies below the onset of dissociation and ionization, absorption will occur to discrete states. low-lying states are usually referred to as valence states, higher- energy ones as Rydberg states. To any given, precisely-specified quantum state of an atomic or molecular ion which can be provided by direct photoionization there corresponds one or more Optically-allowed Rydberg series." The energy levels of the series are given by En = IE - R/(n - 6)? (II—11) where E. is the energy Of the nth Rydberg level, R the Rydberg constant, n the quantum number of the level considered, 6 the quantum defect—a factor which accounts for penetration Of the Rydberg orbital into the ion core and is very nearly constant for large values Of n-and IE is ionization energy Of the molecule to which the series converges. The atomic values Of 6 for first and second row elements are characteristic Of the orbital angular momentum Of the leosely-bound (Rydberg) electron and may be quoted to :t 0.1 as:“ orbital: g - p d f quantum defect (C to Na) 1.1 0.7 0.0 0.0 quantum defect (Si to Ar) 2.0 1.5 0.4 0 The extension Of these quantum defect rules to small molecules has been made mainly by Lindholm.“ The general rule for molecules composed of second row atoms is that ns orbitals are highly penetrating and so have values Of 6 near 1, np orbitals are less penetrating and require a 6 of about 0.6, and nd orbitals are nearly nonpenatrating and have 5 very close to zero.“'” Experimental Rydberg state energies can be utilized with equation (II-11) to Obtain very accurate ionization energies. Photoelectron spectroscopy (PES) provides an initial estimate Of the series limit. The location Of the first excited state Of ‘ an ion is also know from PBS. . The quantum defect, which is a primary consideration in the assignment Of a Rydberg transition, can then be calculated by using the photoelectron ionization energy of that orbital in the Rydberg equation. However, another Option is to fit a series Of transitions to the Rydberg equation with the ionization energy and the quantum defect taken as adjustable parameters. KINFIT 4, a general non-linear, least-squares curve-fitting computer program“ can be used for this purpose. The quantum defect can be assumed tO be constant and independent Of n; the error that results for 1125 for ,a molecule composed of second row atoms is negligible.” Autoionization in Atoms For all systems more complicated (i.a. having more electrons) than a hydrogen atom, there exist states whose energy Of excitation is greater than the first ionization energy. In atomic systems these states can be produced in two ways: two-electron excitation and inner-shell ‘ excitation. An example Of the former is the transition“ He(ls)3thv + He(2s12p1) This transition needs 60.1 eV, whereas the first ionization energy Of He is 24.6 eV, see Figure II-l. An example Of inner-shall excitation is ”(18)3(23)3(29)'(3I)2(3p)° + hv -. ”(ll)’(28)2(29)°3s(3p)‘np requiring an energy between 26.5-29.24 eV for n = 4 to infinity. As to be expected , photon energies required for inner electron excitation are greater than for normal outer electron transitions and these absorption lines are to be found in the vacuum ultraviolet region of the spectrum converging to series limits which correspond to excited ionic states, See Figure II-2. The first ionization energy Of Ar is 15.76 eV. Hence, one can consider a Rydberg series superimposed upon the ionization continuum and converging to the inner-shall excited ion at the 29.24 eV limit?“ see Figure II-3. The ground-state configuration Of the Ar’ ion is split by spin-orbit interaction into a 'Ps/x ground state and a 'Pm state lying 1432 cm" above the former. Optically-allowed Rydberg series converge to both limits. Members of the series converging to the higher 'Pm limit can autoionize if they lie above the 'Pm limit. This is shown in N (I ABSORPTION cosmocm tom" AT STP) Figure II-l O '9' IOO 8 O 70 65 ELECTRON VOLTS 60 55 10 1 ‘ I l I l I l I 160 I90 zoo 'zuo 220 230 240 WAVELEMTH (WM) Autoionization of Two-Electron Excited States in Bali‘- (from reference 24). 11 C .9 .5 ....... O {-3 /—:outoionizotion (l . g //////j/ fi ground state of on . g .9. .3 U _x _— 2 a U 8 hv g "I! § .3 o U > h E "' ground state of atom Figure II-2 Autoionization-Inner Electron Transitions (from reference 3) . 12 02 . . e «to _ man :uooo —H . ’a'. la; 4 5 .- ;! BOOL K 1 l . 4 430 g I ”I 5 600- * I I” ‘ O'imb) y. 15; l I Ill «20 I III a, 200 I III U“ 4° I III . 27.0 ‘ 4“ 75““ ERERGYW) Figure II-3 Autoionization Resulting from Excitation of an Electron from the (inner) 3s Orbital of Argon to up Levels (from reference 25) . 13 Figure 114. X-ray and Auger spectroscopy are based on inner—shall excitation. These two processes are shown schematically in Figure II-5. These kinds Of ionization, in contrast to direct ionization, are called autoionization. Autoionization is an indirect process. Atoms and molecules are first produced in a neutral excited state with energy above an ionization limit, and than spontaneously emit an electron: M + hv 4 M" 9 M+ 9 e. The first step is a resonance process; 4 it can be brought about only by light Of the correct energy and the nature Of accessible excited states is governed by the normal Optical selection rules. Autoionization is the second step, and as no radiation is involved it follows the monopole selection rule that the symmetry species Of the autoionizing state and final state Of the ion plus free electron must be identical. Autoionization in Molecules The absorption Of ultraviolet radiation by molecules is Observed to be considerably more complex than that of atoms. For atoms, interaction with a photon produces either excitation or ionization. If a discrete state in the atom can have an energy identical with an ionization continuum, than, under .certain circumstances, excitation followed by a radiationless transition into the continuous state may take place, leading to autoionization in the photoionization absorption continuum. For molecules, on the other hand, several other processes are also possible which can mask or even completely obliterate the photoionization processes. Thus the interaction Of a photon with a molecule can lead to excited electronic states which in turn dissociate or ionize in competition with direct photoionization. Such effects include predissociation and ENERGY.eV 5.977 I5895 l5.8l4 @754 'I I I ‘r’ I I 1’ I l I I Ar. 2P": Ar. ng/z 15.937 W 5.759 W h A ‘ FI- P a 3" [1" . _i _ l . l I l.‘ _ ,, A I] - g ' ' - P “NM“,JI‘ }\'\///‘.,/ ELI/‘1 _ - R s.... _(_ Figure II—4 Autoionization Converging to the Higher Energy Spin- Orbital Component of the Ground State Configuration of the Ar’ Ion (from reference 24). Figure II-5 15 EXCITATION l —.— Ln ——.——.-— h—. K —o—4— [DE-EXCITATIW { x-mw gm -e—o—+—o- LII! m LU —.—.—— LU r- u] ‘—+—O——’ L; ——.—.—-“I_ \ \ _ I IN ' \‘s Kai : K21Lm\ l l K + + K -e———l- Two alternative modes Of atomic De—Excitation Which Follow Creation Of an Inner Shell Vacancy. (From E. Siegbahn, et a1.; ESCA Almquist and Wiksells, Uppsala, 1967) . 16 preionization, caused by the presence in the same energy region Of overlapping repulsive potential curves. Photoionization Of the molecule may also leave the ion in an excited state, leading to the process Of fluorescence as the ion decays to its own ionic ground state. The number Of excited states Of a neutral polyatomic molecule is large because there are many electrons in many orbitals to excite, and the number Of possible combinations is therefore considerable; in principle, their number is infinite. The number of states Of a polyatomic ion is even larger because the removal Of an electron from any one Of the orbitals can be accompanied by simultaneous excitation of the remaining electrons, thus increasing the number Of possible combinations. Connected with the existence of such a large number of excited states in the neutral molecule and its ion is the phonomenon Of autoionization. Among the excited states of the neutral molecule may be some that will be energetically degenerate with one or more ionic states; that is, a neutral molecule will contain energy in excess Of its first ionization energy. Such states have been called "superaxcited" by Platzman.“ If the potential energy surfaces for the energetically degenerate neutral and ionic states cross, as is likely, the superexcited state may autoionize, i.e., ionize by a radiationless transition; this is a resonance process. Howaver, such a potential energy surface crossing is relatively slow, occurring in approximately 10.11-1042 sec, which amounts to the period Of one or more vibrations. By contrast, direct ionization is essentially immediate (lO'u-IO'“ sec). The excited (neutral) states that autoionize may be those in which two or more electrons are excited, or an inner-shall electron is excited, or in which only one electron is excited but there is at the same time appreciable 17 vibrational excitation. This latter case appears to be the dominant one in polyatomic molecules?" the decay into an ion plus. electron is caused by the exchange Of energy between electronic and vibrational motion, which Of course constitutes a breakdown Of the Born-Oppenheimer approximation. Autoionization mechanisms in molecules are classified according to the nature Of the molecular ion excitation energy which is exchanged for kinetic energy Of the ejected electron. If the excitation energy actually used is in the form of molecular rotation we speak Of rotational autoionization, meaning that only the rotation quantum numbers Of the core change in the transition from molecule to ion. If the vibrational quantum number(s) change we have vibrational autoionization, while if the electronic state Of the core changes we speak Of electronic autoionization. Many excited molecular states are energetically capable Of autoionizing by two or three of these mechanisms which are in competition with one another. The forms Of autoionization which involve nuclear motion rather than electronic motion are generally less rapid than electronic autoionization, and are Observed most frequently in molecular Rydberg states which have electronically ground-state cores. Rotational autoionization has been characterized only in Hz,” while vibrational autoionization has been studied, for example, in 83"“ and non-um in detail. Figure II-6 shows electronic autoionization Of a diatomic molecule,“ and Figure II-7 shows vibrational autoionization in a diatomic molecule.” In Figure II-7, vertical lines from M: to upper states indicate excitation processes. Curved arrows labelled A, B, and C indicate respectively: a b = -3 inter-Rydberg transition from M2‘ to Me”; a D = -1 transition from MP to Ms“ + a; and a two-step energylcv Figure II-6 18 20 In - s ./ \ : / -\ [I I/ : /' . / ‘ 3 / IO 2T0 3.0 mt Autoionization (Thick-Line) and Direct Ionization (Dashed Line) in the Oxygen Molecule (from reference 35). l9 M+M*+eg DO. M+M.. M+M M+M Figure II-7 Schematic Diagram of Potential Curves of Electronic States and Vibronic Transitions in a Diatosic Molecule (fro. reference 27) . 20 process involving a b = -2 transition from M2" to Mz‘, followed by a transition with a» = -1 from 312' to 143* + e. Dissociation and Predissociation Comparison of absolute photoionization with absorption cross- section measurements shows that ‘t/‘s, the photoionization yield, is less than unity and varies with photon energy to converge to unity at high energy. Dissociation into neutral fragments and fluorescence are the two processes that can effectively compete with ionization. However, fluorescence is normally considered as negligible since the coupling with the vacuum radiation field. is usually very small compared with other perturbations. Thus when considering the problem of molecular ion formation one should understand the effects of dissociative channels. For atoms a continuous energy range joins onto each series of electronic states and corresponds to the removal of an electron (ionization) with more or less relative kinetic energy or, conversely, to the capture of an electron by the ion (recombination). In molecules, however, in addition there are continuous ranges that correspond to a splitting of the molecule into smaller components (normal or excited neutral fragments, or positive and negative ions), that is, which correspond to a dissociation. Such a continuous range joins onto the series of vibrational levels of every electronic state and is the only feature present if the electronic state under consideration has no discrete vibrational levels at all (repulsive state). As in a preionization, when discrete electronic states are overlapped by a continuous range of states that correspond to the separation of an electron (ionization), for molecules we have in addition the much more frequent case of overlap 21 by one of the continuous ranges that correspond to a dissociation into atoms or ions and that occur for every electronic state irrespective of whether it is a stable or an unstable state. In the schematic diagram in Figure II-8 the vibrational levels of the upper state B with v > 4 on are overlapped by the continuum of the lower state. In this case the . system can cross, by radiationless transition, from a discrete state into the continuous state which corresponds to dissociation; this process is called predissociation. Herzberg" divides predissociation into three types corresponding to the three forms of energy of the molecule (electronic, vibrational, or rotational). I. Overlapping of a certain electronic state (that is, of its vibrational or rotational levels) by the dissociation continuum belonging to another electronic state; radiationless transition into this other dissociated electronic state may occur. 11. Overlapping of “the higher 9:939:93] levels of an electronic state of a polyatomic molecule by a dissociation continuum joining onto a lower dissociation limit of the same electronic state; radiationless splitting off - of the particular atom or group of atoms (predissociation by vibration) may occur. III. Overlapping of the higher rotational levels of a given vibrational level of a molecule by the dissociation continuum belonging to the same vibronic state; radiationless decomposition of the molecule with no change of electronic or vibrational state (predissociation by rotation). Figure II-9 shows a schematic illustration of several possible types of photodissociation of a polyatomic ion; (a) electronic predissociation; (b) excitation to a dissociative vibrational level of the WillHHWUIUHHIWIIIWHW Figure II-9 ENERGY BOND - LENGTH COORDINATE A Schelatic Illustration of Several Possible Types of Photodissociation of a Polyatonic Ion; (a) Electronic Predissociation; (b) Excitation to a Dissociative Vibrational Level of the Ground Electronic State; (c) Vibrational Predissociation; (d) Direct Photodissociation on a. Repulsive Surface (fro- Reference 39). 23 24 ground electronic state; (c) vibrational predissociation; and, (d) direct photodissociation on a repulsive surface. As a window into the optical spectroscopy of gas phase ions, photodissociation has aroused the most widespread interest. To suggest why the spectroscopist has welcomed this concept so eagerly, one needs only to observe that the maximum concentration of mass-selected ions which can be assembled in' a spectrometer cell with current methods is of the order of 10'“M, which is discouraging in attempts to obtain direct optical absorption spectra. Since photodissociation methods deal usefully with ion populations of tens of ions/cm' or less, their potential spectroscopic importance is clear. Quasiequilibrium Theory The process denoted (b) in Figure II-9 is the direct excitation to a high predissociating vibrational level of the ground electronic state of a molecule. In most cases, this process is unfavorable because of the extreme violation of vibrational selection rules. However, these same levels can be reached as in, process (c) in twosteps by photoexcitation to a bound excited level; followed by internal conversion (dashed arrow) to the vibrationally hot ground state. It appears as if this internal conversion or vibrational predissociation mechanism dominates the dissociation of the great majority of the larger polyatomic ions so far studied.” This situation is usually well described by the energy randomization hypothesis which forms the basis of the quasiequilibrium theory (GET) of unimolecular decay.“ A major advance in our understanding of unimolecular dissociations was the application of this statistical theory (QET) to reactions occurring in a mass spectrometer. At about the same time and 25 independently, Marcus and Rice“ developed an important version of the old RRK‘M' theory, which is now referred to as the RRKM theory. Although the two theories are identical in their fundamental statistical assumptions, the QET expressed the dissociation rate k(E) as a function of the internal energy, E, while the RRKM theory went beyond this form to the thermalized rate, k(T), which is more appropriate to the thermalized neutral experimental systems. Another approach is the Slater“ theory which characterizes a molecule by a potential function having only quadratic terms and internal normal modes of motion which are orthogonal to one other, so that the system does not relax through its vibrational manifold before decomposing. The question of which is the more correct theory of unimolecular dissociation was subsequently resolved in favor of statistical models for molecules having lifetimes longer than 10'“ s.“ If the internal energy is statistically distributed among the vibrational modes both in the excited polyatomic molecule and in the activated complex located at the saddle point of the potential surface, the dissociation rate of a molecule or ion with internal energy, E, and an activation energy of E. is given by equation“ n—n. - k(E) = c z p¥(¢)/hp(n) (II-12) t=o where 9* and p are density of vibrational and internal rotational states of the transition state and the molecule or ion, respectively, s is number of equivalent ways the molecule can dissociate, and h is Planck’s constant. The numerator sums over the number of energetically allowed ways of passing through the region of the transition state. This is because E - E», the energy in excess of that required to pass over the transition state, can be partitioned among the vibrational and internal 26 rotational activation energies for various reactions. In terms of phase space one can visualize this model as in Figure II-IO. For a molecular ion configuration with total energy E there will be some large number of combinations of vibrational quantum numbers (micro states). Each micro state is shown by a square in phase space and for each combination E as 21m (Vi + 1/2). The number of micro states is reduced in the transition state configuration (bottleneck) because only E-Eo energy is available for distribution among vibrational modes. In this model dissociation is a random walk in this phase space. Since the statistical assumption is that all steps between the squares have. equal probability, therefore, the probability for dissociation is related to the ratio of the area of the bottleneck to area of the bottle, which is the same as equation (II-12). In more recent RRKM models a distinction between loose and tight complexes is made.“ A loose complex is one which involves simple bond stretching in the direction of the reaction coordinate and no coupling between radial and internal coordinates exists; distribution of the final products reflects the internal distribution of the complex. In a tight complex coupling between radial and internal coordinates affects the final product energy distribution, and some kind of rearrangement happens before the fragmentation. Although there is extensive experimental support of the fact that ion yield curves are smooth functions of energy alone, that is the molecular ion has reached a state of microcanonical equilibrium, a few counter examples are known which are referred to as "isolated state decay"." CRaOH’ shows both behaviors.“ Most of the ions are produced by a statistical RRKM/QET mechanism, except the 033* ions. This results from a branching in the reaction path brought about by an avoided crossing Figure II-lO 27 Potential Energy Surface Showing a Random Walk in Phase Space leading to Dissociation of a Polyatoaic Ion or Molecule (fros reference 47). 28 between two potential energy surfaces. One of the reactive paths gives rise to GET, whereas the other leads directly to the 033* + OH dissociation asymptote, Figure II—ll. This path which is called ”isolated state decay" takes place by a fast, diatomic—like, non-statistical mechanism which escapes energy redistribution. It has been found” that for polyatomic ions with lifetimes from 10" to 10‘u s randomization of energy takes place prior to decomposition. The early work of Wolfgang” and Henglein“ showed that the time scale for energy randomization in the collision complex in ion-molecule crossed beam experiments is about 10'u s. Kinetic Shift Measurements Metastable decompositions in a mass spectrometer are those which occur after the parent species has been accelerated out of the ion source, and hence, at a later time than usual. If, as is normally the case, the parent species is not homogeneous in energy, and if rate constants for decomposition increase with increasing energy, then those ions undergoing a metastable decomposition will have had less energy than those) decomposing in the ion source itself. That is why the appearance energy (AE), of a metastable ion gives a better estimate of the thermochemical threshold than does the AE of the ion itself.“ The difference between these two appearance energies has been referred to as the "measurable part" of the kinetic shift."I Since even metastable. decompositions need some excess energy to be observable, the translational energy release of this metastable decomposition must be measured to obtain this additional kinetic shift.“ Franklin and his Figure II-ll 29 I / CH§+0H '. \ ‘\ n2 1 \ A A \. I“\ I I ’v * ;/\-_-"}‘ CHZOH .H I \\ I, Fragnentation of the Methanol Cation Showing Both Statistical Deco-position [CRzOH’ + ii] and Non- Statistical Dissociation Through an Isolated Excited Electronic State [Cfla’ + OH] (fros reference 48). 3O coworkers“ have proposed the following relation for linking translational energy release is with total excess energy E a. = E/(0.44N), (II-l3) in which N is the number of vibrational degrees of freedom of the dissociating ion and 0.44N is called the "effective number of oscillators." Kinetic energy release is easily detectable due to the broadening of the corresponding metastable peak in the mass spectrum. Beynon” has shown that the average value of the kinetic energy released, T, is related to the width of the peak by the expression, d = (me/IifluT/eVP/Z (II-14) where d is expressed in _a.m.u. m1 and m are the precursor and daughter ion masses, V the full accelerating voltage the metastable ion crosses before decomposing, and u = (mi-mavma. Ionization Threshold Laws For the first ionization and multiple ionization, Wigner“ and then Geltman“ theoretically showed that the energy dependence of the photoionization and electron impact cross-section in the neighborhood of the threshold follows a power law of the form: s( z) = C(E-Eo)"’1, (II-l5) where E is the ionizing photon or electron energy, E. the threshold energy, 0 a constant, and n the total number of outgoing electrons for the ionizing process.’ According to this equation photoionization will give rise to a step function threshold, Figure II-12b, whereas linear threshold laws are expected for direct electron impact ionization, Figure II-12a, and for photoexcitation and electron attachment, which are resonance absorption processes, (Figure II-lZc). Figure II-l3 shows experimental cross sections near threshold for (a) electron impact 31 . 3a genom— Iruhv guusunoxoouoam. 3v .so«uss«so«oaosm 3v 3.5.5003 .3 seduced-sou aces“: Asv sea-809$ seas-.338 poses—Sousa haves—59$ new soda-Eon nouaoemlssoao Season—F 3030.308»: Amy >0mmzw wihowsoma Nun: Par. 7: By E (3)0 32 . 3H ooseuomou lost seek—003 e3 5:0: mo .uoanousaou usage: Ace .oesxo oases: ea nosuuueaosouoaa 3c .lium no 83833 nous-H 8.583 any new ouoasosAh use: sensuomlssoso ususslwsoaxm Ace «at: 9:5: “>3 >05sz 3:885 mm me we me am _ mm mm em _ a a _ q _ a «— (3)0 33 ionization of helium,“ (b) photoionization of nitric oxide,“ and (c) double ionization of helium by electrons,“ which all agree well with equation (II-l5). Although these experiments are in accord with the theoretical predictions, other studies indicate quite different forms.” Deviations from the step-function threshold law in photoionization are due to numerous factors. In general one can divide these causes in two parts, internal and external. I. Internal Causes (a) Autoionization. Since radiation is absorbed in a transition to a discrete state of the neutral, autoionization is a resonant process and ions will be produced only at the energy of the transition. Autoionization often appears as a sharp, peak-er structure superimposed on the normally featureless direct ionization continuum. However, autoionization exhibits a great variety of asymmetrical line shapes and may appear as a "window resonance" in which there is a decrease or dip in the photoionization cross section.“-'°’“ The lifetimes of autoionizating levels can vary a great deal, ranging from ~lO'“ seconds to ~lO“ seconds.“ For states with short lifetimes, the autoionization structure .may be so broad that it is indistinguishable from the continuum for direct ionization. Autoionization structure on a photoionization efficiency curve (PIE) may also be broadened by emission, in which case it may also be indistinguishable from the continuum. (b) Dissociation and Predissociation. In interpreting ionization efficiency curves, it is necessary to remember that what is measured is 34 the relative amount of various ions present several microseconds after electron or photon impact. In general, this relatively long delay makes the mass spectrometer a much more sensitive instrument for the detection of predissociation than an optical instrument. Predissociation is detectable in a mass spectrometer if the lifetime of the predissociating state is about 10" sec or less. Figure II-14 shows a strong predissociation leading to fragmentation, which causes a dip in the PIE curve.” (c) Internal Energy Distribution. It is frequently possible to prepare an ion with little or no internal energy by photoionization at a wavelength corresponding closely to the ionization threshold. Smaller molecules that are nearly all in the vibrational ground state at room temperature can usually be ionized to give an abundant yield of parent ions which are also nearly all in the vibrational ground state, unless the appropriate Franck-Condon factor is very small. In the case of larger molecules, which contain more thermal vibrational energy, the parent ion can usually be prepared with about the same amount of vibrational energy at threshold. Some "cooling" of the ion can be achieved by photoionization at "hot band" wavelengths corresponding to energies below the ionization energy, but this is usually not very significant.” Figure II-15 shows the internal energy distribution of benzene at 25°C. (d) Franck-Condon Effect. If the Born-Oppenheimer approximation is assumed, p is independent of the nuclear coordinates, and the total time-independent wavefunction is written, v= were (II-l6) Figure II-l4 ENERGY 35 A8 lNTERNUCLEAR DISTANCE Potential Curves for a Diatosic Molecule AB, Showing the Effect of Predissociation on the Electron or Photon Ispact Spectrus (fro. Reference 67). 36 QB 04- 03'- Pit) 02(- llllnllll......---- l l l l Figure II-15 OJ 02 OJ 05 05 03 OJ ENERGY,(¢V) Internal Thermal Energy Distribution for Benzene at 25°C (from Reference 80). 37 where the subscripts signify electronic, vibrational, rotational, and electronic spin wavefunctions. Direct ionization is exactly similar to an electronic transition to a neutral state, except here the final state wavefunction includes both the ion and the photoelectron. Therefore, for a dipole transition one can write: =<¢~ ’w ’u ’s In In“: ”it"s” (II-17) where u is the dipole moment operator, and prime and double prime superscripts are final and initial states, respectively. The total dipole moment is divided into two parts, one due to the electrons and the other due to the nuclei, i.e. u = is»... Since the electronic wavefunctions are orthogonal, (u’ln'fio, equation (II-l7) will reduce to: = (II-18) The separate electronic spin integral is based on the approximation p. = my, where it is assumed that the spin-orbit interaction is small. This approximation is good for light diatomic molecules, but it breaks down for heavier species, for which spin-orbit interaction is strong. For a spin-allowed transition (p’ I.” is equal to one, i.e. B = O. The rotational integral normally gives the general selection rules: AI = 0, 1'1 and M = 0,11. The product separation of the two remaining integrals is based on the assumption that p: = p (q, 0.) .(Q), where q and Q are generalized electronic and nuclear coordinates; Q. corresponds to a fixed nuclear separation at roughly the equilibrium nuclear distance in one of the two electronic states. Based on these assumptions, the first integral of equation (II-18) is called the electronic transition moment matrix; the second is the vibrational overlap integral, whose square is called the Franck-Condon factor.“ Figure 11- 38 16 shows the effect of the Franck-Condon factor on the direct ionization threshold of a diatomic molecule.“ The photoionization cross-section curves for H20 and CH4 shown in Figures II-l'l and II-18 illustrate two quite different cases. The data for H30 show a rather intense step at threshold corresponding to the abundant? production, by direct ionization, of the parent ion in the vibrational ground-state. The data for CH: show a very low probability for forming the ground state of the CH4‘ ion. The reason is that the equilibrium configuration of the ion is considerably distorted from the tetrahedral form of the neutral, and therefore the Franck-Condon factor for the transition to the vibronic ground state of the ion is very small.“ In the case of CH4, the electronic state of the ion is orbitally degenerate and nuclear displacement will occur that destroys the orbital degeneracy. Such displacement is called Jahn-Teller or Renner (for linear molecules) distortion.“ II. External Causes (a) Temperature. The effect of thermal energy on the photoionization threshold has been theoretically investigated by Chupka." He provides .an elegant demonstration that the thermal shift of such a photoionization threshold to lower values is just the average thermal energy of the neutral molecule, provided that the ionization efficiency curve for the molecular or fragment ion is linear in photon energy at absolute zero. The importance of thermal energy for the interpretation of appearance energies of fragment ions determined by photionization techniques has been recognized for some time."'" Methods employed 39 \/’ \ >- \ I , —- g \\/ I g M:: cum. on = \/0.III7I an I. = \/-°N°" M i r: 2 \/ \/0.1207$ m mo 0;: o.“ olIo 0.): 0.“ (NO on 0.1“ 0.1% on 2.20,; l' Ill'" oI'Jai:;;;I.oIII2Ia E I saucy Figure II-16 Franck-condon Factors and Hypothetical Direct Photoionization Yield Curves Expected for Snell, Moderate and Large Bond Length Changes in a Vertical Ionization (fro. Reference 48). 40 |.C) I I I 2 ° H o+ FROM H o t: 2 2 _ L) uI an a) ‘8 0‘ I: 1‘. s‘ . 0 xxx; I...\IV\ \. In I. Z he's. I- 'I“ ‘t a ....I Isl - s a: O 0 0M 1 l l2.00 l3.00 I4.00 PHOTON ENERGY.eV Figure II-l7 Photionization Cross-Section of 320 (froa Reference 15) . 41 I T I f * .~ CH4 :- s z .s V‘I’“ 9 J p. U 3 m 0. O "’ : . U) 0 ' . m 0 o o , u -' .r‘ > : .I I- ' o < .’ : .I , 1‘, DJ 0 e: : l 0.." l2.5 l3.0 ./ L—‘LL l I I l2.0 I4.0 ‘ l6.0 PHOTON ENERGY, eV Figure II-18 Photoionization Cross—Section of CHI 15). (free Reference 42 to minimize thermal contributions include: decreasing the temperature of ion-source to as low as 118K," or introducing the parent molecule as a supersonic molecular beam" with rotational temperatures of just a few degrees above absolute zero and also fairly low vibrational temperatures. (b) Pressure. When the pressure in the ion source is high, so that collision may occur before an ion leaves the source, 'the apparent ionization energy may shift. In the case of p-dioxane (which is discussed in Chapter (V), this shift is toward lower energy. In general, molecules which are excited by photons close to their ionization energy can lose an electron by collisionally-assisted ionization." (c) Instrumental Effects. “ Two examples are the slit effect and the kinetic shift. The former is due to the finite slit width of the photoexcitation monochromator. Its treatment has been discussed by Rider." The latter 'is defined as the excess energy which is required to produce a dewctable yield of photofragmented ions on the time scale of the mass spectrometric measurement, which is about 10" to 10" seconds.""" According to Rosenstock, et al.," in order to observe a fragmentation threshold by mass spectrometry it is necessary that the fragmentation process occur prior to departure of the parent ion from the ion source where it typically resides for ~10“ sec. If the rate constant for decomposition is less than ~10‘s“ no fragments will be detected. Additional excitation energy must be supplied to increase the decomposition rate. Therefore, the apparent threshold energy for fragmentation will be an upper limit of the true activation energy for dissociation. Chupka" has pointed out that for larger molecules the 43 kinetic shift effect is offset to some extent by the distribution of internal thermal energy of the molecule, all of which is assumed to be available for decomposition, in addition to the excitation energy imparted in the ionization process. Vestal” has made calculations which indicate the magnitude of the kinetic shift for a number of typical fragmentation processes. The magnitude of the kinetic shift, apart from rate constants, also depends upon the value of the residence time and the sensitivity of the « mass spectrometer. Rosenstock et al.“ have emphasized that for large molecule fragmentation processes the kinetics of decomposition lead to a gradual increase in fragment ion current, so that there is no well-defined threshold. Kinetic energy release measurements can be used to estimate that part of the kinetic shift and reverse activation energy (energy barrier for fragmented ions and neutral partners to revert back to the transition complex) which is converted to translational energy of the fragment ion.‘1 Recently time- resolved photoionization mass spectrometry in the millisecond range has been employed to study ultraslow unimolecular dissociations having rate coefficients k(E) S 10 e“ near threshold energies.“ Application of this method to the measurement of the threshold energy for fragmentation of bromobenzene to yield the phenyl cation and a neutral bromine atom is shown in Figure II-19. It is clear that the difference in appearance energy with respect to typical conditions is significant (about 0.3 to 0.4 eV). Such experiments can also provide very good estimates for the rate constants for slow decompositions. Heats of Formation for Gas Phase Cations The precise determination of heats of formation for cations in the gas phase has assumed considerable importance in recent years. With 44 .Awm oosezomom lea: economic-can mo semueuieamou—osm 93 cm 8:3 3.5a}: 3238 32.: 58:3 soundness". pee Assesses use 8.3.53 guises—xx vozoeomlolwa. :3 352.”. 208?. gm. V.N_ N N. ON. m : m _ mulum 03: :mnxmu. mzmo L 0m .18 3sz28 34m 0338.20 0. o. o. o. o. _ .o. < flmt a A. b 1 at ML 1 b .b _. q a 00— m m On A): >0".me 442mMFZ. 45 the development of experimental techniques for measurement of equilibrium constants associated with proton-transfer reactions among positive ions, extensive networks of relative gas phase basicities and proton affinities have been compiled.” The assignment of absolute proton affinities relies on the use of suitable reference standards which are usually derived from heats of formation for positive ions. However, these are still not firmly established, with uncertainties in appearance energies, ionization energies, and supplementary thermochemical data giving rise to values over ranges of several kilocalories per mole." The study of proton-transfer reactions involving alkyl ions and alkenes“ has produced a set of relative ionic heats of formation which may assist in their absolute determination. The most precise ionic heats of formation have been obtained from variable-temperature photoionization experiments.‘ These can lead to reliable 0! values, provided that the auxilary thermochemical data at GE is available or can be estimated; in most cases this is the major source of uncertainty in Emacs). Extrapolation of data from experiments performed at room temperature is of limited application because of possible thermal decomposition effects; there is also the likelihood of errors being introduced as a result of extrapolation over a large temperature range. The best technique is to perform low-temperature experiments, although these are not always practicable due to problems with sample condensation and reduced sample pressure. It would thus be useful if a method could be developed to enable ionic heats of formation to be accurately estimated from room-temperature photoionization measurements. 46 For the ionization reaction A v At + e the adiabatic ionization energy (IE) at GE corresponds to the formation of a cation and an electron with zero translational energy, i.e. at rest. For the reaction at a temperature T, it is possible to consider the ”I of an electron to be zero (stationary electron) or 3RT/2, the translational energy (thermal electron). At 298K there will be a difference of 1.48 kcal mol" between the two differently-defined heats of formation (no... - so. = 1.48 kcal mol“ for an unexcited monatomic gas). The stationary electron convention has been used throughout the course of this work. For a generalized process 1 AB + hv + A+ + B + e the potential barrier to the reverse reaction is generally assumed to be zero. Constraints on the reaction (unimolecular decomposition) require that the initial translational momentum of the center of mass (CM) and the angular momentum about the CM of the precursor molecule must be conserved. If the unimolecular decomposition is sufficiently rapid that there is no kinetic shift effect, then the observed appearance energy (AEr experimental) corresponds to the adiabatic appearance energy (AE: adiabatic), which represents the minimum energy to form A‘ + B + e from AB at temperature T. The products of this process (At 4» B + e) are not formed at any well-defined equilibrium thermodynamic temperature because the reaction occurs in isolation (unimolecular decomposition)!“ In reality both reverse activation energy and kinetic shift are present; therefore, a reported appearance energy will be an upper limit to the desired adiabatic value. Figure II-ZO shows the A8 for the reaction M 4 F‘ + N in which all. is the experimental AE, Is is (0 Kinetic Shift ‘L__..‘ )e o h- 0 I: U .‘§ ‘6 2 o 0. Reaction Coordlnole _... Figure II-20 Schematic Diagra- Showing Kinetic Shift and Reverse Activation Energy in the Experiaental Observation of the frag-entation: M + hv -' E’ + N (froa Reference 86). 48 the activation energy, cI‘F is the excess energy, to' is the reverse activation energy and T1: and T' are some portion of :1: and car which appear as kinetic energy.“ From the experimental AE of the above reaction, HI“: of At at room temperature can be calculated from the following equation: mun-(Ft) = ARI- + Elm-(M) - mil-(N) (II-19) where all on." are at 298K. To convert this heat of formation to the heat of formation at 01!, one can use the following equation ”to = flu- + I do, d'l' (II-20) mp is the difference between the heat capacities of products and reactants. The N-F‘ bond energy can be also calculated. for the above reaction (M + hv v F‘ + N) by subtracting the IE of the parent ion from the AE of the M dissociation, i.e., INN-F") = AB (1“, M) - IE(M) (II-21) Complementary Experimental Techniques In addition to photoionization mass spectrometry, which is the subject of the next chapter, two other techniques which are widely used for ion spectroscopy are photoelectron spectroscopy (PBS) and photoelectron-photoion coincidence spectroscopy (PEPICO). The principles of the PBS and PEPICO have been described in detail in several monographs'm‘ and review articles;4| therefore, only an outline will be presented here. A-Photoelectron Spectroscopy When a molecule is irradiated by light with wavelength shorter (energy greater) than the first ionization energy of the molecule, it will ionize: 49 M + hv + M* + e‘. PES is the study of the kinetic energy distribution of the ejected photoelectrons. The HeI resonance source (584 A°, 21.22 eV) has made it possible to obtain ionization energies in the region up to about 21 eV, including 6, r, and non-bonding electrons, by the technique which is sometimes called UVPES. This singling-out of individual molecular orbitals is the most important characteristic of PES. Since the photon energy used in PBS is fixed, the binding (ionization) energy of each molecular orbital, LI, can be obtained by measuring the kinetic energy of the ejected electrons: KE = hv — I). (II-22) The ejected photoelectrons are separated according to their kinetic energies in an electron energy analyzer, detected and recorded. The intensity of the photoelectron current plotted against ionization (binding) energy is called a photoelectron spectrum. The principles of photoelectron spectroscopy are illustrated schematically in Figure II-21 for a normal, closed-shell molecule. Since according to Koopman’s theorem, each ionization energy is equal in magnitude to an orbital energy, each peak in Figure II—21a shows an ionization energy from a different orbital of the neutral molecule as is shown in Figure II-Zlb. (In this figure fine structure such as vibrational structure, spin-orbit splitting, etc. has been ignored for simplicity.) The energy level diagram in Figure II-21c indicates that the ionization energies are given by the energy difference between the final states (the ionized states) and the initial state. Two types of ionization energy, "adiabatic" and "vertical", are defined as ionization from the vibrational ground state of the initial / I Cosme/sec N SO lonlzallon Elle rgv Figure II-Zl I1) (2) (3) (.3) “—‘F 'O— -..- '0.- M (2) __ ”V dre- -e—- .04» IF- -04> -eeF -e—- ll) ____ I.I,I, ‘\\I//' J.J. «ee- 4*.- M ~00- M (b) (c) Principles of UV Photoelectron Spectra. (a) Photoelectron Spectrus, (b) Electronic Configurations of a Molecule and its Ionized States, (c) Energy Level Diagras (fro- Reference 93). 51 state (v' = 0) to v’ = O of the final state, and from v' = 0 to vibrationally-excited levels of the positive ion with maximum photoionization probability, respectively. The schematic PES band in Figure II-22 is drawn such that the vertical-ionization peak appears at v ’ = 5. When there is no vibrational structure in photoelectron spectra, experimental values for the vertical ionization energies are generally determined from the positions of the individual band maxima. Ionization energies reported by PES are generally vertical ionization energies. B—Photoelectron-Photoion Coincidence Spectroscopy (PEPICO) The basic purpose of PEPICO spectroscopy is to study the reactions of molecular ions which contain a defined amount of internal energy. This is done using photoionization, by selecting events in which both the incoming photon energy, hv, and outgoing electron energy, KEe, are known:" M + hv 9 M’ + e'. The excitation energy of the molecular ions of interest, E.‘ can be measured if the original thermal energy content, Eli, and ionization energy, IE, are known E.» = hv - .KE. - IE «I Eu. (II-23) At very low temperature, Eu is assumed to be zero. The quantity defined by the experiment is actually thE., the energy transferred in ionization. Two distinct versions of PEPICO spectroscopy-termed fixed“ and variable” wavelength technique-have been developed. The two approaches are more complementary than competitive, as the following brief comparison reveals. In both of the two current versions of the PEPICO experiment, mass-selected ions are detected in delayed ‘L 113? >-. L 33’ ugl c7 c M' s 3; E l v' =0 M v' - 0 Figure II-22 Vibrational Structure of a Phtoelectron Band and Vertical and Adiabatic Ionization Energies (fro- Reference 93). 52 53 coincidence with energy-analyzed photoelectrons. The coincidence signal identifies photoionization events from which both the ion and the electron have been detected. In the fixed wavelength technique the sample is irradiated by a fixed photon wavelength, usually the Rs! (21.22 eV) live from a discharge lamp, and the molecular ion internal energies are chosen by varying the kinetic energies at which electrons are accepted by the energy analyzer. In the variable wavelength technique a special analyzer is employed for selecting those photoelectrons which have essentially zero kinetic energy and the internal energy of the ion is changed by varying the frequency of the ionizing light. The most promising light source for this kind of coincidence experiment is synchrotron radiation, but a standard laboratory light source such as the hydrogen many-line spectrum or the helium continuum can also be used. The most important difference between these two coincidence experiments arises because indirect autoionizing processes which emit threshold electrons can populate parts of the molecular ion potential energy surface in the variable wavelength option; these processes are not accessible by direct ionization. 54 CHAPTER III EXPERIMENTAL APPARATUS AND PROCEDURES A. Introduction The purposes of this chapter is threefold; to present a brief ' description of the MSU PIMS instrument, to document the experimental conditions under which the data presented in succeeding chapters were recorded, and to discuss experimental problems which were encountered in the course of this work. The instrument has been described in considerable detail elsewhere,“" and a brief overview is included here for completeness and to facilitate the discussion in this chapter. A diagram of the apparatus is shown in Figure III-l. The instrumental problems encountered during the course of this investigation were innumerable, many of them due to lack of experience with the instrument, whose construction was. completed several years ago. Many of these problems have been resolved; however, some are likely to occur ‘ again, and it is hoped that the following discussion will be of help to future users of the instrument. B. The Instrument 1. Light Source The light source is a Hinterregger-type windowless discharge lamp with water-cooled anode and cathode. The discharge tube is a 25 cm long x 4 mm I.D. water-jacketed quartz capillary to which the discharge gas is admitted at the anode after it has passed through a molecular sieve trap immersed in liquid nitrogen. A small pressure differential is P2-\. Pl EN Figure III-l lP Pl lE BA IT let .1— .1 BA-Baffle;. Eli-Entrance Slit; BX-Exit Slit; (IR-Grating; IS-Ion Source; IT-Ion Transducer; LP-Lanp; Pl-First Differential Puaping Port; PZ-Second Differential Pmping Port; P3—Monochroaator Puaping Port; P4-Saeple Chasber Pusping Port; P5-Ouadrupole Chadler Pulping Port; PT-Photon Transducer; OP-Ouadrupole; 08- Quadrupole Support; LE-Ion Dense. 55 56 maintained down the length of the discharge by pumping on the cathode end of the lamp; this stabilizes the discharge and helps eliminate sputtering of the cathode onto the monochromator entrance slits. The cathode vacuum pump is throttled with a needle valve. The rare gas continua can be produced by a high-power pulsed discharge in the pure gas. The M811 instrument utilizes a home—built, single vacuum tube, high-power switching circuit to pulse the output of a (Megavolt) “high voltage d.c. power supply. The switching circuit is in turn driven by a Cober A model 605P high-power pulse generator. The hydrogen many-line pseudo continuum is produced by a d.c. discharge in pure hydrogen at a pressure of 2-4 torr. The d.c. high voltage power supply used in the switching circuit, from which it is disconnected in this application, powers the hydrogen discharge. A 15009, 1500 W resistor limits the current drawn from the supply. The helium discharge produces light in a higher energy range than does the hydrogen lamp. However, the Roots blower which forms part of the lamp gas vacuum system was broken during the course of these experiments. Thus it was not possible to put 70 torr, which is the necessary operating pressure for helium, into the discharge lamp. Even for the hydrogen. lamp, the normal operating pressure of 3 torr . was too high. The pressure was decreased to 2 torr and the Megavolt supply was maintained at 2.75 111! and 700 mA. This gave a negligible decrease in the intensity of the hydrogen many-line spectrum and these operating conditions were used for all of the experiments described in this thesis. Ionization and appearance energies of p-dioxane and 12—crown-4 and their fragments fall in the region where the hydrogen pseudo 57 continuum is a suitable light source (9—13 eV). In the case of the teasers of dibromobenzene threshold PIEs of the parent ions are also easily measured with the hydrogen pseudo continuum, but with this light source only two of the fragments, CoHABr’ and Cell:. could be formed. Other fragments, either due to higher AE’s or lower intensities, were not measurable when the hydrogen lamp was used as the irradiation source. Even the benzyne cation (CsHlt) has very low intensity in the energy range of the hydrogen lamp, and it’s AE is very difficult to measure. 2. Monochromator The monochromator is a McPherson model 225, l-meter, near- normal-incidence instrument. The dispersing element used for this research is a concave osmium-overcoated aluminum grating, ruled with 600 lines/mm and blazed at 900A (Another grating, magnesium fluoride overcoated, ruled with 1200 lines/mm and blazed at 1200A is also available in the laboratory.) The grating has a reciprocal dispersion of 16.8A per mm. Two interchangeable entrance slit plates are available. Each plate has a blank port plus three fixed slit widths: lOum, 50pm and 100nm, or 100m, 300m, and 500m. For a given plate, the entrance slit can be changed without breaking vacuum via a knob located outside of the vacuum changer. Interchangeable exit slit tracks are also available, each of which has four fixed slit widths: 10um, 50pm, lOOum, and 300nm, or 100nm, 300nm, 500nm and lOOOIIm.' Any one of the four slits on a given track can be selected from outside the vacuum via a knob-chain- gear system." The procedure for interchanging entrance and exit slit mounts is given in the PIMS operations manual. The grating is driven 58 by a stepping motor which can be operated manually or with a computer. The grating should be cleaned when the helium lamp spectrum shows significant scattered light below 500A To clean the grating, first rinse with Freon 11 (CFCla), followed by high purity methanol. The methanol rinse is best done by simultaneously applying the methanol from a squeeze bottle and drying the grating with low pressure (10 torr), clean, compressed air, working from top to bottom. The removal, cleaning, replacement, realignment and focusing procedures are listed in the McPherson monochromator manual. Because realignment of the grating requires adjustment of many parameters, a great deal of time can be saved if the grating is cleaned in situ. In this case, care should be taken to direct the cleaning solvents only on the grating (some absorbing material--such as Kimwipesncan be placed around the bottom of the grating mount), and sufficient time thereafter (about one day) should be allotted to allow vaporization of the solvents from the loosely-covered monochromator under moderate nitrogen gas flow. 3. Interaction Region and Sample Inlet System The ion source is a cubical stainless steel box, one inch per side, located one centimeter beyond the exit slit of the monochromator. It has two rectangular apertures on opposite sides to enable transmission of the photon beam. The widths of the rectangular apertures are adjustable to allow uninhibited passage of the photon beam with any chosen exit slit. On the horizontal plane perpendicular to the optic axis is a one-quarter inch diameter ion exit hole. The ion aperture is covered with a high—transmission fine wire grid to minimize penetration into the source of electric fields from the ion focusing lenses located 59 just outside. The repeller, an electrically-insulated stainless steel plate, is positioned inside the ion source across from the ion exit aperture. An adjustable voltage is applied to the repeller to accelerate the ions from the source into an ion lens system. The sample is admitted through a tube located behind the repeller and the sample pressure is measured through a tube in the top of the ion source. The sample pressure is controlled with a Granville-Phillips model 203 adjustable leak valve and measured with a Datamatrics model 1014-A Barocell capacitance manometer. The leak valve and manometer are outside of the high vacuum. 4. Photon Transducer The photon transducer is a sodium salicylate phosphor whose emission intensity is measured with an RCA 8850 photomultiplier tube. Because sodium salicylate possesses a nearly constant quantum yield throughout the helium continuum, and that portion of the hydrogen many-line output of interest in this work, it is superior to ordinary (bare) photomultiplier dynode chains. The sodium salicylate emission intensity can be accurately measured with a standard photomultiplier and the emission band does not shift with the exciting wavelength. The output current of the photomultiplier is amplified with a Keithley 1800 current amplifier and converted to a voltage. Then it is digitized with a voltage-to-frequency converter and a computer-interfaced counter.“ When used with the Hopfield helium continuum, the sodium salicylate transducer detects a great deal of scattered light. This effect is corrected by automatically making repeated measurements of the scattered light during the course of an experiment, then subtracting this contribution from the observed intensities. The scattered light 60 problem is much less bothersome when the hydrogen lamp is used, since this source emits less visible and near ultraviolet radiation and since the many-line spectrum does not begin until wavelengths greater than 800A which is a region where the scattered light distribution has fallen off considerably. 5. Ion Optics and Mass Spectrometer The ions are guided from the ion source and focused onto the quadrupole mass filter with four electrostatic aperture lenses. The electrostatic lens system was supplied with the commercial mass filter, and has been modified by removal of the electron impact source. The mass filter is an Extranuclear 324-9 quadrupole mass filter, with rods 1.9 cm in diameter times 22.0 cm long. It is powered by a model 311 power supply, equipped with a model E I (mass range 0-200 amu) or a model 13 (mass range 0-400 amu) high Q head. The quadrupole (as well as the lenses and ion transducer which are connected to it) is mounted on a support which is easily moved several inches in a direction parallel to the quadrupole axis. This permits placement of the quadrupole at an appropriate distance from the photon beam for use with single or dual chamber ion sources of various depths. The mass of the ions transmitted by the mass filter is controlled by a dial on the front panel of the quadrupole power supply or by a voltage applied to a connector on the back side ”of power supply. To obtain a mass spectrum, the transmitted mass is scanned by a mass scanner which applies a voltage ramp to the connector on the back panel of quadrupole power supply; the mass scanner was designed and constructed by David Rider." 61 When the high Q head is changed or after about six months of continuous use of a given head, the mass filter needs to be "tuned up" to rematch the RF frequency of the quadrupole power supply and the high Q head. The detailed procedures are listed in the Extranuclear quadrupole filter manual. 6. Ion Transducer The ion transducer, located on-axis at the exit hole of the quadrupole, is a Channelth continuous-dynode electron multiplier, model CEM 4816, supplied by Galileo Electro-optics. The channeltron output is measured either in a direct-current or a pulse-counting mode. In the former mode the output is amplified with a Heithley model 417 electrometer. However, since the direct-current method has several disadvantages compared to pulse counting when the output current is small, it was used only for setting up an experiment. All data were collected with a pulse-counting technique, where the output pulses are amplified and discriminated with a circuit developed here. 7. Vacuum System In order to maintain the required pressure differential between the lamp (~10' torr) and the monochromator (~10“ torr), two stages of differential pumping are used. The monochromator is equipped with a McPherson model 820 differential pumping assembly, the first stage of which is- pumped with a 300 liter/sec Roots blower pump. (It is this pump which failed during the course of this work, thus precluding use of the He lamp.) The second stage is pumped by a 300 liter/sec, four- inch diffusion ejector pump, and the monochromator with a 2400 liter/sec, six-inch diffusion pump. The sample and quadrupole chambers 62 are also pumped with six-inch diffusion pumps, rated at 2400 liter/sec and 1800 liter/sec respectively. A baffle around the quadrupole also provides some measure of differential pumping of the sample gas in these chambers. I The pumping speeds quoted above are for untrapped pumps. All four diffusion pumps are trapped with Freon-cooled baffles to reduce the amount of oil back-streaming into the chambers; thus the actual pumping speed in the various chambers will be significantly lower. When the system was pumped down, the best pressure obtained was 9.0x10" torr. Nevertheless, these pumps are sufficient to maintain a pressure on the order of 10" torr in the monochromator with 70 torr of helium in the lamp (100 um entrance slits). The pressure in the quadrupole region can be maintained below 2x104 torr for a sample pressure below 2x10” torr, even with the relatively open ion source currently in use. 8. Interlock System The vacuum and utility service systems in the PIMS instrument 7 are quite complex. Even though each component in the system may be rather dependable, there are so many components involved that the possibility that something will fail in any given time period is quite high. Moreover, even if the individual vacuum components were completely reliable, the necessary electrical and water utilities supplied by“ the University are not. Thus occasional failures of the vacuum system are inevitable. If proper. procedures are not taken, the expensive components inside the vacuum chambers (the grating, channeltron, and quadrupole rods) may be irreparably damaged. All of these components are extremely sensitive to the presence of even very 63 thin contaminating films of pump oil. Thus a security interlock system which monitors the foreline pressure of all diffusion pumps, water pressure of the diffusion pump cooling system, baffles, lamp gas pressure, electrical part of the lamp, and all mechanical pumps has been implementedu to reduce the chance of catastrophic damage to the PIMS apparatus. The interlock system for this instrument was designed as a "stand-alone" system which is not computer operated; it can thus be used to protect the instrument between experiments as well as during them, allowing the instrument to remain safely under full vacuum conditions at all times. Even so, the interlock system is not sufficiently sophisticated to totally automate the pump-down procedure. However, it does mandate the correct sequence of operations during start-up, which greatly reduces operator errors--although it cannot eliminate them. When used to protect the unattended instrument, the action taken by the safety interlock system when an error condition is detected is very simple: any device which, if left on, could conceivably be harmed or cause harm to another device is deactivated. The vacuum system then remains shut down until an operator decides that it is safe to start stain- 9. Instrument Control and Data Acquisition The computer used to control the PIMS instrument during the course of this research was an antiquated PDP 8/F minicomputer (12-bit word length), which has 16K of memory. A PDP 8/M (IE-bit word length, 12K memory) is also available for PIMS experiments. Data and program storage peripherals include a dual floppy disc drive (Sykes 7200 series) and an interchangeable hard disc (WANGCO F—1221). Interaction between 64 the user and the computer takes place via a Heath Z-19 terminal or a Tektronix 4010 graphics display terminal. The data are collected by using a variable integration time technique which permits recording data of the desired quality in an optimum amount of time. The monochromator wavelength is stepped in preset intervals, and, after each step, ion and photon counts are accumulated until the datum has the desired signal-to-noise ratio. Photon and ion count rates, wavelengths and integration times are stored on a floppy or hard disc. The course of the experiment is continuously monitored 'via a computer-interfaced, stepped, stripchart recorder on which the real time PIE is plotted. Periodic measurements of "light" and "dark" photon and ion count rates are made at a reference wavelength so that the data can be corrected for sample pressure and instrument drift. Approximate corrections for stray light are made by measuring the light intensity at two wavelengths where the helium continuum does not emit. Final corrections of the data for sample pressure variations, instrument drift and stray light, as well as the final plot, are discussed in the procedure section of this chapter. C. Experimental Procedure This section presents a flow chart to show how to run a photoionization mass spectrometry experiment. The difference between this section and the operations manual is that this discussion indicates when and why to do each procedure, whereas the operations manual describes how to do it. The experimental procedure introduced will follow the order of an experiment. Sometimes several steps may be started at the same time; higher priority steps will be listed first. To continue an aborted run, 65 or to start a new experiment upon termination of another, the entire sequence may not have to be repeated. Under these circumstances, just skip unnecessary steps and continue to the next one. 1. Samples 1,4-dioxane was obtained from the J. T. Baker Company; para- 040003 was purchased in sealed ampoules from the Stohler Chemical Corp.; 12-crown-4 and ortho- and meta-dibromobenzene were obtained from the Aldrich Chemical Company Inc. Para-dibromobenzene was purchased from the Eastman Kodak Company. All samples were thoroughly degassed by repeated freeze-pump-thaw cycles and used without further purification. For the PIE measurements the sample pressure in the ion source was maintained below 5x10“ torr and 100 um monochromator entrance and exit slit widths were generally employed. For photon energy- and pressure-dependence studies, 500 um slits were employed. The sample pressure was measured with a calibrated ionization gauge which was directly connected to the ion source. 2. System Pump-Down If the PIMS instrument is completely shut down, the following procedure is required. Two parts of the overall system must be initially pumped down, the main part of the system (monochromator, sample and quadrupole chambers) and the lamp gas system. In the main part of the system, after all the diffusion pumps have been turned on and all the gate valves have been opened, the system pressure should be lower than 1.0x10" torr. Normally it is easy to reach 2.0x10" torr with a proper degas; sometimes the pressure can be as low as 9.0x10" torr. After the main chambers have been pumped down, the sample gas line 66 should be evacuated. The purposes of doing this are: ( 1) to pump out the air in the sample gas line, (2) to remove any residue of previous gas samples, and thus avoid contamination of the sample, and (3) to check for leaks in the sample gas line connection. Any leak in the sample gas line will allow air to enter the system, which will change the sample gas purity and lead to extra peaks in the mass spectrum. The lamp gas system is pumped by a mechanical pump only. This part of the system must be absolutely leakproof; this is tested by closing the valve between the line and the mechanical pump, and monitoring the pressure change via a thermocouple gauge. If the rate of pressure increase is less than 1.0x10" torr/min, the lamp gas system is considered sufficiently leakproof. Any leaks in the lamp gas system will contaminate the lamp gas, and produce dirty lamp spectra, including many sharp atomic lines. A dirty lamp gas system (especially when the helium discharge is used) will make data correction harder. The molecular sieve trap in the lamp gas system has to be cleaned periodically when helium gas is used. To clean the sieve, heat the heating tape around it up to ZOO-210°C and continue pumping until the pressure of whole systems return to normal. 3. Lamp Spectra The main purpose of running a lamp spectrum is to check the lamp system quality. Are the emission intensities as expected, and do the wavelength settings in the computer coincide exactly with several calibration lines? From a lamp spectrum, one can determine the condition of the lamp, the calibration of the monochromator and the purity of' the lamp gas. Different observed phenomena infer different kinds of problems. A lamp spectrum provides not only the information 67 mentioned above, but also information about the optimum voltage setting of photomultiplier, the scale of the amplifier, the photon signal, etc., which are parameters to be set before an experiment is initiated. The lamp spectrum can be generated with the grating rotation either under manual or computer control. 4. Mass Spectra Before photoionization efficiency. curves from a given precursor can be obtained, the mass spectrum of the sample gas must be acquired on the PIMS instrument. Photoionization of the sample is accomplished under full illumination of the helium or hydrogen light source. To obtain the mass spectrum, set the monochromator to the "central image", i.e., zero A which is offset by ~100A on the wavelength counter, whereby all of the light from source is reflected from the grating into the reaction region. If the entrance and exit slits are opened to their widest position, maximum photon intensity from the light source impinges on the sample gas. The mass spectrum serves several purposes. (1) It provides a test of the sample gas purity. Because most standard mass spectra are obtained under bombardment of 70eV electrons, one must be careful when comparing PI and EI mass spectra. Although there was generally good agreement between the mass spectra obtained in this work and those tabulated in the literature, the expected variation in the relative intensities of the peaks (both parent and fragment ion) was observed, and highly fragmented daughter ions produced at high ionization energies were sometimes missing from the PI mass spectra. If the photoionization mass spectrum shows several extra peaks compared to the standard mass spectrum, it is probable that the sample gas may be contaminated by some impurity, or that the sample gas inlet line has 68 _ a leak. (2) One can easily determine how many ions can be produced from a given sample gas by the light source. In principle, a PIE curve is obtainable for every ion revealed by a peak in the mass spectrum; the intensity of each peak gives a good idea about how difficult it will be to do that experiment. (3) It is convenient to change the conditions to start the next procedure. Simply change the entrance and exit slits to the desired positions. For a given photon energy range, the lamp gas (helium or hydrogen) can be maintained while PIEs are measured for several different ions from a single sample gas. 5. Experimental Set-Up After the mass spectrum has been obtained, the next step is the choice of an ion for which the photoionization efficiency curve is to be measured. After a particular ion is chosen, the mass filter is set to the corresponding mass region, and the ion optics are adjusted to maximize the ion intensity. Scan the mass in the appropriate region to confirm that the transmitted ion is indeed the one selected; this is facilitated if the ion of interest contains atoms which have characteristic relative isotopic abundance patterns. Sometimes improper adjustment of the ion optics will still allow ions to pass through the mass filter, but the isotopic ratios will not be correct. Once the desired ion is properly identified, reduce the resolution to broaden the peak and thus increase the ion signal. (However, the resolution must not be lowered so far that ions of different m/ z are transmitted.) Find a relatively smooth portion of the broadened mass spectrum and lock the mass filter at this m/z value. Then small drifts in the quadrupole tuning during the course of long experiments will not affect the photoionization efficiency. All the procedures described above can be done either by the computer or 69 manually. Hereafter, all the procedures will be done under computer control. Many parameters must be considered when setting up a photoionization mass spectrometry experiment; these parameters have a great influence on the resulting photoionization efficiency curve. It is necessary to do a quick experiment (or perhaps several) to help decide how to choose these parameters. From the quick run, one can draw conclusions about the following aspects of the final experimental photoionization efficiency curve: (1) the desired signal-to—noise ratio (consistent with time constraints); (2) the duration of the experiment; (3) the energy region which is to be scanned; (4) the wavelength to be chosen for the reference measurement; (5) the desired wavelength resolution; (6) the wavelength interval between successive data points, and (7) the sample pressure. For the actual experiment one also needs to decide: (1) how frequently and for how long to measure the background and reference signals, and (2) where and for how long to measure the stray light. After all these parameters are determined, one can start an experiment. 6. Data Correction Once an experiment is set up, data collection and storage are controlled with a PDP 8/M or PDP 8/F minicomputer. The periodic 4‘ reference measurement is made to correct for sample pressure variation and instrument drift. A shutter is closed in front of the monochromator exit slit under computer control so that the periodic background measurements can be made. Dark counts generated in the detectors are also stored and will be subtracted from the ion and photon signals, which are obtained when the shutter is open, in the later procedures. 70 The stray light measurement is made to correct for the stray light contribution to the photon data. After these three corrections, which are made at the department computer facility (PDP 11/34) or on a PDP 11/23 in the Leroi laboratory, a comparison of the ion signal with the photon signal is made to ascertain whether corrections are necessary for artifacts caused by the finite slit width, scattered light, or detector misalignment. (For more detailed information on these points, see reference 91.) After all of these corrections are made, the final photoionization efficiency curve is plotted out and the experiment is complete. 71 CHAPTIIV PHOTOIONIZATION AND FRAGMENTATION 0F o-, m- AND p-DIBROMOBENZENE A. Introduction In contrast to their monohalo counterparts, very little information is available regarding the ionization and fragmentation of the dihalobenzenes. Thus, although the ionization energies of the three dibromobenzene isomers have been reported from photoelectron spectroscopy (PES) measurementsflm' neither the energetics nor the dynamics of fragmentation processes involving the ortho-, meta- or para-dibromobenzene cations have been studied. The interaction of vacuum ultraviolet radiation with o-, m- and p-dibromebenzene has been investigated at low sample pressures in order to obtain thresholds for their ionization and fragmentation. In addition to monitoring the effects of the substituent geometry, the measurements can be combined with available thermochemical. data for the neutral (fragmentation products) to provide good estimates of bond dissociation energies and of heats of formation for the parent and daughter cations. The interaction of ionizing radiation with gases leads first to molecular ions with a distribution of internal energies. In competition with other deactivation channels, these ions may isomerize and/or dissociate“ if their internal energy exceeds the thresholds for such reactions. Isomerization among dihalobenzenes has been observed both 72 in solution‘ and in rare gas matrices.“ A photoionization mass spectrometric (PIMS) study. of o-, m— and p-dibromobenzene in the gas phase has been carried out, in which both isomerization and dissociation are observed. From the measured appearance energies and ancillary thermochemical data, upper limits for the heats of formation of the parent ions and for CoHaBr‘ and CcHdt have been obtained which are more reliable than previously reported values. The enthalpy of formation of neutral o-benzyne (CsHl), a likely intermediate in may organic reactions,“ is also estimated. B- 3229.19; Photoionization efficiency (PIE) curves in the 9 eV region for the ortho-, meta-, and para—dibromobenzene parent ions are shown in Figure IV-l. The ionization energies (IEs) for the molecular ions were determined by expanding the threshold region and extrapolating the straight-line portion of the appropriate PIE curve just above the ionization onset back to the baseline. The IEs of the isomers are distinct and distinguishable. An expanded plot of the para- dibromobenzene cation PIE in the 8.5-9.5 eV photon energy range is shown in Figure IV-2. ‘ Step-er structure is observed which is attributable to a vibrational progression in the ionic state; the vibrational separation is approximately 0.025 eV as 200 cm". The PIE curves of the CoHIBr’ and CsHI’ daughter ions generated by dissociative ionization of each of .the isomeric dibromobenzenes are collected in Figures IV-3 and IV-4. The appearance energy (AE) of a given fragment cation was chosen as the energy at which the PIE first (rises perceptibly from the baseline. Because of possible kinetic shift, Wavelength (Angstroms) 73 1420.00 1370.00 1 l ‘g 0 Ca e‘ .e a I. a... I .g a. I e . U o O D ’g 0 4'." “f A 0 mu‘ 5 V. '30 g e b '- .13: e’ A: 8 ' ‘ a .4: {an e .' ‘ 3 s‘. a} I. “a. LJJ . B ‘ c‘i : y" f I an 5 p0 Good" me a a a a! $0 #1 ’ n “W fiUIITITT'IUUTIIITITIIITIfiTlT—I—TTTITUI—TTITTIUIIITWITTTIYTTIU 8.70 8.90 9.10 9.30 Energy (eV) Figure IV-l The PIE Curves for the Ortho-, Meta- and Para- Dibromobenzene Parent Ions . 74 Wavelength (Angstroms) 1450.00 1350.00 L l _l l l l l L l J I J l l l I F If I T 7 l A I - I. - I A T l :1 T r U— U 2‘ A * .3 '5 'E . D L >\ e L. O s L- b r: e l» < U; v . l- 0. f. .1} doe—ll} IIIIIUII‘liimlI{IllnnhiUIIIIUUIIIqUIITWIIII‘IIU[TUTTI]HITTIWUITIIIIIIIIIUIIUW‘I[PU"IIIUII 8.50 8.70 8.90 9.10 9.30 9.50 Figure IV-2 Energy (eV) Fine Structure on the PIE Curve of P-Dibroaobenzene. 75 Wavelength (Angstroms) 1050.00 950.00 : l l 4. i I : l I I 1.1 :5 i I i A O D D O D O ’1}? “P9 ’1 ‘ .4: n ‘ O C 0 e 3 0 ° m: ' .>.~ . "ip 9. v ' .4: 8 . .' .D A L n a s, . 0.“: ‘ C be ‘ a’ 0.. A e’ .9 ‘ ° .9 f ' In “ 0’ $2/ a" A J 1 l l _l L L AL I L l A J l l 1 I—I I T 5 I I f | I Iw' I T I“ 11.50 12.00 12.50 13.00 13.50 Energy (eV) Figure IV-3 The PIE Curves of the CeHcBrt Daughter Ions Generated by Photofragaentation of Each of the D ibromobenzenes . Isomeric Wavelength (Angstroms) 76 1050.00 950.00 1 J l l l l L I l l 1 J I l I l 1 I 1 I l I I .W.. l r j I e .0 e n ‘ O I. 00 A 0 O A m e o :1 ° ° 5 c: ' ° u r 3 °. a A. .s >\ p 0 on ma L. a a E ' = : Q: 4.; O 15 °' .3: A“ 2 $ 0 *~ v 43 # e O f LIJ e e 5 ° . . . .l’ . .° 5 .r L L 1 l L J l l I I I l I I I I 11.50 12.00 12.50 13.00 13.50 Energy (eV) Figure IV-4 The PIE Curve of the CsHs’ Daughter Ions Generated by Photofraglentation of Each of the Isoaeric Dibroeobenzenes. 77 reverse activation energy and other factors,” this value must be considered an upper limit to the desired adiabatic AB. The IE and AE values obtained in this work are listed in Table 117-1, where the relative intensities of the cations from each precursor, obtained under full illumination of the hydrogen discharge and. verified by 15-20 all electron bombardment, are also given. In addition to other factors which vary little from one PIE curve to another the error in an extrapolated IE or AE depends on the available photon intensity. With the hydrogen lamp the intensity is high in the vicinity of the parent ion thresholds, and rather low in the AE range of the daughter ions. Wider slits were employed in the latter cases, which leads to larger error in the measurements. Also, the PIE curves of the parent ions rise more sharply and uniformly than those of the daughters, which permits more accurate extrapolation to the baseline. These factors are responsible for the lower error reported for the parent ion threshold energies. The threshold energies obtained in this work were combined with literature thermochemical data to obtain the room temperature enthalpies of formation for the cations studied. The identities of the neutral partners, the values of their heats of formation used in the calculations, and the ionic heats of formation determined therefrom are also listed in Table 111-1. The enthalpies of formation of the neutral parents were calculated from group additivity tables‘;"' those of Br and Br: are rather precisely known!” The bond dissociation energy for removing a Br atom from the benzene ring in camel-v can be determined directly from the difference in the threshold energies of the fragment and molecular ions. [Bond Energy = EMCcHIBrt) + mflBr) - 78 W(CsHeBrx’) = A.E.(CoHcBr*) - I.E.(CoHdBra*)]. These bond energies ‘ are given in the last column of Table IV—l. Within the photon energy range of the hydrogen lamp, the ion signal at m/z = 154 is negligible, and that at m/z = 156 can be accounted for by the natural distribution of ”C in the CsHIBr: precursors. There is no evidence of the production of CoHaBr' under our experimental conditions, probably because the photon energy is below the activation energy for HBr loss. Moreover, the absence of ions at m/z = 77 (CoHo’) indicates that the samples are not contaminated by CeHsBr, a possible impurity. C. Discussion The first ionization process in substituted benzenes involves removal of a I-electron from the ring; the value of the ionization energy is sensitive to the nature and symmetry of the substituents.” The vertical ionization energy of bromobenzene is 9.041 eV, and those of the dibromo derivatives are slightly lower. The order and the numerical values of the IEs of the three isomers determined in this PIMS study are compared with those obtained by PES‘IM.» in Table 111-2. The agreement is excellent. Photoionization gives IEs about 0.05-0.1 eV lower than the vertical photoelectron values. Although the ortho- IE is lower than that of the meta isomer in each report, their difference is within the combined experimental errors of the (determinations. However, the p.-CeHcBh IE is lower than those of the other isomers by more than the experimental error. Its enthalpy of formation (Table IV-I) is also the lowest of the three. The highest occupied I-molecular orbital in the benzene cation is split upon monosubstitution into ’BI and 'An components. Para (1,4) disubstitution is predicted” to give special 79 stabilization to the 'BI orbital (38:; for identical substituents), and this is reflected in our experimental results. Vibrational structure is clearly evident in the threshold region of the p-CeHaBrat PIE curve, Figure IV- 2. Prominent steps, uniformly separated by approximately 0.05 eV, are evident above 8.85 eV. Less pronounced steps can be identified at the mid-points of these intervals, as designated by the vertical lines in the Figure. Similar vibrational structure is seen in the high resolution [Ne(I)] PES spectrum of p-CCHCBPS, where a progression with separation of about 0.025 eV is observed and identified as V. -(aq),1°' the symmetric C-Br stretching vibration in the ground ionic state. We report the LE. of p-CeHIBr: as 8.85 eV. Another step is apparent at 8.80 eV in Figure IV-2. We consider the higher value to be more reliable because the height of the lower-energy step is within the noise level of the data below the threshold, and because it could be due to ionization of the neutral in a thermally excited vibrational state (i.e. due to hot bands). The PIE curves of the CoHIBrt and CeHe’ fragment ions from the isomeric dibromobenzenes (Figure IV-3 and IV-4) have been measured for the first time. Within experimental error, the AEs for the (bromine- loss (M-Br)’ daughter) ions are the same (12.16 i 0.05 eV) for all three isomers (see Table IV-l). This does not necessarily lead to the conclusion that the energy required to break the first Cur-Br bond is independent of the position of the second Br substituent. It is likely that the dissociation energy for the loss of one bromine atom lies above the energy required for substituent isomerization (scrambling) in the ion. (Additional evidence in support of this suggestion is presented 80 below.) Thus, the same ionic precursor would dissociate in each case with the same threshold energy. The calculated bond dissociation energy is the same for each isomer within experimental error. The numerical value, 314 : 7 kJ mol“, matches well the Cue-Br bond energy tabulated1m on the basis of an empirical bond energy scheme. Although decomposition of CoHIBrzt to form CeHsBr’ + Br involves simple bond cleavage and probably passes through a loose transition state with little reverse activation energy,m the experimental threshold may involve a significant kinetic shift." For example, it has recently been shown by elegant time-resolved photoionization mass spectrometry (TRPIMS) measurements that a kinetic shift on the order of 0.3 eV accompanies a normal PIMS determination of the AE of CsHs' from CeHsBr.‘ Thus the CeHdBrt AEs given in Table IV-l must be considered upper limits to the desired 0 K threshold energy. The threshold portions of the PIE curves for Cch’ produced from o-, m- and p-dibromobenzene are shown in Figure VI-4. Due to the weakness of the ion signal, competition with Br-loss fragmentation, and kinetic shift effects, the AEs can be determined less precisely than those of the parent ions and CoHcBrt. Not only are the values for the three isomers the same, well within experimental error (12.4 :1: 0.1 eV), but the appearance energy for production of CoHc‘ from CsHIBr: is only very slightly higher than that for formation of CcHIBr’. This result is consistent with photoionization results from benzene, where the threshold energies of CcHe’ and CoHc’ are the same within experimental uncertaintyflhm as well as with multiphoton ionization studies on benzene.“"1°‘ It is worth noting that compared to AE (CeHIBr’), the experimental AE of cam is likely to be a less reliable upper limit to 81 the threshold energy. ( For CoHa’ from Cell: the experimental- thermochemical difference is about 1 eV.“I1°') With these considerations in mind, it is concluded that within experimental error the activation energies for the formation of CeHsBr" and CcHI’ from each of the isomeric CeHIBr:* cations are the same. Thermochemical information can be used to determine that CeHlt is produced from o-CcHdBrn by the reaction: o-CsHlBra + hv v CsHa’ + Bra + e‘ AE = 12.40 + 0.1 eV. If the neutral partner is Br: we find @4298) for CoHI', most likely the 1,2 dehydrobenzene (benzyne) cation, to be 1283 + 15 kJ mole-1. This is slightly lower than the value suggested by Rosenstock and coworkers;'°v1°' it is probably a better upper limit. If two Br atoms were formed instead, the heat of formation of CcHd’ derived from the experimental AE would be 1088 kJ mol'1, a value well below_ the expected range." The same thermochemical arguments can be applied to meta- and para-dibromobenzene. If one combines the heat of formation for CcHd’ determined from the oeCoHIBra precursor with the experimental AEs and. the enthalpy of formation of the appropriate parent molecule, the heat of formation of the accompanying neutral fragment(s) is calculated to be on the order of 30 kJ. This is consistent only with the formation of Bra, not two Br atoms. The loss of Br: upon photo-fragmentation of CeHIBr: to form Cefie’ thus is not restricted to the isomer initially possessing adjacent Br ring substituents. This result is consistent with the earlier 82 conclusion that isomer scrambling reactions take place prior to or concurrent with the fragmentation. Isomerization accompanying fragmentation of, benzene cations is now well established.‘°“’ Ion fragments produced from deuterated,“ uC-labeled and doubly-labeled speciesm'm retain no "memory" of the position at which the labels were placed. Mass spectrometric measurements on deuterium-labeled chlorobenzenes also show that label scrambling occurs.m The results of the dissociative ionization study of the dibromobenzenes are summarized in Scheme IV-l. The mechanism of the isomer scrambling cannot be determined from this PIMS investigation alone. Photoelectron spectroscopy shows that several excited ionic states exist below the ' photofragmentation thresholds determined here. Their energies and probable symmetry assignments are given in Table IV-3. The state assigned with A—symmetry which lies above 12 eV involves removal of a e C-Br bonding electron, and it is tempting to implicate it as a localized state from which unimolecular decomposition occurs, or at least to consider it a "doorway" state to fragmentation on a lower-lying potential energy surface. However, the true thresholds for Br loss, and probably for Br: loss as well, likely lie at slightly lower energies. Moreover, it is now established from photoelectron- photoion coincidence (PEPICO) measurements and time-resolved PIMS that fragmentation of the mono-halobenzene parent ions to form CsHo’ occurs through a loose transition state on the ground ionic state surface in the cases of Cst 1“ and CsHsBr,”'“5 and perhaps for CoHsCl, also.“'" 1" The ionic fragmentation of benzene to form CoHs’ and CeHII’ occurs i) v.33" 8r 11127512 9V Br 4 $00 \ \‘0 Br Br I \ 83 + 1.7.. + Br2 Br \ + 8r 84 competitively on the ground state CoHs’ surface,” although as noted earlier, the isotopic scrambling of ring and substituent labels reveals that surface crossings have occurred prior to fragmentationM' It is unlikely that photofragmentation of CcHiBrx to form (36343?" and CoHo’ occurs by a different mechanism, although experiments where the threshold kinetic shift and the internal energies of the fragments can be measured are required to confirm this conclusioné“ The vertical ionization of the dibromobenzenes can produce parent ions in any electronic state below the excitation energy. Thus the ions which subsequently fragment initially may be formed with electronic rather than vibrational excitation, and competition among various deactivation channels (e.g direct decomposition, internal conversion, radiation) will ensue. Our results show that the o-, m- and p- dibromobenzene parent ions are distinguishable at the ionization threshold, but that the nature of the ion which fragments to form either CeHABrt or CoHa’ is independent of the substituent distribution on the neutral precursor. Substituent scrambling is likely to occur concomitant with internal conversion surface crossings, and in this case the 'BI ('Bsu for p-CeHIBrit) state [10.64 - 11.08 eV, Table IV-3] may be involved. Elegant photoionization matrix experiments on isomeric di- and tri-chlorobenzenes by the Andrews groupflv" have demonstrated that photochemical rearrangements of. the molecular cations can be induced. For example, o-CoHICls* is converted first to m—CoHIClst and then to p- CeHeCls’ upon continued visible irradiation of an o-CeHIChtlAr matrix; similarly, matrix-isolated m-CoHIClat rearranges to yield the p-CeHeCh’ isomer.” The stability coincides with the relative order of the corresponding ’Bi or 'Bou excited electronic state of the three dichloro 85 isomers. This excited state is formed upon removal of a halogen ‘l- electron from the neutral molecule,“’ which leaves a chlorine lone pair hole in the parent ion. It is suggested that an excited, bridged chloronium cation then forms, and that isomerization occurs via a "chloronium ion walk" around the ring to the most stable isomeric form, which relaxes to the ground cationic state in the matrix.” In the case of the gaseous dibromobenzene cations this mechanism, a bromonium ion walk, could account for the observed isomeric scrambling prior to fragmentation, and for the substituent proximity required for the concerted loss of Br: which accompanies the formation of CsHat. The enthalpy of formation of neutral ortho benzyne, CoHI, can be calculated from the Col-ht appearance energy if the other necessary thermochemical data are known. If we accept 9.45 eV as the adiabatic ionization energy of benzyne,”° then M°nu(CoHd) = AE (CsHI’) - IE (CeHa’) + ”(CsHIBrI) - ”(Brx) is calculated to be 370 :t 20 kJ mol'1. This value is at the low end of the range suggested previously in the literature on the basis of other experimentalut-“M” and theoreticalmlu’ considerations, but we believe that its accuracy is within the quoted uncertainty limits. 86 Table IV-l. Threshold Energies and Thereocbeaical Data Relevant to the Photoionization and Fragsentatiou of p-, s- and o-Dibroaobenzene. Cos-8r Rel. IE or A8 Jpn" Bond Energy‘I Is/z Ion Neutral Int.‘ (eV) (kJ/nol) (kJ/aol) 234-8 p-CsHaBrr 100 8.85 i 0.02 966.9 1' 7 p-CeRIBra 113 :t 5" 155-7 CsHth' 17 12.17 2 0.05 1175.3 1: 10 320 1: 7 Br lll.884¢ 76 CeHdt 6 12.33 i 0.1 1271.8 3: 15 Dr: 30.907‘ 234-8 a-Cslchrz' 100 8.95 10.02 976.5 :l: 7 rCsBsBrz 113 :1: 5° 155~7 CsHaBrt 15 12.15 i 0.05 1173.4 :1: 10 309 i: 7 Br 111.884“ 76 Celict 5 12.44 i 0.1 1282.4 :1: 15 Dr: 30.907“ 234—8 o-CeBsBrz’ 100 8.91 £0.02 976.7 t 7 o-CeBcBrz 117 + 5° 155-7 CeHIBrt 14 12.16 i 0.05 1178.3 1' 10 313 i 7 Br 111.884c 76 CsHs" 5 12.40 t 0.1 1282.5 t 15 Dr: 30.907‘2 ‘Under full illlmination of the hydrogen discharge; 1 53. I’Calculated using group equivalents, Reference 97. °Reference 78‘, pp. I-780-1. 6For the reaction: CelleBrs’ *CeHsBr‘ + Br Table IV-2. Comparison of First Ionization Energies - Dibroaobenzene Isosers Obtained by PIMS and PES. 1.8. (eV) Neutral PIMS‘ PESb PESc PESd o-CchBrz 8.91 8.99 9.02 9.05 l-CeHhBrz 8.95 9.05 ' 9.10 9.07 p-CeHoBrz 8.85 8.90 8.91 8.90 3This work. bCvitas and Klasinc, Reference 7. cStreets and Ceasar, Reference 92. dKilura, et a1, Reference 93. of 87 the 88 Table IV-3. Vertical Ionization Energies (in eV)‘ and Suggested Sy-etry Assign-eatsb of Dibroaobenzene Cation Electronic States. o-CeHsBrz’ 2A2 9.05 281 9.45 382 10.33 281 11.08 2A1 11.3 282 11.38 3A1 12.15 l-CeHaBr2* 2A2 9.07 281 9.57 231 10.73 2A1 10.86 282 11.15 232 11.95 2A1 12.28 p-CeHIBrzt 232. 281; 2Ben 282a 283s 282; 24: 8.91 9.79 10.64 10.94 11.94 12.34 ‘The energy above the ground electronic state of the neutral fer those states below 12.5 eV; ortho- and neta- isomers fro- Reference 93, para- isoner froa Reference 100. I’From Reference 119 and PES peak shapes; sane of the assign-eats differ fra- tbose given in Reference 100 on the basis of SCF calculations. 89 CHAPTER V PBOTOIONIZATION MASS SPECTROMETRIC STUDY OF 1,4-DIOXANE: THERHOC-ISTRY AND ION-MOLECULE REACTIONS I. Introduction The formation of proton-bound dimers in gas phase ion-molecule reactions and the formation of higher charge held aggregates has been the subject of extensive studies. “‘4'" A prominent application of gas- phase ion chemistry is the determination of intrinsic reactivity of molecules free from complicating solvent effects. "‘4" Quantitative measurements of clustering reactions, generalized in equation V-l,m M‘Ba + B "’ M’Bui (V-I) provide a detailed description of solvation energetics and help bridge the gap between gas and condensed phases. Typically 14* is a simple monatomic ion such as H', Li’, Na’, K', F', Cl", Br' or I' and B is H30, NHz, etc.“"'1u In the experiment described in this chapter, M’ is H’, B is p-dioxane (1)-0.3.0.) and n is one. The formation of proton-bound dimers of p-dioxane and their reaction with acetone have been studied by ion cyclotron resonance (ICR) experiments.”' The role of intramolecular hydrogen bonding in dioxane and other di—, tri-, and tetra-others as well as some cyclic 90 crown ethers has also been proposed; from results obtained with a pulsed high-pressure mass spectrometer, a mechanism for cluster formation and fragmentation has been suggested.”’ A study of the hydrogen-bonded dimer ion of p-dioxane by photoionization mass spectrometry has been carried out. In one sense, ion-molecule reactions can be considered as the reverse of ionic fragmentations. Thus a detailed photoionization mass spectrometric investigation of 12-crown-4, the symmetric cyclic ether with formula GM“ has also been undertaken. They will be described in the next Chapter. Here the results of the investigation of p-dioxane are reported. Photoionization efficiency (PIE) curves and appearance energies of p-dioxane and its perdeuterated isomer, CoDsOx, have been obtained. Ion-molecule reactions involving p-dioxane under different sample pressures, photon irradiation energies and instrument repeller voltages have also been fully investigated, and the results are compared with previous reports. II. Results and Discussion The mass spectrum of p-dioxane at a pressure of 1x10“ torr under full illumination of the hydrogen many-line pseudo continuum in the mass range 40-200 amu is shown in Figure V-I. At this pressure the peak at m/z = 89 is due to "Cl-substituted parent ion (4% relative intensity). The intensity of CoHva’ (m/z = 87), the parent ion minus one hydrogen, is very low. However, when the pressure of p-dioxane in the ion source is to ~5xlO" torr the intensities of these two peaks are significantly enhanced. Since the mass of the parent ion is 88 amu, the 91 100 .1 5 pressure=hiO~ Figure V-l hss Spectrus of p-Dioxane Under Full Illusination of the Hydrogen Many-Line Pseudo Continua. 92 ion at m/ z = 89 cannot be formed by a unimolecular process. Moreover, because the m/z = 87 intensity is so low at low pressure, we suspect that ion-molecule reactions are responsible for its formation at higher pressure. Indeed, at low pressures the ion count rates for both m/z = 89 and 87 depend quadratically on the sample pressure over a pressure range of almost one decade; at the highest pressures employable under our experimental conditions, Gill-102* becomes the most abundant ion. The ionization and appearance energy, respectively, of both the parent and protonated parent cations have been measured at room temperature. The PIEs of the molecular ion and the protonated parent in the threshold region (obtained at 2x10”II and 2x104 torr respectively) are shown in Figure V—2. The threshold energies are 9.13 eV and 9.09 eV. That is: p-dioxane -’ p-dioxane+ + e‘ IE = 9.13 1' 0.02eV, and (V-Z) p-dioxane -> p-dioxane+ + e“ followed by p-dioxane+ + p-dioxane + p—dioxane 11* + neutral(s), yields (V-3) . 2 p-dioxane + p-dioxane H+ + neutral(s) + e" AB: 9.09 i‘ 0.02eV (V-4) A. Protonated p-Dioxane The appearance energy for the formation of protonated p-dioxane measured at room temperature is 9.09 i: 0.02 eV which is equal to the IE of p-dioxane within experimental error. The closeness of the AE of protonated dioxane and the IE of the parent ion can be understood in terms of reactions (V-2), (V-3), and (V-4), which show that ionization of p-dioxane is a precondition for the formation of protonated p-dioxane. The similarity in the shapes of the PIE curves for these two ions also 93 Wavelength (Angstroms) 1320.00 1370.00 § : : i + § : wk AA 4° if. 0 O .0 f" A r? . :5. c: D o 00%: so» S‘ . .é g!“ .0 cl! 4 L. O 83 3 ° A 0 Lu ° A — 0 (l on 9.00 9.20 9.40 9.60 Energy (eV) Figure V-2 Photoionization Efficiency Curves For Parent Ion (A) and Protonated Parent Ion (0) . 94 supports the above-mentioned mechanism. The slightly lower AE for protonated p-dioxane also indicates that the potential energy barrier for the ion-molecule reaction (V-3), is negligible. This mechanism is similar to the general reaction,”’ 311* + ether :——¥ ether 11* + B, ‘ (V-5) except the protonating agent here is p-dioxane itself. However, the time scale of our experiment is about 1000 times faster than the experiment in which reaction (V-5) was studied and therefore, equilibrium is not established under the current conditions. The relative intensity of the protonated parent (m/z = 89) in the mass spectra as a function of dioxane sample pressure under two different photon illumination conditions, Lyman-a (1215.6A = 10.20 eV) and central image (full illumination of the hydrogen many-lines psuedo continuum), is given in Figures V-3 through V-7. The relative intensity of protonated p—dioxane first increases with pressure and reaches a maximum, subsequently dropping with increasing pressure. This drop in intensity at higher pressures indicated that the m/z = 89 ion can undergo further reactions, which will be discussed later. B. 9.9.1.1193: At sample pressures below 55le torr, the intensity of Oil-1102* with m/z = 87 is very low. This is consistent with previous reports.““”' Therefore, in order to measure its PIE both entrance and exit slits were increased to 300 pm. This increases the light intensity in the ion source and therefore the number of ions per 9S 0.020 — + l r g.- . . ‘3 0.010 e .3: L . ,///' 0.000c "" V7 1 10'5 10.4 10-3 Pressure (Torr) Figure 7-3 Ion Yield vs. p-Dioxane Sample Pressure Under Full Illunination of the Hydrogen Lamp. 177(0), 133 (A), 116 (U). 73 (O). 59 (9). 96 0.100 - ‘ - i g: N \,_ 0.050< ., a-g b— 0.000 r I I 3 Figure V—4 Pressure (Torr) Ion Yield vs. p-Dioxane Sample Pressure Under Full Illumination of the Hydrogen Lamp. 89 (O), 57 (A), 45 (El). 44 (0). 43 (9)- 97 L000 v :5 ~\ (1500 ~ g: {- 0.000 . . - I ._7. 10"5 10'4 10'3 Pressure(Torr) Figure V-5 Ion Yield vs. p-Dioxane Sample Pressure Under Full Illusination of the Hydrogen Leap. 88 GD), 87 (£0, 58 '(00. .98 0.040 V )- 0.030 b if \ 0.020 1 g: L- 0.010 *- 0.000 10 Pressure (Torr) Figure V-6 Ion Yield vs. p-Dioxane Sample Pressure Under Lylan-s (1215.6 A) Illumination. 177 (0), 133 (A), 116 (a), 73 (0). 44 (O). 43 (A). Ii/ZI.‘ Figure V-7 99 1.000 P- 4 0.800 ’- 0.600 P 0.400 - )— 0.200 - +— 0 ° ‘3’, ‘9 0.000 - ~--—' , 10'5 10'4 10’3 Pressure (Torr) Ion Yield vs. p—Dioxane Sasple Pressure Under Lyssa-s (1215.6 A) Illusination. 89 (0), 88 (A), 87 (I), 45 (0) 100 second, which makes the PIE measurement possible. At the same time the lower irradiation wavelength precision increases the error in the appearance energy measurements. PIE curves for both ConO:’ and CiDtOv have been measured. Their AEs agree well within the combined experimental error of 0.1 eV, and the value (see Table V-l for AE of 0011103’) is about 0.5 eV below the AE previously reportedm following electron bombardment of 1,4-dioxane. Considering the fact that the AEs measured by the photoionization technique for other fragments of p- dioxane are lower by about half an electron volt from the electron bombardment values given in Reference 135, the AE reported here for 0011102’ is a better estimate. Therefore one can write 00300: " Csfivos’ + H + e“ AB = 10.41 eV. (V-6) When the pressure of the p-CoI-hO: in the ion source is raised above 5x104 torr the peak at m/z = 87 is significantly enhanced. Figure V-7 shows that the relative intensity of 0411102’ increases with increasing pressure even when the energy of the ionizing radiation (Lyman-s = 10.20 eV) is below its appearance energy (10.41 eV). This indicates that at higher pressure ion-molecule reactions are responsible for the formation of Gill-103*. Two mechanisms can be postulated for the formation of the CoHva’ cation at higher pressure: 041100: + CcHs02’ -> 043102’ + Fragment(s) (V-7) 041100? + 00310? + H: (V-8) As noted earlier, as the dioxane pressure is increased, the 0610f ion count rate first depends quadratically on the sample pressure over a range of about one decade and then increases even more rapidly. 101 Figures V-5 and V-7 show that at energies both above and below the AE of 008103’ increasing the pressure above 51:10" torr increases the intensity of this ion whereas the intensities of both 0011002: and 0011001: reach a maximum and then drop. Both reactions (V-7) ‘and (V—8) favor the production of CtH102’ at the expense of 0.3-or and CoHnOa’, consistent with the pressure study. In the proposed mechanism for the fragmentation of crown ethers by electron impact, it has been suggested that 0&1va is produced by 33.1088 from protonated p-dioxane, reaction (V—8).1" The cation enthalpies of formation derived in this work (vide infra) indicate that this reaction is endothermic, with endothermicity of 56.8 kJ mol" (0.59 eV). But it is exothermic and spontaneous at the energies under which the pressure study has been carried out (10.2 eV and about 13 eV). C. Unimolecular Dissociation of p-Dioxane at Low Pressure When the sample pressure in the ion source is below 5x10" torr photoionization efficiency curves for the following ions can be measured: CcflsOs’, 0011103’, CaHsO’, CaHsO‘, CzHJOfl 0:11:03 0:11:05 040.01% 0.02035 CaDsO‘, and .CoDsO*. Corrected PIE curves for the cations from 0411-02, obtained over the incident photon energy range 9-13 eV, I are shown (on a fairly coarse scale) in Figure V-8. Each curve has been separately normalized; the approximate relative intensities can be gauged from the mass spectrum observed under full illumination of the hydrogen discharge, indicated in the fourth column of Table V-l. The ionization energies (IEs) of 043.0: and CcDoOa were determined by expanding the threshold region and extrapolating the straight-line portion of the appropriate PIE curve just above onset back to the 102 Wavelength (Angstroms) 1300.00 1200.00 1100.00 1000.00 LllllllllllllllLlllLlLJllll[L111lLlIll1.141 l 1 08! A. T 3 ”fl. ,5 8 0019 A 3 Q U A o A ’0? A . 0:1, A A .4: A . o C I Q A i D 3 O 0 g o q. 3‘ .0 4 £1 E g 53 A A d: .9 A l A A < A1 02 A3 V A LLJ A o A :5 A D- g g A A) 01 f 8° A“ 'WTIIITIIFTUIIIFITITTYITIIIIIYTI 9.00 10.00 11.00 12.00 13.00 Energy (eV) Figure V-8 Photoionization Efficiency Curves for Parent and Daughter Ions of 1,4-Dioxane. C4H002‘(Al), C43102’(.), C3H00*(0), CaflsO (AZ), Czl'lsO‘(0), C2H40’(D), CzflsO’ (A3). ._ 103 baseline. For the fragment ions, the appearance energy (AE) was chosen as the energy at which the PIE first rose perceptibly from the baseline; because of possible kinetic shifts and other factors, even these values must be considered upper limits to the desired adiabatic AE. The values for the deuterated cations are identical within experimental error to their unsubstituted counterparts, and thus AEs for the hydrogen- containing species only are listed in Table V-I. For the parent ion and fragments with m/z = 58, 45, 44 and 43, agreement with the photoionization threshold energies reported by Beer and coworkersm is excellent; well within the combined experimental uncertainty. The values are consistently lower than those obtained from electron bombardment experiments by Collin and Conde,"' by 0.4 eV to as much as 1.7 eV. The photon impact results are clearly more reliable. The AE we measure for the CaHsO’ (m/z = 57) fragment is lower than that given in reference 134 by about 0.3 eV. However, this cation is formed in competition with three lower-energy fragmentations, so its threshold may be subject to significant kinetic shift. As noted above, the observed AB is an upper limit, and perhaps our experimental sensitivity has enabled us to detect CaHAO’ closer to the adiabatic onset. For each ion the experimental AE or IE has been combined in a Hess’ Law calculation with auxiliary thermochemical information for the parent molecule and the presumed neutral partners in the pertinent fragmentation channel to provide the room temperature enthalpy of formation. That value has been adjusted to 0 K using the procedure and vibrational frequency data given in reference 135. Enthalpies of formation at both temperatures are listed for each cation in Table V-l. 104 In order to ascertain whether ions differing in mass by only 1 amu (e.g. m/z 88 and 87, or 58 and 57) were sufficiently well separated in our mass filter, the perdeuterated molecule was studied. It was reassuring that the threshold energies of the well-separated deuterated species (CcDoO:O and CchO’, and CaDsO* and CaDsO’) were the same as those of the hydrogen-containing ions. D. Other Products of Ion-Molecule Reactions The dependence of the p-dioxane mass spectrum on the sample pressure under illumination of the Lyman-s line of hydrogen (1215.611 - 10.20 eV) is shown in Figure V-9. The results are striking. Not only are the protonated parent, the molecular ion, and the H-loss cations formed, but ions at m/z = 177, 133, 114, 116, 73, 45, 44 and 43 are detected. At the highest pressure shown (top frame) the mass spectrum is dominated by the CoH102° cation. From Table V-l it can be seen that only the parent ion(m/z = 88) has. its appearance energy below 10.2 eV. Thus a_l__l of the other species result from ion-molecule reactions involving the parent ion; to form the cation at m/z = 177 at least three molecules must be involved. The mass spectrum is similarly pressure dependent at other photon energies. Regardless of the sample pressure, the ions at m/z = 57, 58, and 59 are not detected under Lyman-s irradiation. However, once the incident energy exceeds about 10.5 eV they occur in the mass spectrum and show pressure dependence which suggests that they arise both from unimolecular decay of CchOz’ and from ion-molecule reactions. At 1050A = 11.81 eV and 5x104 torr dioxane pressure ions at 105 moo, (111100) pressure-1.10‘3 ° 1.! ................... ...1 ............ 1 .................. ...1.1 ........... . .................. , .................. , ...... 1-- 0.300) I oo _ O ‘ pressure=1¢10 4 x10 0 JWJZ ..................................... “I, ......... I ......... I ..... 1. m, ........ Fl ..... , ......... , ......... l ......... , ......... , 100 . _5 0ressure=1s10 o .................. , ......... , ......... , ....... .L. ......... , ......... , ......... , ..... , ......... , ......... , ......... , ......... , ......... . 40 60 80 100 120 '40 160 130 1.1055 (amu) Figure V-9 Mass Spectrum of 1,4-Dioxane, Obtained with Photon Energy of 10.20 eV (1215A), at the Specified Sample Pressures (in torr). 106 m/z = 177, 133, 116, 114, 89, 88, 87 and 73, as well as the usual (low- pressure) fragments from p-dioxane, are detected. Under full illumination of the hydrogen pseudo-continuum (central image) the same cations are observed at pressures as low as 5x10'11 torr. The pressure- dependence of the electron impact mass spectra confirm these results; for example, at p-dioxane pressures in the 5-7x10" torr range 15 eV electrons generate the same mass spectrum as that produced by photoionization by undispersed radiation from the hydrogen lamp. It is worthy of note that the dioxane dimer ion, m/z = 176, was not observed under any conditions with either photon or electron impact. The voltage of the repeller plate in the ion source was varied at selected sample pressures in order to alter the ion residence times. The results are consistent with expectations from the pressure dependence of the dioxane mass spectrum. At a given photon energy and pressure of the neutral molecule, increase of the repeller voltage from its normal, nominal value of 0.1 V leads to mass spectra which are similar to those described previously at lower sample pressures. For a given sample pressure, at the higher voltages an ion has less chance to undergo ion- molecule reactions before it leaves the ion source. E. Ions With m/z Higher Than 89 The intensities of these ions are not sufficiently high to enable their AEs to ~ be measured. This, plus lack of information about both the structures of these cations and their neutral fragmentation partners, as well as their enthalpies of formation, precludes speculation regarding product structures based on the PIMS measurements. In an independent 107 photoionization mass spectrometric study of lZ—crown—4, a similar fragmentation pattern to that of 1,4-dioxane at high pressure is observed (Chapter VI). Moreover, the fragment AEs generally match well under the two sets of conditions. These facts suggest that the ionic fragmentation of p-dioxane at the higher pressures employed in this study may proceed through an intermediate with a chemical formula (0.111006). Whether its structure is cyclic, an open chain, or branched cannot be determined since the ion has not been directly observed in either photoionization or electron impact mass spectrometry. (Indeed the geometry of the parent cation from 12—"crown-4 has not been establishedémm In the case of the dioxane intermediate, the dimer ion (CAHuOc’) could, in fact, be a short-lived, activated ion, which decomposes before it enters the quadrupole mass filter, and mainly produces protonated p-dioxane, which in turn can undergo further reactions. The absence of a peak at m/z = 176 indicates that the dioxane dimer ion is not stable on the time scale of this experiment, which is a) few microseconds. At higher pressures a peak due to m/z = 177 is detected (see Figure V-6), and its intensity increases with further increase in pressure. ~At 10.2 eV only the molecular ion can be formed by direct ionization. Collision of this ion with p-dioxane neutral could produce protonated parent ion which in turn, at higher pressure, could react with a second p-dioxane neutral to form CerrOc’. Production of 0031106 rather than CoHuOc’ on the time scale of the mass spectrometer experiment highlights the necessity of hydrogen bonding in the reaction and the consequent stabilization of the aggregate ion. Two 108 structures are probable for the m/z = 177 cation which are shown schematically belo : m m A. + to -bo d d O\ /O. - - H -- .0 O g-rfiiogane ndifiner O. O B. C ‘Ht 3 protonated 0 ‘0 12—crown-4 \__/ Baer and coworkers have shown that the fragmentation of the p—dioxane cation occurs via a tight activated complex)“ and that the molecular ion keeps its cyclic structure even a few electron volts above its ionization energy. -It has also been shown‘” that, although optimal hydrogen bonding geometry is available in crown ethers, if the interaction is between non-adjacent oxygens such as shown in structure 8, the internal‘hydrogen bond is stable, but its enthalpy in most cases falls substantially short of the strength of the strong hydrogen bonds observed in hydrogen—bridged dimer ions such as structure A, where Hi'n = 30 :1: 2 kcal mol“. On the other hand, it has been proposed” that neutral p-dioxane and protonated p-dioxane are 109 fragments of protonated lZ-crown-4 decomposition, and therefore reaction of protonated p-dioxane with neutral p-dioxane can be considered as the reverse of that fragmentation. Support for both structures A and 8 makes the determination of the enct structure of the ion with m/z = 177 difficult; structure A is more probable due to the stability of p-dioxane’ at energies above its fragmentation threshold and the stronger hydrogen bonding in that structure. Mautner‘” has also shown that the reaction ether 11* + other i ether 11* ether reaches equilibrium within 1-2 ms. Here again the proton-bonded dimer is favored over a possible cyclic structure. Ion with m/z = 133. At 10.20 eV where only the molecular ion can be formed by direct photoionization, the following mechanism can be written for the production of the m/z = 133 cation: CAHsOa’ + CcHsO: " CoHnOa’ + 03330. (V-9) At the highest pressures attainable in these experiments a very small amount of an ion with m/z = 131 was also produced. This cation could be formed by direct ion-molecule interaction, like reaction (V-9), except with CsHiiOs’ + 0:850 as products, or by Hrloss from the ion with m/z : 133, that is 0031003’ " CsHiiOs’ + H3. (V-lO) At higher photon energy, where daughter ions are formed following p—dioxane photoionization, it is also possible that CaHsO’ (m/ z = 45) reacts with neutral p-dioxane to form the m/z = 133 cation: 00300: '1' 031150‘ * CsHisOs*. (V-ll) 110 Figure V-3 shows that the intensity of came first increases with pressure, reaches a maximum and then drops. This suggests that at higher pressure the m/z = 133 cation is involved in further reactions. Ions with m/z = 116 and 114. As in the case of CcHan*, cations with m/z = 116 can be formed from direct reaction of the molecular ion with neutral p-dioxane, and at higher photon energies perhaps also from reaction of the p-dioxane fragment having m/z = 28 with p-dioxane. High resolution mass spectra at various electron impact energies has indicated that at onset .the m/z = 28 ion is 0280*. Therefore, the m/z = 116 ion in the p-dioxane mass spectra is likely to be CsHuDa‘, and the mass 114 ions arise from Fla-loss from 116: p-CsHsOs + C3H0* + CsHizOs’hn/z = 116) + amoorm/z = 114) + m. (V-IZ) One mechanism proposed for the formation of CaHn‘ from the p-dioxane parent ion is a consecutive dissociation process in which the 0084’ grandaughter ion is produced from 0311.0; (m/z = 58) along with 01130.1” The intensity of m/z = 116 increases with pressure in the range of the PIMS experiments while m/ z = 58 constantly decreases, which is consistent with this mechanism. F. Ions With m/z Less Than 88 Cations with m/z = 73, 71 and 59. These ions are detected only at fairly high sample pressure. They are not photo-fragments of p-dioxane at low pressure, even under full illumination of the hydrogen discharge, and they must therefore be products of ion-molecule reactions. Ions 111 with these m/z values are indeed photo-fragments of 12-crown-4, and it has been suggested that the daughter ion at m/z = 133 can undergo further decomposition to produce m/z = 73, 71, 61 and 59.”. Although no ion with mass 61 was detected in this investigation, the observation of ions at m/z = 73, 71 and 59 under conditions where m/z = 133 is maximum, and the increase in their intensity while the intensity of the mass 133 ion subsequently decreases, can be viewed as support for this mechanism. The absence of any further information about these ions precludes speculation regarding their structure. Schemes for the production of the ions discussed in this subsection are given below: Low Energy: direct *’ [C4H90‘ + neutrals],[04870* + neutrals], . a/z = 73 a/z = 71 043st CAHsOf... [CzHaON- + neutrals] / a\z = 59 indirect a Celii:303+ + neutral, [C2H002* + neutrals] a/z = 59 High Energy: CAB-03’ + CxHaO* -' CoHnOa’ -* [same decomposition products] 112 Cations with m/z = 45, 44 and 43. These ions can be produced both by unimolecular fragmentation of the parent cation or from ion- molecule reactions. The mass 43 ion could also be produced from m/z = 45 through 113-loss. In compounds such as the benzo-crowns studied by Gray et a1., metastables are observed which confirm the fragmentation m/z = 89 + 45. This suggests that, in addition to being the product of unimolecular fragmentation of Gill-03*, the cation with m/z = 45 could be produced from protonated dioxane, which itself is a product of an ion- molecule reaction. One possible mechanism for the formation of these ions at high pressure, therefore, will be: 'r——>C21140 + CszO‘ (n/z = 45) t—Hz p-dioxaneli+ -—-——rC21-b0 + CzflaO” (a/z = 43) "——’Czllso + 02HA0*(m/z = 44) The decrease in the m/z = 44 and 45 ion intensities at higher pressure (FigureV-4) is an indication of their further reaction. The increase in intensity of the m/z = 43 cation at pressures where the mass 45 ion. intensity declines can be considered as support for the conversion of CIHAO’ to CzHJO’ via 112-loss. 113 Table V—l. Threshold Energies and Beats of Porestion for Ions Produced Era p- Dioxene. at Low Sqle Pressures. a/z Relative IE: or AE‘ Rana" flro'r' Ion Neutral. Intensity" (eV) (kJ sol“) (10 not“) 88 CAEOI’ 93 9.13 :1: 0.02 564 t 4 595 i: 4 048.02 -316 t 0.8 -285.3 1 0.8 87 Q8102; 3 10.41 t 0.1 5469.7 H 217.8 58 Gama; 100 10.52 i: 0.05 524 :1: 8 0830 -108.8 1 0.8 -105.0 1' 0.8 57 038.0’ 6 10.89 $0.05 $740 18 $752 $8 0800 14.6 21.3 45 CszO' 15 10.46 i: 0.05 736 i" 8 753 :1: 8 C2030 -22.6 -15.9 44 Calls? 8 10.47 t 0.05 890 i: 8 900 :I: 8 CzHcO -165.7 1: 0.4 -155.2 1: 0.4 43 CaitiO’ 6 11.90 10.05 5868 $873 02850 17 -2.1 ' Reference 134 . l’Under full Pressure = 1.2x10'5 torr. CA1: 298°K. illuaination of hydrogen aany-line pseudo-cont inuul. Ion source “Ions are corrected for kinetic energy release (taken from reference 134). Neutrals taken fro. reference 134. 'Ions are corrected for thermal energy. ‘flr(e‘) is taken to be zero. reference 134. Neutral taken froa reference 134. Taperature conversion according to procedure in 114 CHAPTERVI PHOTOIONIZATION .MASS SPECTROMETRIC STUDY OF lZ-CRO'N—4 Introduction Since the first report of the syntheses of macrocych "crown" polyethers by Pederson,1”'“° considerable interest has been shown in the properties of these compounds. Crown ethers have been found to be very useful reagents in organic chemistry because they selectively complex alkali metal ions and can be used to carry them through aqueous non-polar solution boundarieséwm This has been employed as a model for cation transport through membranes.“3 Ultraviolet photoelectron spectra of crown ethers and their electronic structures have been reported, and with the aid of theoretical calculations, the lower IE bands of the spectra have been assigned.” The 70 eV mass spectra of unsubstituted cyclic polyethers, along with their dissociation mechanism and the elemental composition of both ions and neutrals have been suggested."‘ In this work the ionization and fragmentation of lZ-crown-4 (CoHuOo) have been followed by photoionization mass spectrometry. Appearance energies for the protonated parent ion and twelve of its daughter cations are reported. Upon ionization 12-crown-4 produces a rich assortment of fragment ions and neutrals. Some of these, such as 115 CiHsOf, are well-known and stable ions, whose appearance energies can _ be readily calculated because the heats of formation of the ionic and neutral products are established."'1“v“' For these cases comparisons between the new results and previously-reported values are possible. However, in many instances the heats of formation of the fragment cations are not known and even their chemical compositions are not certain, so comparisons with the literature cannot be made. In addition to its inherent interest, the photoionization mass spectrometric study of the fragmentation of .12-crown-4 is also likely to be useful for comparison with ion-molecule reactions of dioxane at higher pressure, which have been described in the preceding chapter. RESULTS The PIEs for the protonated parent and twelve fragments of 12- crown-4 have been measured; PIE curves for the fragments are shown in Figures VI-l and VI-Z. Under .the experimental condition employed in this study the parent ion (m/z = 176) intensity was never above the signal-to-noise level. This result is consistent with mass spectra of 12- crown-4 under electron bombardment.” At higher sample pressures an ion with m/z = 177 was detectable and its intensity was sufficiently high that PIE measurements could be made. At low pressures several fragment ions are detected withAEs in the range 8.86 i 0.05 eV. From this information the ionization energy of lZ-crown-4 was estimated to be 8.86 eV. PES shows four broad peaks around 9.3, 9.9, 11.4 and 12.5 eV;” the reported vertical ionization energy is 9.3 eV, although the onset of this band in the PES spectrum is about 8.6 eV. The threshold energies of the fragment ions determined by PIMS, along with probable 116 Wavelength (Angstroms) PIE (Arbitrary Units) V‘éfimv & 101 o°° 61.13; V W I ”a ‘ I ‘ 9 41‘ '3 '30 I . I . I 5:20 590 30¢. 030 980 ’0- Energy (eV) Figure VI-l Photoionization Efficiency Curves For Protonated and Daughter Ions of lZ-Crow—4. 117 Wavelength (Angstroms) 1300.00 1200.00 :1::§+r.~¢#¢¢¢+14f. . O a O 90 O o ’23 'E :3 >~ L 9 '5 L. 3 E a I I I l l l I I I 1 l I l | I 9.20 9.40 9.50 9.8 10.00 10.20 10.40 10.60 10.80 Energy (eV) Figure VI-2 Photoionization Efficiency Curves For Daughter Ions of lZ-Crown-4. 118 formulas of the ions and neutrals and the relative ion intensities are listed in Table VI-I. DISCUSSION A. PIE Measurement and Thermochemistry A useful property of an ion or molecule is its heat of formation, 51:“. This is a thermodynamic. quantity which, at a nonzero temperature, contains the thermal distribution of internal and translational energies. Photoionization, on the other hand, is a spectroscopic experiment and thus is associated with transitions between specific internal energy states. The dissociation energy refers to the (thermodynamic) energy of the products relative to the reactants. This can be lower than the (spectroscopic) AE because of reverse activation energies and kinetic shifts. Therefore the true value of an ionic heat of formation determined spectroscopically depends on the internal energies of the precursor and the products [ion and neutral(s)], the kinetic shift and reverse activation energy. Where these factors cannot be determined the reported AE value will be only an upper limit. This is in fact the case in this study because no information regarding these matters are available at this time. The AE listed for a given cation in Table VI-lwas chosen as the energy at which the corresponding PIE first rose perceptibly from the baseline. B. Structures and ”Energies of Fragments In order to determine the structures and energies of photo- fragments from PIE measurements one must know the chemical composition of the fragments at the threshold energies. Normal mass 119 Table VI-l. Ionization and Appearance Energies (298K) of lZ-Crown-4 and its Fragment Cations. m/z Ion Relative Ion Intensity IE of AE, eV 176 CeH1004* -- 8.86 1:0.1a 133 CeHana“ 5 8.89 1’ 0.05 117 Gel-11302+ - 1 8.93 i 0.05 101 CdisOs” 2 8.92 i 0.05 89 C48002* 41 8.86 1:0.05 88 Cclis02* 18 9.14 :t 0.05 73 c.Hso+ '66 8.92 .t 0.05 61 Cali-502* 9 9.10 i 0.05 59 Cali'rO+ (or C2H002*) 7 9.70 1:0.05 58 C3850; (or C28202*) 14 9.32 1.0.05 45 C2HsO“ 100 9.70 1' 0.05 44 C2H40’ 70 9.58 i' 0.05 43 C2830+ 22 9.70 :1: 0.05 aInferred from lowest AEs measured. 120 spectrometers give information only about m/z and not chemical composition or structure. For a given nominal m/z, in most cases, one can write several chemical compositions. Their differences lie in the 0.01-0.1 amu range and they may be distinguished only by high resolution mass spectrometry. Moreover, for thermochemical studies one must also know the structure of the neutral fragments which accompany a particular fragmentation, and but little information concerning these neutrals is available. Metastable ions are useful in this regard since they provide information about the precursor of a fragment ion (i.e. whether the fragment ion was produced directly from the molecular ion or is a product of consecutive reactions) and therefore about the neutral counterparts. In addition, it is necessary to know whether the dissociation path is determined by dynamical (or mechanistic) constraints or by thermochemical ones, and thus whether one might expect the formation of fragments which are not necessarily the most stable ones. If different isomers of one compound isomerize to a common ionic structure, it could indicate that the dissociation paths are determined by thermochemical constraints and therefore that the most stable fragments are to be expected. Kinetic shift and reverse activation energy data are also needed in order to obtain information about heats of formation from the differences between activation energies of forward and reverse reactions. In principle, these data can be gathered from time-resolved photoionization mass spectrometry and kinetic energy release measurements. A more accurate onset for a particular decomposition product could be also determined from RRKM/QET statistical theory calculation of the dissociation rate,“'"“‘ which could be compared with PIPECO experiments. If the heat of formation at zero 121 degrees is desired from an experimental value at 298 K, the average thermal energy, (Eu), has to be taken into account. energies. (Eu) and (Em) can be obtained by using the classical values; (Evlb) can be determined by evaluating the vibrational frequencies for each species. For the activated complex these are usually not available and must be estimated. For larger molecules this information becomes less available and therefore the prediction of ionic structures becomes less certain. In this study calculations of heats of formation are limited to those fragment ions whose elemental compositions have been predicted by previous EI mass spectrometry studies.1’°-”" '1“ CaHiGOA’hn/z = 176) At low pressure no ion with m/z = 176 was detectable in our mass spectrometer when 12-crown-4 was irradiated by light from the hydrogen discharge. This result is in agreement with a low pressure, low electron energy study, and with 70 eV mass spectra of crown ethers.”‘ At higher sample pressures an ion with m/z = 177 was detected and its intensity was sufficient to enable measurement of its PIE. The appearance energy of this ion is 8.89 eV. Ions with m/z = 177, 133, 117, 101, 89 and 73 show similar AEs (i.e., within experimental error). The AE of an ion with m/z less than 176 must at least be equal to the IE of the parent ion at low pressure. Therefore the IE of 12- crown—4 must to be 8.86 + 0.1 eV, the minimum AE observed at low pressure. The vertical IE of 12-crown-4 reported by PES is 9.3 eV;“4 the adiabatic value will be somewhat lower. 122 Ionization of the parent neutral is a precondition for fragmentation; if the molecular ion decomposes on a time scale much faster than the mass spectrometry experiment, then the fragment ions formed at the IE will have no kinetic shift. However, because the dissociation products may be more stable than the parent cation, they may be formed with some internal energy. Thus the experimental AEs will give upper limits to the desired thermochemical reaction enthalpy. The ionization energy of lZ-crown-4 is considerably lower (at least 0.27 eV) than that of 1,4-dioxane . In an earlier investigation it was established that the thermochemical stability of polyethers is strongly dependent on the relative positions of the oxygens.“° The apparent strain energies calculated for 1,4-dioxane and 12-crown-4 are 13.8 and 32.1 kJ mol'1 respectively.151 The difference (about 0.19 ev) has been described in terms of differences in the conformations of the structural units that constitute the rings.m The strain has been investigated in the solid phase on the basis of geometrical considerations, and sterically the 1,4 0...0 interaction, which depends on the C-C gauche bond and the 1,4 CH...HC interaction which depends on the oxygen-carbon gauche bond are considered to be” responsible for this strain.153 The interaction energies are in agreement with strain energies determined when similar conformations are assumed for the gas-phase molecule.”1 The relative IEs of . 12-crown-4 and 1,4-dioxane determined by PIMS are consistent with this result; one can interpret both the lower ionization energy and the fragmentation energy of lZ—crown-4 in terms of these steric interactions. From the IE estimated in this work and the known heat of formation of 12—crown—4," an upper limit for Mr" of CsHuOf can be calculated; the value is listed in Table VI-2. l I 123 0011130? (tn/z 2: 133) + C2H30 Similar to the IE, the activation energies for fragmentation of 12- crown-4 are also smaller than those of p-dioxane. - Again, steric effects are likely to be responsible 'for this observation. The AE of the m/z = 133 ion is equal to the estimated parent IE within experimental error and therefore kinetic shift is insignificant. If one assumes the lowest energy structure for the neutral fragment, CH300,“‘ Hess’ law gives AH°mo(CoHisOs*) 5 248.3 kJ/mol. (The inequality arises because the kinetic shift and reverse activation energy are not known.) A cyclic structure has been proposed for this ion."‘ CoH1302*(m/z = 117) + 0211302 The AE of this ion lies only 0.07 eV above the proposed molecular IE and therefore only very little kinetic shift can be involved. No information concerning the structure and thermochemistry of the neutral(s) are available to enable the calculation of the heat of formation of the ion. Qg'HsOWm/z = 101) + (C0H110) This cation has AB = 8.92 eV which is only 0.06 eV above the estimated IE and therefore the kinetic shift is not significant. The heats of formation of the neutral fragment with chemical composition CcHnO is not available, and no information concerning the structure of the neutral (or neutrals) in this reaction has been reported. But if one assumes that this ion is produced through the same mechanism as m/z = 105,1“ by the subsequent loss of two Ha molecules, then the neutrals 124 will be CsHsO + 03H: + 2H:. The heat of formation of CsHsOa’ can be calculated from the experimental AE and Eh values for these neutrals; the result is tabulated in Table VI-Z. CsHsO:*(m/z = 89) + 04810: Among all the fragment ions from 12-crown-4, CiHsOa* has the minimum AE and the probability that this ion is observed with a very small kinetic shift is high. If the neutral has the same structure as in the case of fragmentation of the p-dioxane dimer cation (for which E°moCAH103 = -156.3 kJmol", see previous chapter), then one calculates H‘nuCflsOv = 379.1 kJ mol". This value is in a good agreement with the reported heat of formation of protonated dioxane, and suggests that the m/z = 89 daughter ion from 12-crown-4 has the same structure as protonated p-dioxane. Consider the two processes: (a) ZCAHsOs -' C4390? + CAB-102, and (b) CsHieOc -' CM’ + 001110: The enthalpies of formation of the reactants in both cases are the same" (£011.. 12-Ce4 = 151 kcal mol'1 = zan‘m. p-dioxane). The products of both reactions are also the same. The photon energy at which CcH003’ is observed in the first case (9.09 eV) is slightly higher than. the AE of CoHsOa’ (8.86 eV) from reaction (b). The difference attributed to IEs of the oxygen atoms in p—dioxane and 12-crown-4, indicates that reaction (a) has a somewhat higher reverse activation energy (since both reactions have very little, if any, kinetic shift). In that case, the fragments of reaction (a) will be produced with higher 125 internal or translational energies than the dissociation products in (b). The similarity between the products of reactions (a) and (b) could also indicate a common activated complex for these two reactions. CAHsOi’hn/z = 88) + CcHsOs If one assumes that the neutral in this case is p-dioxane then the heat of formation of the cation calculated from the AE‘ listed in Table VI- 2 is fli°nuCcHsOa* = 364.3 kJmol‘l, which is exactly equal to the heat of formation of the p—dioxane cation measured from the direct ionization of p-dioxane. This reaction can be considered as the reverse of the association of neutral and cationic p-dioxane and the similarity between the IE of p-dioxane at high pressures (~9.I eV) and the AE for production of p-dioxane+ from 12-crown-4 (9.14 :t 0.05 eV) indicates that the latter reaction has almost no reverse activation energy. C0H00’(m/z = 73) + 0:11.30 + 021100: The suggested structure far this cation is a protonated cyclic ether for which the heat of formation is known in the gas phase. With this information, the experimental AE, plus the heat of formation of CaHaO we find m°noo02Hioz = -262.9 kJ mol'1. The suggested structure for this neutral is also a cyclic ether. The calculated enthalpy of formation is considerably higher than MP: of other isomers of 031100: (dimes methyl formate = -354.9 kJ mol'1 and fli": acetic acid = -432.0 kJ mol“); the heat of formation of cyclic 0.3.0. is not known from other work, but one might expect the cyclic forms to be less stable than the well-known isomers. 126 CszO:’(m/z = 61) + 021130 + 041100 Since the heats of formation of the neutrals (considering 041-100 to be a cyclic ether) are known, the heat of formation of the ion can be calculated and the result is given in Table VI-2. No information concerning the heat of formation of this cation is available with which to compare the experimental value. It has been suggested that both C4H90* and CszOa“ cations are produced from a common precursorl" (CoHan’). The results of the PIMS study show that both of these ions have AEs close to that of the CoHnOs’ ion (0.03 and 0.21 eV higher, respectively), which is consistent with this possibility. (If one of the ions with m/z = 73 ’or 61 had a lower AE, then CsHan* could not be the precursor ion.) The results of the p-dioxane pressure study also show that ions with m/z = 73 and 59 (which could be produced from 61‘ -> 59* + H3) start to appear as m/z = 133 reaches its maximum intensity and continue to increase as 133* starts to fall off. These observations support the above mechanism, but some disagreement also exists (see below). C2H302’ (m/z = 59) :1- CzHaO + CIHSO + H2 The AE of this reaction with respect to the threshold for m/z = 61 is higher by 0.6 eV, which indicated either that loss of H: from Cal-1502* is an endothermic process or this ion has been produced with a considerable kinetic shift. Again, since the m/z = 59 ion has a higher AE than CszOzt, the possibility exists that this ion has been produced by decomposition of the m/z = 61 cation. On the other hand, the repeller voltage study made for p-dioxane photoionization over a voltage difference of 0.1 to 15 volts shows that the relative intensity of this ion 127 remains constant with respect to those with m/z = 57 and 58, which indicates that this ion could be the protonated analog of the cation with m/z = 58. If this is the case then the chemical composition of m/z = 59 is CoHvO‘, and the likely neutrals would be 041100: + CHO. For the first case, Hi“: (Cxfiaov) has been calculated and the result is included in Table VI—2. Again, no value is available in the literature for comparison. For the cation with chemical composition CanO‘ (the second case) $111». = 581.9 kJ mol‘1 which is very close to mom. of protonated acetone (i.e. 543.4 kJ mold). CsfisO’ (m/z = 58) + 04%: + 0820 The calculated heat of formation for this ion is 691.4 .kJ mol" which lies between the heats of formation of CH: =C(OH)-CH3* [M'r 1' 660.9 kJ mol") and CHa-CO-CHJ’ (di‘r = 719.0 kJ mol“)]. The closeness of the heat of formation of m/z = 58 to that of acetone might be considered as support for the m/z = 59 ion being protonated acetone. 011130’ (m/z = 45) + 04300: + CaHsO The Hi“: (C:HAO*) = 642.1 kJ mol, which falls between the values for on. - mom-11* (608.2 kJ mol-I) and care-car (660.9 kJ mol"), The heat of formation of the cyclic isomer of CszO’ is 689.7 kJ mol"; and is being considered as it was the suggested the fragment with m/z = 45 in- the electron impact ionization study 12-crown-4.“‘ 0:1140’ (m/z = 44) + 001100: + 03340 The calculated heat of formation of the m/z = 44 ion is 775 kJ mol“, which is very close to HP: = 770.0 kJ mol"1 of CH:=CH-OH‘. 128 CszO’ (m/z = 43) + 04300: + 021150 The calculated heat of formation of this ion based on these neutrals is 602.3 kJ mol“ which is close to dimes (CH3CO’) = 630 kJ moi“. If the neutrals are taken to be 041100: 1» 0:11.10 + H2, then mm“ (Cam-0*) = 642.1 kJ mol“ which is between those of CHrCO’ and CHa=C-OH’ (m0: = 787.1 kJ mol“). As was mentioned at the beginning of this chapter, prediction of the structure of an ion from PIE measurements alone, without complementary structural and thermochemical data, is in most cases impossible. In the photo-fragmentation of 12-crown-4 more information is necessary in most instances in order to choose the exact structure of the fragment ions. The results of this PIMS investigation of lZ-crown-4 indicate that heats of formation of several of the fragments, such as 04H002* and CiHoOs’, are very close to those obtained in the p-dioxane study. On the other hand, different heats of formation are calculated for other ions with the same chemical composition, e.g. CaHoO’, c.n.o+ , C:H40* and CsHaO’. One cannot conclude from these data that the fragmentation products of 12-crown-4 are alike or dissimilar to those from p-dioxane because different choices of neutral fragmentation partners (where inadequate information exists) might provide better agreement between calculated heats of formation for a given cation. 129 Table VI-Z. Thermochemical Data Relevant to the Ionization and Dissociation of 12-Crown-4. a/z ion neutrals ”use in! sol“ 176 CallisOe+ 223' CsHiem -631° 133 CeHmOa“ -6311’ 021130 -23° 117 CeH1302* 0211302 101 C4H503* ‘ 201‘ C2HaO + C2114 + 2112 -23° + 50" 89 048902“ 379' 0011702 -156‘ 88 «1502+ 564' (343802 -316° 73 CellsO+ 514° CzHaO + C2H002 -23°-2631’ 61 C2H502* 453'1 Czlho + CcHsO -23°-184° 59 028302+ . 511' CzlisO + 04890 + H2 -23¢—184° CaH-IO“ 582' 0411802 + C110 -3161’ + 38° 58 Czilaor 617‘ 021100 + CeHsO + H2 -1661’-1841’ CsEO’ 691a 0411st + C1120 -316b-1091’ Table VI-2 continued. 130 s/z ion neutrals ”can kJ sol“ 45 021150l+ 642' 048002 + 02830 -316P-23c 44 C2H’AO+ 775' C4150: + CthO -316P-167° 43 C2330* 642' 04502 + szo + Hz -316°-23° aDetersined fros the present data. bReferences 146 and 153. cReference 145. dFran Table V-I (Chapter V). 1. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20 REFERENCES K.E. McCulloh, Int. J. Mass Spectrom. Ion Phys. 2.1: 333 (1976). .... ..... Y. Maiinovich, R. Arakawa, G. Hasse and C. Lifshitz, J. Phys. Chem. 89, 2253 (1985). J. Dannacher. 9.59:...-M§§§:_.....§12991295-2.- .13: 253 (1984)- T.L. Bunn, A.M. Richard and T. Baer, Jimghemgmghyg. In Press (1985). E.J. Darland, D.M. Rider, R.P. Tully, C.G. Enke and G.E. Leroi, Int. J. Mass Spectrom. Ion Phys. _3__4_, 175 (1980). T. Cvitas and L. Klasinc, Croat. Chem. Acta. 59, 291 (1977). ...... R.H. Levin: ReactiVe Intermediates, ed. by M. Jones, Jr. and R. Am Moss (John Wiley and Sons, New York, 1978). R.Yamdagi and P. Kebarle, J. Am. Chem. Soc. 95, 3504 (1973). G- Herzberg: Atomic Spectra..-999---..Atcmie--.815991952 (Dover Publication, New York, 1945). LN. Levine: 1.491.99th..-..§R991¢r°3992}f (John Wiley and Sons, New York, 1975). G-V- Marr: PhotolomsatlonProcesses mGasses (Academic Press, New York, 1967). V. Fano and J.W. Cooper, RevModPhys 40, 441 (1968). W.A. Chupka, "Ion-Molecule Reactions by Photoionization Techniques" in 19112119199919 Reactions, Vol. 1, ed. by J. L. Franklin (Plenum Press, New York, 1972, pp. 33-77). J. Berkowitz: Photoignizationm"and Photoelectron Spectroscopy (Academic Press, New York, 1979). R.S. Mulliken, Acc. Chem. Res. 9, 7 (1976). J-H-D- Bland: JChImPhxg 7.7.: 613 (1980)- E. Lindholm, Arkiv, Eye. 59, 103 (1969). ”'3' ROM“: 111.8 1195-...-.1311011-39 313193.---.91.... ....-.139.13' atomic MOIPCUIPE (Academic Press, New York, 1974). 131 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 'A.D. Walsh, J. Phys. Radium 1.5, 501 (1954). J.L. Dye and V.A. Nicely, 9.11.9..Qs...§§.° 48, 443 (1971). P. Houchman, H. Wang, W.S. Felps, S. Foster and S.P. McGlyn, "0n the Assignment of Molecular Rydbert Series," in ghemiggl Spectroscopy and Photochemistry in the Vacuum Ultraviolet, ed. by C. Sandorfy, P. J. Ausioos and M. B. Robin (D. Reidel, Boston, 1974,_ pp. 385-393). R.P. Madden and K. Codling, Astra Phys J. 141, 364 (1965). R.P. Madden, D.L. Ederer and K. Coding, Phys. Rev. 777, 136 (1969). R.L. Platzman, Radiat. Res. 17, 419 (1962). R.S. Barry, J. Chem. Phys. 45, 1228 (1966). W.A. Chupka and J. Berkowitz, J. Chem. Phys. 51, 4244 (1969). an... J. Berkowitz and W.A. Chupka, J. Chem. Phys. 5}, 2341 (1969). ‘ W.A. Chupka and J. Berkowitz, J. Chem. Phys. 55, 5726 91968). P.C. Killgoar, G.E. Leroi, J. Berkowitz and W. A. Chupka, 73"""55555 Phys 55, 803 (1973). C.Y. Ng, B. H. Mahan and Y.T. Lee, J. Chem. Phys. 55, 1956 (1976). nun“ J.H.D. Eland, Journal de Chimie Physique 77, 613 (1980). J.H.D. Eland: Photoelectron Spectroscopy (Butterworths, London, 1984). R.S. Berry, J. Chem. Phys. 45, 1228 (1966). G. Herzberg: §Rgctramof Diatomic Molecules (D. Van Nostrand, New York, 1950). R.C. Dunbar, "Ion Photodissociation" in Gas Phase Ion Chemistry, Vol. 2, ed. by M. T. Bower (Academic Press, New York, 1979, pp. 181-220). H.M. Rosenstock, M.B. Wallenstein, A.L. Warhaftingand, H. Eyring, Proc. Natl. Acad. Sci. U.S.A. 55, 667 (1952). R.A. Marcus and O.K. Rice, J. Phys. Colloid Chem. 55, 894 (1951). O.K. Rice and R.C. Ramsperger, J. Am. Chem. Soc. 35, 1617 (1927). 132 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 133 L.S. Kassell, J. Phys. Chem. 32, 225 (1928). N.B. Slater: Theory of Unimolecular Reaction; (Methven, London, 1959). L.D. Spicer and 3.8. Rabinovitch, Annu. Rev. Phys. Chem. 21, 349 (1970). G. Wang and R.A. Marcus, 7179135175...“Phys. 57, 1636 (1977). J.C. Lorquet and B. Leyh, "Potential Energy Surface and Theory of Unimolecular Dissociation" in 1917119....131'0693393 in the Gas Phase, ed. by M. A. Almoster Ferreira (D. Reidel, Boston, 1982, pp 1-6). C. Galloy, C. Lecomte and J.C. Lorquet, JChemPhys 77, 4522, (1982). C. Lifshitz, Adv. Mass Spegtrom. 7, 3 (1977). Z. Herman, A. Lee and R. Wolfgang, g,mcpgg.wgpy§. 51, 452 (1969). 9“ Henglein, g.:.....9119.111:.-.?.11l§° 5.3.! 458 (1970)- I. Herteland and Ch. Ottinger, Z. Naturforsch, A22, 40 (1969). J.H. Beynon, J.A. Hopkinson and G.R. Lester, Int. J. Mass Spectrom. 19.51-311.19 2, 291 (1969). 12.0. Hicklinz and KR. Jennings: OrsMaBBSpectrom .3... 1499 (1970). . ..--.- (1972). M.A. Haney and J.L. Franklin, .12-.-.913991..E1}.X§- 42, 4093 (1968). R.G. Cooks, J.H. Beynon, R.M. Caprioli and G.R. Lester: Metastable 7595 (Elsevier Scient, Amsterdam, 1973). E.P. Wigner, PhngeX 73, 1002 91948). S. Geltman, B.11Z§.9.-_R§X- ”17.0.2, 171 (1956). "a... ,K. Watanable, J. Chem. Phys. 22, 1562 (1954). R.E. Fox, J. Chem. Phys. 22, 200 (1960). R.H. McFarland, Phys. Rev. Letters L0, 397 (1963). G.J. Schultz and R.E. Fox, Phys. Rev. 106, 1179 (1957). 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. C.E. Klots, "Energy Deposition Processes" in Fundamental Process in Radiation Chemistry ed. by P. Ausioos (Interscience, New York, 1968, ppe 1-58)e N.W. Reid, Int. J. Mass Spectrom. Ion Phys. 5, 1 (1971). W.A. Chupka, J. Chem. Phys. 29, 191 (1959). seams-so" G. Herzberg: Molecular Spectra and Molecular Structure III (Van Nostrand, Princeton, New Jersey, 1967). W.A. Chupka, J. Chem. Phys. 54, 1936 (1971). B. Steiner, C.F. Giese and M.G. Ingham, J. Chem. Phys. 24, 189 (1961). V.H. Dibeler and J.A. Walker, Int. J. Mass Spectrom. Ion Phys. 1‘1, 49 (1973). G.R. Parr and J.W. Taylor, Rev. Sci, Instrum. 45, 1578 (1973). P.M. Guyon and J. Berkowitz, J. Chem. Phys. 54, 1814 (1971). D.M. Rider, Ph.D. Thesis, Michigan State University, 1980. L. Friedman, F.A. Long and M. Woifsberg, J Chem. Phys. 25, 714 (1957). C. Liifshitz, Mass Spectrom. Rev. 1, 309 (1982). H.M. Rosenstock, K. Dranl, B.W. Steiner and J.T. Henon, J_.__W13.l'i_y_§: Chem. Ref. Data 5, 1 (1977), Supplement No. 1. M. Vestal "Ionic Fragmentation Processes" in Fundamental Processes in Radiation Chemistry, ed. by P. Ausioos (Interscience, New York, 1968, pp 59—118). H.M. Rosenstock, J.T. Larkins and J.A. Walker, Int. J. Mass Spectrom. Ion Phys. J}, 309 (1973). W.A. Brand and T. Baer, Int. J. Mass Spectrom. Ion Phys. 55, 103 (1983). C. Lifshitz, M. Goldenberg, F.Y. Maiinovich and M. Peres, gggmMgg 55595599. 77, 453 (1982). D.H. Ave and M.T. Bowers: Gas Phase Ion Chemistry, Vol. II, (Academic Press, New York, 1979). S.G. Lies, D.M. Shold and P. Ausloos, J. Am. Chem. Soc. 102, 2540 (1980). 134 85. 86. 87. 88. 89. 96. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 135 J.C. Traeger and R.G. McLoughlin, J. Am. Chem. Soc. 103, 3647 (1981). R.G. Cooks: Collision Spectroscopy (Plenum Press, New York, 1978). D.W. Turner, A.D. Baker, C. Baker and G.R. Brundle: 5457955155: Photogiectron Spectroscopy (Interscience, London, 1970). J.H.D. Eland, Adv. Mass Spect. 7, 17 (1977). J.H.D. Eland, Int. J. Mass Spectrom. Ion Phys. 5, 143, (1972). -------- D.G. Streets, G.T. Ceasar, Mal. Phyg. 25, 1037 (1973). uunnun-uum-uuauu-um~n~ K. Kimura, S. Katsumata, Y. Achiba, T. Yamazaki: Mono Series Res. Inst. Appi. Electricity (Hokkaido University 25, 1978). R.S. Friedman, B.J. Kelsall, L. Andrews, J. Phys. Chem. 55, 1944 (1984). , R.S. Friedman, L. Andrews, J. Am. Chem. Soc. 107, 822 (1985). R.W. Hoffman: Dehygrobenzene and Cycloalkynes (Academic Press, New York, 1967 ). J.L. Franklin, J.G. Dillard, H.M. Rosenstock, J.T. Herron and F.H. Field: Ionization Potentials, Appearnace Potentials, and Heats of Enqrmation of Gaseous Positive ”2.21955, NSRDS-NBS 26, U.S. Government Print Office, Washington, 1969, Appendix 2. A.D. Baker, D.P. May, D.W. Turner, J. Chem. Soc. (B) 22, (1968). A.W. Potts, M.L. Lygus, E.P.F. Lee and G.H. Fattahallah, JChem J.H.S. Green, Spectrochim Acta 26A, 1503 (1970). J.D. Cox, Tetrahedron 15, 1337 (1970). J. Dannacher, H.M. Rosenstock, R. Buff, A.C. Parr and R.L. Stockbover, Chemthhys. 75, 23' (1983). -—u.. ........... Energy Chem (Engl-Transi.) .11, 88 (1977). 1779 (1981). B.D. Koplitz and J.K. McVey, J. Phys. Chem. 55, 2761 (1985). 106. 107. 108. 109. ' 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 136 R.D. Levin and S.G. Lias: Jouqization Potential and Appearance Potential Measurements, 1971-1981 (NSRDS-NBS 71; U.S. Govt. Printing Office, Washington, 1982). H.M. Rosenstock, R.L. Stockbaver and A.C. Parr, .11.:...---.9.1.1.11§1§ mean 2.7.. 745 (1930). H.M. Rosenstock, J. Dannacher, and J.F. Liebman, Radiat. Phys. 9.119.111: 3.9, 7 (1982)- K.R. Jennings, z. Naturforsch. 22A, 454 (1967). J. Harman, A.N. Yea, D.H. Williams, J. Am. Chem. Soc. 22, 2131 (1970). W.O. perry, J.H. Beynon, W.E. Baifinger, J.W. Amy, R.M. Caprioli, K.N. Renaud, L.C. Leitch and S. Meyerson, J. Am. Chem. Soc. 52, 7236 (1970). R.J. Dickinson and D.H. Williams, J. Chem. Soc. B 249 (1971). J.H. Beynon, R.M. Caprioli, W.0. Perry and W.E. Baitinger, JJJAQ, Chem. Soc. 55, 6828 (1972). .n....._~.-m-.on-—mu~~——n H.M. Rosenstock, R.L. Stockbauer and A.C. Parr, J. Chem. Phys. 72, 773 (1980). H.M. Rosenstock,. R.L. Stockbauer and A.C. Parr, J. Chem. Phys. 71, 3708 (1979). stress Such experiments are in progress, T. Baer, personal communication. J.P. Meier and O. Marfbaler, Chem. Phys. 22, 419 (1978). H.F. Grutzmacher and J. Lobemann, Anv. Chem. 795, 81 (1967). P. Natalis and J.L. Franklin, J. Phys. Chem. 69,2935 (1965). “no"... J.O. Noel and M.D. Newton, J. Am. Chem. Soc. 101, 51 (1979). M.J. S. Dewar, G.P. Ford and H.R. Reynolds, J. Am. Chem. Soc. 105, 3162 (1983). J.I. Brauman and L.K. Blair, J. Am. Chem. Soc. 92, 5986 (1970). “-0ss—ss J.L. Beauchamp and M.G. Caserio, J. Am. Chem. Soc. 95, 2638 (1972). 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 1 36. 137. 138. 139. 140. 141. 142. 143. 144. 145. J.B. Briggs, R. Yamdagni and P. Kerbarle, J. Am. Chem. Soc. 94, 5128 (1972). J.E. Barness and ET. McIver, Jr.: "The Gas-Phase Acidity Scale" in Gas Phase Ion Chemistry, Vol. 2, ed. by M.T. Bowers (Academic Press, New York, 1979). A.W. Castleman, Jr., P.M. Holland, D.M. Lindsay and K.I. Peterson, J. Am. Chem. Soc. 100, 6039 (1978). R.L. Woodin and J.L. Beauchamp, Chem. Phys. ill, 1 (1979). J.W. Larson and T.B. McMahon, J. Am. Chem. Soc. 104, 6255 (1982). M. Meot-Ner (Mautner), J. Am. Chem. Soc. 105, 4906 (1983). M.L. Fraser-Monteiro, L. Fraser-Monterio, J.J. Butler, T. Baer and J.H. Haas, J. Phys. Chem. 86, 739 (1982). J.E. Collin and G. Conde, Bull. Cl. Sci. Acad. Rog. Belg. 52,- 978 (1966). Y.C. Lee, A.I. Popov and J. Allison, Int. J. Mass Spectrom. Ion Phys 21.. 267 (1983)- R.T. Gray, D.N. Reinhoudt, K. Spaargaren and J.F. de Bruijn, J. Chem. Soc. Perkins Trans. 2, 206 (1977). G.J. Pederson, J. Am. Chem. Soc. 89, 7017 (1967). C.J. Pederson, J. Am. Chem. Soc. 92, 386 (1970). ...... E. Shchori, J. Jagur-e-Grodzinski and M. Schporer, J. Am. Chem. gun"... N.S. Poonia, J. Am. Chem. Soc. 96, 1012 (1974). A. Pullman, G. Giesstner-Prettre and Yu Y. Kruglyak, Chem. Phys. L953. 35, 156 (1975). M. Kajitani, A. Sugimori, N. Sato, K. Saki, H. Inokuchi and Y. Harada, Bull. Chem. Soc. Japan 32, 2199 (1979). S.W. Bensen: Thermochemical Kinetics (Wiley-Interscience, New York, 1976). 137 146. 147. 148. 149. 150. 151. 152. 153. 138 J.B. Pedley and J. Rylance: SUSSEX-N. P. L. Computer Analyzed Thermochemical Data: Organic and Organo-metallic Compounds (University of Sussex, Sussex, 1977). T. Baer, G.D. Willett, D. Smith and J. S. Phillips, J. Chem. Phys. 70, 4076 (1979). G.D. Willett and T. Baer, J. Am. Chem._Soc. 102, 6774 (1980). D.A. Jaeger and R.H. Whitney, JOrgChem 49, 92 (1975). M- Manson. 13.191199;....Ih9§!r199m- 1.141 (1969). K. Bystrom and M. Mansson, J.nghngSQg. Perkin 2, 565 (1982). F. Podo, G. Nemethy, P.I. Indovina, L. Radics and V. Viti, M91. BPL- 27, 521 (1974). ’ G.L. Sharon, J. F. Liebman and R.D. Levin, J. Phys. Chem. Ref. Data, 13, 695 (1984). 1311311111 1 I H "I H I II II" II. II l1 6 9 1 3 o 3 9 2 3 illlllllllll!