Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Negative ion photoelectron spectroscopy of cobalt tricarbonylnitrosyl Turner, Nicholas James 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_fall_turner_nicholas.pdf [ 2.47MB ]
Metadata
JSON: 831-1.0061771.json
JSON-LD: 831-1.0061771-ld.json
RDF/XML (Pretty): 831-1.0061771-rdf.xml
RDF/JSON: 831-1.0061771-rdf.json
Turtle: 831-1.0061771-turtle.txt
N-Triples: 831-1.0061771-rdf-ntriples.txt
Original Record: 831-1.0061771-source.json
Full Text
831-1.0061771-fulltext.txt
Citation
831-1.0061771.ris

Full Text

NEGATIVE ION PHOTOELECTRON SPECTROSCOPY OFCOBALT TRICARBONYLNITROSYLbyNICHOLAS JAMES TURNERG.R.S.C., The University of Kingston, Surrey, England, 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Chemistry)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1993© Nicholas James Turner, 1993In presenting this thesis in partial fulfillment of the requirements for an advanced degree atthe University of British Columbia, I agree that the Library shall make it freely availablefor reference and study. I further agree that permission for extensive copying of this thesisfor scholarly purposes may be granted by the head of my department or by his or herrepresentatives. It is understood that copying or publication of this thesis for financialgain shall not be allowed without my written permission.Department of ChemistryThe University of British ColumbiaVancouver, CanadaDate 2'f' ^1993DE-6 (2/88)ABSTRACTPhotoelectron spectra are reported for the Co(C0) 3NO- and Co(C0)2NO- anionsproduced from cobalt tricarbonylnitrosyl in a pulsed valve source. The 355 nm spectrawere obtained at a calculated electronic kinetic energy resolution of 96 meV using a newlyconstructed fixed-frequency negative ion photoelectron spectrometer, which is described.Electron affinities of 1.72 (±0.03) eV for Co(C0) 3NO and 1.73 (±0.03) eV forCo(C0)2NO are obtained. The photoelectron spectrum for Co(C0) 3NO- is consistentwith a geometry change from a linear Co—N-0 group in the neutral molecule to a bentconfiguration in the corresponding negative ion.11TABLE OF CONTENTSAbstract^ iiTable of Contents^ iiiList of Tables viList of Figures^ viiAcknowledgements viiiChapter 1 Introduction 11.1 General Introduction 11.2 Selection Rules 41.3 Spectroscopic Data obtained from Negative IonPhotoelectron Spectroscopy. 51.3.1^Electron Affinity 51.3.2^Calculation of Metal-Ligand BondStrengths 61.3.3^Vibrational Structure 81.3.4^Excited Electronic States 91.4 This Work 10Chapter 2 Experimental 112.1 General Description 112.2 Negative Ion Formation 112.3 Mass Selection 142.4 Laser Photodetachment of the Negative Ions 172.5 Timing 19Chapter 3 Characterization of the Negative Ion PhotoelectronSpectrometer 213.1 Characterization of the Mass Spectrometer 213.1.1^Resolution 213.1.2^Spectrometer Path Length 253.1.3^Other Studies 253.1.3a Halogen Mass Spectrum 253.1.3b Beam-Modulation Plate Voltage 263.1.3c Einzel Lens Voltage 263.2 Characterization of the Electron Spectrometer 293.2.1^Resolution 293.2.2^Photoelectron Spectrum of 0 - 303.2.3^Photoelectron Spectrum of 0 2- 333.2.4^Calibration of the Electron Spectrometer 383.2.5^Photoelectron Spectrum of 1- 39Chapter 4 Negative Ion Photoelectron Spectroscopy of CobaltTricarbonylnitrosyl 424.1 Results 424.1.1^Photoelectron Spectrum of Co(C0)3NO- 454.1.2^Photoelectron Spectrum of Co(C0)2NO- 454.2 Discussion 514.2.1^Mass Spectrum 514.2.2^Photoelectron Spectra 52iv4.2.3 Co(C0)3NO 534.3.3a Electron Affinity 534.3.3b Structure and Geometry 544.3.3c Vibrational Structure 594.2.4 Co(C0)2NO 594.2.4a Electron Affinity 594.2.4b Structure and Geometry 604.3^Further Work 63Chapter 5 Conclusion 65Bibliography 66LIST OF TABLES3.1 Mass Spectrometer Resolution for Oxygen Negative Ions 243.2 Mass Spectrometer Resolution for Oxygen Negative IonsOptimized using the Ion Focus 243.3 Effect of Beam-Modulation Plate Voltage on SpectrometerResolution 283.4 Effect of Einzel Lens Voltage on Spectrometer Resolution 283.5 Calibration of the Electron Spectrometer 374.1 Assignment of the Co(C0) 3NO Negative Ion Mass Spectrum 424.2 Effect of Ion Signal on the Co(C0) 3NO - Photoelectron Spectrum 504.3 Effect of Ion Signal on the Co(C0) 2NO- Photoelectron Spectrum 504.4 Fundamental Vibrational Frequencies For Co(CO) 3NO 624.5 Vibrational Frequencies for Cobalt Compounds in Low-TemperatureMatrices 62viLIST OF FIGURES1.1 Negative Ion Photoelectron Spectroscopy 22.1 Negative Ion Photoelectron Spectrometer 122.2 Beam-Modulation Plates 152.3 Spectrometer Timing Sequence 203.1 Oxygen Mass Spectrum 233.2 Halogen Mass Spectrum 273.3 Photoelectron Spectrum of 0 - 313.4 Anion to Neutral Transitions for Oxygen 323.5 Photoelectron Spectrum of 02- 343.6 Energy Level Diagram for the 02- + hv --> 02 + e Reaction 353.7 Photoelectron Spectrum of 1- 414.1 Mass Spectrum for Cobalt Tricarbonylnitrosyl 434.2 Calibrated Mass Spectrum for Cobalt Tricarbonylnitrosyl 444.3 Photoelectron Spectrum of Co(C0) 3NO- at 355 nm 464.4 Effect of laser beam power on the peak centred at 1.76 eV electronenergy in the Co(C0) 3NO- photoelectron spectrum 474.5 Effect of laser beam power on the peak centred at 0.78 eV electronenergy in the Co(C0)3N0- photoelectron spectrum 484.6 Photoelectron Spectrum of Co(C0) 2NO- at 355 nm 494.7 The Structure of Cobalt Tricarbonylnitrosyl 554.8 Correlation Diagram for the linear and bent configurations for theCoNO unit in cobalt tricarbonylnitrosyl 56viiviiiACKNOWLEDGEMENTSI would like to thank Dr. I. Waller for all her help, Cynthia Durance for the kinduse of her computer, Bruce Todd for the graph drawing program, Lui for all his assistanceand Joe for inspiration. Thanks also to Corinne, Bruce, Jamie, Dee, Lucy, Angelo,Penelope and Dave Williams.Finally to Andy and Melody, without whom....1CHAPTER 1. INTRODUCTION1.1 General IntroductionOrganometallic complexes are used extensively in the catalysis of chemicalreactions.' To fully understand the mechanisms by which catalysis occurs, it is necessaryto obtain detailed information about the catalytic species present. This could be used tointerpret the exchange of ligands between the catalyst and substrates, stabilization ofreactive intermediates and high catalytic selectivity.Coordinatively unsaturated organometallic complexes are often intermediates inhomogeneous catalytic pathways. A study of the interaction between the ligands and themetal centre could provide insight into how the complex behaves during the catalyticcycle. There has been limited research carried out on these intermediates since they areoften very reactive and hard to detect. The experimental technique used here, negative ionphotoelectron spectroscopy,' generates negative ions corresponding to unsaturatedorganometallic complexes. The complex of interest is isolated and then examinedspectroscopically to obtain the electron affinity. Vibrational structure and electronicspacings in the neutral complex as well as geometry changes from the negative ion to theneutral species can also be investigated.Negative ion photoelectron spectroscopy involves the formation of gas phasenegative ions. This has the advantage of removing any possible solvent effects, sincesolutions will contain the species coordinated to the solvent and not the "naked" moleculeitself. The negative ion of interest is isolated from all other species using a massspectrometer. This mass-selected ion is then crossed by the output from a fixed-frequencylaser, which induces a transition thus:hvAB- ^ > AB + e-2A small solid angle of the photodetached electrons is collected. The electronbinding energy is equal to the difference between the photon energy of the laser outputand the kinetic energy of the photoelectron. By analyzing the kinetic energy of thephotoelectrons, information about the states in the neutral molecule can be obtained(Figure 1.1). Negative ion photoelectron spectroscopy has been used in the study ofatoms, small clusters, reactive organic species and unsaturated metal carbonyl complexes.Figure 1.1 Negative Ion Photoelectron SpectroscopySchematic showing how a spectrum is obtained.Several other techniques are used in the study of coordinatively unsaturatedorganometallic complexes. Early work involved matrix-isolation techniques, 3 wherephotolysis is used to generate unsaturated metal species, which are then trapped usinginert, low temperature matrices. Alternatively the precursor is trapped in the low3temperature matrix and then irradiated. The complexes are then examined using infra-red,electron spin resonance and other spectroscopic techniques. More recent techniquesinclude ion cyclotron resonance spectroscopy, and transient infra-red absorptionspectroscopy.Transient infra-red spectroscopy has been used to monitor the gas phase productsof photolysis from organometallic complexes, primarily metal carbonyls. 4 An ultra-violetlight source is used in the photolysis of the metal carbonyl, the products are probed usinga laser which monitors modes in the unsaturated fragment by changes in transmission withtime at one infra-red wavelength. Absorption from the parent complex decreases withtime which is followed by a growth in absorption for daughter fragments. A spectrum isbuilt up by tuning the laser to different wavelengths and measuring the transmissionspectrum. The change of products with time can also be measured by comparing a seriesof spectra all measured at the same time after photolysis. In this way informationconcerning unsaturated intermediate photolysis products can be measured.The chemistry of gas-phase transition metal negative ions has also been studied togive valuable thermochemical data as well as metal-ligand bond strengths. Ion cyclotronresonance and its Fourier Transform analogue have been used to investigate the reactionof many unsaturated transition metal complexes.' Briefly, negative ions are formed byelectron impact ionization in a high vacuum cell. The cell is positioned between the polesof a magnet, which causes the ions to be constrained to circular paths. Individual ions arebrought into resonance using a radio frequency pulse and detected by a capacitor plate. Amass spectrum is built up by sweeping across the radio frequency range. In FourierTransform ion cyclotron resonance spectroscopy a complete mass spectrum can becollected with a single radio frequency pulse. Addition of reagent gases to the reactioncell allows ion-molecule reactions to be studied.The flowing afterglow method for negative ion production has also become moreprevalent in the last few years. 6 In these experiments electron impact ionization on the4neutral complex produces the negative ions. These travel down a flow tube using an inertcarrier gas. Reagent gases can be added downstream to induce ion-molecule reactions,the products of which are studied by the use of a quadrupole mass spectrometer.1.2 Selection RulesThe selection rules for negative ion photoelectron spectroscopy are governed bythe dipole approximation:Pabs^[ Wf 11' Wi (it ]2where 11 is the dipole operator and wf and kv i represent the final and initial states in themolecule respectively.A transition is allowed if Pabs # O. Selection rules are obtained by evaluating:Wf Wi cit^ [1]Using the Born-Oppenheimer Approximation, w may be separated into electronic andnuclear contributions:W = Wel WnSubstituting into [1]:f Nice , Wf,n^Wi,n dtWf,el I-1 Wi,e1 dlei • f klitn tifo dtn^[2]The electronic part of equation [2] provides the selection rules for transitions. The nuclearpart contains the Franck-Condon overlap of the vibrational and rotational wave-functions,which determines the intensities of these transitions. Consider the case when kv iei is aneven function. Since p. is an odd operator then w tel must be odd for a transition to occur.However, the final electronic wavefunction corresponds to the continuum wavefunction;5Wfel = tlicontinuumand therefore the "correct" symmetry may always be found, i.e. W Lt Wf W^ i,el (IT is alwaysan even integral and thus non-zero. A similar process occurs when Nri,ei is an odd function.Hence any one-electron transition from the anion to the neutral molecule is allowed, andintensities are governed by Franck-Condon factors.The ground state vibrational level of the anion is always symmetric, and for totallysymmetric modes in the neutral molecule all of the vibrational levels are also symmetric.This can lead to good Franck-Condon overlap to all levels giving large intensities in thephotoelectron spectrum. However, for anti-symmetric modes in the neutral molecule, theeven vibrational levels are symmetric with significant Franck-Condon overlap. Howeverthe odd vibrational levels are anti-symmetric with poor Franck-Condon overlap, and willnot be seen in the photoelectron spectrum.1.3 Spectroscopic Data obtained from Negative Ion Photoelectron Spectroscopy.1.3.1 Electron Affinity.An important fundamental quantity which can be measured using negative ionphotoelectron spectroscopy is the electron affinity. This is defined as the minimum energyrequired to remove an electron from the negative ion of a species: 7A13- ^> AB +where both AB and AB - are in their respective ground vibrational, rotational andelectronic states. The electron is produced with zero kinetic energy. Trends in electronaffinities yield information about the relative stabilities of the anions, as well as theimportance of electron delocalization, metal-ligand bonding and ligand polarizability. Theelectron affinity for a complex may also be useful in the understanding of its reactivity6towards both electrophilic and nucleophilic reagents. Many atomic electron affinities havebeen determined,$ however there have been far fewer values reported for organometalliccompounds containing transition metals. Lineberger and co-workers have investigated thehydrides and dihydrides of some first row transition metals and the unsaturated carbonylseries: Fe(CO)n (n=1 to 4) and Ni(CO)n (n=1 to 3). 9- 13 The measurements for the ironcarbonyl series indicated that in general addition of carbonyl groups to the metal centreincreases the electron affinity of the complex: 12Complex Electron Affinity (eV)Fe O.164(±0.035)FeCO 1.26 (±0.02)Fe(CO)2 1.22 (±0.02)Fe(CO)3 1.8 (±0.2)Fe(CO)4 2.4 (±0.3)This implies that as the size of the molecule increases, the extra charge on the anion ismore widespread causing an increase in electron affinity.1.3.2 Calculation of Metal-Ligand Bond Strengths.The electron affinity can also be used in conjunction with other thermochemicaldata to provide individual metal-ligand bond strengths using thermodynamic cycles, suchas the following for NiCO: Ni (s 2 d 8 ) + CO7D[NiCO] EA [Ni]Ni - (s2 d 9 ) + CONiCONiCOEA[NiCO] D[Ni--CO]where the bond dissociation energy of NiCO is calculated from:D[NiCO] = D[Ni --CO] + EA[Ni] - EA[NiCO]Using their observed electron affinities for the nickel carbonyl series, Ni(C0)„(n=1 to 3) Lineberger and co-workers reported a large variation of metal-ligand bondstrengths for individual complexes:"Complex Metal-ligand bondstrength (kcal mo1 -1 )(CO)3Ni—CO 25 + 2(CO)2Ni—CO 13 + 10(CO)Ni—CO 54 + 15Ni—CO 29 +15These values were calculated using appearance potentials for the appropriate negativeion. 14 The appearance potential is the minimum electron kinetic energy required for theheterolytic dissociation:Ni(CO)n + e-KE^Ni(C0)-n_1 + CO8and gives D[Ni(C0)-n .. 1 —00], the bond strength of the negative ion.Gas-phase acidities" can also be used with electron affinities to calculate bond strengths.The gas-phase acidity (AGacid) is defined by the free energy change of the reaction:AH --->^+ H+^AGacid = AGacid - TASacidand can be calculated for a particular compound from reactions with reference acids.By calculating particular metal-ligand bond strengths a greater understanding ofthe interactions of unsaturated organometallic complexes within catalytic mechanisms canbe gained.1.3.3 Vibrational Structure.The selection rules for vibrational bands in negative ion photoelectronspectroscopy allow any one-electron transition between the anion and the neutral moleculeto occur, with intensities governed by Franck-Condon factors. It is possible to observetransitions which are weak or even forbidden in the infra-red spectrum. Leopold and co-workers, using a high resolution instrument, observed the MCO bending and MCstretching modes of the neutral tricarbonyls of Cr, Mo and W, which had not beendetected by other spectroscopic techniques, as well as values for the C=0 stretchingfrequencies. 16Changes in the equilibrium geometry from the anion to the neutral molecule willcause vibrational progressions which may be visible in the negative ion photoelectronspectra. If the geometry of either state is known, then conclusions about the geometry ofthe other state can be made based on Franck-Condon modelling of the photoelectronspectrum. "Hot bands" in the photoelectron spectrum are caused by large initial statepopulations of excited vibrational levels in the anion, which are not thermally equilibrated9before photodetachment and cause peaks at lower energies than the ground state transitionin the spectrum. These can be used to obtain vibrational frequencies in the anion.The C=0 and N=0 stretching frequencies in unsaturated organometalliccomplexes are very sensitive to changes in geometry between the anion and the neutralmolecule. Shifts in these frequencies can be interpreted in terms of the degree of backbonding into the ligand orbitals and can be used as a test of the electron donating abilityfor a series of different ligand containing complexes. Typical values for the C=0stretching frequencies are 1800-2000 cm -1 and 1700-1900 cm -1 for the N=0 ligand.1.3.4 Excited Electronic States.Negative ion photoelectron spectroscopy may access excited states of transitionmetal complexes to provide information on their geometries and vibrational structures.Leopold and co-workers have reported the photoelectron spectrum for FeCO and haveresolved the vibrational structure of the 5E- excited state as well as the 3E- ground state. 17Charge transfer transitions of nitrosyl-containing species involve the promotion ofan electron from an orbital which is mainly metallic in nature, to one that is localized onthe NO 7c* orbital. The nitrosyl ligand is then reactive to electrophilic attack and the metalcentre to nucleophilic attack, leading to different reaction schemes for the excited state ofthe molecule compared to the ground state. The ligand may change from a linear to a bentconfiguration 18 possibly causing a vibrational progression in the photoelectron spectrum.Since the chemistry of nitrosyl containing organometallic complexes is a large andexpanding field, 19 the ability to access excited states could provide information which willincrease our understanding of how reaction mechanisms proceed.1.4 This WorkAlthough negative ion photoelectron spectroscopy has been used extensively toinvestigate atoms, radicals, clusters and organic species, little work has been reportedconcerning organometallic complexes. The aim of this work was to determine thephotoelectron spectra for nitrosyl containing anions prepared from cobalttricarbonylnitrosyl, Co(C0) 3NO, which is the simplest mono-nitrosyl containingorganometallic complex. Flowing afterglow studies indicated that the negative ionproduced by electron impact was Co(C0)2N0-, which could be re-combined with a COmolecule to form the anion of the parent molecule. 2° No parent anion for anorganometallic complex had been studied using this technique as most processes occur bydissociative attachment:m(C0)„, +  >M(C0),- + (n-x)CO where x = 0 to (n-1).Also there is evidence for a change in geometry from the anion to the neutralspecies for cobalt tricarbonylnitrosyl" and this should be visible in the photoelectronspectrum.The fixed-frequency negative ion photoelectron spectrometer used for thesestudies is described in Chapters Two and Three. Chapter Four contains the resultsobtained from the study of cobalt tricarbonylnitrosyl.1011CHAPTER 2. EXPERIMENTAL2.1 General DescriptionThe experiments reported here were carried out using a fixed-frequency negativeion photoelectron spectrometer (Figure 2.1). Briefly, negative ions are formed andisolated by mass-selection techniques. Species formed may be either the correspondinganion of the parent complex itself, or fragments and clusters formed from it. Theadditional electron is then photodetached using the output from a fixed-frequency laser.The photoelectrons are energy-analysed to provide information about states in the neutralmolecule which are accessible by a one electron transition.The spectrometer consists of three sections; ion formation, mass separation andlaser photodetachment. These are discussed separately in the following sections. Finallythere is a note on the timing sequence of the instrument.2.2 Negative Ion Formation.The design of the ion source is based on that of Johnson et c1. 21 , and has been usedelsewhere in the production of negative ions. 22 It consists of a pulsed molecular beam23,24whose path is intersected by an electron beam. The molecular beam is produced through apulsed valve (General Valve Corporation) to obtain a high density of gas for each pulseentering the source chamber, whilst allowing the chamber to remain at low pressure(2x10-5 Torr) during operation. The chamber is pumped by a 2000 1/sec Edwards Diffstakdiffusion pump.The pulsed valve operates at 301-Iz in conjunction with the laser and has a typicalvalve open time of 200 IAS per pulse. Samples are injected into the chamber through a91210Figure 2.1: Negative Ion Photoelectron Spectrometer.(1) pulsed valve (2) electron gun (3) acceleration plates(4) beam-modulation plates (5) einzel lens (6) ion focus(7) gate valve (8) 5mm aperture (9) electron detector(10) ion detector (+) laser interaction (DP) diffusionpump (TMP) turbo-molecular pump.130.5 mm valve orifice by flowing helium carrier gas at 20 psig (1000 Torr) over the volatilecompound in a sealed tube, producing a 10% mixture for the cobalt tricarbonylnitrosylexperiments reported here. Once in the source chamber, the molecular beam pulse isintersected by a 1 keV electron beam (from a Tektronix 250uA electron gun) close to thevalve nozzle. The fast electrons from this beam (e f-) ionize the carrier gas (A), producingslower secondary electrons (es-). These slow electrons then ionize the sample molecules(B) causing fragmentation (C and D), which can lead to recombinations and clustering:A + of^A+ +Mes + B ----> B -Mes- + B^C- + DThe third body (M) is important as it takes away excess energy from the reactantsthereby stabilizing the products. The ion distribution is dependent upon the position of theelectron gun with respect to the valve nozzle. 21The ions are produced in the continuum flow region of the beam and cool in thesupersonic expansion of the gas into the chamber. Collision-induced transfer between thevibrational and rotational degrees of freedom and the translational energies of the ionsleads to cooling of these degrees of freedom, causing them to occupy mainly the groundlevels. 25 This is necessary for photodetachment experiments, which involve one-electrontransitions from the ground state of the ions, because at room temperature they possesssufficient energy to populate low-lying excited states as well, causing "hot bands" in thephotoelectron spectrum. Measurements for SH - carried out on a similar instrument byKitsopoulos et a/. 26 yielded rotational temperatures of 50-75 K and for IHF a vibrationaltemperature of 100 K has been reported by the same group.2714A small amount of the ions pass out of the chamber through a 1 Omm aperture formass-selection, whilst the main bulk of the gas is extracted by the diffusion pump. Thewhole of this chamber is floated at -1 kV, in order to accelerate the ions to ground oncethey exit through the aperture.2.3 Mass SelectionThe ion pulse emanating from the source chamber passes through a stack of plateelectrodes connected by a resistor chain. This is used to accelerate the ions coaxially to 1kV by raising the voltage from -1 kV in the source chamber to ground.The ion pulse is then mass-selected by means of a beam-modulated time-of-flightmass spectrometer as developed by Bakker. 28 '29 The main advantage of this type of massspectrometer over the more traditional Wiley-McClaren 3° and Wien velocity filter31designs used elsewhere, is that this type produces no spread of kinetic energy in the ionbeam. The other designs depend upon an initial voltage pulse to select a portion of the ionbeam. This introduces a spread in the initial kinetic energies of the ions which lowers theresolution of the spectrometer.The mass spectrometer consists of a pair of flat plates, one centimetre apart,through which the ion beam travels, and a defining slit one metre away. The voltage isheld at a potential of V/2 on one of the plates, whilst the other plate is rapidly switched(10 ns fall time) from V to zero. This effectively changes the polarities of the two platesas the ion pulse passes through them (Figure 2.2). Typical values for the voltage, V, are20 to 50 Volts.Figure 2.2 Beam-Modulation Plates. Pathsof ions from different parts of the pulse showinghow the centre portion is selected. The polaritiesof the plates are switched as indicated.16The energy of the ions as they exit (Eextt) depends upon the forces exerted onthem by the beam-modulation plates, thus:Eexit = Einitial^E+^E_where Einitial is the energy of the ions as they enter the plates, E + is the force exerted onthem by the positive plate, and E_ the repelling force on the ions by the plate after thepotential switches. Any imbalance between E + and E_ will cause deflection of the ions.The middle part of the ion beam is selected by the following mechanism. Figure 2.2ashows what happens to ions in the front portion of the pulse. Initially the ions feel anupward force towards the positive electrode. The polarities of the plates switch when theions are nearly through them, causing the ions to be repelled until they leave the plates.As they have an overall upward force (E+> E_) through the plates, they are expelled at anangle and miss the slit.The ions in the centre of the pulse also feel an initial force towards the positiveplate. However, when they are exactly half-way through the plates, the polarities areswitched and the ions are then repelled with exactly the same force. As they exit theplates there is no net directional force on the ions (E += E_) and they travel through theslit.Figure 2.2c shows the path of the ions at the back of the pulse. Again, these ionsenter the plates and are attracted to the positive plate. The polarity switches and they arethen repelled by the plate. As these ions exit the plates they have a greater downwardforce (E+< E_) and consequently are deflected below the slit.Thus, the mass spectrometer selects a thin slice of the ions at each polarity switch.The ion bunch is focussed along the beam axis using an Einzel lens and ions separate downthe drift tube according to their individual masses. Before they pass through the definingslit into the UHV region of the spectrometer, ions of a specific mass can be temporally17focussed by means of the ion focus. This consists of a pair of metal discs with a 1 cm holethrough which the mass-separated ion beam passes.A negative voltage can be placed on the first disc, while the one furtherdownstream is kept at ground. This bunches the ions together by accelerating the backportion of the ions with respect to the front portion, since they experience the field for alonger time. The timing of the pulse and the magnitude of the voltage (typically -80 to -100 V) applied can be varied for optimal focussing. Once focussed, the ions pass throughthe 5 mm slit into the ultra-high vacuum (UHV) area of the spectrometer.The two regions of the mass spectrometer are differentially pumped using a pair ofEdwards Diffstak diffusion pumps (160 and 100 1/sec respectively). This results in apressure of 1x10 -7 Torr at the entrance to the UHV region when the instrument is in use.2.4 Laser Photodetachment of the Negative IonsThe UHV region of the spectrometer may be isolated by means of a gate valve. Itis maintained at a typical operating pressure of 4x10 -9 T by means of a 400 1/sec. LeyboldTurbovac 340M turbo-molecular pump. The ions pass through the region and aredetected by a chevron mounted pair of microchannel plates (Galileo MCP-025N, activediameter 25 mm). This amplifies the ion signal which is then processed by a 200MHztransient digitiser and stored on a computer. A grid on the front of the detector can beused to block the ion signal during photodetachment by means of a large (-2kV) voltage.The neutrals produced can then be monitored using a 300 MHz oscilloscope (Tektronix).Since they have the same arrival time as the parent ion, this confirms the mass of thephotodetached ions.18Photodetachment is achieved by the radiation from a pulsed Nd:YAG laser(Spectra Physics GCR 3) operated at 30 Hz. For the majority of the experimental workreported here the third harmonic, 355 nm ( 3.492 eV) was used. The laser beam passesinto the spectrometer through an ultra-violet fused silica (UVFS) window. The beamintersects the ions and then passes out of the instrument through another window. Thisallows easy alignment of the laser and helps to eliminate stray photons which can ejectelectrons from metal surfaces. The laser intersects a particular ion mass andphotodetaches electrons. As previously stated, the neutrals produced by this reaction canbe monitored and the laser timing adjusted to optimise the photodetachment. For thiswork the laser was unfocussed and had a beam diameter of 8 mm.The electrons produced enter a second 1 m time-of-flight spectrometer orientatedperpendicular to both the ion beam and the laser output. They are detected by a secondchevron mounted pair of microchannel plates (Galileo MCP-040N), with an activediameter of 4.2 cm. The signal passes through an external 10x pre-amplifier (Ortec 9301)and a 300 MHz discriminator (Phillips 6904) into a 200MHz transient digitiser (LeCroyTR8828D), which is triggered by a photodiode signal from a laser reflection. Thedigitiser is read after every laser shot by a signal averager and periodically transmitted to acomputer.The detector only measures electrons that are photodetached with kinetic energyperpendicular to both the ion beam and the laser path. This helps to minimize degradationof the spectrometer resolution by kinematic effects of the fast-moving ions, due toelectrons with the same centre-of-mass energy having different laboratory frame energies,and allows the effects of laser polarization to be studied. The flight-tube is shielded usingtwo concentric tubes of Conetic AA magnetic shielding (0.030" thick), which aredegaussed to minimize magnetic fields.Other effects on the resolution of the time-of-flight spectrometer are due toinaccuracies in the flight distances of the ions and the timing of the laser. Ions in particular19areas of the beam, or with different trajectories will have to travel slightly differentdistances through the spectrometer depending on when they are photodetached. Thetemporal pulse width of the laser (4-6 ns) is the main defining parameter of thespectrometer resolution. The energy resolution degrades as E 312, hence higher energyelectrons will have poorer resolution than lower energy electrons."To achieve a good signal to noise ratio in most of the experiments, approximatelythree hours of data acquisition at 30 Hz laser repetition rate were necessary. Both themass spectrometer and the electron spectrometer were characterized using the negativeions 0" and Of for the experiments performed at 355 nm.2.5 TimingThe over all timing sequence for the spectrometer is controlled by a StanfordResearch Systems DG535 pulse generator (Figure 2.3). The high voltage supplypretrigger to the Q switch of the laser is triggered first (To) followed by the pulse valve(A). Next the beam modulation plates are fired (B), along with the oscilloscope tomonitor the ion signal. The ion focus is pulsed after the beam modulation plates (C), andfinally the Q switch is triggered to produce the laser pulse (D). All of the timings can beadjusted individually to optimize the instrument for a particular ion species.20To I]^  tA ^n >^< 2ms >B ^ n >^<500,,s><111.s> ----r< 3ms >Figure 2.3 Spectrometer Timing Sequence.Including typical time intervals between pulses.(To) laser pretrigger (A) pulse valve (B) beam-modulation-plates (C) ion focus (D) laser Q switch.21CHAPTER 3 CHARACTERIZATION OF THE NEGATIVE IONPHOTOELECTRON SPECTROMETER3.1 Characterization of the Mass Spectrometer3.1.1 ResolutionThe resolution, or mass resolving power, of the beam-modulated time-of-flightmass spectrometer was determined using the negative ions of oxygen and its clusters. Atypical mass spectrum is shown in Figure 3.1. For this type of mass spectrometer theresolution should not change with time. 28Resolution is defined as the mass of the anion divided by the uncertainty in itsmass:R = m/Am^ [1]As the ions are measured as a function of time of arrival, it is necessary to expressequation [1] in the time domain. Using:E = 1/2mv2 = 1/2mL2/t2where E = energy of the anion, v = velocity of the anion, t = flight time of the anion, andL = length of the spectrometer. Then:t = Lm 1/2/(2E) 1/2^[2].For a slightly larger mass (Am) the difference of arrival times is given by:At = [L/(2E) 1/2 ].((m + Am) 1 "2 - m 1/2)At = [Lm 1/2/(2E) 1/2 ].((1 + Am/m) 1/2 - 1)Since Am/m is small, (1 + Am/m) 1/2 1 + Am/2mAt = [Lm 1/2/(2E) 1/2 ].(Am/2m)From equations [1] and [2]:At = t/2Rhence,R = t/2At.The resolutions for different oxygen anions O n- ( n =1 to 4) are given in Table 3.1,where At is the full width half maximum (FWHM) of each peak. The resolution of thespectrometer is 50 (±5) without the ion focus. However, using the ion focus (Table 3.2)results in up to a three-fold increase in resolution. The values were obtained by measuringthe oxygen mass spectrum, then employing the ion focus to each peak in turn and re-measuring the spectrum.The mass resolution is comparable to other time-of-flight spectrometers using aWien mass filter. 33 However it is lower than spectrometers utilizing beam-modulated orWiley-MacLaren designs, which can achieve resolutions of 150-300. 22,26 This could bedue to an error in the alignment of the spectrometer.More recent work, however, has indicated that there was a fault in the iondetector. Resolutions (t/2At) of 100 have now been measured for 0 2 - , rising to 190 withthe ion focus, making the mass spectrometer resolution comparable with those reportedpreviously.22163248 506410^20^30^40^50^60^70Mass (a.m.u.)Figure 3.1 Oxygen Mass Spectrum The spectrumshows the negative ions formed from oxygen. Thepeak at 50 a.m.u. corresponds to the [02—H20]' ion.23Table 3.1 Mass Spectrometer Resolution for Oxygen Negative IonsNegative Ion Flight time (pts) FWHM (f.ts) Resolution0- 13.14 0.14 472- 18.56 0.17 553- 23.18 0.24 4804 26.20 0.28 47Table 3.2 Mass Spectrometer Resolution for Oxygen Negative Ions Optimizedusing the Ion FocusNegative Ion Flight time (pts) FWHM (pts) Resolution0- 13.28 0.06 11002- 18.90 0.07 13504- 26.48 0.09 14724253.1.2 Spectrometer Path LengthAn important calculation for any time-of-flight mass spectrometer is the length ofthe flight tube. This is necessary for calibration purposes in order to determine the massof unknown sample ions to elucidate their structure. The length of the spectrometer iscalculated from:E = 1/2 mv2 = 1/2 mL2/t2hence:L = (2E t2/m) 1/2where E = the kinetic energy of the ion, m = the mass of the ion, v = the velocity of theion, L = the length of the spectrometer and t = the flight time of the ion.The kinetic energy of the ions is 1 keV and, for oxygen, both the mass and theflight times (Table 3.1) of the ions are known. The data collection program performs aleast squares fit on the values for each of the ions to find the value of L. The spectrometerpathlength is 1.44 m.3.1.3 Other Studies3.1.3a Halogen Mass SpectrumFigure 3.2 shows the mass spectrum for the halide negative ions F -, Cl-, Br-, andF. These were formed from a 5% mixture of each of three halocarbons (CF 2C12, CF3Br,and CF3I) in helium, pulsed directly into the source chamber to produce the halogennegative ions by dissociative attachment. The spectrum also shows the relative amountsof each ion formed. The mass resolving power is such that the isotopes of chlorine andbromine are clearly visible.263.1.3b Beam-Modulation Plate VoltageThe effect of beam-modulation voltage is illustrated in Table 3.3, showing anincrease in resolution with voltage. As the switching time of the plates is kept constant, anincrease in the voltage difference will produce a larger deflection of molecules at theleading and trailing edges of the selected ion packet. This results in less of theseperipheral ions passing through the defining slit and hence increases the resolution. 29However, the time of arrival for the ions is unaffected.3.1.3c Einzel Lens VoltageTable 3.4 shows the effect of Einzel lens voltage on resolution. Increasing thevoltage on the Einzel lens allows more peripheral ions to pass through the defining slit,lowering the resolution of the spectrometer. Again, although resolution changes withvoltage, arrival time is unaffected. Large changes in voltage cannot be determined as ionsare only detected over a small range of Einzel lens voltages.Figure 3.2 Halogen Mass Spectrum The spectrumillustrates the negative ions of the halogens. Theisotopes of bromine ("Br and "Br) and chlorine ("Cland 37C1) are clearly resolved.2728Table 3.3 Effect of Beam-Modulation Plate Voltage on SpectrometerResolution Using the negative ion 0 2 '.B.M.P. (V) Flight time (.ts) FWHM (µs) Resolution20.1 18.56 0.41 2324.8 18.59 0.29 3230.1 18.57 0.29 3234.8 18.58 0.19 4941.4 18.60 0.15 6245.0 18.57 0.16 5850.2 18.56 0.15 62Table 3.4 Effect of Einzel Lens Voltage on Spectrometer Resolution Using thenegative ion 02 '.Einzel Lens (V) Flight time (4s) FWHM (pts) Resolution650 18.60 0.12 78667 18.60 0.15 62680 18.60 0.17 55695 18.60 0.19 49293.2 Characterization of the Electron Spectrometer3.2.1 ResolutionFor a time-of-flight spectrometer, the drift time, T, of an electron in a flight tube oflength L, is related to the electron energy by:E= 1/2 me L2/ T2^[1]Where me is the mass of an electron. The resolution of the spectrometer isdetermined by uncertainties in both the flight time and length of the drift tube. These leadto uncertainties in the measurement of the electron energy, AE. 32 Since both time andlength are squared relative to mass, and energy is linear with respect to mass, equation [1]can be expanded to include these inaccuracies:[AE/E]2 = [20T/T] 2 + [2AL/L] 2^[2]Since only the time-of-flight of the electrons is measured experimentally,uncertainties in the drift tube length can be expressed as additional errors in timing (AT').These are related by:AT' = AL/Vwhere:V = (2E/me)'/230The total time width, AT tot, has components of AT and AT' and equation [2]becomes:From equation [1]:Substituting into equation [3]:[AF/E] 2 = [2ATtot/T] 2^[3]T2 = (1/meE).L2AE2 = [(2ATtot)2me/L2].E.E2^[4]AE oc E312The resolution of the electron spectrometer is thus proportional to the kineticenergy of the electrons to the power of three-halves. The resolution therefore increaseswith electron flight time.3.2.2 The Photoelectron Spectrum of 0 -A typical electron spectrum for 0 - prepared from oxygen gas pulsed into theelectron beam, is illustrated in Figure 3.3. The spectrum consists of 40 000 laser shots ata repetition rate of 30 Hz (22 minutes acquisition time), at a laser beam wavelength of355 nm (3.49 eV) and 1.4 W power with 8 mm beam diameter (laser fluence 73 mJ/cm 2).The full width at half maximum (FWHM) for the peak at 2.031 eV is 136 meV. This canbe compared to the time-of-flight electron spectrometer of Posey et al., which has aresolution of 40 meV (FWHM) for the 0.88 eV electron kinetic energy peak for ()-measured at 532 nm (2.34 eV).34Figure 3.3 Photoelectron Spectrum of O -at 355nm (3.492 eV). The centre of the peakat 2.031 eV corresponds to the electron affinityfor atomic oxygen, 1.461 eV32Since the resolution (AF) of a time-of-flight spectrometer degrades as E312, theFWHM of an electron peak at 0.88 eV can be predicted using the data from the measuredpeak at 2.031 eV and equation [4] above. The calculated value for the resolution at thiselectron energy is 39 meV which is in agreement with that reported by Posey et al.The spectrum for atomic oxygen is common in negative ion photoelectronspectroscopy where it is used as a calibrant due to the single peak observed. Although thespectrum appears as a single peak it consists of six individual peaks which can not beresolved by this technique. These six peaks represent the possible spin-orbit transitions forthe 0- + hv 0 + e reaction, as illustrated in Figure 3.4. It has been shown that thetransition corresponding to the electron affinity for oxygen, 2P312 --> 3P2 , occurs at thecentre of the spectrum peak 35 . This can be used in calibrating the instrument since theelectron affinity of oxygen is accurately known 36 at 1.461 122 eV./ \1 2 3 4 5 6Figure 3.4 Anion to neutralTransitions for OxygenEnergy level diagram for the0-(2P312,112) —> 0(3P2,1,0)reaction showing the six possibletransitions. Peak 3 corresponds tothe electron affinity for 0".3P03P3P22 P1/22 P3/2333.2.3 The Photoelectron Spectrum of 02 -The photoelectron spectrum for the negative ion of diatomic oxygen wasperformed initially in 1972 on the first spectrometer of this type, by Celotta et al. 37 Sincethen Posey et al. have reported a spectrum for 02 on a newly developed pulsedspectrometer34 and more recently Travers et al. have reported a higher resolutionspectrum, and a more accurate value, 0.451 ±0.007 eV for the electron affinity of diatomicoxygen than previously reported. 38For this spectrum, oxygen gas was pulsed directly into the path of the electronbeam to produce 02 in the same way as for 0 - . Figure 3.5 shows the photoelectronspectrum obtained after 40,000 laser shots at a repetition rate of 20 Hz with a laser beamwavelength of 355 nm and 1.4 W power (laser fluence 109 mJ/cm2). The acquisition timefor the spectrum was approximately 35 minutes. A high backing pressure of 0 2 (80 psigc/w 40 psig for the 0- photoelectron spectrum) was used to increase the 02 signal. Thepulsed valve was operated at 20Hz to maintain the stagnation pressure in the sourcechamber.The photoelectron spectrum consists of a series of peaks, compared with the singlepeak for atomic oxygen, due to the vibrational mode present in the oxygen molecule. Acomparison with the previously reported spectra indicates that peaks a-f correspond totransitions to the ground state of 02 , X3 Eg-(v) where v = 0 to 5, as shown in Figure 3.6.The first peak a, corresponds to the electron affinity for diatomic oxygen as it representsthe transition from the ground vibrational, rotational and electronic state of the anion tothe analogous state in the 02 molecule. Peak f probably contains contributions from boththe X3E (5) transition and also the excited state a 1 0g (0) transition (peak g in Figure3.6). The remaining peaks h-j are transitions to the first excited state of 02 , alAg(v).Since it is unclear whether peak f contains the alAg (0) transition it is not possible to assignElectron Energy (eV)Figure 3.5 Photoelectron Spectrum of 02at 355 nm (3.492 eV). The peak labelled acorresponds to the electron affinity of 0 2" at0.451 eV.02*/-•020235the other peaks with any certainty. Finally, the intensity of the transition corresponding topeak i appears high, indicating that it contains some portion of the 0 - transition from amultiphoton reaction such as the two-photon process:02- + hv^0- + 00- + hv --> 0 +Evidence for this was obtained from power-dependence studies carried out on the02 - photoelectron spectrum.32 a Ag0x 3;54320abcd ef3210gh i jFigure 3.6. Energy Level Diagram for the02- hv -+ 02 + e Reaction Peaks a-eare assigned in the 0 2- photoelectron spectrum,whereas peaks f-j are unresolved. Peak arepresents the electron affinity for diatomic oxygen.36It is possible to calculate the energies for each of the assigned peaks (a-e) in thespectrum. This is necessary for calibration purposes, where accurate energies of transitionare needed. The vibrational spacings (6v) between the ground state, v = 0, and the othervibrational levels can be determined using the equation:6v = (v + 1/2) 1/2coe - (v + 1/2)2coexe + (v + 1/2)3coeye ^[1]where w e is the oscillation frequency for oxygen and co exe and w eye are the first andsecond anharmonicity constants respectively. 39 Table 3.5 provides a summary of thecalculations for the vibrational spacings for the levels v = 0 to 4.The electron affinity for 0 2 is 0.451 (±0.007) eV, which corresponds to thetransition from the v'=0 state in the anion to the corresponding state in the neutralmolecule. For this work the laser beam wavelength was 355 nm, equating to a photonenergy of 3.492 eV . Hence the transitions from the X2IIg-(v) state of the anion to theground state of the diatomic oxygen molecule, X 3 -g(v) expel electrons with the followingkinetic energies:Peak Transition v E- v' Electron kineticEnergy (eV)a 0 4— 0 3.041b 1 <--- 0 2.845c 2 <— 0 2.658d 3 <— 0 2.471e 4E-00 2.287These values , in conjunction with the 0- spectrum, can be used to calibrate theelectron spectrometer at 355 nm.Table 3.5 Calibration of the Electron SpectrometerCalculation of the energy difference (AE) for the transitions in the 02-photoelectron spectrum using equation [1], where w e = 1580.361 cm -1 ,coexe = 12.073 cm -1 and coeye = 0.0546 cm -1 (from reference 39).(1 eV = 8065.48 cm').Peak VibrationalLevel (v)c ^(cm-1 ) Transition fromanion to neutralAE (eV)V^v 1a 0 787.18 - -b 1 2343.56 1 4-- 0 0.1932c 2 3876.30 2 +-- 0 0.3830d 3 5385.71 3 <-- 0 0.5701e 4 6872.12 4 <— 0 0.754437383.2.4 Calibration of the Electron Spectrometer.Species of known electron energies are used to calibrate the time-of-flight electronspectrometer. For the results reported here, the output from the laser was 355 nm andthe spectrometer calibrated with the electron affinities of 0 and 0 2 and the vibrationalspacings in the oxygen molecule. The calibration procedure is as follows: the knownrelative energies, Ere, of the calibrants can be converted to the lab energy, Elab, using asmall centre of mass correction factor:Elab = Erel - (me/M).Een,where me is the mass of an electron, M is the mass of the anion and Ecm is the centre ofmass energy of the anion.The kinetic energy of the electrons (Elab) is given by:Elab=- 1/2 Me L2/ T2^[1]where L is the length of the flight tube and T is the flight time of the electron. There ishowever, a finite time delay between pulsing the laser and triggering the transientrecorder, which results in a time offset, To. This is caused by the recorder being triggeredby a reflection from the laser, via a photodiode and also delays due to travel time in thewires. Hence equation [1] becomes:Elab 1/2 me L2/ (T - To)2T = ( 1/2 me/Elab)1/2.1, +.^[2]39Since T is measured and E rei is calculated as above, both the flight tube length andthe time offset can be determined by a least squares fit to equation [2]. The peak areas arekept constant during the transformation by:dT oc dE/E3 /2For the calibration at 355 nm the electron affinity for oxygen is 1.461122 ±0.000003 eV 36 and the energies of the peaks in the 0 2- spectrum are calculated in Section3.2.3. The instrument is calibrated before each experimental run to allow for variations inL and To, typical values for which are 100.4 cm and -70 ns respectively.3.2.5 The Photoelectron Spectrum of I -Some preliminary experiments were carried out using the laser beam wavelength of266 nm (4.661eV). At this energy, the atomic halides are used as calibrants due to theirsimple spectra and known electron affinities. 8 The iodine negative ion 1-, was preparedfrom a mixture of 5% CF3I in helium pulsed into the spectrometer. The r. ion is easilyformed, as shown in the mass spectrum of the halides (figure 3.2). Figure 3.7 illustratesthe photoelectron spectrum obtained at a laser beam power of 1.5 W. The spectrumconsists of 11,000 laser shots at a repetition rate of 20 Hz (laser fluence 117 mJ/cm 2).There is one peak in the spectrum, which corresponds to the transition:ITS () + hv —> I (2P312) + e"This transition represents the ground state to ground state transition andcorresponds to the electron affinity which has been measured at 3.0591 +0.0001 eV.4°40This spectrum, when combined with those for the other halides, can be used to calibratethe electron spectrometer for experiments carried out at 266 nm. Since the peak in Figure3.7 represents only one transition from the iodine anion to the neutral species, theresolution of the peak is the limiting resolution of the spectrometer, which is 82 meV(FWHM) at 1.60 eV.411 .00^1 .25^1 .50^1 .75^2.00^2.25Electron Energy (eV)Figure 3.7 Photoelectron Spectrum of r at 266 nm(4.661 eV). The Full Width at Half Maximum (FWHM)of the peak centred at 1.60 eV electron energy is 82 meV.CHAPTER 4. NEGATIVE ION PHOTOELECTRON SPECTROSCOPY OFCOBALT TRICARBONYLNITROSYL4.1 ResultsThe negative ion mass spectrum obtained for cobalt tricarbonylnitrosyl,Co(C0)3NO, is illustrated in Figure 4.1. A further mass spectrum using a mixture of 10%CF3Br and CF3I in helium as the carrier gas, and utilizing the Br - and I- peaks as internalcalibration standards for the mass spectrometer, is shown in Figure 4.2. The assignmentsfor the major peaks in the spectrum are given in Table 4.1.Table 4.1 Assignment of the Co(CO) 3NO Negative Ion Mass SpectrumResults from Figure 4.1.Peak (amu) Assignment117 Co(CO)NO-145 Co(C0)2NO-173 Co(C0)3NO-262 Co2(C0)3(NO)2-290 Co2(C0)4(NO)2-42145I173117100^150262■....—A—__/200^250^300290IMass (a.m.u.)43Figure 4.1 Mass Spectrum for Cobalt TricarbonylnitrosylU)C=0U441456171173I80^100^120^140^160^180Mass (a.m.u.)Figure 4.2 Calibrated Mass Spectrum for Cobalt TricarbonylnitrosylMass spectrum obtained using a mixture of CF 3Br and CF3I in helium asthe carrier gas. By calibrating the spectrum using the BC and I" peaks, thesample peaks have the masses indicated. Inset is the mass spectrum fordicobalt octacarbonyl (x10) calibrated in the same way. The peakat 171 corresponds to the Co(CO)4 anion.454.1.1 Photoelectron Spectrum of Co(C0)3NO-Figure 4.3 illustrates the photoelectron spectrum for Co(C0) 3NO- at 355 nm and1 W laser beam power (laser fluence = 52 mJ/cm2) calibrated using the oxygen atomic andmolecular anions as described in section 3.4. The spectrum contains a broad peak at 1.76eV, with a full width at half maximum of 233 meV, approximately 2 1/2 times greater thanthe calculated experimental resolution of 96 meV at this electron kinetic energy. There isalso a peak at 0.78 eV. Laser power studies were carried out on this spectrum, and theresults are shown in Figures 4.4 and 4.5. From Figure 4.4 the area of the peak at 1.76 eVincreases linearly with laser power indicative of a one-electron process, whereas for thepeak at 0.78 eV the area is non-linear with respect to energy (Figure 4.5), indicating amultiphoton process. With the electron gun turned off, the peak at 0.78 eV persisted,implying that the electrons are from photodissociation of the neutral background gas. Theeffect of space-charge on the spectrum is shown in Table 4.2. Space-charge effects occurwhen electrons photodetached from near the centre of the ion pulse gain energy from thesurrounding ions, causing a peak shift in the photoelectron spectrum. Changes in the peakcan be monitored by adjusting the ion current. As both the peak time and width do notvary significantly for the main peak, these effects are therefore minimal.4.1.2 Photoelectron Spectrum of Co(C0)2NO-The photoelectron spectrum for Co(C0) 2NO - at 355 nm and 1 W laser beampower is shown in Figure 4.6. It consists of an intense peak, centered at 1.77 eV, with afull width at half maximum of 187 meV, compared with a calculated experimentalresolution of 96 meV at this electron kinetic energy. Table 4.3 illustrates the effect ofvarying the ion current for the peak at 1.76 eV showing that space-charge does not affectthe appearence of the spectrum.0.5^1.0^1.5^2.0^2.5^3.0Electron Energy (eV)46Figure 4.3 Photoelectron Spectrum of Co(C0) 3NO- at 355nm0a):z..-Y0a)470^10^20^30^40^50Loser Energy (mJ/pulse)Figure 4.4 Effect of laser beam power on the peak centred at 1.76 eVelectron energy in the Co(C0) 3NO- photoelectron spectrum The lineargraph indicates a one-electron process.487006005000—x 400Ca)a-3002001 0000^50^100^150^200Laser Energy (mJ/pulse)Figure 4.5 Effect of laser beam power on the peak centred at 0.78 eVelectron energy in the Co(C0) 3NO- photoelectron spectrum Thenon-linear graph indicates a multi-photon reaction.1.00^1.25^1.50^1.75^2.00^2.25^2.50Electron Energy (eV)Figure 4.6 Photoelectron Spectrum of Co(C0) 2NO- at 355nm49Table 4.2 Effect of Ion Signal on the Co(C0)3NO- Photoelectron Spectrumfor the peak at 1.76 eV (1190 ns).Ion Signal (mV) Peak Center (ns) Peak Width (ns)25 1199 5820 1188 6115 1185 5110 1186 57Table 4.3 Effect of Ion Signal on the Co(C0)2NO" Photoelectron Spectrumfor the peak at 1.77 eV (1200 ns).Ion Signal (mV) Peak Center (ns) Peak Width (ns)40 1197 5030 1204 4920 1203 5010 1197 4450514.2 Discussion.4.2.1 Mass Spectrum.The mass spectrum for the negative ions formed from cobalt tricarbonylnitrosylwas first reported by Kiser.'" He found that both Co(C0); and Co(C0), I (NO)- ions,where n=1 to 3, could be detected, but did not observe the parent anion. More recentwork by several groups has shown that both Co(C0) 3NO" and Co(C0)4 (masses 173 and171 a.m.u. respectively) can be formed, as well as fragments of these species, dependingon the experimental conditions. McDonald and Schell found Co(C0) 2N0- was formedfrom Co(C0)3NO. When this ion was reacted with carbon monoxide gas, the anion of theparent molecule, Co(C0) 3N0-, was produced. Trace amounts of Co(C0)4 in theiroriginal mass spectrum were attributed to contamination by dicobalt octacarbonyl,Co2(CO)8, in the original sample. 2°McElvany and Allison found evidence for the breaking of both Co—CO and Co—NO bonds in the cobalt tricarbonylnitrosyl molecule. The relative intensities of theobserved fragments varied with changes in the electron energy of the source. 42 Both theMcDonald and Allison groups used a flowing afterglow source to prepare negative ionsfrom Co(C0)3NO. A mass spectrometric study of cobalt tricarbonylnitrosyl detected theCo(C0)4- ion and the Co2(C0)4(NO)2- cluster. 43The mass spectrum obtained from cobalt tricarbonylnitrosyl (Figure 4.1) wasshown to contain the species listed in Table 4.1 by calibrating the spectrum with thehalogen ions, Br- and 1- (Figure 4.2). This was necessary as the resolution of the originalspectrum, calibrated with oxygen, is such that it was not possible to distinguish betweenthe two possible types of fragment (Co(CO)n- and Co(CO) n_ 1 (NO)- , where n=1 to 4) dueto their similar masses. To unambiguously confirm the assignment, Co 2(CO)8 wasanalysed using the same internal standard procedure and the peak corresponding toCo(C0)4 was detected at 171 a.m.u.524.2.2 Photoelectron Spectra.Figure 4.3 illustrates the photoelectron spectrum for Co(C0) 3N0- . The differencebetween the laser photon energy (3.492 eV) and the electron kinetic energy of the mainpeak centred at 1.77 eV, is 1.72 ±0.03 eV. By assigning the centre of this peak as thetransition from the ground vibrational, rotational and electronic state of the anion to theanalogous state in the neutral molecule, this value represents the electron affinity forcobalt tricarbonylnitrosyl. The laser studies performed on the spectrum show that thepeak at 0.80 eV is due to a multiphoton process and is not considered in the spectralinterpretation. Figure 4.6 shows the photoelectron spectrum for Co(C0) 2N0- . As above,assuming that the centre of the main peak represents the transition from the ground stateof the anion to the ground state of the neutral, the electron affinity for Co(C0) 2NO is 1.73± 0.03 eV.Both of the spectra show one major peak. The Co(C0)3N0- peak is muchbroader than the corresponding Co(C0) 2NO - peak, 233 meV compared with 187 meV,which may represent a geometry change for the parent species from the anion to theneutral. The spectrum for Co(C0) 3NO- may contain unresolved hot bands in the 2.0 to2.5 eV range where there seems to be a long 'tail' to the peak. This could be investigatedfurther using gases such as argon or sulphur hexafluoride (SF 6). Argon is morepolarizable than helium, allowing energy transfer from the vibrationally excited levels inthe Co(C0)3NO- ion to electronic states in the argon matrix. SF 6 contains manyvibrational modes to which energy from vibrationally excited levels in the Co(C0) 3NO-anion can transfer. Both of these gases should reduce the populations of vibrationallyexcited levels which lead to hot bands in the photoelectron spectrum. Since the peak in theCo(C0)2NO- photoelectron spectrum has a much sharper cut-off on the low energy side,the presence of hot bands appears minimal.534.2.3 Co(C0) 3NO4.2.3a Electron AffinityModelli et al. have reported the electron affinity of cobalt tricarbonylnitrosyl at0.5 eV using MS-Xa calculations. 44 According to their calculations the electronegativecharge in the anion is mainly nitrosyl in character but with a significant contribution fromthe metal atom. Values of 34% on the nitrogen atom and 19% on the cobalt atom for theover all charge distribution on the anion were determined. Although the value for theelectron affinity is much lower than the one reported here, 1.72 ±0.03 eV, Modelli et al.did not allow for any possible changes in geometry which may occur in going from theneutral to the anion. The electron affinity reported here is much higher than thatexpected for a complex with a nineteen electron anion. The electron affinity is defined asthe minimum energy required to remove an electron from the Co(C0) 3NO" anion, whichis expected to be small since the corresponding neutral molecule is an eighteen electronstable molecule. This inconsistency could be due to a change from i 3-NO to 1.1 1 -NO in theanion giving an effective seventeen electron complex with a higher electron affinity.This is the first parent complex to be studied using this technique, as the carbonylcomplexes previously reported form negative ions by dissociative attachment and theparent anion is not seen. 12,13,16 Stevens-Miller and Beauchamp 45 have reported theelectron affinity of Co(CO) 4 at 2.3 eV calculated from the gas phase acidity ofCo(CO)4H and the metal-hydrogen homolytic bond energy. This value is higher than theCo(C0)3NO result reported here, even though the NO ligand has a greater electronaffinity than the CO ligand, due to the Co(C0) 4- anion being an eighteen electroncomplex.While no parent anion was found by McDonald and Sche11, 2° the Co(C0)2NO -fragment was detected. They concluded that the electron affinity for cobalttricarbonylnitrosyl must be greater than the Co—CO ligand binding energy. On the basis54that the binding energy is of the order of 30 kcal/mol, from heats of combustionexperiments, 46 then the electron affinity for Co(CO)3NO should be greater than 1.3 eV,which is in agreement with the result reported here. The reaction of Co(C0) 2NO- withcarbon monoxide produced Co(CO)3NO- suggesting that a similar mechanism may occurin the formation of the parent anion formed in our studies.The Co—CO bond may be more labile than the Co NO bond in cobalttricarbonylnitrosyl. The strength of the Co—NO bond has been estimated fromphotodissociation of Co(CO) 3NO by Hellner et al. 47 to be between 2.50 and 3.68 eV,which is approximately twice that of the average Co—CO bond.4.2.3b Structure and GeometryThe electronic structure of cobalt tricarbonylnitrosyl can be considered as aperturbation of the isoelectronic and isostructural nickel tetracarbonyl, Ni(CO)4 with atransfer of charge from the nickel atom to the carbon atom. Accompanying this chargetransfer is a change in symmetry from Ta to C3,. 48 Hedberg et a/. 49 have investigated thestructure of Co(CO) 3NO using gas-phase electron diffraction to measure bond angles andlengths. Their results are shown in Figure 4.7. The molecule has C3v symmetry and theCo—N-0 bond is linear.A comprehensive study of metal-nitrosyl bonds in organometallic complexes hasbeen carried out by Enemark and Feltham. They considered the three atoms as anindividual inorganic structural unit {MNO}x, where x is the number of d electrons in themetal atom plus an additional one from the TC * orbital of each nitrosyl group. 5° Using thisdesignation cobalt tricarbonylnitrosyl is an {MN0} 1 ° complex.55OAII 1.181N1 1.671CO8 \.843cr C CO^11 1.136^00Figure 4.7 The Structure of Cobalt Tricarbonylnitrosylshowing C3v symmetry. Bond lengths are given inAngstroms. Bond angles are LNCoC=107.7° andLCCoC= 1 1 1 .2°.Evans and Zink have proposed that in the lowest excited electronic state of cobalttricarbonylnitrosyl the nitrosyl group changes from a linear to a bent configuration.' 8Figure 4.8 illustrates this using a simplified molecular orbital diagram for the {MNO}'°structural unit. In the ground electronic state the highest occupied molecular orbital(HOMO) is the 4a, orbital, and the lowest unoccupied molecular orbital (LUMO) is the 4eorbital, comprising mainly NO 7C * antibonding character. Addition of an electronfacilitates bending of the nitrosyl bond because the electron can enter the more stable 5a.'orbital which is mainly nitrogen in character. If the nitrosyl group bends then the mostfavourable symmetry change has been calculated to be a reduction to Cs planar thus: 5 '0VNI— Co1C3v56Figure 4.8 Correlation diagram for thelinear and bent geometries for the CoNOunit in cobalt tricarbonylnitrosyl57They obtained evidence for this bending from the reaction of irradiatedCo(C0)3NO with hydrochloric acid. Two reactions could occur depending on theorientation of the nitrosyl group:hv1. Co—NO >^ NOC1Cl-hv2. Co—N [HNO] >^ 1/2 H2O + 1/2 N20The absence of NOC1, but the presence of both water and N 20 from the reactionprovides some evidence for the bending of the nitrosyl group in the lowest excited state ofCo(C0)3NO.Geiger et al. 52 state that in circumstances where the metal in a nitrosyl containingcomplex becomes electron-rich, i.e. it contains more than 18 electrons, then the nitrosylgroup may change from a three electron donor to a one electron donor to relieve theelectronic strain on the molecule:• •LnM—N=0 ^ LnM—N0m e- metal^(m - 2) e - metal13-NO II 1 -NOThis change occurs by the metal donating electrons to the nitrogen sp 2 orbital. Ithas already been shown that there is a correlation between electronic structure and thegeometry of the MNO group and that most bent nitrosyls are attached to electron-richmetals such as the cobalt atom in Co(C0)3N0- . McDonald also proposed that the nitrosylwas reducing its hapticity (the amount of electrons donated to the metal nitrosyl bond)58from TV to in reactions of the 17-electron Co(C0) 2N0- ion with different reagentgases. 2° The formation of Co(C0) 3NO- was considered to produce an r0-NO groupimplying that Co(C0)3NO- is an effective 17-electron complex containing a bent nitrosylgroup:(CO)2 Co-N=0 - + CO = (CO)3 Co-N\0 17 e complex^ 17 e- complexi3-NOThus although the symmetry of the Co(C0) 3NO molecule has been assigned to C3vwith a linear nitrosyl group, there is considerable evidence that attaching an electroncauses the nitrosyl group to bend and that the resultant anion changes symmetry to Cs.This change in geometry from the anion to the neutral would produce a progression in thephotoelectron spectrum for Co(C0)3N0- . The photoelectron spectrum reported here(Figure 4.3) is broader than that for the corresponding Co(C0) 2NO - spectrum, which isconsistent with the proposed symmetry change. The resolution of the spectrum howeveris not high enough to distinguish this progression and confirm the change in geometry.The possible hot bands present in the spectrum at high electron kinetic energies are due totransitions from vibrationally excited states of the anion to various vibrational levels in theneutral complex. This is consistent with a large geometry change in the formation of theanion in the source chamber, when many vibrational modes are excited. These modes coolas the expansion continues, however complete equilibration may not occur.594.2.3c Vibrational StructureThere are fourteen fundamental frequencies expected for Co(C0)3NO with C3vsymmetry. 53 Six will be A 1 , one A2 and seven E symmetry. The A2 mode is infra-redinactive. Several groups have reported vibrational spectra of Co(C0) 3NO in differentmatrices. 54-56 The most comprehensive study was carried out by Jones et al. 54 on sixisotopic species of the molecule. Table 4.4 summarises their results for the infra-red andRaman spectra of cobalt tricarbonylnitrosyl, the results are in good agreement with otherpublished data. 55,56 The major vibrational modes are at approximately 2000 cm -1 (0.25eV) for the CO and NO stretches, the bending modes occur at much lower energies (about500 cm-1= 0.06 eV). The photoelectron spectrum for Co(C0)3NO may contain thesevibrational modes, however the resolution is not high enough to distinguish individualmodes.4.2.4 Co(CO)2NO4.2.4a Electron AffinitySubstitution reactions involving the Co(CO) 2NO - anion carried out by McDonaldand Sche112° indicated that the electron affinity of Co(CO)2NO was greater than theketone (CF3)2C=O ±0.02 eV). Furthermore electron transfer was observedbetween Co(CO)2NO- and tetracyanoethylene (TCNE) indicating that the EA ofCo(CO)2NO was less than TCNE (3.17 ±0.02 eV). This brackets the electron affinity at1.46 eV < EA Co(CO) 2NO < 3.17 eV agreeing with the result reported here, 1.73 eV.By investigating the gas-phase reactions of anions prepared from a series oforganometallic complexes McElvany and Allison 42 concluded that:EA[Cr(CO)3] < EA[Co(C0)2] < EA[Co(CO)NO] < EA[Co(C0) 3] < EA[Fe(C0)3 ]where EA[Cr(CO) 3 ] = 1.349 ±0.006 eV 16 and EA[Fe(C0)3 ] = 1.8 ±0.2 eV.1260Since the electron affinity of the NO ligand is greater than that of the CO ligand then it isexpected that EA[Co(C0) 2NO] > EA[Co(C0)3]. The upper limit for the electron affinityof the cobalt series is now 1.73 ±0.03 eV. An over all ordering of electron affinities forthe cobalt carbonyl/mono-nitrosylcarbonyl series can now be written:1.35 eV < EA[Co(C0)2] < EA[Co(CO)NO] < EA[Co(C0)3] < EA[Co(C0)2NO]EA[Co(CO)3NO] =1.73 eV < EA[Co(C0) 4] 2.3 eV.4.2.4b Structure and Geometry.Very little information has been reported about Co(C0) 2NO. Crichton and Resthave measured the infra-red spectrum for Co(C0) 2NO, prepared by the u.v. photolysis ofCo(C0)3NO in low temperature matrices of argon and methane (about 20 K). 57 Theirresults indicated that Co(C0) 2NO possesses C2v symmetry:This structure has fifteen fundamental vibrational modes with the followingsymmetries: 6A 1 , A2, 5B 1 and 3B 2 . Again the A2 mode will be infra-red inactive.Crichton and Rest measured three of these fundamentals, and compared them to theparent molecule in the same matrices. The results obtained for both molecules are listed inTable 4.5. A comparison of the two compounds shows that the vibrational energies arevery similar. Although there has been no structure for the Co(C0) 2NO- anion reported, it61is isoelectronic with Ni(C0) 3- and Cu(CO)3 , both of which have planar n-3hgeometries. 58,59 Therefore it is very likely that Co(C0)2N0" will be planar also.The photoelectron spectrum for Co(C0) 2NO may contain the vibrational modeslisted above, but individual modes can not be distinguished due to the broadness of thepeak. The relative intensities of vibrational transitions are determined by the Franck-Condon overlap of the vibrational wave functions between the negative ion and the neutralmolecule. Vibrational levels corresponding to a significant change in geometry will beactive and appear as intense peaks in the spectrum. Bengali et a/. 16 have reported thephotoelectron spectra for the series M(CO) 3 (where M = Cr, Mo and W) which containonly weak vibrational peaks from the C=0 stretch and other vibrational modes, leadingthem to conclude that there is no geometry change from the anion to the neutral molecule.Since there is also no significant geometry change expected between the neutral and theanion for Co(C0)2NO, then the vibrational modes may not be excited and will not appearas intense peaks in the photoelectron spectrum. Furthermore, as both the anion and theneutral molecule are expected to have C2, symmetry, the antisymmetric modes (A 2, B 1 andB 2) will not be visible in the spectrum due to zero or very small Franck-Condonoverlaps.°62Table 4.4 Fundamental Vibrational Frequencies for Co(C0)3NO (from Reference 54).Symmetry Assignment Description Frequency (cm-1)Ai v1 sym. C—O stretch 2108v2 N-0 stretch 1822v3 Co—N stretch 596.5v4 sym. C—O bend 466V5 Co C 390sym.^stretchv6 C—Co C def 80.4sym.A2 v7 C-0 twist –408E v8 deg. CO- 0 stretch 2046.5v9 Co—N-0 bend 565.5v 10 Co^C--0 bend 483V 11 CO C-0 bend 441.2V 12 stretch 310.4deg. Co^CV 13 C— Co^deform.-C 85.2V14 C—Co--N deform. 64.9Table 4.5 Vibrational Frequencies for Cobalt species in Low Temperature Matrices(from Reference 57).Co(C0)3NOArgon (cm -1)Co(C0)2NOMethane(cm - ')Mode Argon (cm') Methane(cm - ')Ai [v(C0)] 2106 2103 2071 2064E [v(CO)] 2040, 2037 2035 1997 1993*Ai[v(NO)] 1819, 1815 1813, 1810 1785, 1783 1781, 1776,1774* Bands are split due to matrix perturbations.634.3 Further WorkThe work reported here could be extended to include the other members of thecobalt series: Co(CO)NO and Co(CO)n, where n = 1 to 4, to give values for their electronaffinities. The Co(C0)4 anion is an eighteen electron complex with a correspondinglyhigh electron affinity (2.3 eV). An accurate value for this species could be determined.The electron affinity values could be used to calculate individual metal ligand bondstrengths, if the corresponding heterolytic bond dissociation energies are measured usingother techniques. For example:Measurement of DRCO) 2N0Co --CO] would allow the calculation of the Co—CO bondstrength for Co(C0)3NO. Similar cycles could be used to calculate the corresponding Co—NO bond strength, and bond strengths for both Co—NO and Co—CO in the otherunsaturated cobalt complexes. Comparisons could be drawn between the relativestrengths of the carbonyl and nitrosyl ligands in different environments around the cobaltmetal centre.The photoelectron spectrum for Co(C0) 3NO - is the first reported for a parentanion for an organometallic complex. Future work could include the investigation of64other nitrosyl-containing molecules such as Fe(CO) 2(NO)2 , to see if other parent anionscan be formed, due to a change in hapticity of the NO ligand.The clusters observed in the mass spectrum for cobalt tricarbonylnitrosyl could beexamined to give electron affinities and spectroscopic information about their structures.This could be expanded to include mixed-metal clusters with the addition of othercomplexes (e.g. Fe(CO) 5) into the gas mixture.Finally, the extension of the current instrumental design to include a highresolution electron spectrometer will enable individual vibrational modes to be resolved inthe photoelectron spectrum, yielding detailed structural information about the complexesstudied.65CHAPTER 5. CONCLUSIONThe photoelectron spectra of the complexes, Co(C0) 3NO- and Co(C0)2N0- havebeen reported at 355 nm, using a newly constructed fixed-frequency negative ionphotoelectron spectrometer. The spectrometer has been characterized using the negativeions of oxygen and a preliminary experiment using iodine has been carried out using alaser beam wavelength of 266 nm.The electron affinity for Co(C0) 3NO has been determined at 1.72 ±0.03 eV andfor Co(C0)2NO, 1.73 ±0.03 eV. The electron affinity for Co(C0) 3NO is consistent withthe corresponding anion changing from a nineteen electron complex (which would beexpected to have a low electron affinity) to a seventeen electron complex by means of theNO group becoming rl 1 from ri 3 in the neutral molecule There appears to be a significantgeometry change from the anion to the neutral for Co(C0) 3NO which is reflected in thebroader spectrum compared to Co(C0)2N0- . This geometry change is consistent with thechange from a linear nitrosyl group in the neutral molecule and a bent nitrosyl group in thenegative ion. The low resolution of the spectrum does not allow this to be confirmed. Forthe Co(C0)2NO- spectrum, the sharper peak obtained indicates a smaller change ingeometry from the anion to the neutral molecule.BIBLIOGRAPHY.1. G.W. Parshall and S.D. Ittel Homogeneous Catalysis Wiley Interscience (1992).2. R. Mead, A.E. Stevens and W.C. Lineberger in Gas Phase Ion Chemistry VolumeThree Ed. M.T. Bowers. pp 214-248. Academic Press (1984).3. M. Poliakoff and J.J. Turner J. Chem. Soc. Dalton Trans. 1351 (1973), ibid.2276 (1974).4. E. Weitz J. Phys Chem. 91 3945 (1987).5. R.R. Squires Chem. Rev. 87 623 (1987).6. D. Smith and N.G. Adams in Gas Phase Ion Chemistry Volume One Ed. M.T.Bowers pp 53-86 Academic Press (1979).7^B.K. Janousek and J.I. Brauman in Gas Phase Ion Chemistry Volume Two Ed.M.T. Bowers pp 53-86 Academic Press (1979).8.^H. Hotop and W.C. Lineberger J. Phys. Chem. Ref Data. 14 731 (1985).9^A.E. Stevens, C.S. Feigerle and W.C. Lineberger J. Chem. Phys. 78 5420(1983).10^A.E. Stevens-Miller, C.S. Feigerle and W.C. Lineberger J. Chem. Phys. 87 1549(1987).11^A.E. Stevens-Miller, C.S. Feigerle and W.C. Lineberger J. Chem. Phys. 84 4127(1986).12^P.C. Engelking and W.C. Lineberger J. Am. Chem. Soc. 101 5569 (1979).13^A.E. Stevens, C.S. Feigerle and W.C. Lineberger J. Am. Chem. Soc. 104 5026(1982).14. R.N. Compton and J.A.D. Stockdale Mt. J. Mass Spectrom. Ion Phys. 22 47(1976).15. J.E. Bartmess and R.T. McIver in Gas Phase Ion Chemistry Volume Two ed.M.T.Bowers pp 87-121 Academic Press (1979).666716. A.A. Bengali, S.M. Casey, C. Cheng, J.P. Dick, P.T. Fenn, P.W. Villalta and D.G.Leopold J. Am. Chem. Soc. 114 5257 (1992).17. P.W. Villalta and D.G. Leopold J. Chem. Phys. 98 7730 (1993).18. W. Evans and J. Zink J. Am. Chem. Soc. 103 2635 (1981).19. G.B. Richter-Addo and P. Legzdins Chem. Rev. 88 991 (1988).20. R.N. McDonald and P.L. Schell Organometallics 7 1806 (1988).21. M.A. Johnson, M.L. Alexander and W.C. Lineberger Chem. Phys. Lett. 112285 (1984).22. R.B. Metz, A. Weaver, S.E. Bradforth, T.N. Kitsopoulos and D.M. Neumark J.Phys. Chem. 94 1377 (1990).23. T.E. Adams, B.H. Rockney, R.J.S. Morrison, and E.R. Grant Rev. Sci. Instrum.52 1469 (1981).24. W.R. Gentry and C.F. Giese Rev. Sci. Instrum. 49 595 (1978).25. D.H. Levy Sci. Am. 250 96 (1984).26. T.N. Kitsopoulos, I.M. Waller, J.G. Loeser and D.M. Neumark Chem. Phys.Lett. 159 300 (1989).27. I.M. Waller, T.N. Kitsopoulos, and D.M. Neumark J. Phys. Chem. 94 2240(1990).28. J.M.B. Bakker J. Phys. E. 6 785 (1973).29. J.M.B. Bakker J. Phys. E. 7 364 (1974).30. W.C. Wiley and I.H. MacLaren Rev. Sci. Instrum. 26 1150 (1955).31. L. Wahlin Nucl. Instrum. Methods. 27 55 (1964).32. D. Wilden, P. Hicks and J. Corner J. Phys. B 9 1959 (1976).33. H.B. Ellis and G.B. Ellison J. Chem. Phys. 78 6541 (1983).34. L.A. Posey, M.J. Deluca and M.A. Johnson Chem. Phys. Lett. 131 170(1986).35. F. Breyer, P. Frey and H. Hotop Z. Phys. A. 286 133 (1978).36.^D.M. Neumark, K.R. Lykke, T.Andersen and W.C.Lineberger Phys. Rev. A.32 1890 (1985).6837. R.J. Celotta, R.A. Bennett, J.L. Hall, M.W. Siegel and J. Levine Phys. Rev. A.6 631 (1972).38. M.J. Travers, D.C. Cowles and G.B. Ellison Chem. Phys. Lett. 164 449(1989).39. G. Herzberg Spectra of Diatomic Molecules Van Nostrand Reinhold (1950).40. C.R. Webster, I.S. McDermid and C.T. Rettner J. Chem. Phys. 78 646(1983).41. R. W. Kiser Recent Advances in Mass Spectrometry Ed. K. Ogata and T.Hayakawa pp 844-49, University Park Press (1970).42. S. W. McElvany and J. Allison Organometallics 5 416 (1986).43. T.B. McMahon Private Communication.44. A. Modelli, A. Foffani, F. Scagnolari, S. Torroni, M. Guerra and D. Jones J. Am.Chem. Soc. 111 6040 (1989).45. A.E. Stevens Miller and J.L. Beauchamp J. Am. Chem. Soc. 113 8765 (1991).46. F.A. Cotton, A.K. Fisher and G. Wilkinson J. Am. Chem. Soc. 81 800 (1959).47. L. Hellner, J. Masanet and C. Vermeil Chem. Phys. Lett. 83 474 (1981).48. B. Bursten, J.R. Jensen, D. Gordon, P.Treichel and R.F. Fenske J. Am. ChemSoc. 103 5226 (1981).49. K. Hedberg, L. Hedberg, K. Hagen, R.R.Ryan and L. H. Jones Inorg. Chem.24 2771 (1985).50. J.H. Enemark and R.D. Feltham Coord. Chem. Rev. 13 339 (1974).51. D.M.P. Mingos Inorg. Chem. 12 1209 (1973).52. W.E. Geiger, P Rieger, B. Tulyathan and M. Rausch J. Am. Chem. Soc. 1067000 (1984).53. R.S. McDowell, W.D. Horrocks Jr. and J.T. Yates J. Chem. Phys. 34 530(1961).54. L.H. Jones, R.S. McDowell and B.I. Swanson J. Chem Phys. 58 3757 (1973).55.^R.J. Clark Inorg. Chem. 6 299 (1967).6956. A. Poletti, A. Foffani and R. Cataliotti Spectrochimica Acta 26A 1063 (1970).57. 0. Crichton and A.J. Rest J. Chem. Soc. Dalton. 536 (1977).58. P.A. Breeze, J.K. Burdett and J.J. Turner Inorg. Chem. 20 3369 (1981).59. H. Huber, E.P. Kundig, M. Moskovits and G.A. Ozin J. Am. Chem. Soc. 972097 (1975).60^S.E. Novick, P.C. Engelking, P.L. Jones, J.H. Futrell and W.C. Lineberger J.Chem. Phys. 70 2652 (1979).

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061771/manifest

Comment

Related Items