La caractérisation complète d’intermédiaires réactionnels intervenants dans des procédés de catalyse homogène est une tâche ardue en raison de leur réactivité et de leur faible concentration. Ceci est particulièrement vrai pour les espèces radicalaires telles que les complexes organométalliques réduits, qui sont des intermédiaires en photocatalyse ou lorsque ces complexes possèdent des ligands non-innocents. Par conséquent, leur structure électronique est encore mal comprise, sachant que l'électron ajouté peut être situé sur différents sites de la molécule.Dans ce contexte, nous avons développé une méthode d'analyse pour étudier en phase gazeuse des complexes organométalliques radicalaires. Des complexes organométalliques multichargés du zinc et du ruthénium avec des ligands bidentes de type bipyridine ou tridente de type bis(imino)pyridine ont d’abord été obtenus et isolés en phase gazeuse. Ils sont ensuite réduits avec les méthodes d’activation par un électron spécifiques à la spectrométrie de masse, la dissociation par capture ou transfert d’électron (ECD/ETD), permettant de former des espèces métalliques radicalaires monochargées. Celles-ci sont enfin isolés et leur spectre infrarouge est obtenu à l’aide de la spectroscopie d’action basée sur la dissociation induite par l’absorption de plusieurs photons dans l’infrarouge (IRMPD). Les méthodes DFT fournissent un complément pour modéliser la structure électronique et le spectre IR de ces espèces.Les challenges à relever pour développer ce nouvel outil d'analyse étaient de deux ordres. Tout d'abord, nous devions être en mesure d'obtenir les complexes souhaités en phase gazeuse. Ceci nous a conduit à examiner de multiples paramètres, tels que la nature des ligands ou l’énergie interne déposée lors de l’étape de réduction. Le deuxième défi portait sur l'utilisation des méthodes de modélisation. Nous avons montré l’absence de fiabilité des méthodes standards de modélisation pour décrire à la fois la structure électronique et le spectre infrarouge des complexes réduits. Les données expérimentales obtenues durant ce travail ont donc été utilisées comme références pour identifier les fonctionnelles DFT les plus appropriées pour l’étude de ces complexes radicalaires. / The complete characterization of reaction intermediates in homogeneous catalytic processes is often a difficult task owing to their reactivity and low concentration. This is particularly true for radical species such as reduced organometallic complexes, which are intermediates in photocatalysis, or when these complexes included non-innocent ligands. Consequently, their electronic structure in the ground state is still poorly understood, knowing that the added electron can be located on different sites of the molecule.In this contect, we developed an analytical method to study radical organometallic complexes in the gas phase. We started with formation of suitable multi-charged zinc organometallic complexes in the gas phase from mixture of zinc metal cation and bipyridine-type bidentate or bis(imino)pyridine tridentate ligands. Multicharged ruthenium complexes with similar ligands have also been studied. Under ideal circumstances these complexes were isolated and reduced in the gas phase to form monocationic metal species. Electron activated methods such as electron capture dissociation (ECD) and electron transferred dissociation (ETD) techniques, available in FT-ICR mass spectrometers, have been used to that end. The resulting Zn and Ru radical cation complexes are then isolated in the gas phase and probed via infrared multi photon dissociation (IRMPD) action spectroscopy. In support, DFT theoretical calculations were performed to model their electronic structure and IR spectra.Two main issues were faced during the development of this new analytical tool. First, we had to be able to obtain the desired complexes in the gas phase. This has lead to monitor various parameters, such as the nature of the ligands or the internal energy provided by the reduction step. The second challenge dealt with the use of modeling methods. We have shown that standard modelling tools lack the accuracy to predict both electronic structure and spectral signatures of reduced complexes. The experimental data gathered in this work have therefore been used as benchmarks for the identification of DFT functionals that are most appropriate for the study of these radical complexes.

Les adduits cuivre-oxygène dans les métallo-enzymes ont été proposés comme étant responsables de l'activation de liaisons C-H, processus qui ont un intérêt pour des applications industrielles potentielles. La première partie de ce travail est consacrée à une présentation de différentes mono-oxygénases à cuivre et de leurs complexes modèles. Récemment, des intermédiaires réactionnels ont émergé et parmi ceux-ci, des espèces de valence mixte CuIICuIII ont été proposées comme étant des espèces réactives clés pour l'activation de liaisons C-H fortes.Dans ce travail, à partir de ligands binucléants basés sur un espaceur 1,8-naphtyridine, la stabilisation et les caractérisations spectroscopiques de ce type d’intermédiaires à haut degré d’oxydation sont explorées. La préparation d’espèces Cu2:O2 à partir de l'activation du dioxygène par les complexes CuI2 est discutée. Deux complexes µ-ɳ2:ɳ2-peroxo-CuII2 ont été préparés à -80°C et caractérisés par différentes méthodes spectroscopiques associées à des calculs par la théorie de la fonctionnelle de la densité (DFT). A partir de nouveaux ligands dissymétriques possédant une fonction amide, nos tentatives pour contrôler la préparation des complexes binucléaires associés sont également présentées. Puis, les caractérisations des espèces à valence mixte CuIICuIII obtenues par mono-oxydation électronique des complexes CuII2 sont décrites (voltammétrie cyclique, résonance paramagnétique électronique, UV-visible, proche infrarouge et DFT).Enfin, ce travail est complété par l’étude de la réactivité des espèces CuIICuIII, pour lesquelles la littérature est presque inexistante. Lorsque des ligands stériquement encombrés sont utilisés dans les espèces à valence mixte, des oxydations intramoléculaires sont observées, alors que l’espèce CuIICuIII possédant un ligand moins encombré oxyde le toluène. Il est à noter que l'ajout d'une base rend le système catalytique. / Copper-oxygen adducts in enzymes have been proposed to be responsible for the activation of C-H bonds, a process that has industrial applications. The first part of this thesis is therefore dedicated to a discussion on various copper oxygenases and their model complexes. Recently, key reactive intermediates have emerged and among them mixed valent CuIICuIII species have been proposed to be responsible for strong C-H bond activation.In this work the stabilisation and spectroscopic characterisation of high valent intermediates using dinucleating ligands based on a 1,8-naphthyridine spacer are explored. The generation of Cu2:O2 species from the activation of O2 by CuI2 complexes is discussed. Two µ-ɳ2:ɳ2-peroxo-CuII2 complexes have been prepared at -80°C and characterised by spectroscopy and density functional theory (DFT). Our attempts at generating dinuclear systems using new dissymmetric ligands with an amide function are also discussed. Finally the successful characterisation of mixed valent CuIICuIII species by mono-electronic oxidation of CuII2 complexes is described (cyclic voltammetry, electron paramagnetic resonance, UV-visible, near infrared and DFT).The last part focusses on probing the reactivity of CuIICuIII species, for which the literature is almost inexistent. When sterically congested ligands are used to support the mixed valent system, intramolecular aliphatic C-H oxidation was observed, whether as the CuIICuIII species supported by a less bulky ligand was able to oxidise toluene. Interestingly the addition of a base made the system catalytic.

Doctor of Philosophy / Department of Chemistry / Stefan Bossmann / My thesis is divided into two parts. The first part is focused on studies of N-heterocyclic carbene (NHC) palladium(IV) intermediates, which are involved in oxidative addition mediated C-C, and C-O bond formation processes as well as in C-Cl bond forming reactions via a reductive elimination process. Bis-NHC-Pd(II) complexes have been reported as effective catalysts to mediate direct conversion of methane into methanol. However, a H-D exchange study revealed that the bis-NHC-Pd(II) complexes are not the active species responsible for the C-H bond activation reaction. This unexpected result implies that the high oxidation state bis- NHC-Pd(IV) species may be the real catalyst! The oxidative addition of methyl iodide to the bis- NHC-Pd(II)-Me2 complex led to the successful observation of the formation of a transient trimethyl bis-NHC-Pd(IV) intermediate by both 1H-NMR and 13C-NMR spectroscopy. Different oxidants such as O2, PhI(OAc)2, PhI(OTFA)2 and Cl2 reacted with the bis-NHC-Pd(II)-Me2 complex, and competitive C-C and C-O bond formations, as well as C-C and C-Cl bond formations were observed. Dioxygen triggered C-C bond formation under dry condition and both C-C and C-O bond formation in the presence of H2O gave strong indications that the bis-NHCPd( II)-Me2 complex can be oxidized to a bis-NHC-Pd(IV) intermediate by dioxygen. The reaction between the hypervalent iodine regents PhI(OAc)2 and PhI(OTFA)2 and the bis-NHCPd( II)-Me2 complex gave only reductive elimination products. Therefore, this system can act as a model system, which is able to providing valuable information of the product forming (functionalization) step of the C-H bond activation system. The reaction between chlorine and the bis-NHC-Pd(II)-Me2 complex resulted in a relatively stable bis-NHC-Pd(IV)-Cl4 complex, which was characterized by 1H-NMR spectroscopy and mass spectroscopy. The structure of bis- NHC-Pd(IV)-Cl4 was unambiguously established by X-ray crystallography. The second part of this thesis describes the synthesis of functionalized bimagnetic core/shell iron/iron oxide nanoparticles for the treatment of cancer. Biocompatible dopamineoligoethylene glycol functionalized bimagnetic core/shell Fe/Fe3O4 nanoparticles were prepared via ligand exchange, and purified by repeated dispersion/magneto-precipitation cycles. A porphyrin (TCPP) has been tethered to the stealth nanoparticles to enhance their uptake by tumor cells and (neural) stem cells. The stealth nanoparticles have been delivered in a mouse model to tumor sites intravenously by using the EPR (enhanced permeation and retention) effect. Magnetic hyperthermia proved to be very effective against B16-F10 mouse melanomas in Charles River black mice. After hyperthermia, the nanoparticles have shown a significant effect on the growth of tumor (up to 78% growth inhibition).

In this study, an attempt was made to identify the electronic and steric properties of the precatalyst ligands that determine the characteristics of phosphine ligated Grubbs-type precatalysts for alkene metathesis by means of molecular modelling. It was found from studying the literature that the possibilities for synthesising a wide range of phosphine ligands are almost unlimited. Additionally, it was found that there is no easy method to determine the electronic and steric properties of the precatalyst ligands in existence. Molecular modelling might provide a method to study potential ligands and precatalysts before tedious synthesis methods are attempted. It was found that the theoretically calculated structures of the commercially available precatalysts compared well with the experimental data reported in literature. It is also shown that the energy profiles for alkene metathesis of simplified model systems do not compare well with non-simplified systems. Correlations between these simplified model systems and experimental work have to be regarded as serendipitous at best. When the energy profiles of the various new and commercially available precatalysts are compared, similarities in the energy trends for 1-octene metathesis are observed. These similarities raise questions about the significance of the differences in the energy barriers. In an effort to better understand this, two low activity precatalysts were also investigated in an attempt to identify the area or trend of poor catalysis. Instead of providing the desired different result, trends very similar to that of the highly active precatalysts were observed. This led to the observation that, without a sufficiently large dataset, great care should be taken before conclusions are drawn from theoretical work. Since the electronic investigation did not provide the desired result of finding a fast and effective method of determining which ligand merits further investigation, some steric aspects were studied. Once again, the precatalysts proved to be remarkably similar and no definitive answer for the observed differences in the various precatalysts could be determined. A preliminary experimental study into the feasibility of the synthesis of the new potential ligands was done. The multi-step synthesis route resulted in low yields in some cases, with the need for large volumes of solvents to purify the products. The toxicity of phenylphosphine also has to be taken into account when considering these types of ligands. A new precatalyst obtained by using a new ligand should show a remarkable improvement over the current commercially available precatalysts to justify the additional cost to synthesise a new ligand. It would seem that for future projects more consideration should be given to the deactivation mechanism of the Grubbs-type precatalysts, since this seems to be the logical starting point to look for the answers to the experimentally observed differences. A deeper understanding of the mechanism of alkene metathesis can only be obtained if all aspects are investigated in as much detail as possible. While the results did not provide the initially expected outcome, some valuable insights were gained that challenge the current way of thinking about the alkene metathesis mechanism. It is also clear that to oversimplify a very complex reaction and using limited data will lead to false assumptions being made. / PhD (Chemistry), North-West University, Potchefstroom Campus, 2014

In this study, an attempt was made to identify the electronic and steric properties of the precatalyst ligands that determine the characteristics of phosphine ligated Grubbs-type precatalysts for alkene metathesis by means of molecular modelling. It was found from studying the literature that the possibilities for synthesising a wide range of phosphine ligands are almost unlimited. Additionally, it was found that there is no easy method to determine the electronic and steric properties of the precatalyst ligands in existence. Molecular modelling might provide a method to study potential ligands and precatalysts before tedious synthesis methods are attempted. It was found that the theoretically calculated structures of the commercially available precatalysts compared well with the experimental data reported in literature. It is also shown that the energy profiles for alkene metathesis of simplified model systems do not compare well with non-simplified systems. Correlations between these simplified model systems and experimental work have to be regarded as serendipitous at best. When the energy profiles of the various new and commercially available precatalysts are compared, similarities in the energy trends for 1-octene metathesis are observed. These similarities raise questions about the significance of the differences in the energy barriers. In an effort to better understand this, two low activity precatalysts were also investigated in an attempt to identify the area or trend of poor catalysis. Instead of providing the desired different result, trends very similar to that of the highly active precatalysts were observed. This led to the observation that, without a sufficiently large dataset, great care should be taken before conclusions are drawn from theoretical work. Since the electronic investigation did not provide the desired result of finding a fast and effective method of determining which ligand merits further investigation, some steric aspects were studied. Once again, the precatalysts proved to be remarkably similar and no definitive answer for the observed differences in the various precatalysts could be determined. A preliminary experimental study into the feasibility of the synthesis of the new potential ligands was done. The multi-step synthesis route resulted in low yields in some cases, with the need for large volumes of solvents to purify the products. The toxicity of phenylphosphine also has to be taken into account when considering these types of ligands. A new precatalyst obtained by using a new ligand should show a remarkable improvement over the current commercially available precatalysts to justify the additional cost to synthesise a new ligand. It would seem that for future projects more consideration should be given to the deactivation mechanism of the Grubbs-type precatalysts, since this seems to be the logical starting point to look for the answers to the experimentally observed differences. A deeper understanding of the mechanism of alkene metathesis can only be obtained if all aspects are investigated in as much detail as possible. While the results did not provide the initially expected outcome, some valuable insights were gained that challenge the current way of thinking about the alkene metathesis mechanism. It is also clear that to oversimplify a very complex reaction and using limited data will lead to false assumptions being made. / PhD (Chemistry), North-West University, Potchefstroom Campus, 2014

The reaction of the lithium tris(2-pyridyl)stannate [LiSn(2-py6OtBu)3] (py6OtBu = C5H3N-6-OtBu),1, with the element(II) amides E{N(SiMe3)2}2 (E = Sn, Pb) afforded complexes [LiE{Sn(2 py6OtBu)3}3] for E = Sn (2) and E = Pb (3), which reveal three Sn–E bonds each. Compounds 2 and 3 have been characterized by solution NMR spectroscopy and X-ray crystallographic studies. Large 1J(119Sn–119/117Sn) as well as 1J(207Pb–119/117Sn) coupling constants confirm their structural integrity in solution. However, contrary to 2, complex 3 slowly disintegrates in solution to give elemental lead and the hexaheteroarylditin [Sn(2-py6OtBu)3]2 (4).

The dependence of tumour growth and metastasis on blood vessels makes tumour angiogenesis a rational target for therapy. Imaging of αvβ3 expression could potentially be used as a biomarker and an early indicator of efficacy of antiangiogenic treatments at a molecular level. Research efforts have mainly focused on the development of RGD-based radiolabelled αvβ3 inhibitors suitable for PET and SPECT imaging modalities that, owing to their high sensitivity, represent the most powerful tool for monitoring in vivo tumour angiogenesis. The aim of this multidisciplinary project was the design, synthesis and biological evaluation of novel αvβ3 ligands as molecular imaging probes. Three classes of integrin antagonists were designed: 1) triazole-based RGD mimetics that can be isotopically-labelled with tritium, fluorine and iodine radioisotopes by means of highly practical procedures, 2) RGD peptidomimetics incorporating the metabolically stable 2,2,2-trifluoroethylamine function as a peptide bond bioisostere and 3) RGD cyclopeptides conjugated with FDR, a novel prosthetic group allowing glycosylation and 18F-fluorination of aminooxy-functionalised molecules in one synthetic step. RGD-based strategies have also been used for selective tumour delivery of chemotherapeutic agents. A number of cytotoxic drugs have been conjugated to RGD peptides, providing experimental evidence that αvβ3 targeted chemotherapy strategies could be used as a powerful tool to reduce the toxicity and augment the therapeutic window of existing cytotoxic agents. In this work, we described the rational design of a novel targeted cytotoxic conjugate containing a triazole-based RGD peptidomimetic as tumour-homing motif of the potent antimitotic agent, paclitaxel. Preliminary in vitro studies were performed to assess the therapeutic potential of this targeted cytotoxic construct.

Achiral wide bite angle ligands have been shown to be highly active and to induce excellent chemo- and regioselectivities in many homogeneously catalyzed reactions. However, only a few examples of chiral wide bite angle ligands are known so far. A diphenyl ether backbone was selected to allow maximum synthetic versatility and potential for a modular approach to design and synthesize such chiral diphosphorus ligands. Three synthetic strategies have been explored in this thesis: i) introduction of chiral substituents in the ligand backbone, ii) the use of P-stereogenic donor atoms and iii) the synthesis of chiral mixed-donor ligands bearing chiral auxiliary groups on the phosphorus atoms. Functionalization of the 3,3'-positions of 2,2'-bis(diphenylphosphino)diphenyl ether by carboxylic acid or ether auxiliaries was achieved via straightforward four-step routes to generate a library of ligands that were tested in various catalytic reactions. In the Pd-catalyzed asymmetric allylic alkylation of l,3-diphenyl-2-propenyl acetate and cyclohexyl-2-enyl acetate with dimethyl malonate the enantioselectivity was found to depend on the size of the chiral auxiliary introduced within the diphenyl ether backbone and its proximity to the phosphorus donor groups and hence to the active metal centre. Two types of mixed donor bidentate diphosphorus ligands based on the diphenylether backbone have been established, i.e. phosphine-phosphite and phosphine-phosphonite derivatives. A small ligand library bearing different chiral auxiliaries was accomplished via straightforward syntheses that enable derivatization of the respective phosphite and phosphonite moieties in the final step. In the Rh-catalysed hydrogenation of several benchmark substrates high conversion and moderate to high enantioselectivities (up to 97% for dimethyl itaconate) were obtained. The enantioselectivity was influenced by the size of the ortho-substituent on the chiral auxiliary group of the phosphite or phosphonite fragment. Two modular synthetic approaches for the preparation of novel wide bite angle diphosphine ligands containing stereogenic P-atoms have been developed. Both protocols involved diphenylether as backbone and the chiral ephedrine based precursor (2R[subscript(P)],4S[subscript(C)],5R[subscript(C)])-oxazaphospholidine borane as initial auxiliary to induce chirality at phosphorus. Various novel diphosphines were isolated as highly enantioenriched compounds with dr-ratios up to 95:5.

A peptide nucleic acid (PNA) encoded 1296 member peptide library was synthesised and incubated with a variety of cell types. Library members entering cells were extracted, hybridised onto DNA microarrays and the peptide identity was determined via deconvolution. Global consensus analysis highlighted the tetrapepide, Glu-Llp- Glu-Glu (Llp is 6-hexamine-N-aminoacetic acid), a surprise in view of the basic residues typically observed in cell penetrating peptides. When evaluated, Glu-Llp- Glu-Glu revealed cellular uptake comparable to a known basic peptide (tetraLlp). In depth delineation via clustering analysis allowed assessment of differential cellular uptake, with the identified peptides showing clear cellular specificity. This was verified by peptide synthesis and cellular uptake analysis by flow-cytometry, and in all cases an endocytic uptake mechanism was confirmed. This approach establishes a strategy for the identification of short peptides as tools for selective delivery into specific cell types. The incubation of a 10,000 member PNA-encoded peptide library with D54 and HEK293T transfected with CCR6 cells followed by microarray analysis allowed detailed information on the interaction between peptide-ligands and cell surface receptors to be extracted. This allowed the identification of new ligands for integrins and G-protein coupled receptors and offers a novel approach to ligand discovery allowing the comparative analysis of different cell types for the identification of differences in surface-receptor ligands and/or receptor expression between various cell types. In addition, this work included the development of a novel method for the indirect amplification of a PNA library by amplification of a complementary DNA library hybridised to the PNA. The generation of 10,000 defined pieces of DNA would have a myriad of applications, not least in the area of defined or directed sequencing and synthetic biology, but also in applications associated with encoding and tagging. By this approach DNA microarrays were used to allow the linear amplification of immobilised DNA sequences on an array followed by PCR amplification. Arrays of increasing sophistication (1; 10; 3875; 10,000 defined oligonucleotides) were used to validate the process, with sequences verified by selective hybridisation to a complementary DNA microarray with DNA sequencing demonstrating error rates of ca ≈ 0.2%. This technique offers an economical and efficient way of producing hundreds to thousands of specific DNA primers, while the DNA-arrays can be used as “factories” allowing specific DNA oligonucleotide pools to be generated with or without masking. This study also demonstrated a significant variance observed between the sequence frequencies found via Solexa sequencing compared to microarray analysis.

Oxaliplatin is one of the three FDA-approved platinum anticancer drugs and considered a third generation drug, discovered after the first generation drug cisplatin and second generation drug carboplatin. It is known to react with proteins and DNA nucleotides in the body. Reaction with DNA occurs primarily at guanosine residues and secondarily at adenine residues for oxaliplatin and other platinum drugs. We have previously studied oxaliplatin and an analog with additional steric hindrance in the amine ligand and found that the analog had different reactivity with methionine. Now, we have prepared oxaliplatin and its three analogs Pt(Me2dach)(ox), Pt(en)(ox) and Pt(Me4en)(ox) and have reacted each platinum compound with both guanine and adenine nucleotides at pH 4 and pH 7 at different molar ratios. These reactions have been characterized by Nuclear Magnetic Resonance (NMR) spectroscopy equipment over time to observe the formation of products and compare them on the basis of their kinetics and binding affinities. NMR has shown that even under the conditions of excess platinum, the dominant products are usually those with two nucleotides coordinated to one platinum center. Reactions are faster at pH 7 than pH 4 due to deprotonation of phosphate group. Reactions of GMP with a platinum center are faster than reaction with AMP because of the chelate formed by the oxalate ligand. The extra methyl groups on the oxaliplatin analogs do not appear to slow down the reactions with nucleotides considerably. The pH generally affects the rate but does not substantially affect the product distribution.