Despite previous extensive study of the widely-employed ruthenium-catalysed olefin metathesis reaction, the finer mechanistic details have not been elucidated. An area that is noticeably lacking is spectroscopic exploration of the relevant complexes. In this work, organometallic ruthenium complexes of importance to olefin metathesis have been investigated using Ru K-edge X-ray absorption spectroscopy. The lowest energy feature in the Ru K-edge spectrum has been unambiguously assigned as due to Ru 4d←1s transitions. These electric-dipole-forbidden transitions are extremely sensitive to geometry. For centrosymmetric complexes, the pre-edge feature has very low intensity because it is limited by the weak electric quadrupole mechanism. By contrast, non-centrosymmetric complexes exhibit a substantial increase in pre-edge intensity because Ru 5p-4d mixing introduces electric-dipole-allowed character to the Ru 4d←1s transitions. The energy of the edge feature in the Ru K-edge spectrum corresponds to ionisation of 1s electrons and is a good indicator of the charge on the metal centre. Unexpectedly, we found that the first-generation (L = PCy₃) Grubbs precatalyst (1) has a higher 1s ionisation energy than the second-generation (L = H₂IMes) complex (2). This effect provides a compelling rationale for the unexplained differences in phosphine dissociation kinetics for complexes 1 and 2: the phosphine dissociation rate of 2 is slower than 1 because the metal centre is more electron-deficient in 2. Density functional theory calculations confirm the charge differences and offer some insight into the nature of bonding in these complexes, particularly with regard to the N-heterocyclic carbene and trialkylphosphine ligands. On the basis of these results, we propose that, for this system, the NHC ligand is a weaker σ-charge donor than the phosphine ligand, and that the NHC accepts significant π-electron density from the metal; both interactions function to reduce the electron density on the ruthenium centre. An ultimate goal is to investigate reactive species in the olefin metathesis mechanism; accordingly, we have made considerable progress toward collecting XAS data for a metallacyclobutane species, and we are pursuing methods to trap the four-coordinate intermediate in the metathesis cycle.

Cyclic peroxides are abundant in Nature and synthetic manipulation of the peroxide linkage and alkene portion of 1,2-dioxines has established 1,2-dioxines as important compounds as both chemical building blocks and bioactive compounds. Much of the chemistry performed thus far utilizing 1,2-dioxines involves the initial rearrangement of 1,2-dioxines to γ-hydroxyenones allowing for the generation of structural motifs such as cyclopropanes, THF’s, THP’s, 1,4-diketones and natural sugars. Herein describes the synthesis of novel 1,2-dioxines with a variety of tethered functionalities and their transformations to afford novel cyclic compounds whilst maintaining the peroxide linkage intact. Chapter two outlines the intramolecular cyclisations of tethered hydroxyl and carboxylic acid moieties onto the olefin of 1,2- dioxines to generate both tetrahydrofurans and dihydrofuran-2(3H)-ones, whilst maintaining the peroxide linkage. This work presents the first examples of syn fused cyclic peroxide furans through intramolecular cyclisation of tethered hydroxyl groups. Improved methods for the oxidation of hydroxyl tethered 1,2-dioxines to carboxylic acid moieties are also reported. In addition, improved methods for intramolecular cyclisation of carboxylic acid moieties where developed to afford syn fused cyclic peroxide lactones. Furthermore, reduction of the peroxide bond enabled generation of functionalized tetrahydrofurans and dihydrofuran-2(3H)-ones which have previously been utilized as synthetic building blocks for several natural products. Chapter three reports the first examples of carbenoid insertion into the peroxide linkage of 1,2-dioxines allowing for the generation of novel bicyclic hemiacetals. Alternatively novel tricyclic cyclopropyl peroxides where generated through insertion into the olefin whilst maintaining the peroxide linage intact. Additionally, the attempted intramolecular cyclisation of diazoketone tethered bicyclic 1,2-dioxines was also probed. Furthermore, the attempted intermolecular insertion of diazoketones onto 1,2-dioxines are presented within this chapter. Finally, Chapter four outlines the intramolecular cyclisations of bromo-alkyl tethered 1,2-dioxines to furnish novel cyclic cyclopentyl peroxides whilst maintaining the peroxide linkage intact. The work presented in this chapter represents the first examples of the synthesis of syn fused cyclic cyclopentyl peroxides. In summary, this thesis outlines methodology towards the synthesis of novel cyclic peroxides from 1,2-dioxines containing tethered functional groups. / Thesis (Ph.D.) -- University of Adelaide, School of Chemistry and Physics, 2009

Zugl.: Marburg, Univ., Diss., 2008

Zugl.: München, Techn. Univ., Diss., 2008

Techn. Hochsch., Diss., 2001--Aachen.

Universiẗat, Diss., 2002--Tübingen.

Fluorocarbons are versatile molecules that are used in multiple industries ranging from pharmaceuticals to refrigerants, insecticides and advanced materials. More particularly hydrofluorocarbons (HFCs) and hydrofluoroolefins (HFOs) are current replacements for ozone-depleting chlorofluorocarbons (CFCs) that were used for decades as refrigerants, propellants, solvents and blowing agents. However, syntheses of HFCs and HFOs involve energy-intensive processes and toxic compounds such as heavy metals and anhydrous HF. Development of more sustainable, energy efficient and "greener" synthesis of small fluorocarbons is needed, which draws attention to organometallic catalysis, especially with abundant, inexpensive and non-toxic transition metals. One approach to new organometallic routes to hydrofluorocarbons involves the formation and functionalization of fluorometallacycles. Previous work in the 1990’s by Baker et al. demonstrated the catalytic hydrodimerization of tetrafluoroethylene (TFE) using Ni catalysts with π-acidic phosphite ligands. They also demonstrated the hydrogenolysis of the d6 ferracyclopentane, Fe(CO)4(1,4-C4F8), 2-1, under high pressure and temperature with different additives to give mixtures of different hydrofluorocarbons. Since that time the reactivity of d8 fluorometallacyles has been extensively studied, leading to fundamental understanding and new catalytic applications. However less attention has been paid to d6 systems, the synthesis and reactivity of which are the focus of this Thesis. Following introduction and background in Chapter 1, Chapter 2 presents the synthesis and characterization of a series of new NHC-, phosphine- and nitrogen-ligand-substituted Fe(II) perfluorometallacycles derived from complex 2-1. This led to the discovery of the first example of a fluorinated metallacyclocarbene obtained from in situ Cα–F bond activation that afforded FeF(triphos)(1,4-C4F7), 2-6, (triphos = bis(2-diphenylphosphinoethyl)phenylphosphine) during the P-based linear tridentate ligand substitution reaction. [Fe(triphos)(1,4-C4F7)(NCMe)]+BPh4-, 2-7, and Fe(OTf)(triphos)(1,4-C4F7), 2-8, were derived from 2-6 by treatment with NaBPh4 in acetonitrile and Me3Si-OTf, respectively (Tf = triflate, SO2CF3). The same phenomenon was not observed with hard-donor N-based linear tridentate ligand, terpy’, (terpy’ = 4′-(4-methylphenyl)-2,2′:6′,2′′-terpyridine), presumably because of the less Lewis acidic metal center. Fluoride abstraction from Fe(terpy’)(CO)(1,4-C4F8), 2-9, by a Lewis acid, however allowed for Cα–F bond activation to give the cationic iron monocarbonyl carbene complex, [Fe(terpy’)(CO)(1,4-C4F7)]+OTf–, 2-10. Chapter 3 investigates further the reactivity of these new Fe(II) perfluorometallacycle complexes. The lack of reactivity of the mono- and di-substituted Fe carbonyl perfluorometallacycles with Lewis acids confirmed that Cα–F bond activation only occurs when there is enough π-backbonding into the Cα–F anti-bonding orbital, as π-acceptor phosphines and carbonyl ligands can compete for the metal back-bonding. Indeed, Cα–F abstraction is only observed with Fe(terpy’)(CO)(1,4-C4F8), 2-9, due to the poor acceptor ability of the nitrogen ligand. On the other hand, the lack of electron density on the metal center can cause the Fe center to act as an internal Lewis acid, promoting Cα–F migration as observed in situ during the triphos substitution reaction. These results show that d6 [Fe] perfluorometallacycles do not share similar reactivity with d8 [Ni] perfluorometallacycles. Moreover, the study of the character of the Fe=CF bonds suggests a nucleophilic carbene for 2-6, while 2-7, 2-8 and 2-10 all displayed electrophilic carbene character. Furthermore, hydrogenolysis of Fe(OTf)(triphos)(1,4-C4F7), 2-8, and [Fe(triphos)(1,4-C4F7)(NCMe)]+BPh4-, 2-7, at low pressure and room temperature, generated exclusively H(CF2)3CFH2, HFC-347pcc, and iron hydrides, confirming a previous hypothesis that attributed formation of this hydrofluoroalkane to an Fe carbene intermediate. In contrast, [Fe(terpy’)(CO)(1,4-C4F7)]+OTf–, 2-10, reacts with H2 to yield HF and an unidentified iron complex, showing that the nature of the ancillary ligands greatly influences the reactivity. Chapter 4 explores the reactivity of phosphine-substituted cobalt(I) carbonyl hydride complexes towards TFE to expand our work on d6 perfluorometallacycles. The most electron-rich ligands prevented metallacycle formation or slowed it down possibly due to strong π-backbonding into the CO ligands, making it harder to generate an open coordination site. Indeed, a mixture of the Co-tetrafluoroethyl complex, derived from insertion of TFE into Co–H, and the zerovalent dimer/hydrogenated TFE products, derived from the reaction of the Co–H with the 16e- CoLn(CO)3-n(CF2CF2H) intermediate, were obtained with the bulkiest ligands, CoH(dcppe)(CO)2 and CoH(Pcp3)(CO)3 (dcppe = 1,2-bis(dicyclopentylphosphino)ethane, cp = cyclopentyl). With the slightly less bulky PiBu3 ligand, further reactivity of the insertion product with TFE slowly formed a d6 metallacycle hydride complex. In contrast, with the dppe and tripod cobalt carbonyl hydrides, metallacycle product formation was evident even at short reaction times with insertion/hydrogenation ratios of 1:1, showing that using less electron-rich, steric bulky ligands prevented the bimolecular Co dimer formation, but left enough room for binding a second equivalent of TFE for metallacycle formation. Finally, Chapter 5 summarizes the findings of this Thesis and discusses future directions based on this work.

Despite previous extensive study of the widely-employed ruthenium-catalysed olefin metathesis reaction, the finer mechanistic details have not been elucidated. An area that is noticeably lacking is spectroscopic exploration of the relevant complexes. In this work, organometallic ruthenium complexes of importance to olefin metathesis have been investigated using Ru K-edge X-ray absorption spectroscopy. The lowest energy feature in the Ru K-edge spectrum has been unambiguously assigned as due to Ru 4d←1s transitions. These electric-dipole-forbidden transitions are extremely sensitive to geometry. For centrosymmetric complexes, the pre-edge feature has very low intensity because it is limited by the weak electric quadrupole mechanism. By contrast, non-centrosymmetric complexes exhibit a substantial increase in pre-edge intensity because Ru 5p-4d mixing introduces electric-dipole-allowed character to the Ru 4d←1s transitions. The energy of the edge feature in the Ru K-edge spectrum corresponds to ionisation of 1s electrons and is a good indicator of the charge on the metal centre. Unexpectedly, we found that the first-generation (L = PCy₃) Grubbs precatalyst (1) has a higher 1s ionisation energy than the second-generation (L = H₂IMes) complex (2). This effect provides a compelling rationale for the unexplained differences in phosphine dissociation kinetics for complexes 1 and 2: the phosphine dissociation rate of 2 is slower than 1 because the metal centre is more electron-deficient in 2. Density functional theory calculations confirm the charge differences and offer some insight into the nature of bonding in these complexes, particularly with regard to the N-heterocyclic carbene and trialkylphosphine ligands. On the basis of these results, we propose that, for this system, the NHC ligand is a weaker σ-charge donor than the phosphine ligand, and that the NHC accepts significant π-electron density from the metal; both interactions function to reduce the electron density on the ruthenium centre. An ultimate goal is to investigate reactive species in the olefin metathesis mechanism; accordingly, we have made considerable progress toward collecting XAS data for a metallacyclobutane species, and we are pursuing methods to trap the four-coordinate intermediate in the metathesis cycle. / Science, Faculty of / Chemistry, Department of / Graduate

Ruthenium - catalyzed olefin metathesis is now an invaluable tool in organic synthesis. However, routes to the dominant metathesis catalysts, the second - generation Grubbs and Hoveyda catalysts (RuCl 2 (PCy 3 )(H 2 IMes)(=CHPh) and RuCl 2 (H 2 IMes)[= CH( o - O i Pr)C 6 H 4 ], respectively) are plagued with problems. The common reliance on in situ methods to generate the N - heterocyclic carbene H 2 IMes severely limits stoichiometric control, and results in contamination by byproducts, some of which are readily overlooked, and some of which are difficult to remove. Both can affect batch - to - batch reproducibility in catalysis. This thesis work demonstrated that widespread perceptions of the instability of free H 2 IMes are erroneous, and that the free carbene is readily handled under water - free conditions. Clean, convenient, near - quantitative routes were developed to these second - generation catalysts by ligand exchange of their first - gen eration counterparts RuCl 2 (PCy 3 ) 2 (=CHPh), RuCl 2 (PCy 3 )[= CH( o - O i Pr)C 6 H 4 ] with free H 2 IMes, with sequestration of the liberated phosphine by an ion - exchange resin. A second focus was examination of a much - debated hypothesis in olefin metathesis: that is, the extent to which the high productivity of the Hoveyda catalysts reflects re - uptake of the styrenyl ether functionality released in the initial cycle of metathesis. Current evidence for and against this "boomerang" hypothesis is critically examined, and new approaches to examining its operation are described. Specifically, the rate of decomposition, vs. re - uptake, is examined for the active species RuCl 2 (PCy 3 )(=CH 2 ), and background exchange of the parent catalyst with free styrenyl ether is measured by use of a 13 C - labelled styrenyl ether. These studies confirm the relevance of the boomerang mechanism for first - generation Hoveyda catalysts.

For the preparation of semiconducting polymers often ‘standard’ catalytic systems are used without further optimisation. New ligands, such as N-heterocyclic carbenes have shown excellent activity in cross-coupling reactions (e.g. Suzuki-Miyaura reaction, or Hartwig-Buchwald amination). These systems show excellent conversions under mild conditions and even allow the use of aryl chlorides as reagents. Nevertheless, previously no system has been reported for the synthesis of conjugated polymers, e.g. the Suzuki polycondensation or Buchwald-Hartwig type polycondensation using these catalysts. A NHC-Pd based catalytic system was optimised for a polyamination reaction and the catalyst [(IPr)Pd(allyl)Cl] was found to be the most active. Polytriarylamines were synthesised using the optimised catalytic system and tested in organic field-effect transistors. Mobilities found were low which was found to be attributed to the presence of high molecular weight fractions. Molecular weights were controlled using an in situ end-capping approach and polymers tested in the semiconducting layer of OFETs gave similar mobilities tothose reported earlier. Several polytriarylamines, which have not been reported previously, were synthesised using NHC-chemistry and the in situ end-cappingapproach, as well as polytriarylamines that have been reported previously using Pd/phosphine catalysts. One library containing polymers based on biphenyles andbridged biphenyles and another library containing polymers with bridged oligoarenes were synthesised. Suzuki polycondensation was also studied besides the polyamination protocol and low catalyst loadings and reaction temperatures could be realised using a NHC-Pd catalyst. Sulfur containing monomers that could not be polymerised using the polyamination were polymerised successfully. All polymers were fully characterised and studied as the active layer in organic field-effect transistors. The highest mobilities determined for these polymers (~10-2 cm2/Vs) is close to the highest reported for this class of polymer reported to date.