Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Exploring reactivity of four square planar beta-ketoaminato cobalt compounds in the [+1, +2, +3] oxidation… Sherwood, Roger 2013

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

Item Metadata

Download

Media
24-ubc_2013_spring_sherwood_roger.pdf [ 4.34MB ]
Metadata
JSON: 24-1.0071923.json
JSON-LD: 24-1.0071923-ld.json
RDF/XML (Pretty): 24-1.0071923-rdf.xml
RDF/JSON: 24-1.0071923-rdf.json
Turtle: 24-1.0071923-turtle.txt
N-Triples: 24-1.0071923-rdf-ntriples.txt
Original Record: 24-1.0071923-source.json
Full Text
24-1.0071923-fulltext.txt
Citation
24-1.0071923.ris

Full Text

EXPLORING REACTIVITY OF FOUR SQUARE PLANAR BETA-KETOAMINATO COBALT COMPOUNDS IN THE [+1, +2, +3] OXIDATION STATES WITH APPLICATIONS TO CONTROLLED RADICAL POLYMERIZATION AND LIGNIN DEGRADATION  by  Roger Sherwood  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  THE COLLEGE OF GRADUATE STUDIES  (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA  (Okanagan)  April 2013  © Roger Sherwood, 2013  Abstract Four Schiff-base square planar β‐ketoaminato cobalt(II) compounds and their cobalt(III) ethyl derivatives were synthesized. The relative cobalt-carbon bond strengths of the latter were assessed in order to predict their relative reactivities and potential applicability as organometallic catalysts and as mediators for the radical polymerization of vinyl acetate, methyl acrylate, and acrylonitrile. 1H NMR, 19F NMR, UV-vis spectroscopy, and X-ray crystallography were used to characterize all novel Co(III)Et compounds, and previously-synthesized Co(II) compounds not fully characterized. Cyclic voltammetry of all cobalt(II) compounds revealed Co(II/III) redox processes in dichloromethane, and reversible Co(I/II) couplings in donor-solvent acetonitrile. Electronic modifications of the equatorial ligands afforded a series of complexes with varying degrees of reactivity. Relative cobalt-carbon bond strengths of cobalt(III) ethyl compounds were found by monitoring their thermal decompositions in the presence of a radical trap by UV-vis, and by observing equilibrium constants for persistent radical trapping of the ethyl radicals between cobalt(II) compounds by 19F NMR. Larger bond dissociation energies were found to be largely dependent on stabilization of the Co(III) species by donor ligands; however, solvent stabilization is also thought to play a significant role. Reactions explored included simple oxidation-reductions, Heck-type cross-coupling, lignin model degradation, electrocatalytic evolution of hydrogen, and mediation of radical polymerization of functionalized alkenes. Relative affinities for axial neutral donors and dioxygen were also explored in more detail.  ii  Preface The near entirety of research contributions presented in this document is the sole work of Roger Sherwood. X-ray quality crystals of the CoIII(Me/CF3)Et compound along with six runs of raw polymerization data were performed by Justin Black. Some references have been made to a communication submitted two years prior to the commencement of this M.Sc. degree. The crystal structure and synthesis of “CoIII(Me/Ph)Et” was presented in that work, but is included here for a more comprehensive thesis. Research in that communication was conducted by Dr. McNeil and myself, which laid the groundwork for Chapter 5.  Sherwood, R. K.; Kent, C. L.; Patrick, B. O.; McNeil, W. S. Chem. Commun. 2010, 46, 2456– 2458.  iii  Table of Contents Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iii Table of Contents ......................................................................................................................... iv List of Tables ............................................................................................................................... vii List of Figures ............................................................................................................................. viii List of Schemes ........................................................................................................................... xiii List of Symbols and Abbreviations ........................................................................................... xv Acknowledgements ..................................................................................................................... xx Chapter 1: Introduction ............................................................................................................... 1 1.1 Organometallic (1st row transition metal) radical reactivity and potential applications to organic synthesis and catalysis ............................................................................................. 1 1.2 Square planar cobalt complexes as a model for vitamin B12 ................................................ 3 1.3 Cross-coupling and other reactivity ..................................................................................... 7 1.4 Measuring bond dissociation energies ................................................................................. 8 1.5 Organometallic mediated radical polymerization .............................................................. 11 1.5.1 Free radical polymerization (FRP) ............................................................................ 11 1.5.2 Organometallic mediated radical polymerization (OMRP) ....................................... 12 1.5.3 Reversible termination (OMRP-RT) ......................................................................... 14 1.5.4 Degenerative transfer (OMRP-DT) ........................................................................... 15 1.5.5 Catalytic chain transfer polymerization (CCTP) ....................................................... 16 1.6 Reactivity towards dioxygen .............................................................................................. 17 1.7 Catalytic lignin degradation ............................................................................................... 20 Chapter 2: Synthesis and Structural Characterization of Co(II) and Co(III)Et Compounds .................................................................................................................................. 22 2.1 Preface ................................................................................................................................ 22 2.2 Synthesis ............................................................................................................................. 22 2.3 X-ray, 1H NMR, 19F NMR, and UV-vis characterization................................................... 25 2.4 Experimental ...................................................................................................................... 31 2.4.1 General considerations............................................................................................... 31 2.4.2 Synthesis of Co(III) alkyl compounds ....................................................................... 32 2.4.3 Yields ......................................................................................................................... 33 iv  Chapter 3: Stoichiometric and Catalytic Reactivity of Co(II) Complexes ............................ 34 3.1 Preface ................................................................................................................................ 34 3.2 Cyclic voltammetry ............................................................................................................ 34 3.3 Reduction and subsequent by reaction with alkyl halides .................................................. 38 3.4 Attempts for electrocatalytic evolution of H2 gas .............................................................. 43 3.5 Oxidative addition and subsequent reaction with Grignard reagents ................................. 44 3.6 Coupling reactions .............................................................................................................. 46 3.6.1 CoIII(Me/CF3)Et reactivity with alkenes .................................................................... 46 3.6.2 (E)-Stilbene ................................................................................................................ 47 3.6.3 (E)-Chalcone .............................................................................................................. 49 3.7 Reactivity with Dioxygen ................................................................................................... 54 3.7.1 Axial donor influence ................................................................................................ 55 3.8 Oxidation of veratryl alcohol ............................................................................................. 57 3.9 Experimental ...................................................................................................................... 63 3.9.1 General Considerations .............................................................................................. 63 3.9.2 Electrochemistry ........................................................................................................ 64 3.9.3 Preparation of CoIII(Me/CF3)Et via reduction and subsequent reaction with ethyl iodide ................................................................................................................ 64 3.9.4 Reactions of Co(II) compounds with iodine .............................................................. 65 3.9.5 Attempted synthesis of CoIII(Me/CF3)Et by oxidative addition of I2, followed by reaction with MgBrEt ........................................................................................... 65 3.9.6 Attempted reaction of CoII(CF3) with various acids to evolve H2 gas ...................... 65 3.9.7 Reactions of CoIII(Me/CF3)Et with (E)-stilbene ........................................................ 66 3.9.8 Reactions of CoIII(Me/CF3)Et with (E)-chalcone ...................................................... 67 3.9.9 Reactions of veratryl alcohol under 1atm O2 ............................................................. 67 Chapter 4: Studies of Cobalt-Carbon Bond Homolysis .......................................................... 69 4.1 Preface ................................................................................................................................ 69 4.2 Investigation of neutral axial donors on Co(II) and Co(III)Et compounds ........................ 69 4.2.2 Binding constants for THF-Co(III)Et compounds ..................................................... 70 4.2.3 Binding constants for solvents to Co(II) .................................................................... 74 4.3 Kinetic determination of the bond dissociation energy for Co(III)Et compounds using a radical trap. ............................................................................................................ 77 4.4 Determining the BDE for Co(III)Et compounds using 19F NMR ...................................... 82 v  4.5 Summary ............................................................................................................................ 86 4.6 Experimental ....................................................................................................................... 88 4.6.1 Kinetics experiments ................................................................................................. 88 4.6.2 Axial base experiments .............................................................................................. 88 4.6.3  19  F NMR experiments ................................................................................................ 89  Chapter 5: Organometallic Mediated Radical Polymerization .............................................. 90 5.1 Preface ................................................................................................................................ 90 5.2 Reversible termination (RT) experiments .......................................................................... 90 5.2.1 Vinyl acetate .............................................................................................................. 91 5.2.2 Methyl acrylate .......................................................................................................... 92 5.2.3 Acrylonitrile ............................................................................................................... 93 5.3 Degenerative transfer (DT) experiments in vinyl acetate .................................................. 94 5.3.1 Mechanistic studies .................................................................................................... 98 5.3.2 Varying temperature ................................................................................................ 102 5.3.3 Varying cobalt to monomer ratios ........................................................................... 104 5.4 Co(III)Et as radical initiator ............................................................................................. 106 5.5 Discussion ........................................................................................................................ 107 5.6 Experimental .................................................................................................................... 112 5.6.1 General considerations............................................................................................. 112 5.6.2 UV-vis monitoring of DT experiment in VAc ........................................................ 112 5.6.3 Polymerization reactions: degenerative transfer ...................................................... 112 5.6.4 Polymerization reactions: reversible termination .................................................... 112 Chapter 6: Conclusion .............................................................................................................. 114 References .................................................................................................................................. 118 Appendices ................................................................................................................................. 125 Appendix A: UV vis of Co(II) in various solvents ................................................................. 125 Appendix B: X-ray crystallographic data ............................................................................... 127 Appendix C: NMR spectra ..................................................................................................... 128  vi  List of Tables Table 1:  Factors affecting O2 binding in square planar cobalt compounds ........................ 19  Table 2:  Imine, β-ketone, and cobalt-carbon bond lengths for Co(III)Et structures ........... 27  Table 3:  Bond angles and degrees of planarity for Co(III)Et structures ............................. 27  Table 4:  19  F NMR and UV-vis characterization data for Co(II) and Co(III)Et  compounds ............................................................................................................ 28 Table 5:  1  Table 6:  Redox potentials of Co complexes determined from cyclic voltammetry  H NMR data of Co(III)Et compounds ................................................................. 30  measurements ........................................................................................................ 35 Table 7:  Percent conversions of veratryl alcohol to veratryl aldehyde under various conditions .............................................................................................................. 58  Table 8:  Screening tests for catalyst loading, pH, and axial base pyridine for conversions of veratryl alcohol to veratryl aldehyde under various conditions using CoII(CF3/Ph) ................................................................................................ 59  Table 9:  Conversion of veratryl alcohol to veratryl aldehyde for two separate bases ........ 62  Table 10:  Summary of thermodynamic values for binding THF to Co(III)Et compounds. . 73  Table 11:  Summary of kinetic data ....................................................................................... 81  Table 12:  Summary of enthalpies from experiments for the reactions shown in Fig. 33 ..... 87  Table 13:  Molecular weight data for polymer samples isolated from DT experiments at approximately 30% conversion............................................................................. 97  Table 14:  Polymer data for selected conversions in temperature-varying experiments with CoII(CF3/Ph) as catalyst .............................................................................. 103  Table 15:  Polymer data for varying catalyst concentration at ~50% conversion ............... 106  Table B.16:  X-ray crystallographic data ................................................................................. 127  vii  List of Figures Figure 1:  Vitamin B12 and four examples of simple model complexes in their Co(II) oxidation states........................................................................................................ 4  Figure 2:  Examples of some square planar compounds and their coordination to axial solvents ................................................................................................................... 6  Figure 3:  Energy diagram for Co-C bond homolysis ........................................................... 10  Figure 4:  Orbital diagrams for highest energy orbitals of square planar cobalt(II) d7 without axial base, axial base coordinated, and molecular oxygen coordinated with axial base....................................................................................................... 18  Figure 5:  Cobalt(II) and ethyl cobalt(III) compounds investigated in this study ................. 22  Figure 6:  ORTEP models for complexes CoII(CF3), CoIII(Me/CF3)Et, CoIII(CF3/Ph)Et, (THF)CoIII(CF3)Et, and CoIII(Me/Ph)Et................................................................ 26  Figure 7:  UV-vis of CoII(CF3) in THF and hexanes ............................................................. 29  Figure 8:  UV-vis of CoIII(Me/CF3)Et in diethyl ether, benzene (λmax = 652 nm), and THF ................................................................................................................ 29  Figure 9:  Cyclic voltammogram of CoII(Me/CF3): 5 mM in DCM, 50 mV/s; 1 mM, MeCN, 100 mV/s; 1 mM in MeCN, 100 mV/s..................................................... 35  Figure 10:  Cyclic voltammograms for attempts at reduction in DCM for all Co(II) compounds ............................................................................................................ 38  Figure 11:  Reaction of CoII(Me/Ph) with Na metal followed by cyclohexyl chloride in THF. .................................................................................................................. 41  Figure 12:  SN2 mechanism to form cobalt(III)alkyl vs. oxidative addition to form the octahedral structure via a less rigid square planar structure ................................. 42  viii  Figure 13:  Reactions of all cobalt(II) compounds with I2 in various solvents: THF, diethyl ether, and hexanes. .................................................................................... 45  Figure 14:  Monitoring cobalt species at selected wavelengths when reacted with (E)-stilbene.. .......................................................................................................... 48  Figure 15:  1  H NMR of the extract from reaction of CoIII(Me/CF3)Et with (E)-stilbene  with the unaccounted-for triplet/quartet outlined. ................................................ 49 Figure 16:  1  H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with  (E)-chalcone with a compact fluorescent bulb...................................................... 50 Figure 17:  Aromatic region 1H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with (E)-chalcone with a compact fluorescent bulb.. ................ 51  Figure 18:  Upfield region 1H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with (E)-chalcone with a compact fluorescent bulb.. ................ 52  Figure 19:  1  H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with  (E)-chalcone with a tungsten full spectrum bulb (200 W).. .................................. 53 Figure 20:  UV-vis of reaction of Co(II) compounds to atmospheric conditions ................... 55  Figure 21:  Short term reaction of CoII(CF3/Ph) and CoII(Me/CF3) with O2 in THF/tBuOK/DMSO solutions............................................................................... 56  Figure 22:  UV-vis spectra of CoIII(CF3/Ph)Et (5 μmol) in toluene with increasing THF. ..... 72  Figure 23:  Plots of absorbance vs. mmol of THF added to CoIII(CF3/Ph)Et (5 μmol) in toluene at three different wavelengths .................................................................. 72  Figure 24:  Van‟t Hoff plot of CoIII(CF3/Ph)Et (5 μmol) in toluene with 3.7 mmol THF ...... 73  Figure 25:  Calculations for finding solvent binding constants (e.g. KTHF) by simple arithmetic with known values.. ............................................................................. 75  Figure 26:  UV-vis spectra of CoII(CF3) in hexanes and various amounts of DMSO.. ........... 76 ix  Figure 27:  Plot of  -  vs  monitoring two different wavelengths for  solvent DMSO ...................................................................................................... 77 Figure 28:  Co(Me/Ph) species at the start/finish of an experiment in THF, and pseudo first-order rate plot for CoIII(Me/Ph)Et decomposition in TEMPO. ..................... 79  Figure 29:  Kinetics of the decomposition of CoIII(Me/Ph)Et at 65 °C in the presence of various equivalents of Co(II) (0 – 6) and radical trap TEMPO (2 – 40), and an Eyring plot assuming ko = 0, where TEMPO/Co(III)Et = 20 with no Co(II) added. ......................................................................................................... 81  Figure 30:  19  F NMR for a mixture of CoIII(Me/CF3)Et and CoII(CF3) in DMSO-d6. ............. 83  Figure 31:  Energy diagram and equilibrium equation illustrating solvent effects of DMSO and acetone on bond strengths .............................................................................. 85  Figure 32:  Bond dissociation energies for Co(III)Et compounds and corresponding oxidation potentials.. ............................................................................................. 86  Figure 33:  Equilibrium equations for which thermodynamic data were obtained experimentally....................................................................................................... 87  Figure 34:  First order rate plot for RT polymerization experiments in vinyl acetate and methyl acrylate ...................................................................................................... 91  Figure 35:  Rate plots for DT experiments at 1.1 and 1.3 equivalents of AIBN ..................... 95  Figure 36:  Plots of Mn and PDI (Mw / Mn) vs. % conversion for all catalysts (AIBN/Co = 1.3). .................................................................................................. 96  Figure 37:  Absorbance of OMRP under DT conditions for catalyst CoII(Me/CF3) before excess radicals are present. ................................................................................... 98  Figure 38:  Absorbance of OMRP under DT conditions for catalyst CoII(Me/CF3) after excess radicals are present. ................................................................................... 99 x  Figure 39:  Changes in absorbance for selected wavelengths before and after excess radicals are present. ......................................................................... 100  Figure 40:  UV-vis spectrum of CoIII(Me/CF3)Et in 20 % PVAc medium superimposed with 1.3 equiv. of AIBN/Co mechanistic data at 40 min. and 1 hour ................. 102  Figure 41:  Rate plots for varying temperature experimental data under degenerative transfer conditions, where AIBN/CoII(CF3/Ph)/VAc = 1:1.3:1000. ................... 103  Figure 42:  Dependence of PVAc molar mass Mn on monomer conversion (%) for bulk polymerization of vinyl acetate at 75 °C and [AIBN]/[CoII(CF3/Ph)] = 1.3. [VAc]/[CoII(CF3/Ph)] = 500 and 1000 ................................................................ 105  Figure 43:  Time dependence of ln([M]o/[M]) for the bulk polymerization of vinyl acetate at 75 °C. [VAc]/[CoII(CF3/Ph)] = 500 and 1000, and [AIBN]/[CoII(CF3/Ph)] = 1.3 .............................................................................. 105  Figure 44:  Rate plots for DT experiments with CoII(Me/CF3) where [Co(II)/AIBN/VAc] = 1:1.3:1000 and [Co(III)/Co(III)Et/AIBN/VAc] = 1:1:1.78:1000 ....................... 107  Figure 45:  Molecular weight data for DT experiments with Co(Me/CF3) where [Co(II)/AIBN/VAc] = 1:1.3:1000 and [Co(II)/Co(III)Et/AIBN/VAc] = 1:1:1.78:1000. ..................................................................................................... 107  Figure A.46: UV-vis of CoII(Me/Ph) in THF and toluene. ...................................................... 125 Figure A.47: UV-vis of CoII(CF3/Ph) in THF, diethyl ether, and hexanes. ............................. 125 Figure A.48: UV-vis of CoII(Me/CF3) in THF, diethyl ether, and toluene. ............................. 126 Figure A.49: UV-vis of CoII(CF3) in THF and hexanes ........................................................... 126 Figure C.50:  1  H NMR of CoIII(Me/Ph)Et in acetone-d6. .......................................................... 128  Figure C.51:  1  H NMR of CoIII(CF3/Ph)Et in acetone-d6. ......................................................... 129  Figure C.52:  1  H NMR of CoIII(Me/CF3)Et in acetone-d6. ........................................................ 130 xi  Figure C.53:  1  H NMR of CoIII(salphen)Et in acetone-d6. ........................................................ 131  Figure C.54:  1  H NMR of (THF)CoIII(CF3)Et in acetone-d6. Using integration from THF:..... 132  Figure C.55:  1  H NMR of (THF)CoIII(CF3)Et in dmso-d6. ....................................................... 133  Figure C.56:  1  H NMR of CoIII(CF3)Et in chloroform-d. .......................................................... 134  Figure C.57:  1  H NMR with presence of CoIII(Me/CF3)Et in acetone-d6 after reaction  with I2 and MgBrCH2CH3................................................................................... 135 Figure C.58:  19  F NMR for a mixture of CoIII(Me/CF3)Et and CoII(CF3/Ph) in  acetone-d6, 50 °C. ............................................................................................... 136 Figure C.59:  19  F NMR for a mixture of CoIII(Me/CF3)Et and CoII(CF3/Ph) in  acetone-d6, 50 °C. ............................................................................................... 137  xii  List of Schemes Scheme 1:  Homolytic vs. two-electron heterolytic bond cleavage........................................... 1  Scheme 2:  Example of both single electron transfer (SET) and 2-electron processes working in tandem in a generic cobalt(I) catalyzed radical cyclization reaction by vit.B12-type complexes. ........................................................................ 2  Scheme 3:  Previous use of organocobalt reagents/catalysts in alkyl Heck-type couplings ..... 8  Scheme 4:  Reaction equilibrium between an organic radical and Co(II).. ............................. 10  Scheme 5:  Ideal pathway for free radical polymerization of some alkene monomers initiated by radical initiator AIBN ....................................................................... 12  Scheme 6:  Mechanism for the [Co(acac)2]-mediated OMRP of VAc .................................... 14  Scheme 7:  Reversible termination mechanism showing a cobalt dormant species and its free form which accompanies liberation of a free radical. .................................... 15  Scheme 8:  Degenerative transfer mechanism ......................................................................... 16  Scheme 9:  Catalytic chain transfer polymerization mechanism ............................................. 16  Scheme 10:  Proposed mechanism for the oxidation of propenoidic phenols (R-OH) using Co(salen) ..................................................................................................... 20  Scheme 11:  Suggested mechanism for oxidation of veratryl alcohol by Co(salen) and dioxygen ............................................................................................................... 21  Scheme 12:  Synthetic routes to all Co(II) compounds ............................................................. 23  Scheme 13:  Reaction for synthesis of all ethyl cobalt(III) compounds .................................... 25  Scheme 14:  A summary of redox processes from cyclic voltammetry experiments ................ 38  Scheme 15:  Reduction followed by reaction with alkyl halide “R-X” (where X = Cl, Br, I) as a possible route to Co(III)R complexes (where R = alkyl group) ....................................................................................... 39  xiii  Scheme 16:  Reaction scheme attempting to synthesize a cobalt(III) hydride.. ........................ 39  Scheme 17:  Oxidative addition followed by reaction with Grignard reagent (where X = Cl, Br, I) as a possible route to Co(III)R complexes (where R = alkyl group) ........................................................................................ 44  Scheme 18:  Proposed stoichiometric cross coupling mechanism between Co(III)Et and (E)-chalcone/stilbene ............................................................................................ 47  Scheme 19:  Oxidation of veratryl alcohol to veratryl aldehyde. .............................................. 58  Scheme 20:  Proposed mechanism for the oxidation of benzylic alcohol veratryl alcohol by catalyst CoII(CF3/Ph)........................................................................................ 62  Scheme 21:  Equilibrium between solvent-free and THF-bound species of Co(II). ................. 70  Scheme 22:  Mechanistic pathways for decomposition of Co(III)Et and equations deduced from rate laws ......................................................................................... 78  Scheme 23:  Ethyl radical transfer between cobalt compounds Co(Me/CF3) and Co(CF3/Ph) and the derivation for Keq and  Scheme 24:  ................................................... 82  Proposed mechanism of a DT process with CoII(Me/CF3) and UV-vis bands corresponding to each catalytic intermediate ........................................... 101  Scheme 25:  Illustration of the concept of radical interchange for the degenerative transfer mechanism in Co(porphyrin).. ............................................................... 110  Scheme 26:  Mechanism showing how CCT occurring in DT could take away radicals in solution and regenerate Co(II) to slow down the overall polymerization process................................................................................................................. 111  xiv  List of Symbols and Abbreviations A  angstrom  Abs.  absorbance  acac  acetylacetonato  AIBN  2,2′-azobis(isobutyronitrile)  AN  acrylonitrile  atm.  atmospheric  B  base  bae  bis(acetylacetonate)ethylenediamine  BDE  bond dissociation energy  beta  beta  cat.  catalyst  CCT  catalytic chain transfer  CCTP  catalytic chain transfer polymerization  CF3  trifluoromethyl  Chel  chelate  Co(II)  cobalt (compound) in the +2 oxidation state  Co(III)R  alkyl bound cobalt (compound) in the +3 oxidation state  Co-C  cobalt-carbon  DCM  dichloromethane  ΔG  Gibb's free energy  ΔH  enthalpy enthalpy of activation  ΔS  entropy xv  entropy of activation DFT  density functional theory  dh2  bis(dimethylgloximato)  DMF  N,N‐dimethylformamide, HCON(CH3)2  DMSO  dimethyl sulphoxide, (CH3)2SO  dpp  1,3-bis(diphenyl-phosphanyl)alkyl  DT  degenerative transfer  ε  molar extinction coefficient (L∙mol-1cm-1)  E1/2  half-wave redox potential  Epc  cathodic peak potential  equiv.  equivalents  Et  ethyl  EWG  electron withdrawing group  Fc+/0  ferrocenium/ferrocene (redox couple)  FRP  free radical polymerization  GC  gas chromatography  GPC  gel permeation chromatography  gCOSY  gradient-selected correlation spectroscopy  h  Plank's constant (6.626 × 10-34 m2 kg∙s-1)  hrs  hours  i  Pr  isopropyl  IR  infrared (spectroscopy)  k  rate constant  kb  Boltzmann constant (1.381 × 10-23 m2 kg s-2 K-1) xvi  kcal∙mol-1  kilocalories per mol  kDa  kilodalton (dalton = 1000 gmol-1)  Keq  equilibrium constant  kobs  observed pseudo first-order rate constant  kp  rate constant for polymer propagation  krad  kobs – k1  Ksolv  solvent-bound equilibrium constant, (  L  ligand  λ  wavelength  λmax  wavelength at which a local maximum in absorbance  )  occurs M  moles per liter (mM = millimoles per liter)  M(n)  metal (n = oxidation state)  MA  methyl acrylate  Me  methyl  MeCN  acetonitrile  min  minutes  mL  milliliters  mM  millimoles per liter  Mn  number average molecular weight  M-R  metal-alkyl bond  MS  mass spectroscopy  Mtheo  theoretical molecular weight  xvii  MW  molecular weight  Mw  weight average molecular weight  NMR  nuclear magnetic resonance spectroscopy  OAc  acetate group  °C  degrees Celcius  OMRP  organometallic mediated polymerization  P  polymer chain  PDI  polydispersity index (Mw/Mn)  Ph  phenyl  ppm  parts per million  PRE  persistent radical effect  PVAc  polyvinylacetate  py  pyridine  R  refers either to an alkyl group or the gas constant (8.314 J∙K-1∙mol-1)  RSD  relative standard deviation  RT  reversible termination  salen  N,N‟-bis(salicylaldehyde)-ethylenediamine  salphen  N,N‟-bis(salicyidene)-o-phenylenediamine  sec  seconds  SET  single electron transfer  δ  chemical shift, or referring to a "sigma" bond.  SN2  bimolecular substitution reaction  sulphosalen  bis[(5-sulphonatosalicylaldehyde)ethylenediiminato] xviii  T  temperature  TAP  tetraanisylporphyrinato  TBDMSCl  tertbutyldimethylsillychloride, Si(Me2tBu)2(Cl)  t  Bu  tert-butyl  TEMPO  2,2,6,6-tetramethyl-piperidin-1-yl)oxyl  THF  tetrahydrofuran  TMP  tetramesitylporphyrin  μ-oxo  end-on binding mode of dioxygen to a metal in a 1:1 ratio  UV-vis  UV/visual (spectroscopy)  V  volts  V-70  2,2‟-Azobis(2,4-dimethyl-4-methoxyvaleronitrile)  VAc  vinyl acetate  W  Watt  X  halogen  xix  Acknowledgements I would like to express my gratitude to all the previous graduate and Honours degree students whom have aided me in any way in the chemistry research lab: Julia Conway, Jordan Sanders, Cory MacLeod, Addison Desnoyer, Alex Pickering, Lydia Gurley, Elizabeth Strohm, and of course Wen Zhou, without whom I would surely be confounded to all ends with indecipherable failures, buried beneath a sea of my own dirty glassware, and minus a good friend. I would also like to thank Carlos Jule for his contribution to acquiring data on the GC/MS, Dr. Brian Patrick for obtaining and refining X-ray crystallographic data, and Justin Black for his significant contribution in the synthesis of cobalt compounds and ligands, and in polymerization data acquisitions. His contributions are much appreciated in helping to make this a more comprehensive work. Professors and lab technicians I must thank for exceeding their duties as instructors and mentors and have made my degree a worthwhile endeavor include: Judit Moldovan, Dr. Susan Murch, Dr. James Bailey, Dr. Sandra Mecklenburg, Dr. Paul Shipley, Dr. Edward Neeland, Dr. David Jack, and Dr. Kevin Smith. Kevin embodies the true scientist, whose energy and boundless enthusiasm for research chemistry has been an inspiration to me throughout my degree. Lastly I would like to thank my supervisor, Dr. Stephen McNeil. He is the best teacher I have ever had, and the most patient and sympathetic person I have ever known. His advice has been finishing my own thoughts for years, and my deepest appreciation for contributing to my degree and for influencing my life as a student goes most affectionately to him.  xx  Chapter 1: Introduction This chapter briefly outlines important background information on chemical reactions involving catalysts similar to the specific β-ketoaminato cobalt(II) compounds under study in this thesis. Experimental results, detailed procedures, and discussions are presented in more detail in subsequent chapters.  1.1 Organometallic (1st row transition metal) radical reactivity and potential applications to organic synthesis and catalysis Organometallic compounds of first-row transition metals (i.e. row four in the periodic table) such as Cr, Mn, and Fe have relatively low energy barriers for single-electron oxidations connecting between adjacent oxidation states, e.g. Cr2+ and Cr3+; Fe2+, Fe3+, and Fe4+, etc.1 As a result, their metal-carbon and metal-hydrogen bonds are more susceptible to homolytic bond cleavage reactions in addition to the heterolytic reactions more commonly observed in organometallic chemistry. The former involves single-electron decoupling of a covalent bond, resulting in a formal one-electron change to the metal oxidation state, which is a trait uncharacteristic of the strictly two-electron (heterolytic) chemistry exhibited by most organic compounds or heavier 4d and 5d metals (Scheme 1).  Scheme 1: Homolytic (top) vs. two-electron heterolytic bond cleavage (boxed)  1  While characterization of paramagnetic intermediates2 via NMR techniques may be more onerous for single electron transfer processes, the advantage in developing such chemistry lies in the potential to construct alternative routes via radical processes to attain organic synthetics.3 Reaction mechanisms requiring metal compounds to undergo a variety of oxidation states via both single and two-electron processes are potentially achievable with cobalt, since cobalt compounds with oxidation states of -1, 0, +1, +2, +3, and even +44 are possible. Access to two-electron reduction/oxidations of cobalt (e.g. cobalt(I/III)), as well as adjacent oxidation states (e.g. cobalt (II/III)) by homolysis or single-electron reductions is exemplified by the various steps in the general mechanism5 for radical cyclization reactions (e.g. of alkyl bromides to activated alkenes)6 by vitamin B12-type complexes as shown below in Scheme 2.  Scheme 2: Example of both single electron transfer (SET) and 2-electron processes working in tandem in a generic cobalt(I) catalyzed radical cyclization reaction by vit.B12-type complexes (anionic species are omitted). 2  The reactions shown in Scheme 2 are similar to those explored in our model compounds. Single electron reduction of Co(II) by reaction with metals creates a Co(I) species, reaction of an alkyl halide with Co(I) affords a cobalt-bound Co(III)alkyl compound, and homolysis releases the carbon-based radical to react with an alkene and liberate the Co(II). This quickly forms a new Co(III)alkyl species with the new radical, and subsequent β-H elimination yields the new alkene along with a reactive hydridocobalt that rapidly reforms the Co(II) species. In order to propose viable mechanisms, predict new reactions, and postulate potential intermediates for catalyst systems, it is important to study the validity of a catalyst to undergo such intermediate processes as laid out in Scheme 2. How readily the various oxidation states are achieved can also influence the behaviour of the more labile ligands bound to cobalt (e.g. Co-C bonds), and this is related to the electronic nature and geometry of the ligand set.7 Studying electrochemical potentials by cyclic voltammetry and metal-alkyl bond dissociation energies of catalysts is therefore useful for gaining insight into the nature of the catalyst.8  1.2 Square planar cobalt complexes as a model for vitamin B12 Many metal compounds with four-coordinate chelating ligands are biologically inspired “model compounds” in that their electronic environments and reactivity mimic the active sites of enzymes involving metal cofactors, such as coenzyme B12, or “vitamin B12” (Fig. 1).  3  Figure 1: Vitamin B129 (right) (a) cyanocobalamin (b) adenosylcobalamin, and four examples of simple model complexes in their Co(II) oxidation states (left)  Coenzyme B12 is the most well studied biological system involving cobalt, the enzymatic activity for which is predicated on the ability of the cobalt cofactor to undergo heterolytic or homolytic Co-C bond cleavage,8,10 with the latter process generating an alkyl radical. Indeed, Co-C bonds in model compounds are also well known to undergo homolytic cleavage under the appropriate conditions of heat and/or light, and bond strengths for a variety of model compounds have been consolidated by Halpern and others, typically in the presence of nitrogen donor bases. Some ranges as examples are 17-25 kcal·mol-1 for the (py)Co(salphen)R system11 (salphen = N,N‟-bis(salicyidene)-o-phenylenediamine), and 17-22 kcal·mol-1 for (B)(dh2)Co(CH(CH3)C6H5)12,13 (dh2 = bis(dimethylgloximato)). In general, Co-C bond strengths of model compounds fall in the ranges of 19-32 kcal·mol-1,8,9 though for (H2O)Co(salen)CH3 (salen = N,N‟- bis(salicylaldehyde)- ethylenediamine) an incredibly high 40 kcal·mol-1 was measured.8 4  Schiff base complexes of cobalt with N/O donor atoms are exemplified by the Co(salen) and Co(bae) (bae = bis(acetylacetonate)ethylenediamine) systems, and reversible binding of dioxygen,14 carbon based radicals, and neutral axial bases15 are long-standing phenomena for these compounds. These characteristics were also explored in our compounds, which contain structurally-similar Schiff-base β-ketoaminato ligands. The effect of varying degrees of ligand substitution on their catalytic abilities were also explored.16,17,18 Applications to catalytic behavior, namely oxidation of the lignin model compound veratryl alcohol, and organometallic mediated radical polymerization, are some of the goals of this project. Square pyramidal five-coordinate cobalt(III) alkyl complexes often become axially coordinated opposite the Co-C bond in the presence of Lewis base donors to form six-coordinate octahedral 18 electron structures.17 In a vitamin B12 coenzyme analogue it has been suggested that catalytic activity associated with Co-C bond homolysis can be dramatically decreased by lowering the basicity of the axial nucleotide.10 Increasing the basicity of axial neutral donors to Co(II) has been correlated with increasing oxygen uptake19 to form dioxygen-bound Co(III)O2− species, as well as increasing Co-C bond strengths in decomposition studies with C6H5CH(CH3)Co(dh2)B.13 Both observations were attributed to the effect of stabilizing the electron-deficient cobalt to increase the barrier for Co-C or Co-O dissociation. However, extending this idea to all Co-C bonds in model complexes does not always work, as the Co-C bond in the water adduct of Co(salen)CH3 has been shown to be stronger than those with various N-donors.8 Short chain Co-C bonds from polymers to Co(acac)2 ((acac) = (acetylacetonato)) have been shown to be more readily released upon addition of axial donors; however, this was an unforeseen consequence of the formation of a stabilized CoII(acac)2(L)2 species,20 and not due to weaker bonds. In any case, the potential of solvents to axially coordinate is information worth investigating. A polymerization medium, for example, must contain monomer, and in the cases 5  of vinyl acetate, methyl acrylate, and acrylonitrile, each can potentially behave as a neutral axial donor to cobalt compounds in both the +2 and +3 oxidation states. For square planar cobalt(II) compounds, axial coordination by neutral donor ligands is possible in either one or two locations (Fig. 2). The tendency to coordinate 0, 1, or 2 axial bases can depend on the concentration of the ligand in solution,21 and is not always predictable; however, it has been suggested in cobalt porphyrin models that metal d-orbital energies should be more sensitive to axial ligation in electron-deficient porphyrin systems with weaker field ligands.22 This makes sense from inductive reasoning: weaker field donor ligands result in a more electron-deficient metal center.  Figure 2: Examples of some square planar compounds and their coordination to axial solvents13,15,22,23  6  1.3 Cross-coupling and other reactivity A metal-catalyzed cross-coupling reaction involves the coupling of molecular fragments via elimination from a transient organometallic or coordination complex. Palladium and nickel are the most common metal sources for cross-coupling reactions, and their use in carbon-carbon bond formation of sp, sp2, and sp3 carbons, as well as C-N bond formation,24 has drastically increased the number of conceivable routes for organic synthesis.25 The use of cobalt in crosscoupling reactions has ecological and environmental benefits over palladium (expensive) and nickel (toxic) catalysts,26 and its use in Heck-type reactions is complementary to the palladium system since decomposition by β-H elimination of organometallic intermediates is not necessarily a limitation.23,25,27 Unlike catalytic cycles involving Co(dpp) systems (dpp =1,3bis(diphenyl-phosphanyl)alkyl), which contain neutral-bound tridentate ligands and Co(0) intermediates,28 oxidative addition of alkyl halides to Co(II) are not feasible with our system. This is due to the relatively electron-poor nature of the Co(II) species, and reaction of alkyl halides must therefore be preceded by reduction to Co(I). Fortunately, mechanisms for coupling reactions involving vitamin B12 model cobalt compounds rarely invoke oxidative addition to Co(II).5 Previous work from Carreira, et.al29 involved intramolecular cyclization reactions, making use of the “cobaloxime” system (Co(dh2)) and base abstraction of the putative Co(III)-H species to regenerate Co(I) to provide turnover. Incidentally, this same catalyst, is the earliest known to be utilized for organometallic radical polymerization (OMRP),30 a process described in more detail later. Coupling has also been observed between styrene and various alkyl halides using zinc as a reducing agent.31 Heck-type coupling reactions where vitamin B12 model compounds (including vitamin B12 itself) that undergo Co(I)-Co(II)-Co(III) oxidation states typically follow some variation of the steps outlined in Scheme 3, such as the radical cyclization reactions shown in Scheme 2. Regeneration of Co(I) is accomplished with the use of stoichiometric reductants. 7  Scheme 3: Previous use of organocobalt reagents/catalysts in alkyl Heck-type couplings (modified from Carreira, et.al.)29  When Co(III)R species decompose homolytically under influence of heat or light in the presence of alkenes, product distribution depends on the substrate, the persistent radical (i.e. the Co(II) catalyst), and the concentrations of each in solution.32 Radical-radical coupling, β-H elimination, and radical polymerization are all possible reactions, the latter option resulting when significant alkene is present.  1.4 Measuring bond dissociation energies The bond dissociation energy (BDE) is the overall change in enthalpy (  ) between the  associated (lower enthalpy), and dissociated (higher enthalpy) species when a bond is broken.33 The relation between the Gibbs free energy of the reaction ( (  ) to  ) and the equilibrium constant  is shown in equations 1-4.  8  (equation 1) (equation 2) (equation 3) (equation 4)  If equilibrium constants can be measured at different temperatures, then the slopes from plots of ln term,  vs. can be used to find the enthalpy, and the intercept used to find the entropy  . This method was used for finding association constants (  ) of THF to cobalt  compounds in this thesis. For Co-C bond homolysis, this method for obtaining  cannot be employed because a  mixture of Co(II) with Co(III)R monitored at various temperatures would be virtually unchanging. As the initial radicals that become released from cobalt rapidly react in solution, they leave behind a small quantity of stable Co(II), and because re-trapping of the radical by Co(II) is so fast, the alkyl group is effectively suppressed as a free radical, and the species persists as a cobalt-bound or solvent caged species (dotted box in Scheme 4). This rapid trapping of a reactive radical by a stable one is one aspect of the “persistent radical effect” (PRE).34 Slow decomposition of organic radicals at higher temperatures might give a continuous very slow buildup of Co(II), but no valuable thermodynamic information can be extracted − only a vague idea of how fast a Co(III)-R decomposes.  9  Scheme 4: Reaction equilibrium between an organic radical and Co(II). Rate constants are approximate34 and are used to demonstrate the role of the PRE.  Kinetic studies can be carried out in order to reliably calculate the BDE for Co-C bonds. The addition of a radical trap such as TEMPO (2,2,6,6-tetramethyl-piperidin-1-yl)oxyl) to rapidly sequester R● is required in order to overcome Co(II) recombination.35 If the reversible binding by Co(II) process (i.e. k1) approaches the diffusion controlled limit (k1 > 1010 M-1s-1),36 the kinetic barrier for recombination of radical in low-viscosity solvents to Co(II) is sufficiently low ( as  ≤ 2 kcal·mol-1) that the bond dissociation energy (BDECo-R )11 can be approximated − 2 kcal·mol-1 = BDECo-R. The energy diagram is outlined below in Fig. 3.  Figure 3: Energy diagram for Co-C bond homolysis  10  The relationship between the overall observed rate constant (  ) and  is given by  equations 5-7:  (equation 5) (equation 6) (  )  ( )  * ( )  +  (equation 7)  Eyring equation demonstrating relation between pseudo first order rate constant, and activation enthalpy,  The Eyring equation shown in equation 7, and the values extracted from rate plots of (  ) vs. ( ) lead to determination of kinetic values (  ) and therefore  (i.e. the BDE).  1.5 Organometallic mediated radical polymerization 1.5.1 Free radical polymerization (FRP) Free radical polymerization (FRP) of olefins involves rapid propagation of a growing chain (polymer) constructed from similar molecular units (monomers), instigated and reactivated by highly reactive radical initiators such as AIBN (2,2′-azobis(isobutyronitrile) (Scheme 5). Oxygen reactivity with radicals may preclude air exposure under the reaction conditions, but the advantage lies in the applicability of this method to numerous sorts of monomers.37  11  Scheme 5: Ideal pathway for free radical polymerization of some alkene monomers initiated by radical initiator AIBN  Unfortunately, free radical polymerization results in branched polymers with undesirable macroscopic properties and high polydispersity values. This is the result of unavoidable competitive side reactions38 occurring at diffusion controlled rates, such as termination and disproportionation.37 A phenomenon called „backbiting‟ can also occur. This is where abstraction of a hydrogen atom from the middle of the polymer chain by another unstable terminal radical effectively rearranges the location of the radical about the polymer, and results in extensive branching.39 One can imagine how radical coupling reactions that terminate polymer chains indiscriminately can form polymer chains with unpredictable and disparate molecular weights.40 For this reason free radical polymerization is not considered controlled. 1.5.2 Organometallic mediated radical polymerization (OMRP) The basis of organometallic radical polymerization (OMRP) relies on rapid deactivation (trapping) of polymer radicals via a covalent bond to a metal. This is accompanied by slow re12  release of those radicals in solution in order to maintain a low concentration of reactive species while allowing the polymer to propagate.41 By limiting free radical polymerization and termination pathways, distribution of polymer chains with varying degrees of molecular weight is effectively reduced, or “narrowed,” and the process is therefore under “control.” Controlling the molecular weight distribution at the molecular level allows chemists to explore otherwise unobserved macroscopic properties of the polymer, such as flexibility and strength, and this is achieved when the following criteria are met42: (1) high degree of linearity in the polymer, characterized by a low molecular weight distribution (Mw/Mn or PDI), and (2) linear growth in the number average molecular weight (Mn) as a function of the % conversion of monomer into polymeric units. The ability of a system to yield polymers of predictable chain lengths by varying initiator to monomer ratios is also a desirable property. An apparent first-order reaction is characteristic of both controlled polymerization and FRP.43 The most successful OMRP of alkene monomers via cobalt mediation employ the cheap and readily available Co(acac)2 compounds.20,44,45,46,47,48,49 Besides being the most successful in narrowing molecular weight distribution, this system is also the most thoroughly studied mechanistically and computationally and therefore serves as a good model from which to understand radical polymerization using new cobalt catalysts. Decomposition of radical initiator V-70 (2,2‟-Azobis(2,4-dimethyl-4-methoxyvaleronitrile) in the presence of vinyl acetate monomer (VAc) and potential neutral axial donors (L) produces the mechanism outlined in Scheme 6 for this system, which was reproduced from Poli.50  13  Scheme 6: Mechanism for the [Co(acac)2]-mediated OMRP of VAc  1.5.3 Reversible termination (OMRP-RT) Reversible termination (RT) is perhaps the simplest model for OMRP (Scheme 7). The process self-regulates polymer growth by reversibly activating and releasing ζ-bonded polymer chain radicals with a metal M(n+1)-P complex.50 The efficiency of the M(n) trap ensures that a low concentration of radicals is maintained in order to dissuade radical-radical coupling reactions,45 effectively controlling the outcome of the polymer. In cobalt systems, Co2+ serves as the trap, M(n), forming a M(n+1)-P (or Co(III)-P) dormant species ζ-bonded to a polymer chain. Only upon temporary release of the radical polymer chain can propagation occur.  14  Scheme 7: Reversible termination mechanism showing a cobalt dormant species (left) and its free form (right, Co(II)) which accompanies liberation of a free radical  1.5.4 Degenerative transfer (OMRP-DT) Degenerative transfer (DT) is where organic radicals in solution exchange with those dormant-bound to a metal (Scheme 8). When equilibrium lies too much in favour of the dormant species M(n+1)-P, the RT process fails due to the lack of radicals in solution. A DT mechanism was therefore proposed in the literature to explain the successful control and rapid polymerization rates that were observed using the square planar Co(TMP) (TMP = tetramesitylporphyrin) system with an excess of radicals (vs. cobalt metal, M(n)) being injected into solution.43  15  Scheme 8: Degenerative transfer mechanism  1.5.5 Catalytic chain transfer polymerization (CCTP) Fast β-H elimination and subsequent addition of the M-H complex to an alkene to reform an M-R complex is referred to as catalytic chain transfer (CCT).43 Regeneration of radicals is required in order to undergo “living” polymerization, and so the reaction must be reversible for CCTP (Scheme 9).  Scheme 9: Catalytic chain transfer polymerization mechanism (modified from Poli50)  16  High yields of low molecular weight polymers with terminal alkene groups are the result of this chemical process, since more chains are produced than are cobalt species.50 CCTP is observed as a mechanism in Co(porphyrin) and Co(dh2) systems,51 though interplay with DT43,52 and RT42 is also possible. The fate of all polymer chains in CCTP are terminal alkene ends („dead chains‟), and monomers more likely to β-H eliminate are those containing hydrogen atoms, such as methacrylates, methylstyrene and methacrylonitrile; acrylates and vinyl acetate are less prone to undergo β-H elimination.23  1.6 Reactivity towards dioxygen Organic molecules are typically found in the singlet ground state and therefore have large thermal barriers for reaction with dioxygen. Transition metals, on the other hand, have the ability to adopt multiple oxidation states and paramagnetic electronic configurations, which allows them to spin-pair and interact with the ubiquitous and life-preserving diatomic molecule.53 This can be unfortunate in cases where catalysts are sensitive to decomposition, as dioxygen abounds outside the confines of a glove box. However, the propensity to reversibly bind dioxygen can also prove useful. If dioxygen interacts with the catalyst, its complete electronic nature changes and can become “active” toward organic substrates. Dioxygen binding to metal centers is a well-studied phenomenon, particularly for model compounds of monooxygenase enzymes involving iron,54 where reversible oxygen-binding is crucial. The β-ketoaminato complexes under study are structurally similar to well-known oxygen carriers of cobalt in the literature55 and because the first step in the catalytic mechanism for lignin degradation as well as other processes requires dioxygen binding, some insight into their reactivity with dioxygen is important.  17  Because cobalt in its +2 oxidation state acts as a nucleophile to electrophilic oxygen,15 only those compounds electron-rich enough can bind molecular oxygen to an appreciable degree. Factors influencing dioxygen binding are those that act on the metal center, either directly with the d-orbitals, or inductively via the ligand peripheries. More than often, axial bases in the form of donor solvents (usually amines) are therefore required in order to contribute the extra electron density needed to bind dioxygen. It does so by raising the  orbital of the metal that is oriented  axially towards dioxygen high enough in energy to spin-pair and bind the molecule19 (Figure 4).  Figure 4: Orbital diagrams for highest energy orbitals of square planar cobalt(II) d7 without axial base (left), axial base coordinated (middle), and molecular oxygen coordinated(middle) with axial base  It is worthy to note that increasing axial base strength tends to raise the bond dissociation energy of both Co-O as well as Co-C bonds along the same axis,56 presumably for the same reason: stabilization of a more electron-deficient cobalt that results from the axial bond. Table 1 summarizes the factors affecting O2 binding in square planar cobalt(II) complexes. 18  Table 1: Factors affecting O2 binding to square planar cobalt(II) compounds Factor Equatorial ligands19,57 Lewis Base adducts57 Solvent Effects Pi-donors/acceptors of Axial Bases  Sterics19,58  Temperature  O2 bonding affinity (↑ or ↓) EWG ↓ O2 affinity and vice versa; main factor in determining affinity stabilize Co-O2 bond, ↑ O2 affinity with ↑ base strength more polar solvent stabilizes Co(III)-O2 compound; ↑ O2 affinity pi-acceptors destabilize Co(III)-O2- , whereas pi-donors stabilize electron- deficient cobalt and ↑ O2 affinity 2:1 Co:O2 adducts can occur if steric bulk does not preclude it (e.g. bulkier porphyrins19 do not form 2:1 adducts). May influence axial base coordination as well Higher temperatures have been shown to produce 2:1 Co:O2 μ-oxo adducts more favourably with Co(salphen/salen)59  Model compounds known to bind dioxygen that are most structurally similar to ours are the Co(salen) and Co(bae) systems. The former is a good candidate for lignin degradation from preliminary studies in the oxidation of veratryl alcohol60 despite its known tendency to lock dioxygen into a dimeric structure when sufficient steric bulk is not present on the ligand.59 Electron paramagnetic resonance, X-ray and IR spectroscopic data on O-O bond lengths seem to support a significant transfer of electron density from cobalt onto bound dioxygen.14 Oxygen adducts of cobalt are therefore often represented as Co(III)-(O2)−, whereupon a bound “superoxo” resides and has the potential to react with susceptible electrophiles. Although the ability of Co(salen) complexes to bind dioxygen is well studied, degradation to species incapable of performing the task is a major barrier for grander scale applications.59,61 A similar problem exists in utilizing metalloporphyrins for oxidation reactions.62  19  1.7 Catalytic lignin degradation Lignin is an abundant and ubiquitous amorphous phenolic organic network polymer.63 As a major constituent of wood, it is generated in massive quantities in the pulp and paper industry as a waste product.64 Limited uses for lignin can be attributed to its inertness to chemical degradation due to the vast web of C-C and C-O bonds in its chemical structure, and methods to break the material down into predictable and useful subunits63 have therefore been the subject of many years of research.54,62,65,66 Conventional harsh oxidation methods to remove propenoidic phenols in lignin have prompted investigation of catalysis using well-known oxygen-activating Co(salen) compounds from various research groups.65,67,68 The affinity of cobalt for phenolic groups,69 as well as its ability to reversibly bind dioxygen are responsible for its potential use in aerobic oxidation of model compounds, a first step in the depolymerization of lignin. The reaction with phenolic substrates is believed to occur as shown in Scheme 10, with facile homolytic release of phenoxy radical RO●,70 and subsequent reaction with dioxygen at the substrate.65  Scheme 10: Proposed mechanism for the oxidation of propenoidic phenols (R-OH) using Co(salen)65,68  20  A study for the oxidation of veratryl alcohol (a benzylic alcohol lignin model substrate), using a series of Co(salen) type catalysts showed that although Co(salen) was the most effective under certain reaction conditions, that Co(bae), a β-ketoamine of cobalt more similar to those in our study, was a very close second.60 For this reason, the same substrate and variables were investigated, only using our cobalt system instead of Co(salen). The mechanism in Scheme 10 only applies to phenolic substrates, but activation of dioxygen followed by hydrogen abstraction from the substrate is still likely occurring, resulting in exclusive formation of the aldehyde. The proposed mechanism for oxidation of benzylic substrates from Repo and Weckhuysen67 is shown in Scheme 11, and differs from Scheme 10 in that it does not invoke a superoxo cobalt intermediate, but rather a μ-peroxo bridge between Co(salen) molecules.  Scheme 11: Suggested mechanism for oxidation of veratryl alcohol by Co(salen) and dioxygen.67  21  Chapter 2: Synthesis and Structural Characterization of Co(II) and Co(III)Et Compounds 2.1 Preface Structures were validated in this chapter, along with discussion of synthetic techniques employed in order to obtain all Co(II) and Co(III) compounds. Various spectroscopic methods were used for characterization and UV-vis spectra in various solvents for all Co(II) and Co(III) ethyl (Co(III)Et) compounds is especially important for applications to experiments in subsequent chapters. All main structural data are tabulated in this section, with X-ray crystallographic and UV-vis spectra of Co(II) compounds found in Appendix A and B.  2.2 Synthesis The cobalt(II) compounds under study are previously established square planar “Schiffbase” complexes with varying degrees of electron withdrawing substituents (phenyl, methyl, and fluoro groups) on the peripheries of the equatorial ligand.71 Alkylations of these compounds afforded their corresponding Co(III)Et compounds. The structures and abbreviations used throughout this work for these eight compounds are found in Figure 5.  Figure 5: Cobalt(II) and ethyl cobalt(III) compounds investigated in this study 22  Synthesis of the Co(II) compounds were performed according to literature methods71 with samples of CoII(CF3) submitted for the first time for X-ray crystallographic analysis. The tetradentate β-ketoaminato ligands are formed by reaction of ethylene diamine with two equivalents of the β-ketone via a Schiff-base condensation reaction. Regioselectivity to dissuade self-cyclization is achieved in the case of the (Ph/CF3) and (CF3) ligands by first protecting one side of the β-ketone using tert-butyldimethylsilyl chloride (Scheme 12). Co(II) compounds are made either by subsequently reacting deprotonated ligands with the appropriate cobalt salt via salt metathesis, or by aminolysis when cobalt(II)(NSiMe3)2(THF) is utilized as the cobalt source.  Scheme 12: Synthetic routes to all Co(II) compounds 23  Synthetic design of novel square pyramidal Co(III)Et compounds was approached using numerous techniques, the original intention being to find a more widely applicable method where a variety of alkyl groups could be incorporated onto the cobalt metal center. Ideally, this would be accomplished by reaction of cobalt(III)(X) (X = halide) with a Grignard reagent (RMgX, where R = alkyl group) or alkyl lithium reagent (RLi), for which numerous alkyl groups can be easily selected. Organocobalt compounds of similar RCo(bae) compounds (including CH3CH2Co(bae)) have been obtained this way,72 along with unique examples of Co(I) alkyls;73 however, no such success was found in isolating Co(III)R using this method for our complexes. Attempts at oxidative addition of silver triflate followed by reaction with Grignard reagents gave low yields of product when reacted with CoII(Me/Ph).74 Generating reduced species or cobalt hydrides by sodium borohydride or H2 followed by reaction with activated alkenes or alkyl iodides is also a well-known method for obtaining alkylcobalamins,13 RCo(salphen) compounds,75 and alkyl cobaloximes.76 This method (using sodium borohydride) was unsuccessful when attempted with CoII(Me/Ph), and was therefore abandoned as a plausible route for alkylations of other compounds. In the future, the use of PdCl2 with sodium borohydride before addition of the alkyl iodide should be attempted, as this has been shown to be effective for methylations of Co(salen) and Co(bae) compounds.77 The implementation of a new method using triethyl borane as a radical carbon source was instead employed for obtaining Co(III)Et compounds, as immediate colour changes were observed upon trapping of liberated ethyl radicals from boron to cobalt.78 This reaction was observed in previous work for preparation of CoIII(Me/Ph)Et,79 and the method was shown to be applicable to the rest of our compounds as well (Scheme 13). Reaction of ethyl catechol borane with O2 and CoII(Me/Ph) was unsuccessful.  24  Scheme 13: Reaction for synthesis of all ethyl cobalt(III) compounds  There are a few disadvantages of this protocol that limit its versatility. Preparation of organoborane precursors without β-hydrogens can be difficult,80 and they are not as readily available from chemical suppliers as Grignard reagents or alkyl halides. More importantly, the use of O2 is required to react with triethylboron, and so the product must be air stable to some degree, or at least less reactive towards oxygen than BEt3. Such is the case for our compounds, which are relatively air stable. As an example of its potential use with similar compounds, CoIII(salphen)Et was synthesized successfully using this protocol and yielded clean 1H NMR of the product in good yield.  2.3 X-ray, 1H NMR, 19F NMR, and UV-vis characterization X-ray structures for the CoII(CF3) compound and all novel Co(III)Et compounds were obtained, data from which are summarized in Appendix B (Table 16), with ORTEP models shown in Fig. 6.  25  Figure 6: ORTEP models for complexes CoII(CF3) (top), CoIII(Me/CF3)Et (top left), CoIII(CF3/Ph)Et (top right), (THF)CoIII(CF3)Et (bottom left),and CoIII(Me/Ph)Et (bottom right)  For the alkyl complexes, relevant bond lengths for the imine and carbonyl groups (from the ligand) and for the Co-C bond of solid state structures were tabulated (Table 2).  26  Table 2: Imine, β-ketone, and cobalt-carbon bond lengths for Co(III)Et structures Bond  Co(Me/Ph)Et  Co(CF3/Ph)Et  1.33(4) 1.24(3)  *  C=N/Å  Co(Me/CF3)Et  1.300(9) 1.316(9)  (THF)Co(CF3)Et  1.307(2) 1.313(2)  1.287(9) 1.275(8) 1.2929(18) 1.289(9) 1.280(8) 1.2881(19) Co-C/Å 1.991(6) 1.983(7) 1.9815(17) *C N and C O bond lengths are listed from each side of the ligand. *  C=O/Å  1.302(5) 1.311(5) 1.285(4) 1.285(4) 2.001(4)  Deviations in bond lengths between structures were no more than ten picometers for imine/C=O bond lengths, and within two picometers for cobalt-carbon bonds, and as such the structures are remarkably similar. This is consistent with only small differences in experimentally observed cobalt-carbon bond strengths for these compounds (± 5 kcal∙mol-1). Bond angles also support a high degree of planarity that is consistent in all structures (Table 3). Table 3: Bond angles and degrees of planarity for Co(III)Et structures Complex  Co(Me/Ph)  Co(CF3/Ph)  Co(Me/CF3)  (THF)Co(CF3)  N-Co-C angle (°)  96.5(8) 88.4(17)  95.1(3) 96.8(3)  94.11(7) 92.31(6)  90.57(15) 91.46(16)  O-Co-C angle (°)  89.7(11) 99.3(7)  94.5(3) 87.9(3)  97.10(6) 91.07(6)  94.98(13) 89.68(15)  1.0391  1.040  1.041  1.019  Planarity ∑  {  }  19  F NMR and UV-vis characterization was acquired for all Co(II) and Co(III)  compounds, data from which were used for further experiments and are summarized in Table 4.  27  Table 4: 19F NMR and UV-vis characterization data for Co(II) and Co(III)Et compounds Compound  19  F NMRa (ppm)  UV-vis (THF) λmax, nm (ε, 103L·mol-1cm-1)  UV-vis (non-coord. solv.) (λmax nm, 103L·mol-1cm-1)  CoII(Me/Ph)  n/a  Toluene: 328(19); 415(7.8); 480(4.9)  CoII(CF3/Ph)  -97.84  CoII(Me/CF3)  -111.23  CoII(CF3)  -50.86, -57.59  328(20.7); 415(8.2); 478 (3.7) 294(12.5); 342(6.6) shoulder; 405(2.4) 283(14.3); 335(5.3) shoulder; 382(4.4) 308(9.1); 335(7.3) shoulder  CoIII(Me/Ph)Et  n/a  CoIII(CF3/Ph)Et  -73.33  CoIII(Me/CF3)Et  -73.21  CoIII(CF3)Et  322(11.0); 407(3.1 shoulder); 472(1.5) shoulder 367(3.2); 400(2.8) shoulder; 460(1.0) shoulder; 509(0.51) shoulder 358(3.2); 385(2.6) shoulder; 445(1.0) shoulder; 519(0.38) shoulder 306(9.8); 415(2.6); 480, (1.2) shoulder; 542(0.76) shoulder  Hexanes: 308(12); 340(9.1) shoulder; 388(4.8); 457(4.9) Toluene: 299(9.0); 381(5.0); 446(1.5) Hexanes: 299(10.); 314(9.7); 352(7.0); 399(5.5); 418(5.3); 474(1.8) Toluene: 323(4.8); 402(4.5) shoulder; 472(2.8) shoulder; 668(1.4) Benzene: 371(4.1); 460(1.3) shoulder; 650(0.44) Benzene: 363(3.1); 450(0.87) shoulder; 652(0.32)  -65.92, Hexanes: 382(2.5); 411(2.5); 472(1.2); -74.09; b-64.78, 540(0.58) shoulder; 650(0.36) -72.74 (a) 19F NMR data referenced to external standard trifluoroacetic acid (δ -78.5 ppm), solvent: acetone-d6, 25 °C (b) DMSO-d6, 25 °C  UV-vis spectra suggest that THF coordination to CoII(CF3/Ph) and CoII(Me/CF3) may be occurring to some extent (possibly in equilibrium) as there is a slight broadening of bands compared with toluene spectra. The CoII(CF3) on the other hand, shows distinct spectra when dissolved in THF compared to hexanes, suggesting coordination of THF to this compound (Fig. 7). Virtually identical toluene and THF spectra for CoII(Me/Ph) are evidence for a noncoordinated species in THF. Outcomes are explained by the electron withdrawing CF3 groups resulting in more electron-deficient and Lewis acidic cobalt centers.  28  14  Co(II)(CF3)  ε (x103Lmol-1cm-1)  12 THF  10  Hexane 8 6 4 2 0 290  340  390  440 490 Wavelength (nm)  540  590  640  Figure 7: UV-vis of CoII(CF3) in THF and hexanes  The most salient feature of the UV-vis spectra for Co(III)Et compounds is a band at 600700 nm that only appears in non-coordinating solvents for all Co(III)Et compounds, with an example shown in Figure 8 for CoIII(Me/CF3)Et. This band was exploited in calculating binding constants for THF, discussed in Chapter 4.  Figure 8: UV-vis of CoIII(Me/CF3)Et in diethyl ether, benzene (λmax = 652 nm), and THF  29  1  H NMR spectra for the Co(III)Et compounds are found in Appendix C (Fig. 50-56), with  chemical shifts summarized in Table 5. As expected, more deshielded methylene protons (CoCH2-CH3) are observed as the electron withdrawing character of the ligand increases. Table 5: 1H NMR data of Co(III)Et compounds Compound and Solvent CoIII(Me/Ph)Et (acetone-d6) CoIII(CF3/Ph)Et (acetone-d6) CoIII(Me/CF3)Et (acetone-d6) CoIII(CF3)Et* (acetone-d6) CoIII(CF3)Et** (chloroform-d) CoIII(CF3)Et* (dmso-d6) CoIII(salphen)Et (chloroform-d)  CoCH3 CH2CH3  CoR(NCH2-)2 CH2CH3  CH  Ph ring protons  -0.17 (t, 3H)  2.22 (s, 6H)  3.47 (q, 2H)  3.63 (m, 4H)  5.96 (s, 2H)  7.38 (m, 6H); 8.02 (m, 4H)  -0.27 (t, 3H)  n/a  4.25 (q, 2H)  3.32 (m, 4H)  5.53 (s, 2H)  7.29 (m, 4H); 7.49 (m, 6H)  -0.04 (t, 3H)  2.23 (s, 6H)  3.84 (q, 2H)  3.65 (m, 4H)  5.58 (s, 2H)  n/a  0.24 (t, 3H)  n/a  4.46 (q, 2H)  3.78 (m, 4H)  5.88 (s, 2H)  n/a  -0.11 (t, 3H)  n/a  4.41 (q, 2H)  3.69 (m, 4H)  6.07 (s, 2H)  n/a  0.25 (t, 3H)  n/a  4.13 (q, 2H)  3.45 (m, 4H)  5.70 (s, 2H)  n/a  -0.54 (t, 3H)  n/a  3.57 (q, 2H)  n/a  n/a  6.72 (t, 2H J = 7.2 Hz);7.35 (m, 6H); 7.47 (d, 2H, J = 8.8 Hz); 7.96 (m, 2H), 8.75 (s, 2H) *THF adduct. **THF absent in synthesis. See appendix for spectra. All J values for ethyl couplings are 7.6 Hz.  Synthesis of CoIII(CF3)Et was first carried out in THF, and although the compound was dried in vacuo before redissolving in ether/hexanes, and again after recrystallization, 1H NMR spectra always showed a residual signal for THF in a 1:2 ratio to cobalt, and UV-vis in noncoordinating solvent showed no band in the 600-700 nm region (i.e. same as the THF spectrum). The 1:2 ratio is likely just a coincidence in its exact stoichiometry, as some THF may have evaporated. X-ray crystallographic data later confirmed that THF is coordinating to this compound in the solid state in a 1:1 ratio (Fig. 6). THF coordination in the solid state is not observed for any of the other ethyl complexes or any of the Co(II) species. Synthesis was thereafter carried out in hexanes (THF not present) and though the recrystallization solvent diethyl ether is contaminating the NMR spectrum, its quantity is negligible, and is not thought to 30  be coordinating to cobalt. UV-vis data were acquired using the latter compound, where no THF is coordinated.  2.4 Experimental 2.4.1 General considerations Unless otherwise stated, for this chapter and all subsequent work, solvents and reagents were used as follows: hexanes, toluene, diethyl ether, dichloromethane, and THF were purified by passage through activated alumina and deoxygenizer columns from Glass Contour Co. (Laguna Beach, CA, USA). Anhydrous benzene and acetonitrile were purchased from Aldrich, stored in the glove box, and used directly. Vinyl acetate, methyl acrylate, and acrylonitrile were distilled from calcium hydride and filtered through alumina. C6D6 (Na metal and benzophenone), CDCl3 (CaH2), and acetone-d6 (CaSO4) were distilled off of corresponding drying agents. DMSO-d6 was used freshly from the bottle. All NMR solvents and alkene monomers were degassed by three freeze-vacuum-thaw cycles before use. All other reagents and solvents were purchased from Aldrich and used as received. Celite (Aldrich) was dried overnight at 120 °C before being evacuated and then stored under nitrogen. 19  F and 1H NMR spectra were recorded on a Varian Mercury Plus 400 spectrometer.  Chemical shifts were referenced to the solvent peak for 1H NMR spectra, and 19F NMR spectra were referenced to external standard trifluoroacetic acid (25 °C) between each run by first locking and shimming the solvent used for the sample, inserting and referencing neat TFA to 78.5 ppm (with the sample NMR solvent selected; not TFA), and acquiring the 19F NMR spectrum for the sample. UV/Vis spectroscopic data were collected on either a Varian Cary 100 Bio UV-Visible spectrophotometer or a Shimadzu UV 2550 UV-vis spectrophotometer in a specially constructed cell for air-sensitive samples: a Kontes Hi-Vac Valve with PTFE plug was attached by a professional glassblower to a Hellma 10 mm path length quartz absorption cell 31  with a quartz-to-glass graded seal. All handling of reagents were carried out under nitrogen in a glove box. The synthesis of all ligands and Co(II) starting material compounds were carried out using known literature methods.71 2.4.2 Synthesis of Co(III) alkyl compounds In a typical preparation of ethyl-cobalt compounds, cobalt(II) was dissolved in THF or diethyl ether and added with 0.67 equivalents triethylborane (from a 1.0 M solution in hexanes) to a sealed glass vessel. Typical scales were on the order of 0.1 to 1 mmol of Co(II). To the stirring orange/brown solution was added slowly via glass syringe approximately 1.5 equivalents of dioxygen from air (assuming ideal gas law at 20 °C; 20% oxygen content, ~ 170 mL air per mmol of Co(II)). Immediately after addition of air, solutions turned from orange-brown to dark brown in THF and to dark green when a non-donor solvent is utilized. Solutions were reduced in vacuo to a dark brown/green crude solid, taken into the glove box, filtered through silica/ether, concentrated, and recrystallized using the appropriate solvents: CoIII(Me/Ph)Et: THF/hexanes; CoIII(Me/CF3)Et: ether/hexanes (1:1); CoIII(CF3/Ph)Et and CoIII(CF3)Et: hexanes or slow evaporation from ether (into hexanes) for X-ray. Note: THF must not be employed as the reaction solvent for CoIII(CF3)Et unless the corresponding THF adduct is desired. The synthesis of CoIII(CF3/Ph)Et is described below as an example: In a glove box, CoII(CF3/Ph) (174.0 mg; 0.3390 mmol) was dissolved in 5 mL THF in a 50 mL bomb. Triethylborane (0.23 mL of a 1.0 M solution in hexanes) was added, and the bomb sealed and removed from the glove box. Air (56 mL) was syringed into the solution with stirring via a septum and syringe, and an observed colour change was noticed within the first 20 mL addition of air (from orange-red to brown). The solvent was removed under vacuum, leaving behind a light brown powder, which was returned to the glove box, redissolved in ether/hexanes  32  (1:1, ~2 mL; dark green), and pipette-silica filtered into a 5 dram vial and placed in a freezer (-35 °C) overnight to yield 122.2 mg (61.0 %) of dark green-black crystals. 2.4.3 Yields CoIII(Me/Ph)Et: Yield: 89.0%. CoIII(CF3/Ph)Et: Yield: 61.0%. CoIII(Me/CF3)Et: Yield 80.3%. CoIII(CF3)Et: Yield 79.1%. CoIII(CF3)Et: Yield 55.9% (assuming THF/Co = [0.5:1] from 1H NMR). CoIII(salphen)Et: Yield 79.3 % of black powder; washed with hexanes/ether yields clean compound with few impurities visible by 1H NMR.  33  Chapter 3: Stoichiometric and Catalytic Reactivity of Co(II) Complexes 3.1 Preface Plausible reaction pathways requiring access to +1 and +3 oxidation states of cobalt are explored in this section. Cyclic voltammetry was utilized to give insight into the propensities or reluctances for the cobalt(II) compounds to undergo these oxidation states, and stoichiometric reactions at accessing Co(I) and Co(III) species were performed thereafter with some success. Attempts at electrocatalytic hydrogen evolution for the compound with the highest reduction potential, CoII(CF3), stoichiometric coupling of the ethyl radical from CoIII(Me/CF3)Et to (E)stilbene and (E)-chalcone, reactivity with dioxygen, and veratryl alcohol oxidation is also included here.  3.2 Cyclic voltammetry Cyclic voltammetry was used to determine Co(II/III) and Co(I/II) redox couplings in order to explore the electronic effects of equatorial ligands on the metal center. Implications for reactivity and bond strength are discussed in succeeding chapters. Sample voltammograms of CoII(Me/CF3) are shown in Fig. 9, and redox potentials are summarized in Table 6.  34  0.07  I (mA)  0.05 0.03 0.01 -1.60 -0.02  -1.20  -0.80  -0.40  0.00  0.40  0.80  1.20  1.60  2.00  V (Volts)  -0.04 -0.06  Figure 9: Cyclic voltammogram of CoII(Me/CF3): 5 mM in DCM, 50 mV/s (red); 1 mM, MeCN, 100 mV/s, (blue); 1 mM in MeCN, 100 mV/s (green)  Table 6: Redox potentials of Co complexes (mV vs. Fc+/0) determined from cyclic voltammetry measurements  Complex  E1/2 (CoII/CoIII)a  Ep,a (CoII/CoIII)b  E1/2 (CoII/CoI)c  CoII(Me/Ph) CoII(CF3/Ph) CoII(Me/CF3) CoII(CF3)  459 821 856 1076  116 490 523 961  -1485 (irrev. anodic) -1180 -1235 -858  E1/2 (CoII/CoIII)d  E1/2 (CoII/CoI)d  45 80  -1775 -1610  Literature Values CoII(Me/Me, bae) CoII(Me/Ph)  (a) 5 mM in DCM, NBu4PF6 (0.1 M), 50 mV/s (b) 1 mM in MeCN, NBu4PF6 (0.1 M), 100 mV/s (irrev. cathodic waves) (c) 1 mM in MeCN, NBu4PF6 (0.1 M), 100 mV/s (d) DMF solution of Et4NClO4 (0.1 M), E1/2 + 0.001/V from Kasuga, 1984.15  Three of the four compounds were found to undergo at least partially reversible Co(I/II) couplings carried out in acetonitrile (MeCN) under the experimental conditions, the exception being for CoII(Me/Ph) where only the anodic peak was observed. Irreversible Co(II/III) oxidations observed in MeCN led to implementation of DCM as the solvent thereafter for these Co(II/III) couplings; however, no Co(I/II) couple was present in DCM.  35  Not surprisingly, Co(II) compounds with more electron withdrawing substituents on the equatorial ligands are generally more difficult to oxidize since the electron-deficient metal center becomes destabilized in high oxidation states. The only exception is observed when substitution of a phenyl group for a CH3 moiety (i.e. going from CoII(Me/CF3) to CoII(CF3/Ph)) seems to increase the oxidation potential of the Co(II) (E1/2(CoII/CoIII) = +856 mV → + 821 mV). The fact that CoII(CF3/Ph) is both easier to reduce and oxidize than CoII(Me/CF3) is a testament to the captodative effects of its ligand. Stabilization of both an electron-rich and electron-deficient metal center through extended conjugation may well be occurring, despite the phenyl group being relatively far away from the metal center. Half reduction potentials where methyl groups are substituted for phenyl groups does not always produce the same effect as can be seen from redox potentials from Kasuga.15 For example, going from Co(bae) to CoII(Me/Ph) oxidation is less favourable (E1/2(CoII/CoIII) = +45 mV → + 80 mV), whereas reduction is more favourable with addition of the phenyl group (E1/2(CoII/CoI) = -1.775 V → -1.610 V), as expected. This may suggest that the electron withdrawing effect of a CF3 group is necessary in order to inductively increase the donating tendency of the phenyl ring to stabilize Co(III). Irreversible oxidations of Co(II) complexes in the donor solvent MeCN can be attributed to stabilized octahedral or square pyramidal solvent-coordinated [CoIII-(CH3CN)n]+ species, where n = 1 or 2. Similar isolable Co(III) compounds with the formula [L2CoIII(Chel)]ClO4 (where Chel = salen, salphen, dh2, bae, and L = various N-donors) and their redox chemistry has been studied elsewhere in the literature.81 Our results show the redox processes for oxidation follow similar trends in both MeCN (irreversible) and DCM (reversible) solvents, with increasing oxidation potentials in the order: CoII(CF3) > CoII(CF3/Me) > CoII(CF3/Ph) > CoII(Me/Ph).  36  It should be noted that peak separations are unusually large for Co(II/III) redox couples in DCM, though not to the same extent as those run in MeCN. This could be due to solvent interactions and/or structural rearrangements upon oxidation, or by slow electrode kinetics.82 Rerunning Co(II/III) oxidations of CoII(CF3) with compensated internal resistance and situating the electrodes closer together did not seem to help in minimizing the separation; however, reversible redox processes are still presumed to be occurring since the peak currents on the cathodic and anodic waves are nearly equal to each other.15 Reversible reduction of Co(II) was observed in acetonitrile, with our result for the reduction of CoII(Me/CF3) at 1.235 V agreeing well with the literature value83 (1.215 V), discrepancies from which may be attributed to using different solvents/electrolyte combinations (DMF/TEAP vs. DCM/ NBu4PF6). Destabilization of an electron-rich species with donor group ligands accounts for the increasing reduction potential to [Co(I)]− in the following order: CoII(Me/Ph) > CoII(CF3/Me) > CoII(CF3/Ph) > CoII(CF3). Approaching – 2V in DCM consistently shows dramatic current surges for all compounds occurring well before the point at which DCM starts to become reduced (Fig. 10). The fact that reversible features are present in MeCN and not observed in DCM is consistent with the idea that DCM is reacting rapidly with any [Co(I)] − generated in solution and possibly generating (CH2Cl)Co(III). Precedent for this sort of reaction has been shown in a known preparation of (CH2Cl)Co(bae) involving reaction of Co(bae) in a sodium amalgam in THF with dichloromethane.84 A summary of redox processes is described in Scheme 14.  37  Figure 10: Cyclic voltammograms for attempts at reduction in DCM for all Co(II) compounds (5.0 mM, 25 °C, 50 mV/s, NBu4PF6 (0.1 M))  Scheme 14: A summary of redox processes from cyclic voltammetry experiments  3.3 Reduction and subsequent by reaction with alkyl halides Results from the previous section suggest that a viable route for generating Co(III) alkyls would be to first reduce Co(II) compounds and then react the electron-rich [Co(I)] − nucleophiles with alkyl halides (Scheme 15).  38  Scheme 15: Reduction followed by reaction with alkyl halide “R-X” (where X = Cl, Br, I) as a possible route to Co(III)R complexes (where R = alkyl group)  Reduction of all Co(II) complexes proceeds in the presence of sodium metal and is typically signified by a dramatic colour change: CoI(Me/Ph)− is a vibrant maroon, and the other Co(I)− compounds intense turquoise and blue-green colours. Low energy bands (700-806 nm) in the UV-vis spectra were also observed accompanying reductions, which has also been observed for ligand-based radicals of chromium(II).85 Exposure of the postulated [Co(I)]− species to atmospheric oxygen or to the weak proton source diisopropylethylamine hydrochloride in a vain attempt to generate sustainable Co(III)-H species yields rapid conversion back to the Co(II) (Scheme 16).  Scheme 16: Reaction scheme attempting to synthesize a cobalt(III) hydride. The reaction failed and instead yielded starting material, Co(II).  Cobalt hydrides have been postulated as intermediates in similar compounds; however, few examples have been isolated in the literature, and are typically observed at best as transients.86 [HCo(CO)4] is one of the rare isolable hydride compounds of cobalt.87 Usually, cobalt hydrides undergo rapid decomposition to H2 gas and Co(II), as observed for vitamin B12 analogues.  39  In contrast to sodium reactions, no changes in electronic spectra were observed using the gentler reducing agent magnesium for all Co(II) compounds. Subsequent reaction of the reduced species [CoI(CF3/Ph)]Na or [CoI(Me/CF3)]Na with ethyl iodide shows presence of the Co(III)Et by UV-vis; however, the alkyl cobalt could only be isolated cleanly (as evidenced by 1H NMR) from the CoIII(Me/CF3)Et compound. Reactivity of CoII(CF3) and CoII(Me/Ph) was observed following sodium reduction with cyclohexyl halides, where halide „X‟ = I, Br, and Cl. Evidence supports a reaction proceeded to some degree using any of the cyclohexyl halides in the case of CoII(Me/Ph), but only with cyclohexyl iodide for CoII(CF3). The latter reaction showed exclusively the presence of Co(II) when X = Br and Cl by UV-vis for this compound, and presumably a mixture of Co(II) and product when cyclohexyl iodide was used, with a small shoulder at ~684 nm appearing that is not indicative of Co(II). For reactions with CoII(Me/Ph), some contamination with unreacted starting material is possible in the cases of the bromo and chloro substrates, as only 2 hr reaction time was employed, and there are similarities in spectra to Co(II). Although CoII(CF3) is easiest to reduce, the electron-withdrawing effect of the ligand may delocalize the electron away from the metal center, creating a more stable complex as well as a poorer nucleophilic cobalt center. Vitamin B12 and Co(dh2) (cobaloximes) also produce poor [Co(I)]− nucleophiles for this reason, reactions with aryl iodides yielding only the unchanged Co(II) starting material;9 however, even in these compounds, reduction followed by reaction with methyl or ethyl iodide is a known method.77 Hexane extracts taken from the reaction mixture gave spectra distinct from starting material for CoII(Me/Ph) (insol. in hexanes), and were observed by UV-vis in THF to exhibit features similar to those for the corresponding Co(III)Et compounds, but with wavelengths 40  shifted from 668 nm (in toluene) to ~700 nm (THF). Superimposed UV-vis spectra (in THF) of CoII(Me/Ph), CoII(Me/Ph) reacted with sodium metal, and CoII(Me/Ph) reacted with sodium metal followed by reaction with cyclohexyl chloride are shown in Figure 11. The Co(II) starting material spectrum (green) looks similar to the product (red), except for a small band at 700 nm in the latter, possibly indicating the presence of some alkylated species.  Figure 11: Reaction of CoII(Me/Ph) (green) with Na metal (blue) followed by cyclohexyl chloride (red) in THF  The UV-vis spectrum of the postulated CoIII(Me/Ph)(cyclohexyl) compound in THF looked identical to that in hexanes, suggesting either negligible interaction with THF for this alkyl derivative, or that „[(cyclohexyl)CoIIIX]−‟ or „ICoIII‟ is being formed via oxidative addition, rather than the expected 5-coordinate CoIII(cyclohexyl) complex. Work on related salen and salphen systems has suggested that the formation of such oxidative addition products requires some flexibility in the chelating ligand. In a less flexible ligand system, primarily a fivecoordinate alkyl complex results (Fig. 12, top), while ligands similar to our ketoaminato groups  41  might accommodate a concerted oxidative addition of RX, affording a [(cyclohexyl)CoIIIX]− product (Fig. 12, bottom).88  Figure 12: SN2 mechanism to form cobalt(III)alkyl (top) vs. oxidative addition (bottom) to form the octahedral structure via a less rigid square planar structure Auspicious spectral results for reaction of CoII(Me/Ph) with sodium and cyclohexyl chloride (Fig. 11) prompted an attempt at a stoichiometric synthesis on a greater scale by implementing the same protocol as for preparation of CoIII(Me/CF3)Et with sodium metal, only with cyclohexyl chloride in place of ethyl iodide. This is a potentially more versatile synthetic route to alkyl-cobalt(III) compounds than reaction with O2 and trialkylboranes (BR3) since the latter reagents are not readily available, and are more difficult to prepare.80 Alkyl chlorides, on the other hand, are inexpensive and more readily available. The most electron-rich CoII(Me/Ph) compound should provide a more aggressive nucleophile upon reduction, which explains its greater observed reactivity with the alkyl chloride. Unfortunately, isolation of clean product could not be obtained, and this is perhaps due to some of the cobalt(II) undergoing irreversible reduction and perhaps decomposition, consistent with its irreversible reduction in MeCN. Also, 42  the radical stability and steric demand of the secondary carbon of a cyclohexyl may be enough to thermodynamically dissuade its long-term adherence to cobalt.  3.4 Attempts for electrocatalytic evolution of H2 gas Conventional electrolysis of water to yield O2 and H2 gas is slow without use of a catalyst,89 and therefore inefficient. Most pathways proposed for evolution of dihydrogen using cobalt catalysts involve generation of a Co(III)H species immediately following formation of Co(I),86 ideally at potentials only slightly greater than those needed for self-reduction (i.e. a low overpotential). Protonation from an acidic medium, or bimolecular combination reductively eliminates and liberates H2 from solution.86,90 Precedent for such a process has been shown for nickel and cobalt diimine-dioxime compounds (E½Co(I/II) = -0.82 to -1.48 V),91 as well as simpler cobaloxime catalysts in aqueous media.92 With successful electrocatalytic generation of H2, cyclic voltammograms should show a peak-shape catalytic wave that approaches a plateau as the acid concentration becomes high. This is caused by a rapid catalytic reaction resulting in a suppressed concentration and diffusion of acid to the electrode, which controls the current.93 A relatively favourable reduction potential for the reduction CoII(CF3) compound (-0.858 V) led to a brief investigation in its ability to evolving hydrogen gas from acidic media. H2 formation, however, was not observed in the presence of the hydrochloride salt of N,Ndiisopropylethylamine or p-toluenesulphonic acid, as immediate colour loss observed in the former, and gradual loss upon subsequent additions in the latter suggested decomposition. Although the Co(II) appears stable in benzoic acid, with no noticeable loss in colour in solution, no shift in reduction potential or increase in current indicating a catalytic wave or a shift in  43  potential was observed when up to a 4-fold excess of the acid was used. The reaction was therefore abandoned.  3.5 Oxidative addition and subsequent reaction with Grignard reagents Access to the alkyl cobalt(III) via oxidative addition of I2 followed by salt metathesis using CH3CH2MgI was also explored as a possible synthetic route to cobalt(III) alkyl complexes. Alkylated cobalt compounds are typically obtained by reacting Grignard reagents to Co(III) halides,72,94 and this can be an important step for catalytic cross coupling reactions (Scheme 17).  Scheme 17: Oxidative addition followed by reaction with Grignard reagent (where X = Cl, Br, I) as a possible route to Co(III)R complexes (where R = alkyl group)  Reaction of I2 with either CoII(Me/CF3) and CoII(Me/Ph) in THF caused a colour change from orange to purple in the former, and from red-orange to light brown in the latter; UV-vis spectra suggest formation of a new species that we attribute to the Co(III)I complex (Fig. 31). Subsequent reaction with the Grignard reagent, though never successful in yielding clean Co(III)Et product, gave a promising UV-vis spectrum indicative of the previously isolated alkyl compound for CoIII(Me/CF3)Et.  44  Figure 13: Reactions of all cobalt(II) compounds with I2 in various solvents: THF (top left/right), diethyl ether (bottom left), and hexanes (bottom right).  The reaction of I2 with CoII(CF3) and CoII(CF3/Ph) were not carried out in THF, as better solubility permits their study in non-donor solvents. The main advantage of using non-donor solvents is that more diagnostic UV-vis spectra for the starting material Co(II) can be detected should the reactions fail. Results suggest that a donor solvent at least as good as THF is required for oxidative addition of I2 to occur, as diethyl ether was insufficient for use with CoII(CF3/Ph), and hexanes 45  not suitable for CoII(CF3). The latter observation is not surprising, as this compound is also stable in air when dissolved in hexanes over weeks, but reacts over the course of a day under the same atmospheric conditions in THF. THF is not presumed to coordinate to CoII(Me/Ph) as UV-vis in donor and non-donor solvents are the same, but solvent stabilization of the Co(III) species is enough to permit oxidative addition of I2, since a reaction is clearly observed when I2 is mixed with CoII(Me/Ph) in THF (Fig. 13). Reaction of I2 with catalyst CoII(Me/Ph) in THF and subsequent reaction with EtMgBr did not yield product, but very unusual UV-vis spectra. The THF spectrum looks similar to the Co(II) when I2 is added, as does the resulting orange solid when the solvent is evaporated. However, redissolving the solid in toluene gives a green solution (λmax = 593 nm), and air exposure yields a blue/purple solution (λmax = 573 nm). Since neither of the products are indicative of the desired ethylated Co(III) compound (λmax ~ 668 nm), the reaction was abandoned.  3.6 Coupling reactions The transfer of an alkyl group from a Co(R) species to an alkene,26 followed by dehydrocobaltion29 is a key step in cross coupling reactions that has not yet been explored with our compounds. This process been shown to occur in acyl Co(salphen)95 and alkyl Co(dh2)96 compounds in vinylation reactions with styrene and light exposure. In this section, reactivity of CoIII(Me/CF3)Et with some alkenes was therefore explored. 3.6.1 CoIII(Me/CF3)Et reactivity with alkenes Decomposition of CoIII(Me/CF3)Et by light or heat in the presence of alkenes was attempted and the byproducts analyzed by 1H NMR to search for cross coupling products: compounds resulting from reaction between the ethyl radical and the substrate alkene. The CoIII(Me/CF3)Et compound was utilized for its easy work up due to the relatively poor solubility 46  of its corresponding Co(II) byproduct in the extraction/reaction solvent (diethyl ether). The removal of Co(II) from solution by precipitation would presumably have the added benefit of dissuading β-H elimination of the ethyl radical.32 Similar techniques with Co(II) sequestering agents have been successfully employed in the Co(dh2) system in stoichiometric reactions of alkyl cobalt compounds with activated alkenes.32 The substrates tested were (E)-chalcone and (E)-stilbene, which were reacted stoichiometrically (1:1) with respect to CoIII(Me/CF3)Et and the proposed reaction mechanism is found in Scheme 18.  Scheme 18: Proposed stoichiometric cross coupling mechanism between Co(III)Et and (E)-chalcone/stilbene  3.6.2 (E)-Stilbene UV-vis spectra of a few drops of the reaction solutions were recorded in tandem in order to monitor Co(III)R decomposition when heated to 65°C. They showed disappearance of the Co(III) and emergence of the Co(II) species (Fig. 14). Light-induced homolysis was not 47  attempted due to tendency of (E)-stilbene to undergo isomerization and dimerization upon exposure to light.97  Figure 14: Monitoring cobalt species at selected wavelengths when reacted with (E)-stilbene. Co(II) (381 nm, red), and Co(III)Et (649 nm, blue).  Decomposition to Co(II) occurred, albeit very slowly, by no discernable rate law, and 1H NMR of the extract showed mainly the presence of starting materials (E)-stilbene and CoIII(Me/CF3)Et (Fig. 15). The only unaccounted for peaks are a coupled triplet (δ 1.32; 3H; J = 7.136 Hz) and quartet (δ 4.19; 2H; J = 7.132 Hz) that are due neither to solvent impurity nor to the ethyl group of the CoIII(Me/CF3)Et compound (Fig. 15). If a cross-coupled product is forming, integration shows only 9.9 % formation relative to the remaining (E)-stilbene, and so self-termination of the ethyl radical or β-H elimination of ethylene77 is most likely the main reaction accompanying the decomposition of cobalt(III) to cobalt(II). It could be that the concentration of the substrate was perhaps not high enough to enable a competitive insertion, or that this particular alkene, self-stabilized by extended conjugation and 48  non-activated by functional groups, is less susceptible to reaction with the ethyl radical. In any case, Figure 14 gives an idea as to the behavior and speed that self-termination and/or β-H elimination is occurring, with a rapid onset and about halfway completion at 23 hours. The reaction slows down as it progresses, likely due to the buildup in Co(II) to sustain a persistent radical effect. An anomalous sudden decrease in Co(II) and accelerated loss of Co(III)Et after ~80 hours was also observed.  Figure 15: 1H NMR of the extract from reaction of CoIII(Me/CF3)Et with (E)-stilbene with the unaccounted-for triplet/quartet outlined (red). The starting materials are superimposed: (E)stilbene (green) and CoIII(Me/CF3)Et (blue). 3.6.3 (E)-Chalcone (E)-Chalcone contains a double bond that is adjacent two disparate chemical environments: a carbonyl carbon on one side, and a phenyl group on the other. This compound 49  was reacted similarly to (E)-stilbene, except under a compact fluorescent bulb for 24 hours (instead of heat) to induce homolysis of the Co-C bond. Insoluble red/brown solids were apparent in the reaction mixture (presumably Co(II) in ether/hexanes) and were filtered off. Organics were rinsed through with hexanes, and after ether-washing, some starting material CoIII(Me/CF3)Et was isolated (18.4% unreacted; confirmed by 1H NMR). 1  H NMR of the crude organic product showed presence of mostly starting material in the  aromatic region (Fig. 16 and Fig. 17). At first glance, some upfield triplets (0.7-1.0 ppm) and quartets (2.5-4.0 ppm) seemed to suggest ethyl group RCH2CH3 protons coupled onto the alkene (Fig. 16); however, closer examination of the gCOSY spectrum showed no observable coupling between these signals (Fig. 18). The reaction did not proceed in the dark as evidenced by UV-vis of a small portion of the reaction mixture.  Figure 16: 1H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with (E)-chalcone with a compact fluorescent bulb. The starting material, (E)-chalcone is superimposed in blue. 50  Figure 17: Aromatic region 1H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with (E)-chalcone with a compact fluorescent bulb. The starting material, (E)-chalcone is superimposed in blue and is clearly present.  51  Figure 18: Upfield region 1H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with (E)-chalcone with a compact fluorescent bulb. gCOSY couplings are situated atop coupled peaks.  The reaction was attempted again, only with a full spectrum tungsten bulb (200 W). After 24 hours the solution was vibrant orange with solid orange precipitates, indicative of Co(II), and after 28 hours the reaction was worked up similarly to the previous experiment, only a dark green-brown solid was isolated. 1H NMR of the crude solid showed a complex mixture with clearly all the (E)-chalcone reacted (Fig. 19). The abundance of peaks in the aromatic region may indicate polymeric species, but since peaks are well resolved small oligomers or a combination of several different isomers of many different compounds is more likely.  52  Figure 19: 1H NMR (CDCl3) of the extract from reaction of CoIII(Me/CF3)Et with (E)-chalcone with a tungsten full spectrum bulb (200 W). The starting material, (E)-chalcone is superimposed in blue.  Since reactions with heat and under a compact fluorescent bulb produced mostly starting materials, and reactions under a tungsten bulb produced a complex mixture, this stoichiometric reaction was abandoned, and for this specific compound under the experimental conditions, Heck-type couplings of alkyl groups to alkenes may not be feasible. Any hope of a catalytic cycle would require this step to yield clean isolable products, as well as the use of a stoichiometric reductant to regenerate the Co(I) to react with an alkyl halide, but as shown in section 3.3, a reductant better than magnesium is required. Many parameters can still be adjusted, such as temperature, solvent, and axial bases. Reactions carried out with higher equivalents of substrate95 may also help improve results. 53  3.7 Reactivity with Dioxygen Short-term and long-term exposure of our Co(II) compounds to air was examined by UVvis in order to observe reactivity under atmospheric conditions. This was accomplished either by bubbling in O2 directly or letting the compounds stand in THF in small vials with pinholes on the caps (so as to minimize evaporation), respectively. Bubbling dioxygen gas for 3-6 minutes effected no change to the UV-vis spectra of THF solutions of the Co(II) compounds. The reaction needs time to proceed, which is not surprising as synthesis and evolution of O2 adducts in similar compounds often require longer reaction times and cold temperatures (-20 to -175°C).14,55,98 Upon long term exposure (i.e. standing in air, THF) compounds CoII(Me/Ph) and CoII(CF3) experienced two distinct chemical changes: once after 22 hours, and another at 66 hours, possibly signifying formation of O2 adducts of 1:1 or 2:1 ratios to cobalt, axial water ligation from trace amounts in the atmosphere,99 or perhaps eventual decomposition (Fig. 20). CoII(CF3/Ph) and CoII(Me/CF3) took two days to realize any noticeable changes in spectra, until conforming to those observed after 66 hours (Fig. 20). In any case, the cobalt(II) compounds underwent chemical changes in atmospheric conditions in THF, and further reactivity towards veratryl alcohol under strictly aerobic atmospheres supports a mechanism wherein some activation of dioxygen is occurring.  54  Figure 20: UV-vis of reaction of Co(II) compounds to atmospheric conditions: (A) CoII(Me/Ph), (B) CoII(CF3/Ph), (C) CoII(Me/CF3), (D) CoII(CF3)  3.7.1 Axial donor influence Early studies in the literature have shown that the CoII(Me/Ph) compound readily binds dioxygen in the presence of N-donor axial base pyridine at 0 °C, and only a very small amount in toluene.19,100 Similarly, CoII(CF3), the compound in this study with the highest oxidation potential, is stable when left air exposed in hexanes for weeks, but its UV-vis spectrum changes dramatically within 22 hours when the compound is dissolved in THF (see Fig. 20). Axial base 55  studies later showed a relatively large affinity of CoII(CF3) for THF (Chapter 4). These observations suggest that solvent coordination plays a role in promoting reaction with dioxygen. The short-term behaviour of Co(II) compounds in the presence of oxygen and stronger axial donors was explored. Co(II) compounds were first dissolved in THF, then added DMSO (280 equiv.), followed by a slight excess of tBuOK, with oxygen bubbled through for each solvent combination in 30 second intervals until no noticeable changes were observed in spectra. Although using THF while bubbling in dioxygen for all Co(II) compounds over a span of 3-6 minutes produced no spectra changes, when DMSO was added (followed by O2 bubbling), the UV-vis spectrum of CoII(CF3) appeared to change, albeit subtly, to a new species, with a shoulder appearing at 330-335 nm. The spectrum obtained by oxygen addition to a tBuOK/THF solution of this compound also changed similarly to that with THF/DMSO. The other three cobalt(II) compounds required the presence of anionic base tBuOK in order to see noticeable immediate changes upon the addition of oxygen, with compounds CoII(CF3/Ph) and CoII(Me/CF3) exhibiting the most abrupt conspicuous changes (Fig. 21).  Figure 21: Short term reaction of CoII(CF3/Ph) and CoII(Me/CF3) with O2 in THF/tBuOK/DMSO solutions 56  Decomposition cannot be ruled out, since evaporating the solvent in vacuo and redissolving in THF does not yield the original Co(II) spectra. This could also indicate irreversible binding of O2, or conversion to a (tBuO)Co(III)/(tBuO∙+)Co(II) mixture.101 For the CoII(Me/Ph) compound, no spectral change was apparent after 3 min of O2 bubbling; however, after being under the headspace of air over 24 hours, there was an obvious loss of the 413 nm band to a shoulder. In any case, given the appropriate base, all compounds seem to react with oxygen in the presence of the anionic donor tBuOK, the CoII(CF3) also reacting immediately with dioxygen when dissolved in donor solvent DMSO.  3.8 Oxidation of veratryl alcohol Exploring the ability to functionalize lignin with oxygen or to oxidize the substrates following depolymerization is a step toward its ultimate biorefinery and selective degradation.102 Simple Co(salen) compounds show promising results in the literature for lignin model oxidation, and EPR studies support a mechanism involving activation of oxygen by the catalyst, followed by hydrogen abstraction of the substrate to generate a phenoxy Co(III) radical species.65,68 Because the square planar (β-ketoaminato)cobalt(II) compounds used in our lab have been shown to react with O2, probably via formation of similar cobalt-oxo intermediates, the possibility of reacting such compounds in the presence of lignin model substrates was considered. Preliminary experiments under various conditions were carried out to evaluate the effectiveness of each Co(II) compound as a possible catalyst for oxidation of lignin model compound 3,4-dimethoxybenzyl alcohol (a.k.a. veratryl alcohol) and the results are summarized in Table 7. The oxidation of veratryl alcohol to veratryl aldehyde is shown in Scheme 19.  57  Scheme 19: Oxidation of veratryl alcohol (left) to veratryl aldehyde (right).  Table 7: Percent conversions of veratryl alcohol to veratryl aldehyde under various conditions  Compound  % Conversion by 1H NMR A 4.48 5.21 0.79  Co(CF3/Ph) Co(CF3/Me) Co(CF3) B  2.23 4.04 1.48 0.78  Co(Me/Ph) Co(CF3/Ph) Co(CF3/Me) Co(CF3) C Co(CF3/Ph) Co(CF3/Me) Co(CF3)  2.47 1.89 0.65  (A) sealed vials, 0.0025 mmol (1.45%) Co(II); 0.4 mL THF, 2.0 mL 0.1 M NaOH; 0.173 mmol veratryl alcohol; 18 hrs; 65 °C (B) test tubes with pinholes, 0.0075 mmol (5%) Co(II); 2 mL THF; 0.15 mmol veratryl alcohol; 2.0 mL 0.1 M NaOH; 24 hrs; 80 °C; evaporated to ~1/2 volume over time. (C) test tubes with pinholes 0.0075 mmol (5%) Co(II); 2 mL MeCN; 0.15 mmol veratryl alcohol; 2.0 mL 0.1 M NaOH; 67 hrs; 80 °C  The starting material and product veratryl aldehyde were detected either by 1H NMR or by GC/MS. Integrations of the veratryl aldehyde proton (product; 9.85 ppm) vs. those of the methylene (4.60 ppm) and aromatic protons of the starting material (6.80 – 6.92 ppm) were used to calculate average percent conversions, and were found to agree to ± 0.2 %. Experiments carried out in the absence of cobalt and with experiments using cobalt(II) acetate both showed negligible conversion of the alcohol. The cobalt(II) acetate catalyst was converted almost immediately to an insoluble black material, likely a kinetically inert cobalt(III) hydroxide 58  compound, upon addition of 0.1 M hydroxide solution. Nearly complete colour loss is observed when CoII(CF3) is used as catalyst and negligible conversion using this compound is likely attributed to immediate catalyst decomposition. Screening tests for CoII(Me/Ph) were abandoned due to solubility issues and low conversions for this compound. With the exception of the sealed-vial experiment (Table 7; entry A), where oxygen is limited to the headspace in the vessel, the CoII(CF3/Ph) compound seemed to give the highest product to starting material ratios, and was selected as the most promising catalyst for further study. Catalyst loading, pH, and axial base pyridine were the variables tested for this compound, and results are summarized in Table 8. Table 8: Screening tests for catalyst loading, pH, and axial base pyridine for conversions of veratryl alcohol to veratryl aldehyde under various conditions using CoII(CF3/Ph) Conditions Cat. 10 % Cat. Loading 1 % Cat. Loading 5 %, 67 hours Cat loading 0.05% no NaOH pH 14 saturated NaOH pyridine 1.9 equiv. pyridine 10 equiv. pyridine 100 equiv.  1  H NMR (%) 2.68 0.88 2.47 0.61 -  GC/MS (%) 3.07 1.20 1.19 1.36 1.55 0.32 0.38 0.87 0.19  Average (%) 2.81 0.99 " 0.8 " " " " " "  Unless otherwise stated, cat. loading is 10%, T = 80°C, veratryl alcohol = 0.15 mmol, 48 hrs, and V = 4 mL (2 mL 0.1 M NaOH, 2 mL MeCN). All conversion % values determined by GC/MS and/or 1H NMR.  Looking at all screening tests, it was apparent that MeCN, originally utilized to improve miscibility of the catalyst with water/NaOH, could be responsible for generally low conversions. As a solvent, it could be hindering axial coordination of the substrate to an appreciable degree.18 A more likely possibility is that the catalysts are converting to inactive forms with acetonitrile strongly stabilizing axial ligation of substrate,65 as suggested by irreversible oxidations observed 59  in previous voltammograms when oxidations were carried out in acetonitrile. All experiments carried out in THF yielded higher conversions despite being biphasic with the alkaline aqueous phase, and so acetonitrile was abandoned as a suitable solvent for catalysis. High catalyst loading afforded highest conversion, though little variance was observed between 1 and 0.05% loading and 5 and 10% loading. A saturated solution of NaOH led to negligible conversion, most likely due to catalyst decomposition under the harsh alkaline conditions. Whether sodium hydroxide is increased in concentration (pH 14) or not used at all (water/MeCN only), the conversions are both lower compared with using pH of 12.7. Sodium hydroxide must therefore not be too high or low, as a saturated solution gives negligible conversion, and a 0.05 M solution shows higher conversion at 10% catalyst loading. This result is consistent with similar studies done on the Co(salen) compound, where it is suggested that there is a balance in pH: too high means decomposition of catalyst, and too low means insufficient deprotonation of the substrate to promote the reaction.103 Compounds CoII(CF3/Ph) and CoII(Me/CF3) were tested further since they typically gave the most promising results in preliminary studies using THF as solvent. All observations led to the implementation of the following optimized conditions: Solvent: THF (2 mL), Temperature: 80 °C, Vessel: 50 mL sealed bomb, NaOH(aq): 2 mL 0.15 M, O2: 1 atm, Cat. loading: 10%, no axial base, 48 hours. Dramatic increases in product conversion were observed for both compounds: 31.3 % using CoII(CF3/Ph) as catalyst and 33.3 % using CoII(Me/CF3), with the control experiment (no catalyst present) showing less conversion to the aldehyde (12.7 %). For reference, using the more water soluble Co(sulphosalen) (bis[(5-sulphonatosalicylaldehyde)ethylenediiminato]) compound, only 15% conversion was observed in a separate study by Floriani, et. al under similar conditions.104 Indeed, increasing oxygen concentration was a major factor, even without catalyst  60  present. The deposition of a dark black residue on the side of the reaction flasks became increasingly apparent under these new conditions, along with a decrease in yellow colour as the reaction progresses. The black solid was insoluble in acetone and in concentrated nitric acid. These observations, along with a decrease in pH to 5-6 of the aqueous phase are evidence of catalyst decomposition to cobalt oxides, where hydroxyl anions are being sequestered to an oxidized cobalt(III) hydroxide. It cannot be ruled out that peroxo-bridged species are forming species that are too stable, effectively deactivating the catalyst. This has been observed as a consequence of lowered pH solutions of cobalt under aerobic conditions,103 with the product being insoluble until liberated with addition of more hydroxide ions. Triethylamine, a non-coordinating base, was substituted for sodium hydroxide, along with the removal of water, in order to assuage harsh alkoxide/hydroxide reaction conditions that might be causing catalyst decomposition, and to see whether NaOH plays a significant role. The reaction shuts down for CoII(Me/CF3) to a low conversion of 11.5 % (< 12.7% observed with no cat.), and little change was observed using triethylamine with CoII(CF3/Ph). The rate limiting step using CoII(CF3/Ph) is likely product dissociation,67 since both weak and strong bases (NEt3 and NaOH, respectively) give the same result, but because NEt3 fails in the CoII(Me/CF3) system, deprotonation is likely the main barrier for turnover for this compound. A proposed mechanism is postulated for our system in Scheme 20 and is similar to that of Scheme 11.  61  Scheme 20: Proposed mechanism for the oxidation of benzylic alcohol veratryl alcohol by catalyst CoII(CF3/Ph)  Another possible explanation is that whereas THF is a sufficient axial base for dioxygen activation or substrate binding to CoII(CF3/Ph), sodium hydroxide is required for CoII(Me/CF3). Peroxo-bound dimers might also be becoming locked down in the less sterically-encumbered CoII(Me/CF3) complex unless NaOH is present. Results are summarized in Table 9. Table 9: Conversion of veratryl alcohol to veratryl aldehyde for two separate bases  Catalyst  Base  Conv. (%) by 1H NMR  CoII(CF3/Ph)  NaOH NEt3  26.7 33.5  CoII(CF3/Me)  NaOH NEt3  32.3 11.5  no catalyst  NaOH  12.7 62  3.9 Experimental 3.9.1 General Considerations Cyclic voltammetry was carried out under N2 using a three-electrode configuration (glassy carbon working electrode, Pt wire counter electrode, Ag/AgCl reference) and a PARSTAT 2273 potentiostat and function generator. The ferrocene/ferrocenium couple served as internal reference.105  GC/MS (for oxidation of veratryl alcohol experiments ): A Varian Saturn GC/MS/MS with Orbitrap mass analyzer CTC Analytics was combined with a Varian CP-3800 gas chromatograph equipped with a 1177 (middle) injector and a Combi PAL autosampler with headspace technology. The capillary column was a Varian Chromapack CP-Sil 8 CB column, with dimensions: 30.0m x 0.25m ID x 0.25μm (coating).  Screening tests for reactions of veratryl alcohol carried out in test tubes and vials were prepared in air from stock solutions of Co(II) prepared in the glove box. NaOH stock solutions were prepared in water in 50 mL volumetric flasks. Samples were heated in a heating block or in an oil bath to the desired temperatures after all reagents were added in the following order: Co(II), NaOH, py (if used), veratryl alcohol. All experiments were stirred in test tubes with small punctured holes in the caps, except for entry in Table 7 (A), where sealed vials in a heating block were used with no stirring. Experiments thereafter with CoII(CF3/Ph) and CoII(Me/CF3) under 1 atm of O2 were carried out in bombs (see “Reactions of veratryl alcohol under 1atm O2”). All extractions from test tube and vial experiments followed a similar protocol: a half volume of water was added, the solutions were neutralized with HCl, and solutions were extracted with ethyl acetate (3 x 5 mL). The combined extracts were dried over MgSO4 and concentrated in vacuum to crude off-white solids which were evaluated by 1H NMR.  63  3.9.2 Electrochemistry Into an electrochemical cell was placed 20 mL of the appropriate anhydrous solvent, either DCM or MeCN, along with 775 mg (2.00 mmol) of NBu4PF6 and a stir bar. The solution was purged with vigorous stirring for 2 minutes and a solvent background spectrum acquired at 100 mV/s and 50 mV/s. If the baseline was not satisfactory (i.e. within ± 2 μA), the solution was purged for an additional minute and reran until no background signals were present. Cobalt(II) (0.020 mmol for 1 mM or 0.1 mmol for 5 mM) was then added, followed by additional purging (~1 min), and scans were performed from -0.5 → 2 →-0.5V for oxidations, and 0 → -2 → 0 V for reductions at scan rates of 50 and 100 mV/s. After sufficient voltammograms were acquired, a small spatula tips worth of ferrocene was added until it persisted as the dominant redox couple in the voltammogram. The value from the half-reduction potential (mV) of ferrocene was then subtracted from the Co(II) data to yield corrected potentials for voltammograms. 3.9.3 Preparation of CoIII(Me/CF3)Et via reduction and subsequent reaction with ethyl iodide To a solution of the cobalt(II)(Me/CF3) (38.9 mg 0.100 mmol) in 2 mL THF was slowly added a suspended solution of crushed sodium metal (3.0 mg, 0.13 mmol) in THF (10 mL). The solution became dark after 10 minutes and after 24 hours of stirring a deep blue colour persisted. Air exposure of this blue solution yields the Co(II) as evidenced by UV-vis. The solution was filtered by a Kim wipe/pipette into a new vial (Celite filtration changes the highly reactive blue solution to red-brown, presumably starting material). With stirring, ethyl iodide (8 μL, 0.1 mmol) was added dropwise and the blue filtrate and the solution became dark brown. After 2 hours the mixture was reduced to a solid, redissolved in minimal ether (~1 mL), and Celite-pipette filtered into a new vial. The filter was rinsed with 1 mL hexanes and the combined filtrates were placed in the freezer for one week. The isolated solid was dark purple while wet and green-gold after drying. Yield = 33.8 mg, 80.8%. 64  3.9.4 Reactions of Co(II) compounds with iodine For CoII(Me/Ph) and CoII(CF3) 0.015 mmol of each Co(II) was taken from stock solutions made from THF (2 mM Co(II)) and hexanes (5 mM Co(II)), respectively. To the stirring solutions were added 3.8 mg (0.015 mmol) of iodine and after 1 hour of stirring UV-vis spectra were acquired. For CoII(CF3/Ph), iodine (13.0 mg, 0.0512 mmol) dissolved in 3 mL diethyl ether was added to a stirred solution of CoII(CF3/Ph) (51.2 mg, 0.100 mmol) in 10 mL of diethyl ether. UVvis was acquired from 2 drops of the solution in diethyl ether (starting material). 3.9.5 Attempted synthesis of CoIII(Me/CF3)Et by oxidative addition of I2, followed by reaction with MgBrEt Iodine (13.0 mg, 0.0512 mmol) dissolved in 3 mL diethyl ether was added to a stirred solution of CoII(CF3/Ph) (51.2 mg, 0.100 mmol) in 10 mL of diethyl ether. UV-vis was acquired from 2 drops of the solution in diethyl ether (starting material). A similar procedure was carried out for CoII(Me/CF3) (58.3 mg, 0.150 mmol of Co(II), and 0.0406 mmol of I2) in THF, only ethyl magnesium bromide (0.050 mL of 3.0 M solution in diethyl ether), was added dropwise after addition of iodine and let stir for 24 hours in the dark. The solution was filtered, reduced to a brown solid, redissolved in diethyl ether to yield a green solution and placed in freezer, however no product could be obtained. When reduced back to a solid (brown), 1H NMR showed a mixture of impurities and free ligand, though some CoIII(Me/CF3)Et was also detected (see Appendix C, Fig. 57). 3.9.6 Attempted reaction of CoII(CF3) with various acids to evolve H2 gas The protocol for these experiments is identical to other Co(II) reduction experiments, with the exception of adding small quantities of acid into 1.0 mM of CoII(CF3) after one scan was carried out in the absence of acid. The following quantities and observations were observed: 65  p-Toluenesulphonic acid monohydrate: after 7.60 mg (0.0400 mmol) of the acid was added to 0.02 mmol of CoII(CF3) in 20 mL MeCN, some colour loss and no change in voltammogram (50 mV/s) was observed. An additional 7.66 mg (0.0403 mmol) was added, followed by an additional scan (100 mV/s) resulting in complete colour loss in the solution and an irreversible redox wave. Scan range: (+0.3→ -1.5→ +0.3 V). Benzoic acid: added acid (0, 2, 4, 8 mM) with subsequent scans yielded no changes in voltammograms and no colour changes in a 1.0 mM solution of CoII(CF3). UV-vis also confirms stability of CoII(CF3) (0.002 mmol) in benzoic acid (6.3 mmol) in 3 mL MeCN. Scan range: (0→ -1.5→ 0 V). N,N-diisopropylethylamine hydrochloride: 0.010 mmol mixed with 0.002 mmol of CoII(CF3) in a UV-vis cell with 3 mL of MeCN shows complete loss of colour, and a loss in the shoulder at 445 nm in UV-vis. No electrochemistry was performed with addition of this acid to CoII(CF3). 3.9.7 Reactions of CoIII(Me/CF3)Et with (E)-stilbene CoIII(Me/CF3)Et (34.8 mg; 0.0832 mmol) was placed in a 50 mL sealable glass vessel along with (E)-stilbene (15.0 mg; 0.0832 mmol) and 6 mL of benzene. A few drops of the reaction mixture were placed in an air-sensitive UV-vis cell diluted in benzene (~ 3 mL). Both reaction mixtures were sealed and heated to 65 °C, with UV-vis acquisitions taken periodically over the span of 120 hours. After 5 days, the bulk solution was concentrated in vacuo to a solid, extracted in hexanes (10 mL), Celite-filtered, and the solvent removed in vacuum to a green solid (40.75 mg) that contained starting materials (E)-stilbene and CoIII(Me/CF3)Et by 1H NMR.  66  3.9.8 Reactions of CoIII(Me/CF3)Et with (E)-chalcone CoIII(Me/CF3)Et (70.9 mg; 0.170 mmol) was placed in a 50 mL sealable glass vessel along with (E)-chalcone (35.4 mg; 0.170 mmol) and 20 mL of ether and 20 mL of hexanes. The vessel was sealed and the reaction mixture stirred for 24 hours under a compact fluorescent bulb (150 W), after which time a red-brown precipitate was observed in a light green coloured solution. The reaction mixture was concentrated in vacuum to a small volume (~ 5 mL) and filtered through a pipette silica-hexanes filter in air. The residue was washed with hexanes (15 mL) and the filtrate (clear and lightly tinged yellow) was evaporated to yield 7.05 mg of a whiteyellow solid, which was air-dried and analyzed by 1H NMR and COSY. The solid was washed through with diethyl ether and this fraction kept separately in the freezer overnight, eventually yielding dark maroon crystals of the starting material, CoIII(Me/CF3)Et (13.1 mg = 18.4 % unreacted). 3.9.9 Reactions of veratryl alcohol under 1atm O2 Co(II) (0.0200 mmol) was placed in a 50 mL sealable glass vessel and a stir bar. If NaOH was used, 2 mL of a 0.15 M aqueous solution (0.30 mmol) was added with 2 mL THF. If triethylamine was used instead of NaOH, 4 mL of THF dissolved the Co(II), and 45 μL (0.32 mmol) of triethylamine was added thereafter. Veratryl alcohol (29.1 μL; 0.200 mmol) was added lastly to the reaction mixtures. The vessel was purged with O2 with the cap slightly open for 2 minutes to flush the system and fill the headspace. After turning off the O2 slowly (bubbler stopped bubbling), the cap was immediately sealed, and the reaction heated to 80 °C in an oil bath with stirring. CoII(Me/CF3) typically turns from orange to dark red, and CoII(CF3/Ph) from dark red to brown after 3 hours of reaction time. After 48 hours, 15 mL of water was added, the solutions were neutralized with HCl, and extracted with ethyl acetate (3 x 5 mL). The combined  67  extracts were dried over MgSO4 and concentrated in vacuum to yellow/brown oil and solids which were evaluated by 1H NMR.  68  Chapter 4: Studies of Cobalt-Carbon Bond Homolysis 4.1 Preface This purpose of this portion of the project was to investigate which compounds in the cobalt(III) oxidation state, while bound to alkyl substituents, would exhibit the highest Co-C bond strengths. Solvent binding constants for THF were calculated in order to see which compounds would have highest affinities for neutral axial donors. The Co-C bond lengths in our Co(III)Et complexes are more or less the same by X-ray crystallography, meriting closer examination into factors dictating their Co-C lability. Values of ~2 Å are also virtually identical to those observed in alkyl cobaloxime and Co(salen/salphen) systems.75 The only clue yet as to the relative strength of such bonds lies in the voltammograms since a more favourable Co(II/III) couple should correlate with a more stabilized Co(III) species. The expected order of increasing bond strength is thus expected to be the same as the order of increasing oxidation potentials: RCoIII(CF3) < RCoIII(CF3/Me) < RCoIII(CF3/Ph) < RCoIII(Me/Ph) The relation of BDE to OMRP is that a lower trapping efficiency of the cobalt(II) species with more electron withdrawing groups would be expected to result in higher concentration of radicals and a higher rate of polymerization (Chapter 5). A more thorough evaluation of relative Co-C bond strengths is offered in this chapter and evidence of solvent influence turns out to be a major factor. Information on relative strengths of Co-C bonds in the Co(III)Et complexes of our model compounds are of potential importance to many catalytic reactions, especially OMRP, and our values typically fall in the upper ranges of similar compounds, signifying relatively strong Co-C bonds.  4.2 Investigation of neutral axial donors on Co(II) and Co(III)Et compounds As crystals, the compounds under study are isolated as either Co(II) square planar compounds, or 5-coordinate square pyramidal Co(III) structures. However, in solutions of 69  increasing donor strength, Co(II) compounds tend to darken, and more significantly for some than others. A significant clue as to whether or not Co(III) is 5-coordinate or 6-coordinate (i.e. solvent-coordinated) is colour: compounds behave similarly to 5-coordinate alkyl cobalt(III)(salen)9 and RCo(bae)84 complexes in that they are characteristically green when unsolvated, but tend toward more brown colours when they are six-coordinate species. UV-vis experiments were carried out to explore the tendency of our compounds to axially coordinate oxygen donor solvents. Where less disparate spectra seem to suggest little influence of THF on Co(II), all Co(III)Et compounds were found to coordinate THF to some degree, and a significant darkening from green to brown is observed. Formation constants (  ) of THF to all  Co(III)Et compounds and DMSO and THF to CoII(CF3) were measured according to equation 8. (equation 8)  The axial coordination of THF to Co(II) occurs similarly to the Co(III)Et compounds, and is illustrated below in Scheme 21.  Scheme 21: Equilibrium between solvent-free and THF-bound species of Co(II).  4.2.2 Binding constants for THF-Co(III)Et compounds Axial coordination is known to stabilize octahedral Co(III) species and the cobalt-carbon bond strength is influenced by the electron-richness of the metal center. Alkyl cobalt models for vitamin B12 with equatorial nitrogen donor atoms (e.g. CH3Co(dh2)) typically contain more 70  acidic metal centers and are therefore more prone to exist as six-coordinate in donor solvents compared to cobalt compounds containing the β-ketoaminato (salen/salphen/bae) N/O donors.84,106 In fact, CH3Co(dh2) forms dimers in dichloromethane.106 If the solvent is an oxygen donor, such as water or methanol, these tend to have low formation constants toward alkyl(chelate)cobalt in solution.107 It was therefore surprising to see our Co(III)Et compounds show evidence of binding to some oxygen donors. The first indication that electron-deficient Co(III)Et species were more likely to coordinate axial donors was from UV-vis experiments in vinyl acetate, where the CoIII(CF3)Et compound produced electronic spectra that suggested solvent interaction. The tendency to be solvent-coordinated octahedral Co(III)Et complexes in THF was measured by exploiting the presence of bands at ~650 nm present when the compounds were dissolved in non-coordinating solvents (either toluene or hexanes). Subsequent addition of aliquots of THF decreases the absorbance of this band until it virtually disappears. An example is shown for CoIII(CF3/Ph)Et in Figure 22, with absorbance plots shown at three different wavelengths for this compound when more and more THF is added shown in Figure 23.  71  Figure 22: UV-vis spectra of CoIII(CF3/Ph)Et (5 μmol) in toluene with increasing THF.  Figure 23: Plots of absorbance vs. mmol of THF added to CoIII(CF3/Ph)Et (5 μmol) in toluene at three different wavelengths (400, 535, and 648 nm); T = 25 °C  Equation 4 described in the introduction (Chapter 1) shows how varying temperature and monitoring changes in wavelength can yield useful thermodynamic data, namely  , and  72  from plots of ln(K) vs. . This Van‟t Hoff plot is shown in Figure 24 for CoIII(CF3/Ph)Et with thermodynamic data summarized for all compounds found in Table 10.  Figure 24: Van‟t Hoff plot of CoIII(CF3/Ph)Et (5 μmol) in toluene with 3.7 mmol THF  Table 10: Summary of thermodynamic values for binding THF to Co(III)Et compounds. Compound CoIII(Me/Ph)Eta  K (± RSD %); ∆G° (cal·mol-1, 25 °C) 0.03 (± 42); -17  CoIII(CF3/Ph)Etb 0.18 (± 13); -109 (400 nm) 0.19 (± 18); -112 (535 nm) 0.20 (± 16); -117 (648 nm) III Co (Me/CF3)Et 0.10 (± 28); -57 CoIII(CF3)Etc 8.7; -5200  (kcal·mol-1 ± RSD %) -3.6 (± 8) -1.4 (± 9) -4.0 (± 7) -3.3 (± 14)  ∆So (cal·mol-1K-1 ± RSD %) -18 (± 6) -12 (± 4) -16 (± 6) -14 (± 3)  -1.34 (± 4) -8.07 (± 13)  -1.3 (± 18) -22 (± 16)  (a) Two values for and were found using two different experiments with varying amounts of THF. (b) As (a), only for K and ∆G°, values were found by monitoring three different wavelengths (λmax = 400, 535 and 648 nm). (c) Non-coordinating solvent: hexanes.  73  Equilibrium constants at 25°C reveal that axial THF coordination is generally more favoured as the electron withdrawing effect of the equatorial ligand increases, unless entropy effects are a major factor. The observed entropy changes for the association of THF to all Co(III)Et compounds are negative as expected; however, this entropy decrease is larger for some compounds over others: about -15 cal·mol-1K-1 for the two compounds with phenyl groups, but only about -1 cal·mol-1K-1 for CoII(Me/CF3). This may be due to a disruption of stabilizing solvent pi stacking effects between toluene solvent and the aromatic ligand groups, resulting in a higher entropy increase upon THF dissociation. Any relative changes in molecular motion between solvated and unsolvated forms of the compound results in changes to the binding entropy.108 It is therefore not surprising that CoIII(CF3)Et which has the highest solubility in hexanes, and therefore the strongest intermolecular interactions with the solvent, also exhibits a large entropy increase (22 cal·mol-1K-1) when THF is released into hexanes. The binding affinities for THF were shown to have little variation between the first three compounds (  = 0.03 – 0.19, Table 10). Only in the CoIII(CF3)Et compound does a more  substantial increase in the tendency to coordinate THF become apparent, and this is evidence for a more electron-poor Co(III) metal center as a square pyramidal species. A more favourable binding constant for THF to CoIII(CF3/Ph)Et over CoIII(Me/CF3)Et is inconsistent with electrochemical data, since the latter consistently shows a more electron-deficient cobalt center as Co3+, but this can be attributed to the large differences in entropy between these complexes.  4.2.3 Binding constants for solvents to Co(II) The binding constants for DMSO and THF to CoII(CF3) were found using UV-vis, monitoring wavelengths 400 and 418 nm where peaks are prominent in non-donor solvent 74  hexanes, and not present in DMSO or THF. This technique is essentially identical to the one used for finding THF binding constants to Co(III)Et compounds. values at room temperature were established for the CoII(CF3) compound in both THF and DMSO to illustrate its propensity to coordinate in donor solvents. Beer‟s law was used to first extract a value for the amount of free cobalt present in solution, followed by basic arithmetic to find the concentration of the solvent-bound species (Fig. 25).  Figure 25: Calculations for finding solvent binding constants (e.g. KTHF) by simple arithmetic with known values (blue)  The  values are 159 ± 36 (% RSD) and 2.2 ± 0.9 (± at 95% confidence) for DMSO  and THF respectively. Acetone is also thought to coordinate if used as a solvent since its spectrum is similar to that of THF and DMSO, however its binding constant must be significantly lower, as no changes in spectra were observed with a 9-fold excess of acetone to cobalt.  75  Figure 26: UV-vis spectra of CoII(CF3) in hexanes and various amounts of DMSO. Wavelength maxima monitored at 418 and 400 nm for calculation of KDMSO..  is 140 ± 13 (average % RSD) when averaging values at 418 and 400 nm for DMSO using a similar technique employed for coordination of pyridine to Schiff base compounds,15 where plots of  vs  yield straight lines from which  was deduced (  , Fig. 27). The straight line is also evidence that solvent coordinates 1:1 with cobalt.15  76  Figure 27: Plot of  vs  monitoring two different  wavelengths (418 nm, blue; 400 nm, red) for solvent DMSO. Unfortunately, for other compounds spectra in THF and non-donor solvents are less disparate, with only a slight broadening of peaks. This could be due to an equilibrium existing between coordinated and non-coordinated states in THF. Solubility issues with CoII(Me/Ph) limits its study in non-donor solvents; however, UV-vis spectra in toluene and THF are virtually identical, indicating that THF is not a donor solvent for this compound, which is not surprising considering its low affinity for THF in the form of Co(III)Et (K25°C = 0.03). It is important to note that the binding constants for THF are larger in the Co(III) oxidation than the Co(II) for the Co(CF3) compound, and this trend is likely true for all compounds. The result also implies that if axial coordination is not occurring in the Co(III)Et compound, then it is unlikely to be occurring in the corresponding Co(II).  4.3 Kinetic determination of the bond dissociation energy for Co(III)Et compounds using a radical trap. Cobalt(III)alkyl compounds are known to undergo bond homolysis in the presence of the radical trap (2,2,6,6-tetramethyl-piperidin-1-yl)oxyl (TEMPO) at elevated temperatures.13 TEMPO is a sterically bulky nitroxyl-based persistent radical that does not self-couple to form a 77  peroxo bond, but instead reacts quickly with carbon-based radicals.36 Since β-hydrogen atoms are present in the alkyl substituent bound to cobalt, elimination of the corresponding alkene can occur in addition to Co-C homolysis. Halpern has shown that β-H elimination has a barrier approximately 2 kcal∙mol-1 higher in energy than cobalt-carbon bond homolysis in the (py)(dh2)Co(CH(CH3)C6H5) system using equilibrium methods.12,56 Assuming the change in radical concentration over time is zero (i.e. steady state approximation), and that disappearance of Co(III)R in the presence of trap occurs as the sum of two pathways (β-hydrogen elimination and trapping of R●), equations in Scheme 22 can be deduced,11 and this allows for the extraction from kinetic plots to yield rate constants such as ko, k−1, and k1 in order to evaluate  for Co-C bond homolysis.  Scheme 22: Mechanistic pathways for decomposition of Co(III)Et and equations deduced from rate laws.11  Kinetic runs were carried out for CoIII(Me/Ph)Et and the CoIII(Me/CF3)Et at 65 °C by keeping their concentrations constant while varying Co(II) and TEMPO concentrations (0-6 and 78  0-40 equivalents, respectively). Disappearance of the CoIIIEt (652 nm) for CoIII(Me/CF3)Et and emergence of the Co(II) complex (413 nm) for CoIII(Me/Ph)Et decomposition was monitored by UV-vis (Fig. 28) and the observed results conformed to expected pseudo first order behaviour under these experimental conditions. The reactions were monitored to >3 half-lives in every case.  Figure 28: Co(Me/Ph) species at the start/finish of an experiment in THF (left), and pseudo firstorder rate plot for CoIII(Me/Ph)Et decomposition in TEMPO; (Co(III)Et/Co(II)/TEMPO = 1:1:10).  Some assumptions and variations to Halpern‟s method11,56 are worth noting since these could affect the accuracy of results. Whereas diethyl ether, a non-coordinating solvent, was used for experiments with CoIII(Me/CF3)Et for calculating the dissociation of the ethyl radical from the square pyramidal structure only, solubility problems resulted in using THF as a solvent for CoIII(Me/Ph)Et. THF has been shown to bind to the alkyl complex, albeit rather poorly (  = 1-  4 kcal·mol-1) and so it may be affecting the bond strength to some degree. Also, the rate constant ko (at 65°C) was estimated by minimizing the standard deviations of the intersection points from plots of (kobs – k1) (or “krad”) vs. [Co(II)]. Ordinarily, the value of ko is obtained by running experiments under limiting high Co(II) and low radical trap TEMPO.11 The experimental value 79  for ko using the CoIII(Me/CF3)Et compound was found to agree reasonably well with ko using this method (6.3 vs. 9.8 x 10-5 s-1), and deviation is attributed to a high relative standard deviation (16%) in the experimental as a result of exponential decay not fully reaching baseline after 17 hours reaction time in the spectrometer. Having to evaluate ko at every temperature for Eyring plots would be tedious, and so the assumption that k1 = kobs - ko(65°C)/k1(65°C) x kobs was utilized. Values for  are similar when  assuming ko is negligible (i.e. k1 = kobs) since ko/k1 is very low (0.03 and 0.07 for 1 Co(Me/Ph) and Co(CF3/CH3), respectively). The β-H elimination step, if indeed it is proceeding, is thus comparatively slow. Values of k-1/k2 from the slopes of plots such as those in Figure 29 range from 3-5, and with a known value of k2 on the order of ~109 M-1s-1 at 18 °C for primary alkyl substituents to TEMPO,36,109 k-1 seems to be in range of the diffusion controlled limit. A sample Eyring plot with an excess of TEMPO is shown in Figure 29, the slope from which was used to calculate  .  80  Figure 29: Kinetics of the decomposition of CoIII(Me/Ph)Et at 65 °C in the presence of various equivalents of Co(II) (0 – 6) and radical trap TEMPO (2 – 40) (left), and an Eyring plot assuming ko = 0 (right), where TEMPO/Co(III)Et = 20 with no Co(II) added.  Conditions for Eyring plots (Fig. 29, right) were arranged so that no additional Co(II) was added, and that an excess of 20 equiv. TEMPO was used to ensure that kobs = ko + k1. varied only slightly (+ 1.5 %) when an additional two equivalents of Co(II) were added, showing that the excess amount of TEMPO is adequate. Table 11: Summary of kinetic data Compound  ko (65°C), s-1 a CoIII(Me/Ph)Et 5.6 x 10-5 d CoIII(Me/CF3)Et a9.8 x 10-5 e CoIII(Me/CF3)Et 6.3 x 10-5  k1 (65°C), s-1 1.9 x 10-3 8.6 x 10-3 8.5 x 10-3  k-1/k2 (65°C) 2.8 4.9 4.0  ∆H‡, kcal·mol-1 29.2b /(29.9)c 25.6 27.0  ∆S‡, cal·mol-1 ·K 15.3b/(28.2)c 3.4 30.3  BDECo-R, kcal·mol-1 27b/(28)c 24 25  In all cases . ∆H‡ and ∆S‡ is ≤ ± 3% RSD (a) values estimated from minimal standard deviations in k1 (65 °C). (b) assumes ko= 0 (c) ko/k1 = 0.029 for all temperatures (d) ko/k1 = 0.074 for all temperatures; ∆H‡ and ∆S‡ are identical when ko = 0 is assumed. (e) experimentally obtained: 7:2 ratio Co(II):TEMPO; k o ± 16% RSD; ko/k1 = 0.114 is used for all temperatures.  81  4.4 Determining the BDE for Co(III)Et compounds using 19F NMR Decomposition studies of the cobalt-alkyl species are tedious, kinetic runs from which require about 3 hours to display an unchanging spectrum. However, by allowing equilibrium to establish by mixing an alkyl species of one cobalt compound with the cobalt(II) species of another,110,111 an approximation of bond dissociation energies could be obtained for compounds CoIII(CF3/Ph)Et and CoIII(CF3)Et. The entropy term for this equilibrium is assumed to be negligible due to the structural similarity of the starting reactants and products. The reaction is shown in Scheme 23, with thermodynamic equations (equations 9-14) and a sample 19F NMR spectrum for a mixture of CoIII(Me/CF3)Et and CoII(CF3) in DMSO shown (Fig. 30). Other 19F NMR spectra can be found in Appendix C (Fig. 58 and Fig. 59).  Scheme 23: Ethyl radical transfer between cobalt compounds Co(Me/CF3) and Co(CF3/Ph) and the derivation for Keq and  (equation 9) 82  (equation 10) (equation 11) (equation 12) (equation 13) (equation 14)  Figure 30: 19F NMR for a mixture of CoIII(Me/CF3)Et and CoII(CF3) in DMSO-d6.  The equilibrium constant found by using integration values from signals arising from Co(II) and Co(III)Et species in the19F NMR, and was converted to  for the process (  ). This value was then added to the measured BDE reference for CoIII(CF3/Me)Et (  = 24.3 kcal·mol-1) thereby obtaining bond energies for CoIII(CF3/Ph)Et and CoIII(CF3)Et.  83  It is important to emphasize that the Co-C BDE for CoIII(Me/CF3)Et of 24.3 kcal·mol-1 is taken from kinetic experiments in non-coordinating solvent diethyl ether, and is assumed to be the same as (or at least very close to) the Co-C BDE of this compound in acetone.112 UV-vis confirms that, like diethyl ether, acetone does not coordinate to Co(Me/CF3) in either the Co(II) or Co(III)Et forms, and so this assumption is likely sound. The same holds true for Co(CF3/Ph), but unfortunately not for Co(CF3), since it is electron-poor enough to coordinate acetone. An acetone mixture of the reference compound CoIII(Me/CF3)Et with CoII(CF3), favours the ethyl bound as CoIII(CF3)Et, which is unexpected since this compound is the most unstable as Co(III) according to electrochemical data. The explanation is simple: acetone only interacts with the CoIII(CF3) compound (in both oxidation states, as confirmed by UV-vis) and not with the reference, CoIII(Me/CF3) in either oxidation state, and so CoIII(CF3)Et experiences solvent stabilization from acetone, rendering the Co-C bond effectively stronger in this solvent. Relative stabilization of the Co(II) compared to Co(III) must not be as great, but this is expected since axial donors are less prone to coordinate Co(II) as evidenced by the THF binding studies. Conversely, a mixture of the reference compound, CoIII(Me/CF3)Et, with CoII(CF3) in dimethyl sulphoxide shows a stronger Co-C bond for CoIII(Me/CF3)Et as expected, since coordination of all species occurs in DMSO. DMSO is thought to coordinate all Co(II/III) species much more strongly (DMSO > THF > acetone)112 as evidenced by dramatic changes in UV-vis spectra with a loss of higher wavelength bands. Unfortunately, assuming the DMSObound reference (DMSO-CoIII(Me/CF3)Et) has the same Co-C BDE in diethyl ether (noncoordinating) is not a good assumption, and so the absolute BDE‟s cannot be known for either complex in DMSO; only the relative BDE difference can be defended. The concepts described are illustrated in Fig. 31 and summarized in Fig. 32.  84  Figure 31: Energy diagram and equilibrium equation illustrating solvent effects of DMSO and acetone on bond strengths  Although actual BDE values cannot be extracted reliably from DMSO or acetone data for CoIII(CF3)Et, all species are believed to be coordinated in DMSO and so the enthalpy value for this equilibrium (equilibrium in Fig. 31, bottom) is thought to best represent the relative bond strength. In other words, if solvent is stabilizing all species in solution to the same degree, then the observed weaker BDE for CoIII(CF3)Et compared to CoIII(Me/CF3)Et is the result. This is in agreement with results from electrochemical data, where a favourable Co(II)→Co(III) redox couple reflects an increase in bond strength for corresponding alkyl cobalt compounds. The correlation can be found in Fig. 32. 85  Figure 32: Bond dissociation energies for Co(III)Et compounds and corresponding oxidation potentials. The asterisk * indicates relative BDE‟s of this compound to CoIII(Me/CF3)Et in the solvents shown.  The idea that the relative strength of the cobalt-carbon bond for the CoIII(CF3)Et compound (compared to CoIII(Me/CF3)Et) could be reversed by switching solvent to acetone, is an observation that could have drastic implications for controlled polymerization reactions, since a coordinating medium could alter the bond strengths of the catalyst. This is not surprising, as adding nitrogen donors or coordinating solvents has been shown to influence reaction rates in other cobalt mediated OMRP.20  4.5 Summary A summary of experimental results is found in Table 12 for the reactions described in Chapter 4 (summarized in Fig. 33).  86  Figure 33: Equilibrium equations for which thermodynamic data were obtained experimentally (Table 12)  Table 12: Summary of enthalpies from experiments for the reactions shown in Fig. 33 Reaction i ii iii iv v vi vii viii ix  a  c  (± RSD %) -2.51 (± 9) 29.2-29.9 -3.66 (± 5)  −1.48 (24.9)d -1.34 (± 4) 25.6-27.0 −0.73 +3.11 -8.34 (± 13)  *all units are in kcal•mol-1 (a) obtained from UV-vis equilibrium studies (b) obtained from kinetic measurements − 2 kcal·mol-1) (c) obtained from equilibrium studies by 19FNMR III and are relative to Co (Me/CF3)Et; a negative value indicates exothermic in the forward direction (d) BDE for CoIII(CF3/Ph)Et, square pyramidal (no axial base).  87  4.6 Experimental 4.6.1 Kinetics experiments In a typical experiment, stock solutions were prepared for Co(III)Et (1.00 mM), the corresponding Co(II) (1.0-5.0 mM), and TEMPO (4.00-20.0 mM) in volumetric flasks. The volume of Co(III)Et stock solution was consistent for every experiment (1.00 mL), with varying amounts of Co(II) and TEMPO ranging from 0-5.00 mL, placed (with the Co(III)Et, 1.00 mL) into new 10 mL volumetric flasks and made to the mark with solvent. THF was used with CoIII(Me/Ph)Et and diethyl ether for CoIII(Me/CF3)Et experiments. These dilutions were then used for UV-vis analysis at 65 °C until the spectrum was virtually unchanging. When not in use, stock solutions were stored in the freezer (-35 °C) for up to two days, after which time new stock solutions were made. 4.6.2 Axial base experiments CoII(CF3) with THF and DMSO: Stock solution of CoII(CF3) was made using hexanes as solvent (0.016 mM) and diluted by 1/10 with varying amounts of THF added (0 - 2 mL). Extinction coefficients for maxima at 400 and 418 nm were found when no THF was used and when sufficient THF was used so as no change in spectra was observed upon further addition. Subtracted solvent backgrounds were obtained for all runs (hexanes for THF < 1.0 mL, otherwise appropriate THF/hexanes mixtures were used). A similar procedure was invoked for use with DMSO, only a stock solution of DMSO (in hexanes) was needed in order to accurately add smaller quantities. Co(III)Et with THF: More concentrated solutions were used with ethyl cobalt(III) compounds (0.5-1.0 mM) in order to observe the disappearance of the smaller band at ~650 nm upon addition of THF. Specific wavelengths (λmax ~ 650 nm) monitored were: [CoIII(Me/Ph)Et, 668 nm]; [CoIII(Ph/CF3)Et, 648 nm]; [CoIII(Me/CF3)Et, 650 nm], and [CoIII(Me/CF3)Et, 650 nm]. 88  Toluene was used as the non-donor solvent for all experiments except for when CoIII(CF3)Et was the ethyl cobalt(III) being analyzed, in which case hexanes were used. Thermodynamic information for all solvent-complex formations were deduced by evaluating equilibrium constant dependencies on varying temperature. Temperatures were changed by ± 5 °C with approximately 5 minutes equilibration time before acquiring. This was found to be satisfactory, as no further changes in spectra are observed after this time. 4.6.3 19F NMR experiments Co(II) and Co(III)Et (0.025 mmol each) compounds were combined in a sealed 1.0 mL NMR solvent (DMSO-d6 or acetone-d6), and heated overnight at 90 °C in an oil bath (DMSO) or 75 °C (acetone) for 3 hours. The tubes were then heated in the NMR spectrometer and allowed to equilibrate for 10 minutes before acquiring spectra at 50 °C.  89  Chapter 5: Organometallic Mediated Radical Polymerization 5.1 Preface The bulk of experimental work is contained in this chapter, and was devoted to applying our square planar (β-ketoaminato)Co(II) compounds to the mediation of radical polymerizations, while trying to elucidate mechanistically how the process functions. Experiments generally showed low conversions and slow reactions under RT conditions; however, DT polymerizations of vinyl acetate provided more rapid polymerization rates while maintaining reasonable degrees of polymer control. Although a perfectly controlled polymerization of vinyl acetate was never fully attained, relatively low polydispersity values suggest experimental improvements in the future could lead to more promising results in utilizing these catalysts for OMRP.  5.2 Reversible termination (RT) experiments Experiments under reversible termination conditions were carried out for radical polymerizations of vinyl acetate (VAc), methyl acrylate (MA), and acrylonitrile (AN). Under these conditions, the number of molecules of Co(II) radical traps employed is never exceeded by radicals generated from decomposition of the initiator, V-70 or AIBN. Since the corresponding Co(III)Et compounds already have been shown to undergo homolysis when heated to 65 °C, it was assumed that similar homolytic trap and release would also ensure RT under the appropriate conditions. Decreasing rates of polymerization should correlate with increasing Co-C bond strengths of the catalysts since radical release enables propagation. Rapid trapping and slow release of the propagating polymer is responsible for keeping the radical concentrations and polydispersity low, and earlier work has showed promising results using CoII(Me/Ph) as catalyst and methyl acrylate as monomer.79  90  5.2.1 Vinyl acetate Experiments using 0.8 equiv. AIBN as the radical initiator and 1000 equiv. vinyl acetate were attempted; however, conversions were generally low, and polymer growth rates were exceedingly slow, erratic, and irreproducible. As an example, using CoII(CF3/Ph) as the catalyst at 75 °C, conversions never exceeded 4 % upon repeated experiments. However, when V-70 was utilized in place of AIBN and the monomer/Co ratio was increased from 1000:1 to 15000:1, some conversion was observed, and the new simplified protocol (see experimental) allowed for identical VAc and MA polymerization conditions to be carried out with non-negligible conversions for each monomer. Outcomes were examined using the four cobalt compounds as mediators and plots under first-order consumption of monomer are found in Fig. 34.  Figure 34: First order rate plot for RT polymerization experiments in vinyl acetate (left) and methyl acrylate (right). [Co(II)/V-70/monomer] = 1:0.75:75000, T = 65 °C  Catalysts CoII(Me/CF3) and CoII(CF3) exhibit the most significant polymer growth in vinyl acetate (29.3 and 32.8% conversion after 71 hours, respectively), whereas compounds 91  CoII(Me/Ph) and CoII(CF3/Ph) show very little growth (7.6 and 9.2 % conversion, respectively). This is not surprising since these latter compounds are most easily oxidized and are presumed to form the strongest Co-C bonds of the trapped species. Thus homolytic release and RT ceases to persist. It has already been shown that in acetone a stronger Co-C bond is observed for CoIII(CF3)Et over CoIII(Me/CF3)Et (Chapter 4). Because a stronger Co-C bond means less radical release, the rate of growth in VAc is comparable to, and even slightly higher using CoII(Me/CF3). UV-vis shows that VAc is a non-coordinating solvent, as VAc spectra are identical for all Co(II) compounds as in non-coordinating solvents. Therefore, relative to acetone the VAc medium would be expected to increase the trapping efficiency of the polymer by CoII(CF3) since the Co(II) species is no longer solvent stabilized for this compound. Overall, the results are as expected, but unfortunately long reaction times to reach minimal conversions led to implementing higher radical initiator concentrations and degenerative transfer conditions in order to speed up the reactions. 5.2.2 Methyl acrylate The induction period for MA was much shorter than for VAc, with conversions consistently above 30% after 20 minutes (VAc was < 8% after 4 hours). After this initial rapid influx of radicals, however, growth stabilized to measurable rates that seemed to be consistent for all catalysts, and also very similar to those experiments in VAc where appreciable growth rates were observed (using CoII(CF3) and CoII(Me/CF3)). The greater stability of a MA localized radical compared to that of VAc113 explains the initial rapid growth; however, the independence of catalyst selection and monomer (VAc vs. MA) on rates of propagation is a bit unsettling. Similar observations were observed for various Co(bis-aminopyrrole) catalysts in VAc,114 and 92  unfortunately no correlation to Co-C bond strength could be revealed. Experimental conditions were different than those reported from an earlier communication79 in that the monomer/cobalt ratios were different (500:1 was changed to 15000:1), and very little toluene was used. 5.2.3 Acrylonitrile Polymerization of acrylonitrile initiated by V-70 at 65 °C under RT conditions yielded immediate precipitation of white solids for catalysts CoII(CF3/Ph), CoII(Me/CF3), and CoII(CF3). The reactions were ceased after 2 hours. These catalysts likely have a low trapping efficiency for the propagating polymer compared to propagation, as the latter is clearly faster. The compounds may also be undergoing deactivation as nitrile-coordinated species to Co(II/III) under the reaction conditions. Using CoII(Me/Ph) as catalyst, conversion remained negligible for 71 hrs, and only at 2 % after 95 hrs. Precipitation of a brown solid was observed in solution and the residue (after 21 hrs) examined by UV-vis (in THF) did not show presence of the catalyst, which could be decomposing under the reaction conditions. Radicals may be terminating immediately following radical-trapping by cobalt, or immediately upon contact with the catalyst. In an attempt to alleviate problems with the former three catalysts where trapping was not taking place, coordinating solvent DMSO in a large excess with respect to cobalt, was added to the reactions. DMSO has successfully been employed in the OMRP of acrylonitrile using the Co(acac) system by Debuigne, et. al.20 and low polydispersities were attributed to both the solvent effect and the solvation of the polymer. Unfortunately, precipitation of a white solid polymer was visible at 30 minutes for all solutions, albeit with more mobility in solution than the thick white “blocks” formed in vials when DMSO was not employed. The reaction was halted  93  after 2 hours. The most notable difference when using DMSO as co-solvent was polymer formation using CoII(Me/Ph) as catalyst, which did not occur in the absence of DMSO. Either these catalysts are not ideal as mediators for propagation of this monomer, or significant experimental alterations need to be explored further. Solubility is a major issue for the resulting polymer, though this in itself may be indication of significant amount of free radical polymerization yielding primarily large insoluble, and highly branched polymers.  5.3 Degenerative transfer (DT) experiments in vinyl acetate Intercepts of first order monomer consumption plots under DT conditions show a radical release efficiency of 0.60 (±0.02) in all cases. The assumption is that one molecule of initiator (V-70 or AIBN), after self-coupling some of its own radicals, effectively produces 1.2 equivalents of active radicals in RT experiments (i.e. 60 % efficiency), and as such, a conservative amount (0.89 equiv.) of radical species per cobalt was used in RT experiments to ensure RT conditions were met. Trends from 1.1 and 1.3 equivalents of AIBN show polymer growth rates are influenced by the catalyst and plots of Mn vs. % conversion show gradual loss of control after 20 % conversion in most experiments with 1.1 or 1.3 equivalents AIBN/Co for all catalysts (Fig. 36). First order rate plots are shown in Figure 35 and polymer samples analyzed with ~30 % conversions are summarized in Table 13.  94  Figure 35: Rate plots for DT experiments at 1.1 (top) and 1.3 (bottom) equivalents of AIBN. Legend shows catalysts in order of increasing rate of reaction from top to bottom  95  Figure 36: Plots of Mn and PDI (Mw / Mn) vs. % conversion for all catalysts (AIBN/Co = 1.3).  96  Table 13: Molecular weight data for polymer samples isolated from DT experiments at approximately 30% conversion  Co(CF3/Ph) equiv. AIBN 1.1  % Conv. 31.26  Mn  Mw/Mn  16209  1.3  27.52  22138  equiv. AIBN 1.1  % Conv. 31.44  Mn  Mw/Mn  15116  1.3  32.36  19888  Co(Me/CF3) equiv. AIBN 1.1  % Conv. 31.25  Mn  Mw/Mn  1.19  Time (h) 6.50  21341  1.07  Time (h) 5.38  1.08  4.42  1.3  33.79  18900  1.19  4.72  equiv. AIBN 1.1  % Conv. 29.79  Mn  Mw/Mn  1.25  Time (h) 7.25  17735  1.31  Time (h) 7.43  1.16  4.42  1.3  31.95  19284  1.19  4.67  Co(Me/Ph)  Co(CF3)  The order of rates, with respect to the catalysts used, are not consistent for each set of experimental conditions (either 1.1 or 1.3 equiv. AIBN), possibly due to inaccurately weighed sample/initiator or contaminated samples, as reflected by variation in induction period. However, the main observation holds true: the catalysts with either the most electron withdrawing (CoII(CF3)) or donating groups (CoII(Me/Ph)) yield slower reactions than the other two catalysts. This is not the same observation as seen in the RT results and a counterintuitive one since CoII(CF3) is expected to have the weakest Co-C bond in the absence of a donor solvent. Although vinyl acetate coordination to alkyl compounds of Co(CF3) may be stabilizing the Co-C bond of cobalt-capped polymers, it does not explain why disparate rates are observed when the overall radical concentration should remain constant once all of the initiator is consumed. One hypothesis is that when cobalt holds onto its Co(III) state more rigorously under DT conditions, radical coupling and termination of excess radicals (opposed to degenerative transfer), is more prevalent as the initiator decomposes, and so the rate is effectively slower. Mechanistic studies using CoII(Me/CF3) were carried out in an attempt to provide insight into the reactions.  97  5.3.1 Mechanistic studies OMRP of vinyl acetate under DT conditions for catalyst CoII(Me/CF3) was monitored using UV-vis in order to observe reactivity of the catalyst during the course of the reaction. Figure 37 outlines the initial reaction before the cobalt is saturated with radicals and Figure 38 shows the reaction when excess radical is present.  Figure 37: Absorbance of OMRP under DT conditions for catalyst CoII(Me/CF3) before excess radicals are present.  98  Figure 38: Absorbance of OMRP under DT conditions for catalyst CoII(Me/CF3) after excess radicals are present.  In Figure 39 the changes in absorbance for selected wavelengths before and after excess radicals are present is plotted. A straight line when the Y-axis is (Ao-A)/(Ao-A∞) shows zeroorder kinetics with respect to the cobalt species being monitored.  99  Figure 39: Changes in absorbance for selected wavelengths before (blue diamonds) and after (red squares) excess radicals are present.  Rates of disappearance of Co(II) (λmax = 378 and 450 nm) and emergence of a “CoIII(Me/CF3)Et-like” band (λmax = 650 nm; attributed to short-chain Co(III)-P intermediates) were equivalent until the amount of radical exceeded that of the catalyst, at which point the latter band began to diminish and emergence of a new compound (λ = 490 nm) was observed. Formation of this “mystery compound” likely signifies the conversion of short-chain cobalt(III)alkyl compounds to longer cobalt-capped polymer chains43 as the rate of its emergence 100  is equal to that of the disappearance of the short-chain band. Also, the band at 490 nm only appears once radical concentration exceeds that of the catalyst, and we already know that any reversible termination until that point produces insignificant conversions of monomer under the time frame of the reaction. Scheme 24 summarizes the wavelengths monitored and their corresponding hypothesized molecular identities.  Scheme 24: Proposed mechanism of a DT process with CoII(Me/CF3) and UV-vis bands corresponding to each catalytic intermediate  To show that the presence of the band at 490 nm was not simply a consequence of a more viscous medium, perhaps forcing itself into the coordination sphere of a cobalt(III)-alkyl square pyramidal compound, a UV-vis spectrum of CoIII(Me/CF3)Et was acquired in a 20% polymer (Mw ~ 100 000, polymer beads) solution in vinyl acetate. The spectrum was similar to the 40 min  101  elapsed-time spectrum (presence of a high-wavelength band at 627 nm) but significantly different from that of the new species arising when R∙/Co > 1:1, and t > 40 min. (Fig. 40).  Figure 40: UV-vis spectrum of CoIII(Me/CF3)Et in 20 % PVAc medium (blue) superimposed with 1.3 equiv. of AIBN/Co mechanistic data at 40 min. (green, R∙/Co ≈ 1:1) and 1 hour (R∙/Co > 1:1, red)  A separate reaction wherein the best radical trap, the CoII(Me/Ph) compound, was reacted with AIBN (1 equiv.) without the presence of monomer yielded no reaction, signifying that in neat monomer, cobalt does not immediately trap the initiator, but rather a short-chain polymer radical, post-initiation. 5.3.2 Varying temperature Using the CoII(CF3/Ph) as catalyst under degenerative transfer conditions (AIBN/Co = 1.3), conversions at ~13 and ~50 % were compared across a range of temperatures (Table 14).  102  Table 14: Polymer data for selected conversions in temperature-varying experiments with CoII(CF3/Ph) as catalyst Temperature % (oC) Conversion 63 75 97  13.8 13.0 49.6 13.4 49.9 51.1  Total Time (h) 22.7 23.4 4.92 0.40 0.58 0.65  Time after induction (h) 5.51 6.23 1.03 0.10 0.29 0.35  Mn; [Mw / Mn]; Mtheo  Mn/Mtheo  12972; [1.25]; 12331 14151; [1.28]; 11664 19688; [1.35]; 43062 7815; [1.09]; 12014 26846; [1.30]; 43395 32961;[1.29]; 44538  1.05 1.21 0.46 0.65 0.62 0.74  Rate plots for the three different temperatures monitored (63, 75, and 97°C) are found in Figure 41, with increasing rates following with increasing temperatures as expected.  Figure 41: Rate plots for varying temperature experimental data under degenerative transfer conditions, where AIBN/CoII(CF3/Ph)/VAc = 1:1.3:1000.  High temperature was actually shown to exhibit better control in general; however during the initial phase of the reaction (at low conversions) polymer control was best maintained at 103  lower temperatures. At higher percent conversions, control is best attained at higher temperatures. Higher conversions typically afford higher molecular weight distributions. 5.3.3 Varying cobalt to monomer ratios Using CoII(CF3/Ph) under degenerative transfer conditions, higher catalyst loading afforded better control. We see that when more cobalt is available to trap (relative to monomer) the rate of propagation is also slower, despite the ratio of radical initiator to cobalt being the same in each experiment. Molecular weight data for the 1000 equiv. (monomer/cobalt) show that after the induction period the polymer immediately grows up to about 20 kDa and no larger, and with continuous broadening in molecular weight distribution as the monomer consumption proceeds. But with higher catalyst loading (500 equiv. monomer/cobalt) the average molecular weight of the polymer gradually becomes larger and follows more closely with the theoretical (one cobalt per growing chain), and this is more characteristic of a “living” system. Also, the molar mass increase as a function of monomer conversion yields a straight line, indicative of a more controlled process (Fig. 42).45 More cobalt available to undergo DT to suppress the timespan of free radical propagation in the surrounding medium is likely responsible for a higher degree of control. This is also consistent with a slower observed rate of polymerization (Fig. 43). Molecular weight data for 50% conversion are summarized in Table 15.  104  Figure 42: Dependence of PVAc molar mass Mn on monomer conversion (%) for bulk polymerization of vinyl acetate at 75 °C and [AIBN]/[CoII(CF3/Ph)] = 1.3. [VAc]/[CoII(CF3/Ph)] = 500 (red squares) and 1000 (blue diamonds)  Figure 43: Time dependence of ln([M]o/[M]) for the bulk polymerization of vinyl acetate at 75 °C. [VAc]/[CoII(CF3/Ph)] = 500 (red squares) and 1000 (blue diamonds), and [AIBN]/[CoII(CF3/Ph)] = 1.3 105  Table 15: Polymer data for varying catalyst concentration at ~50% conversion Co/AIBN/VAc % Conversion 1:1.3:500 1:1.3:1000  49.8 49.6  Total Time (h) 5.50 4.92  Time after induction (h)  Mn; [Mw / Mn]; Mtheo Mn/Mtheo  1.62 1.03  14990; [1.21]; 21918 19688; [1.35]; 43062  0.77 0.46  5.4 Co(III)Et as radical initiator Previous work with CoIII(Me/Ph)Et showed it could be used as an effective radical initiator for polymerization of MA since subsequent formation of the corresponding Co(II) mediator could ensure reversible termination.79 Well-defined Co(acac)-P initiators (where P is a short chain polymer, or “oligomer”) have also been successfully employed for this purpose in VAc.44,46 Unfortunately, this technique did not work for alkyl compounds in VAc under RT conditions when Co(III)Et was used as the sole radical source, and negligible conversions were observed over several days. This can be attributed in part to initiator deactivation due to the persistent radical effect, but also to the stability, and therefore dormancy of this compound.46 Similar results were obtained under RT conditions when AIBN was used as the radical initiator. An experiment was conducted where CoIII(Me/CF3)Et, CoII(Me/CF3), and AIBN were reacted together, and the amount of carbon-based radicals (including potential ethyl radicals) was arranged to be in slightly greater excess than when reacted in a 1.3:1 ratio of AIBN to cobalt. The experimental induction period (R∙/Co ≥ 1 ) was much shorter, as expected, since half of the cobalt is initially being locked down as ethyl-bound (CoIII(Me/CF3)Et), and was found to agree with the theoretical (2.3 vs. 2.4 hours, respectively) assuming a radical efficiency of 0.6 for AIBN. The results showed a rate of polymer growth that was significantly reduced despite the greater radical excess (Fig. 44), a polydispersity somewhat closer to unity, and molecular weights that were slightly greater than the experiment wherein only AIBN was used as the sole radical initiator (Fig. 45). 106  Figure 44: Rate plots for DT experiments with CoII(Me/CF3) where [Co(II)/AIBN/VAc] = 1:1.3:1000 (blue) and [Co(III)/Co(III)Et/AIBN/VAc] = 1:1:1.78:1000 (red). Data points with negligible conversions to polymer are omitted for clarity.  Figure 45: Molecular weight data for DT experiments with Co(Me/CF3) where [Co(II)/AIBN/VAc] = 1:1.3:1000 (left) and [Co(II)/Co(III)Et/AIBN/VAc] = 1:1:1.78:1000 (right).  5.5 Discussion The natures of the β-ketoaminato catalysts in this study compared to the more successful Co(acac)2 catalysts used in the OMRP of vinyl acetate, can be used to explain the somewhat 107  discouraging results from polymerization experiments. Even with 6.5x the cobalt to radical initiator added and at a mild 30 °C, the Co(acac)2 system undergoes rapid polymerization after a short induction period.47 However, generally slow RT experiments along with early loss of control after ~20 % conversion in most DT experiments, and results from Chapter 4 showing relatively high BDE values for Co-C bonds of ethyl complexes, suggest that the cobalt is simply binding too tightly to polymer chains using our catalysts. Unsuccessful attempts at separating the cobalt from the polymer via irradiation of light or air exposure followed by silica and/or alumina filtration is further evidence for this. It would seem at this point that Co-C bonds must be sufficiently weak in order for the controlled process to work well. An example of a more successful system exhibiting weaker Co-C bonds is the Co(porphyrin) system, where the alkyl bond of Co(TAP) (TAP = tetraanisylporphyrinato) has a  value of 18 kcal∙mol-1;115 ~5  kcal∙mol-1 lower (weaker) than all of our β-ketoaminato compounds. However, even in the Co(acac)2 system, the limit can be reached, as a lack of control of acrylates was attributed to weak bonding to polymer chains.47 This latter problem was cleverly overcome in a later paper by using a larger excess of the trap, or by optimizing conditions and using a mix of Co(acac)2 and CoIII(acac)2(PVAc)n species.46 Like the porphyrin systems, cobalt structures are rigid; unable to self-stabilize to tetrahedral geometries in the Co(II) oxidation state. Also, axial coordination to Co(II) from the solvent is not considered a factor here, and therefore sequestering of the trap in this in manner to promote homolysis is not an option. With the Co(acac)2 system, addition of donor solvents effectively weakens the Co-C bonds since it out-competes intramolecular stabilization from the polymer OAc groups to the Co(III)P by coordinating axially 2:1 with cobalt as Co(II), and this method fortuitously also happens to keep the trapping efficiency sufficient for control.20  108  Altering the experimental conditions may improve the polymerization outcomes (i.e. better control and high molecular weights) using the β-ketoaminato mediators in this study. In retrospect, the most reasonable future experiment adjustment would involve RT experiments at higher temperatures than 75 °C to promote homolysis. This was attempted only under DT conditions in this study (97 °C) for the CoII(CF3/Ph) compounds, and better control at high conversion (50 %) compared to lower temperature experiments was the auspicious outcome. One could also employ the method used with the Co(dh2) system, that is, constant irradiation of the medium with light to encourage homolysis for RT.30 Nearly every DT experiment run at 75 °C has shown loss of control when the concentration of radicals exceeds that of cobalt. That is, a continual increase in conversion follows with increasing shortcomings in molecular weight compared to the theoretical. However, when implementing more catalyst relative to monomer, better control was maintained for longer (Fig. 42), and so 500 equiv. of VAc perhaps should be used instead of 1000 equiv. for future experiments. With a greater chance of radical species finding a cobalt center vs. another radical with which to terminate, „one chain per cobalt‟ may be easier achieved. DFT calculations and kinetic experimental data in the porphyrin system43 supports an associative mechanism for DT whereby mediation of living radical polymerization of vinyl acetate involves radical interchange through a three-centered three-electron transition state, (Scheme 25).  109  Scheme 25: Illustration of the concept of radical interchange for the degenerative transfer mechanism in Co(porphyrin). The radical species •CH(OAc)CH3 may undergo radical polymerization when released into solution.  Should this transition state instead be an actual energetic minimum in our system, it would explain why the disparate polymerization rates depend on the catalyst under DT conditions. Those cobalt compounds that could harbour the solvent-caged three-electron [P---Co---P] species more effectively would be expected to produce lower polymerization rates. For example, CoII(CF3) produces fastest polymer growth rates of VAc relative to the other catalysts under RT conditions, but slowest growth rates were observed under DT conditions. Also for this compound, there is less steric repulsion for incoming radicals and a higher affinity for polar groups on the polymer to coax the radicals into the coordination sphere, and so a higher tendency to sequester radicals would be expected. The suggestion of interplay between two radicals about the metal center under DT conditions is also consistent with the experimental observation that when short-chain polymer radicals are replaced in part by ethyl radicals, the polymerization rate decreases, since the rate of exchange from [P---Co-Et] to [P-Co----Et] would also not be expected to be the same. DFT calculations are recommended in the future to support this general idea.  110  The appearance of red solutions (from green or brown) in RT experiments indicates that regeneration of Co(II) from self-termination and CCT reactions may be occurring to some degree once significant time is passed. Disparate DT rates could simply be indicating that different catalysts simply undergo CCT at different rates, taking radicals out of the equation in the process, and slowing down the polymerization reaction. Though hydrogen atom transfer can reinitiate growth in the Co(porphyrin) systems, decomposition of Co-H species to H2 and Co(II) might be occurring in our system too rapidly for any CCT polymerization reactions to take place. A buildup of the radical trap (Co(II)) would result, and diffusion of radicals into solution for propagation becomes more and more attenuated (Scheme 26).  Scheme 26: Mechanism showing how CCT occurring in DT could take away radicals in solution and regenerate Co(II) to slow down the overall polymerization process.  111  5.6 Experimental 5.6.1 General considerations Molecular weight data on polymer samples were characterized by gel permeation chromatography (GPC), performed on a Polymer Laboratories PL-GPC 50+ with a PL-AS RT autosampler and PL-RI detector, a 5.0 m PLgel guard column, and two PLgel 5 m MIXED-C columns in series. The eluent was THF flowing at 1 mL/min at 30 °C. Polymer molecular weights were calculated against PS-H polystyrene standards (Polymer Laboratories). 5.6.2 UV-vis monitoring of DT experiment in VAc Catalyst CoII(Me/CF3) (0.014 mmol) and V-70 (0.021 mmol) were dissolved in 5 ml vinyl acetate. 0.200 ml of this solution was diluted into 2.8 ml vinyl acetate (3 ml total; 32.5 mmol) in an airtight UV-vis cell and heated to 50°C, at which time spectra were acquired every 3 minutes. 5.6.3 Polymerization reactions: degenerative transfer In a typical DT reaction, Co(II) was loaded into a 50 mL bomb (0.050 mmol) with AIBN (9.0 mg; 0.055 mmol for 1.1 equiv. or 10.7 mg; 0.065 mmol for 1.3 equiv. experiments), along with 4.6 mL (50 mmol) vinyl acetate and a stir flea. The red solution was placed in an oil bath at 75 °C and stirred with consistent stirring rate. Aliquots were removed periodically (~ 0.3 mL) with nitrogen backpressure to avoid possible oxygen contamination from air, and placed in a preweighed flask. The sample was weighed prior to concentrating to a solid, and then reweighed for gravimetric analysis. Samples were redissolved in THF, filtered through an alumina pad and injected into the GPC for analysis. 5.6.4 Polymerization reactions: reversible termination Reversible termination reactions were prepared by first making stock solutions of Co(II) (5.0 mM in THF) and V-70 (15.1 mg in 0.5 mL toluene to dissolve, and made to the mark with hexanes; 4.9 mM), aliquots from which were distributed in vials (4 per sample), with 0.325 mL 112  from the Co(II) stock (0.0016 mmol) and 0.25 mL (0.0012 mmol) of V-70. Monomer was added lastly to the vials (12.2 mmol AN; 24.8 mmol MA; 24.4 VAc) and DMSO in reruns of polymerizations of acrylonitrile (0.29 mL; 4.1 mmol) only. Vials were sealed with Teflon tape and placed in a heating block (16 vials total) at 65 °C, and removed at various times for aliquot removal (~0.3 mL) and gravimetric analysis.  113  Chapter 6: Conclusion The reactivity of our cobalt compounds for steps (i-vi) listed in the catalytic cycle from Scheme 3 (reproduced below for convenience) are summarized as follows:  Step (i): Reduction of Co(II) → Co(I) can be accomplished with Na metal but not with Mg metal, and air/moisture must be excluded from the reaction. The Co(I/II) couple was shown to be reversible for all compounds by cyclic voltammetry except for CoII(Me/Ph); however, the reluctance to react with the weaker reducing agent magnesium may preclude turnovers in catalytic reactions under mild conditions where Co(I) is required. Step (ii): the most promising results found were from reacting [CoI(Me/CF3)]Na with ethyl iodide, as the product was isolated cleanly. CoII(CF3) may not be a suitable catalyst based on poor nucleophilicity of its Co(I)−, and secondary alkyl halides (cyclohexyl halides) react more reluctantly, with evidence of Co(II) being produced. Other compounds show promising UV-vis with evidence of product formation. Step (iii and iv): Heat induces Co-C homolysis in the presence of TEMPO but little, if any cross coupling product was observed with reaction of CoIII(Me/CF3)Et with (E)-stilbene; the reaction could be limited to more activated alkenes, or require either higher temperatures or lower 114  temperatures reacted for longer. Light induces rapid homolysis and reaction with (E)-chalcone, yielding unpredictable mixtures of products using CoIII(Me/CF3)Et. (v): β-H elimination does not appear to play a significant role in the presence of radical traps (TEMPO); however, formation of polymers of lower than expected molecular weights suggests termination pathways via irreversible catalytic chain transfer to form terminated alkene ends may be occurring in DT experiments. A rapid step (vi) gives back the Co(II), which hinders homolysis and therefore shuts down polymerizations under RT conditions. Exposing [Co(I)]− to proton sources also yields the Co(II), demonstrating the instability of cobalt(III) hydrides; abstraction of β-hydrogen atoms from alkyls bound to Co(III) is not thought to occur reversibly. A full catalytic process involving Heck-type couplings is likely not suitable for these compounds, especially since sodium metal, a harsh reductant, is required for conversion of Co(II) to Co(I); however, further experimentation is warranted in reactions involving Co(I)Co(II)-Co(III) intermediates since these oxidation states can be accessed, and are often reversible in our compounds. Screening tests for various functionalized alkenes could potentially reveal catalytic activity with this cobalt system. Cobalt-carbon bond dissociation energies of Co(III)Et compounds increase in the order of increasing oxidation potentials for the corresponding Co(II) compounds. Oxygen donor solvent THF has the ability to bind Co(III) compounds to form octahedral structures in solution (and in the solid state in the case of CoIII(CF3)Et and reactivity of Co(II) and Co(III) has been shown in nearly every chapter to be influenced by the presence of axial binding solvents. This recurring idea is perhaps best exemplified by 19F NMR studies where the relative Co-C bond strength order could be reversed by simply changing the solvent (acetone vs. DMSO).  115  The cobalt compound with the most electron-withdrawing groups (Co(CF3)) displays the highest propensity for axial coordination to THF in both oxidation states, but less so as Co(II), as evidenced by binding studies with DMSO and THF. This affinity may be coupled with additional stability of Co-C or Co-O2 bonds trans to the neutral donor. Large entropic values in the Co(III)Et compounds with aromatic groups about the equatorial ligand account for unexpected trends in binding constants to oxygen donor solvents such as THF,84 and the nature of this effect is still not fully understood. Oxidation of veratryl alcohol in the presence of catalyst requires pressurized O2 atmospheres for appreciable conversion. Acetonitrile should not be used as solvent, and a specific recommended pH should be found for future experimentation. Results show somewhere above 7 and below 14 is best. Best results were obtained using CoII(CF3/Ph) with triethylamine as base (33.5 % conversion), and CoII(Me/CF3) with sodium hydroxide (32.3 %). Triethylamine should not be used with CoII(Me/CF3), as lower conversion than the control experiment shows lack of catalytic activity. OMRP experiments using cobalt displayed moderate control of molecular weights under DT conditions for vinyl acetate, but RT experiments were sluggish for this monomer. Polymer growth rates and overall conversions were correlated with Co-C bond strengths to our catalysts: insignificant growth over 71 hours was observed for those compounds showing highest Co-C BDE‟s for ethyl radical (< 10 % for CoII(Me/Ph) and CoII(CF3/Ph) compared with compounds exhibiting weaker bonds (CoII(Me/CF3) and CoII(CF3) ~30 % conversion). Precipitated solid white polymers were observed for acrylonitrile, and using methyl acrylate as monomer showed virtually identical rates of growth for all catalysts.  116  Varying temperature and monomer/catalyst ratio experiments suggest 500:1 ratios of monomer/cobalt and higher temperatures should be used in future experiments.  117  References (1)  Smith, K. M. Organometallics 2005, 24, 778–784.  (2)  Ullrich, J. Top Curr. Chem. 2012, 320, 323–452.  (3)  Ullrich, J. Top Curr. Chem. 2012, 320, 121–190.  (4)  Pfaff, F. F.; Kundu, S.; Risch, M.; Pandian, S.; Heims, F.; Pryjomska-Ray, I.; Haack, P.; Metzinger, R.; Bill, E.; Dau, H.; Comba, P.; Ray, K. Angew. Chem., Int. Ed. Engl. 2011, 50, 1711–5.  (5)  Ullrich, J. Top. Curr. Chem. 2012, 320, 191–322.  (6)  Giese, B.; Hartung, J.; He, J.; Hüter, O.; Koch, A. Angew. Chem., Int. Ed. Engl. 1989, 28, 325–327.  (7)  Costes, J.-P.; Cros, G.; Darbieu, M.-H.; Laurent, J.-P. Transition Met. Chem. 1982, 7, 219–221.  (8)  Li, G.; Zhang, F. F.; Chen, H.; Yin, H. F.; Chen, H. L.; Zhang, S. Y. J. Chem. Soc., Dalton Trans. 2002, 105–110.  (9)  Bhandal, H.; Patel, V. F.; Pattenden, G.; Russel, J. J. J. Chem. Soc. Perkin Trans. 1990, 2691–2701.  (10)  Brown, K. L. Dalton Trans. 2006, 1123–1133.  (11)  Tsou, T.; Loots, M.; Halpern, J. J. Am. Chem. Soc. 1982, 104, 623–624.  (12)  Halpern, J.; Flora, T. T. N.; Rempel, G. L. J. Am. Chem. Soc. 1979, 101, 7124– 7126.  (13)  Ng, F. T. T.; Rempel, G. L.; Mancuseo, C.; Halpern, J. Organometallics 1990, 9, 2762–2772.  (14)  Crumbliss, A. L.; Basolo, F. J. Am. Chem. Soc. 1970, 92, 55–60.  (15)  Kasuga, K.; Iida, Y.; Yamamoto, Y.; Aihara, M.; Kudo, M. Inorg. Chim. Acta 1984, 84, 113–116.  (16)  Li, B.; Bai, S.; Wang, X.; Zhong, M.; Yang, Q.; Li, C. Angew. Chem. Int. Ed. 2012, 51, 11517–11521. 118  (17)  Omae, I. Appl. Organometal. Chem. 2007, 21, 318–344.  (18)  Frostin-rio, M.; Pujol, D.; Bied-charreton, C.; Fauvet, M. P.-F.; Gaudemer, A. J. Chem. Soc. Perkin Trans. 1984, 1971–1979.  (19)  Jones, R. D.; Summerville, D. A.; Basolo, F. Chem. Rev. 1979, 79, 139–179.  (20)  Debuigne, A.; Michaux, C.; Jérôme, C.; Jérôme, R.; Poli, R.; Detrembleur, C. Chem. Eur. J. 2008, 14, 7623–37.  (21)  Gomes, J.; De Castro, B. Inorg. Chem. 2000, 39, 1994–1997.  (22)  Smirnov, V. V.; Woller, E. K.; DiMagno, S. G. Inorg. Chem. 1998, 37, 4971–4978.  (23)  Debuigne, A.; Poli, R.; Jérôme, C.; Jérôme, R.; Detrembleur, C. Prog. Polym. Sci. 2009, 34, 211–239.  (24)  Fors, B. P.; Buchwald, S. L. J. Am. Chem. Soc. 2010, 132, 15914–15917.  (25)  Tsuji, T.; Yorimitsu, H.; Oshima, K. Angew. Chem. Int. Ed. 2002, 41, 4137–4139.  (26)  Gosmini, C.; Bégouin, J.-M.; Moncomble, A. Chem. Comm. 2008, 3221–33.  (27)  Cahiez, G.; Moyeux, A. Chem. Rev. 2010, 110, 1435–62.  (28)  Affo, W.; Ohmiya, H.; Fujioka, T.; Ikeda, Y.; Nakamura, T.; Yorimitsu, H.; Oshima, K.; Imamura, Y.; Mizuta, T.; Miyoshi, K. J. Am. Chem. Soc. 2006, 128, 8068–77.  (29)  Weiss, M. E.; Kreis, L. M.; Lauber, A.; Carreira, E. M. Angew. Chem., Int. Ed. Engl. 2011, 50, 11125–8.  (30)  Arvanitopooulos, L. D.; Greuel, M. P.; Harwood, H. J. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 1994, 35, 549–550.  (31)  Detlefsen, W. D.; Branchaud, B. P. Tetrahedron Lett. 1991, 32, 6273–6276.  (32)  Branchaud, B. P.; Yu, G.-X. Organometallics 1993, 12, 4262–4264.  (33)  Blanksby, S. J.; Ellison, G. B. Acc. Chem. Res. 2003, 36, 255–63.  (34)  Daikb, B. E.; Finke, R. G. J. Am. Chem. Soc. 1992, 114, 2938–2943.  (35)  Halpern, J. Polyhedron 1988, 7, 1483–1490. 119  (36)  Newcomb, M. Tetrahedron 1993, 49, 1151–1176.  (37)  Matyjaszewski, K.; Patten, T. E.; Xia, J. J. Am. Chem. Soc. 1997, 119, 674–680.  (38)  Gitsov, I.; Weimer, M. W.; Fr chet, J. M. J. J. Polym. Sci. A: Polym. Chem. 1998, 36, 955–970.  (39)  Dossi, M.; Storti, G.; Moscatelli, D. Macromol. Symp. 2010, 289, 119–123.  (40)  Webster, O. W. Science 1991, 251, 887.  (41)  Hurtgen, M.; Detrembleur, C.; Jerome, C.; Debuigne, A. Polym. Rev. 2011, 51, 188–213.  (42)  Le Grognec, E.; Claverie, J.; Poli, R. J. Am. Chem. Soc. 2001, 123, 9513–24.  (43)  Li, S.; De Bruin, B.; Peng, C.-H.; Fryd, M.; Wayland, B. B. J. Am. Chem. Soc. 2008, 130, 13373–13381.  (44)  Debuigne, A.; Champouret, Y.; Jérôme, R.; Poli, R.; Detrembleur, C. Chem. Eur. J. 2008, 14, 4046–59.  (45)  Debuigne, A.; Caille, J.-R.; Jérôme, R. Angew. Chem., Int. Ed. Engl. 2005, 44, 1101–1104.  (46)  Hurtgen, M.; Debuigne, A.; J r me, C.; Detrembleur, C. Macromolecules 2010, 43, 886–894.  (47)  Kaneyoshi, H.; Matyjaszewski, K. Macromolecules 2005, 38, 8163–8169.  (48)  Debuigne, A.; Caille, J.-R.; Jérôme, R. Macromolecules 2005, 38, 5452–5458.  (49)  Piette, Y.; Debuigne, A.; Jérôme, C.; Bodart, V.; Poli, R.; Detrembleur, C. J. Polym. Sci., Part A: Polym. Chem. 2012, 3, 2880–2891.  (50)  Poli, R. Eur. J. Inorg. Chem. 2011, 2011, 1513–1530.  (51)  Gridnev, A. A.; Ittel, S. D. Chem. Rev. 2001, 101, 3611–3659.  (52)  Peng, C.; Scricco, J.; Li, S.; Fryd, M.; Wayland, B. B. Macromolecules 2008, 2368–2373.  (53)  Hanzlik, R. P.; Williamson, D. J. Am. Chem. Soc. 1976, 98, 6570–6573.  120  (54)  Mlodnicka, T. J. Mol. Catal. 1986, 36, 205–242.  (55)  Schaefer, W. P.; Huie, B. T.; Kurilla, M. G.; Ealick, S. E. Inorg. Chem. 1980, 19, 340–344.  (56)  Halpern, J. Acc. Chem. Res. 1982, 15, 238–244.  (57)  Chen, D.; Martell, A. E. Inorg. Chem. 1987, 26, 1026–1030.  (58)  Tovrog, B. S.; Kitko, D. J.; Drago, R. S. J. Am. Chem. Soc. 1974, 1461, 5144–5153.  (59)  Chen, D.; Martell, A. E.; Sun, Y. Inorg. Chem. 1989, 28, 2647–2652.  (60)  Kervinen, K.; Lahtinen, P.; Repo, T.; Svahn, M.; Leskelä, M. Catal. Today 2002, 75, 183–188.  (61)  Wilmarth, W. K.; Aranoff, S.; Calvin, M. J. Am. Chem. Soc. 1946, 68, 2254–2256.  (62)  Ford, W. T.; Zhu, W. J. Mol. Catal. 1993, 78, 367–378.  (63)  Zakzeski, J.; Bruijnincx, P. C. a; Jongerius, A. L.; Weckhuysen, B. M. Chem. Rev. 2010, 110, 3552–3599.  (64)  Wu, A.; Patrick, B. O.; Chung, E.; James, B. R. Dalton Trans. 2012, 41, 11093– 106.  (65)  Bolzacchini, E.; Canevali, C.; Morazzoni, F.; Orlandi, M.; Scotti, R. J. Chem. Soc., Dalton Trans. 1997, 4695–4699.  (66)  Hanson, S. K.; Wu, R.; Silks, L. A. “Pete” Angew. Chem., Int. Ed. Engl. 2012, 51, 3410–3413.  (67)  Kervinen, K.; Korpi, H.; Gerbrand Mesu, J.; Soulimani, F.; Repo, T.; Rieger, B.; Leskelä, M.; Weckhuysen, B. M. European Journal of Inorganic Chemistry 2005, 2591–2599.  (68)  Canevali, C.; Orlandi, M.; Pardi, L.; Rindone, B.; Scotti, R.; Sipila, J.; Morazzoni, F. J. Chem. Soc., Dalton Trans. 2002, 3007–3014.  (69)  Zombeck, A.; Drago, R. S.; Corden, B. B.; Gaul, J. H. J. Am. Chem. Soc. 1981, 103, 7580–7585.  (70)  Chavez, F. A.; Mascharak, P. K. Acc. Chem. Res. 2000, 33, 539–545. 121  (71)  Gurley, L.; Beloukhina, N.; Boudreau, K.; Klegeris, A.; McNeil, W. S. J. Inorg. Biochem. 2011, 105, 858–866.  (72)  Costa, G.; Mestroni, G.; Tauzher, G.; Stefani, L. J. Organomet. Chem. 1966, 6, 181–187.  (73)  Tellmann, K. P.; Humphries, M. J.; Rzepa, H. S.; Gibson, V. C. Organometallics 2004, 23, 5503–5513.  (74)  Wence, D. L. K. Synthesis and Characterization of Square Pyramidal Low Spin Co(III) Ketoiminato Alkyl Complexes. B.Sc. Thesis, University of British Columbia Okanagan, 2009.  (75)  Summers, M. F.; Marzilli, L. G.; Bresciani-pahor, N.; Randaccio, L. J. Am. Chem. Soc. 1984, 106, 4478–4485.  (76)  Schrauzer, G. N.; Holland, R. J. J. Am. Chem. Soc. 1979, 93, 1505–1506.  (77)  Schrauzer, G. N.; Silbert, J. W.; Windgassen, R. J. J. Am. Chem. Soc. 1968, 90, 6681–6688.  (78)  Ollivier, C.; Renaud, P. Chem. Rev. 2001, 101, 3415–34.  (79)  Sherwood, R. K.; Kent, C. L.; Patrick, B. O.; McNeil, W. S. Chem. Commun. 2010, 46, 2456–2458.  (80)  Ryznar, J. W.; Grange, L.; Premo, J. G. Preparation of Organoboranes 1963, 3100181.  (81)  Costa, G.; Mestroni, G.; Puxeddu, A.; Reisenhofer, E. J. Chem. Soc. (A) 1970, 2870–2784.  (82)  Amélie, K.; Kanso, H.; Baptiste, B.; Arora, H.; Philouze, C.; Olivier, J.; Herve, V.; Luneau, D.; Orio, M.; Thomas, F. Inorg. Chem. 2012, 51, 10557–10571.  (83)  Brockway, D. J.; West, B. O.; Bergen, V. den J. Organomet. Chem. 1983, 249, 205–225.  (84)  Hemphill, W. D.; Brown, D. G. Inorg. Chem. 1977, 16, 766–769.  (85)  Zhou, W.; Chiang, L.; Patrick, B. O.; Storr, T.; Smith, K. M. Dalton Trans. 2012, 41, 7920–7930.  122  (86)  Marinescu, S. C.; Winkler, J. R.; Gray, H. B. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 15127–15131.  (87)  Ungávry, F.; Markó, L. Organometallics 1982, 1, 1120–1125.  (88)  Bergen, A. Van den; Murray, K. S.; West, B. O. Journ. Organometal. Chem. 1971, 33, 89–96.  (89)  Hu, X.; Brunschwig, B. S.; Peters, J. C. J. Am. Chem. Soc. 2007, 129, 8988–98.  (90)  Valdez, C. N.; Dempsey, J. L.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 15589–15593.  (91)  Jacques, P.-A.; Artero, V.; Pécaut, J.; Fontecave, M. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 20627–20632.  (92)  McCormick, T. M.; Han, Z.; Weinberg, D. J.; Brennessel, W. W.; Holland, P. L.; Eisenberg, R. Inorg. Chem. 2011, 50, 10660–6.  (93)  Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; 2nd ed.; John Wiley & Sons, Inc., New York, 2001.  (94)  Lelj, F.; Morelli, G.; Ricciardi, G.; Rosa, A. Inorg. Chim. Acta 1990, 176, 189–194.  (95)  Coveney, D. J.; Patel, V. F.; Pattenden, G. Tetrahedron Lett. 1987, 28, 5949–5952.  (96)  Ghosez, A.; Göbel, T.; Giese, B. Chem. Ber. 1988, 121, 1807–1811.  (97)  Sigman, M. E.; Barbas, J. T.; Corbett, S.; Chen, Y.; Ivanov, I.; Dabestani, R. J. Photochem. Photobiol., A 2001, 138, 269–274.  (98)  Abel, E. W.; Pratee, J. M.; Whelan, R. Inorg. Nucl. Chem. Letters 1971, 7, 901– 904.  (99)  Hoffman, B. M.; Diemente, D. L.; Basolo, F. J. Am. Chem. Soc. 1969, 92, 61–65.  (100) Rodley, G. A.; Robinson, W. T. Nature (London) 1972, 438. (101) Kurahashi, T.; Fujii, H. Inorg. Chem. 2013, 52, 3908–3919. (102) Zakzeski, J.; Jongerius, A. L.; Weckhuysen, B. M. Green Chem. 2010, 12, 1225. (103) Kervinen, K.; Korpi, H.; Leskelä, M.; Repo, T. J. Mol. Catal. A: Chem. 2003, 203, 9–19. 123  (104) Sippola, V. O.; Krause, A. O. I. J. Mol. Catal. A: Chem. 2003, 194, 89–97. (105) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877–910. (106) Ludwick, L. M.; Brown, T. L. J. Am. Chem. Soc. 1969, 91, 5188–5189. (107) Randolph, J. G.; Brown, T. L. Inorg. Chem. 1973, 12, 2815–2819. (108) Bronowska, A. K. In Thermodynamics - Interaction Studies - Solids, Liquids and Gases; Piraján, D. J. C. M., Ed.; InTech, 2011; pp. 1–49. (109) Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1992, 114, 4992–4996. (110) Bergen, A. van den; West, B. O. Chem. Commun. 1971, 1971, 52–53. (111) Van Den Bergen, A.; West, B. O. J. Organomet. Chem. 1974, 64, 125–134. (112) Gutmann, V. Coord. Chem. Rev. 1976, 18, 225. (113) Moad, G.; Solomon, D. H. The Chemistry of Free radical polymerization; Pergamon: Oxford, 1995. (114) Black, J. O. B.Sc. Thesis, University of British Columbia Okanagan, 2012. (115) Wayland, B. B.; Gridnev, A. A.; Ittel, S. D.; Fryd, M. Inorg.Chem 1994, 33, 3830– 3833.  124  Appendices  ε (x103Lmol-1cm-1)  Appendix A: UV vis of Co(II) in various solvents  Co(II)(Me/Ph)  24 22 20 18 16 14 12 10 8 6 4 2 0  THF Toluene  290  340  390  440 490 Wavelength (nm)  540  590  640  Figure A.46: UV-vis of CoII(Me/Ph) in THF and toluene; spectra are nearly identical.  Co(II)(CF3/Ph)  14  ε (x103Lmol-1cm-1)  12  THF  10  Ether  8  Hexane  6 4 2 0 290  340  390  440 490 Wavelength (nm)  540  590  640  Figure A.47: UV-vis of CoII(CF3/Ph) in THF, diethyl ether, and hexanes.  125  16  Co(II)(Me/CF3)  14 ε (x103Lmol-1cm-1)  12  THF  10  Ether Toluene  8 6 4 2 0 290  340  390  440 490 Wavelength (nm)  540  590  640  590  640  Figure A.48: UV-vis of CoII(Me/CF3) in THF, diethyl ether, and toluene.  14  Co(II)(CF3)  ε (x103Lmol-1cm-1)  12 THF  10  Hexane 8 6 4 2 0 290  340  390  440 490 Wavelength (nm)  540  Figure A.49: UV-vis of CoII(CF3) in THF and hexanes.  126  Appendix B: X-ray crystallographic data Table B.16: X-ray crystallographic data Complex Formula Mw Colour, Habit Crystal System Space Group Cryst. dimen. (mm) Lattice Type Lattice Parameters  CoIII(Me/Ph)Et C24H27CoN2O2 434.41 brown, rod monoclinic P21/n (#14) 0.12 x 0.14 x 0.60 primitive a = 7.5525(2) Å b = 17.4630(6) Å c = 15.6446(5) Å α = 90 ° β = 96.808(2)° γ = 90° V = 2048.8(1) Å3  CoIII(CF3/Ph)Et  CoIII(Me/CF3)Et  THF-CoIII(CF3)Et  C24H21O2N2F6Co C14H17O2N2 F6Co 542.36 418.23 black, irregular black, prism triclinic monoclinic P-1 (#2) P21/c (#14) 0.14 x 0.22 x 0.28 0.10 x 0.21 x 0.23 primitive primitive primitive  C18H19N2O3F12Co 598.28 red, needle monoclinic P21/c (#14) 0.03 x 0.11 x 0.53 primitive  a = 9.8704(8) Å b = 10.753(1) Å c = 12.508(1) Å α = 113.707(4)° β = 105.035(4)° γ = 90.398(4)° V = 1164.6(2) Å3  a = 14.5355(9) Å b = 15.885(1) Å c = 14.071(1) Å α = 90° β = 92.560(3)° γ = 90° V = 3245.8(4) Å3  a = 17.648(3) Å b = 17.885(3) Å c = 14.794(2) Å α = 90° β = 103.562(6)° γ = 90° V = 4539(1) Å3  CoII(CF3) C12H8N2O2F12Co 497.12 red, irregular monoclinic C2/c (#2) 0.18 x 0.20 x 0.30 C-centered a = 23.201(2) Å b = 11.5412(9) Å c = 15.066(2) Å α = 90° β = 129.369(1)° γ = 90° V = 14735.4(13) Å3  Z Dcalc M(MoKα)  4 3 1.408 g/cm 8.60 cm-1  2 3 1.547 g/cm 8.09 cm-1  8 3 1.712 g/cm 11.32 cm-1  8 3 1.751 g/cm 8.78 cm-1  8 3 2.117 g/cm 12.50 cm-1  2Θmax Tot. # Reflections Unique Reflc, Rint Observed data (I>2ζ(I)) R1, wR2 (F2 all data) R1, wR2 (F2,I>2ζ(I)) Goodness to Fit  61.1° 31367 6288; 0.061 4667 0.125; 0.139 0.087; 0.127 1.18  52.2° 43467 4591, 0.073 3221 0.139; 0.293 0.099; 0.245 1.08  60.1° 56318 9521, 0.048 7509 0.047; 0.075 0.032; 0.069 1.02  50.5° 91359 8314, 0.073  60.1° 25379 4574, 0.021 4152 0.026; 0.054 0.022; 0.052 1.05  0.082; 0.134 0.047; 0.114 1.01  127  Appendix C: NMR spectra  Figure C.50: 1H NMR of CoIII(Me/Ph)Et in acetone-d6.  128  Figure C.51: 1H NMR of CoIII(CF3/Ph)Et in acetone-d6.  129  Figure C.52: 1H NMR of CoIII(Me/CF3)Et in acetone-d6.  130  Figure C.53:1H NMR of CoIII(salphen)Et in acetone-d6.  131  Figure C.54: 1H NMR of (THF)CoIII(CF3)Et in acetone-d6. Using integration from THF: [THF/Co] = [0.45:1] ≈ [0.5:1].  132  Figure C.55: 1H NMR of (THF)CoIII(CF3)Et in dmso-d6.  133  Figure C.56:1H NMR of CoIII(CF3)Et in chloroform-d.  134  Figure C.57: 1H NMR with presence of CoIII(Me/CF3)Et in acetone-d6 after reaction with I2 and MgBrCH2CH3.  135  Figure C.58: 19F NMR for a mixture of CoIII(Me/CF3)Et and CoII(CF3/Ph) in acetone-d6, 50 °C.  136  Figure C.59: 19F NMR for a mixture of CoIII(Me/CF3)Et and CoII(CF3/Ph) in acetone-d6, 50 °C.  137  

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.24.1-0071923/manifest

Comment

Related Items