Equilibrium and kinetic data for the solvent extraction of germanium by three impure commercial 7-alkylated-8hydroxyquinoline extractants which vary in structure at the 7alkyl group, are evaluated in order to elucidate an holistic kinetic extraction model which accounts for the various reactions and partition effects occurring during the metal-ion chelation process. It is proposed that for the extraction process, which is first order in germanium concentration, by the ligand reagents Lix 26, TN 01787 and TN 02181, the rate-determining step, on stereochemical grounds, is the attachment of either a neutral ligand or a protonated ligand species-to the biligand intermediate GeL2 2+ (L:ligand) at the interface. In high speed shaking/mixing assemblies the extraction process was observed to occur in two discrete reaction regimes : a fast initial rate for which the orders with respect to ligand reagent are 1,06, 2,10 and 1,77 for TN 02181, Lix 26 and TN 0178-7 respectively, and a slower subsequent rate for which the apparent reaction orders with respect to ligand concentration are 1,12, 2,70 and 3,08 for TN 02181, Lix 26 and TN 01787 respectively. For the slower reaction regime, orders between 1 and 3 are explicable if the steady state approximation is invoked for the intermediate germanium species GeL3+ and GeL2 2+ formed at the interface. In the fast reaction regime, it is proposed that the accelerated extraction rates are a function of (i) the speciation of germanium and (ii) participation in the rate-determining step by the protonated ligand moiety H2L+HS04- which is rapidly formed after phase contact. At low ligand concentration, the following order of ligand efficacy has been observed : TN 01787 < Lix 26 < TN 02181 whereas at high concentration ligand efficacies are similar because the interface is saturated with ligand. Orders with respect to [H+] for the reagents vary from -1 to - 3 during the course of reaction, indicating complex mixedorder behaviour. The effects upon extraction of ionic strength, temperature, the addition of organic modifiers and diluent nature are investigated as well as the kinetics of germanium stripping by aqueous hydroxide. The physical effects of interfacial tension, viscosity and relative dielectric constant are also reported and suggestions are made as to their effect upon the extraction characteristics. Computer modelling of the extractants has been used as an aid in describing size, structure and stereochemical considerations of the ligands and the chelate products. / Thesis (Ph.D.)-University of Natal, Durban, 1990.

A series of 88 novel tetrahydroisoquinoline (TIQ) compounds have been synthesised for applications in asymmetric catalysis. Several chiral TIQ ligands, possessing N,O and N,N donor atoms, have been prepared and evaluated for the catalytic asymmetric transfer hydrogenation (ATH) of pro-chiral ketones. The highest selectivity obtained for the asymmetric transfer hydrogenation of acetophenone with the N,O donor atom ligands was >99 % ee at low temperatures in iso-propanol with [Ru(p-cymene)Cl2]2 as a pre-catalyst. The observed enantioselectivity was supported by theoretical calculations using the Jaguar interphase program (Paper I). An enantioselectivity of 70 % was obtained with the ligands possessing the N,N donor atoms with the observation that water played a significant role in the enantioselectivity of the ATH reaction of acetophenone (Paper II). An investigation into the usefulness of the TIQ scaffold with other donor atoms was also undertaken. A series of novel P,N oxazoline ligands were synthesised and coordinated to Iridium BArF. These complexes were screened as chiral catalysts for the high pressure asymmetric hydrogenation of unsymmetrical olefins. The reactions proceeded readily at ambient temperature and provided selectivities of up to >91 % ee with excellent conversion rates (>99 %) for the benchmark reactions. Based upon these favorable results, the ligands providing the best results were further screened on a variety of functionalized and unfunctionalised olefins (Paper III). The success of the TIQ backbone in ATH reactions prompted an investigation into its applications in carbon-carbon (C-C) bond forming reactions. The Henry (nitroaldol) reaction is an important C-C bond forming reaction with the chiral oxazoline class of ligands being widely utilised. A series of novel, chiral TIQ oxazoline ligands were synthesised and complexed to various metals (Cu, Sc, Co, Zn, Ni and Mn). These complexes were screened as chiral catalysts in the asymmetric Henry reaction. The highest enantioselectivity (>77 %) was obtained when Cu(OAc)2was employed as a pre-catalyst and iso-propanol as a solvent (Paper IV). The final chapter deals with carbon-sulphur bond formations facilitated by the conjugate addition of thioglycolate to various chalcones. A series of novel, chiral TIQ N-oxide ligands were synthesised and complexed to lanthanum. The complexes were screened for activity against the benchmark reaction and an enantioselectivity of 81 % was obtained. / Thesis (Ph.D.)-University of KwaZulu-Natal, Westville, 2010.

Kraussianone 1 and kraussianone 2 were previously isolated as active compounds from the roots of Eriosema kraussianum Meisn., a plant used for the treatment of male impotence and urinary complaints in KwaZulu-Natal. The objectives of this study were firstly, to develop a method for the analysis of metabolites from E. kraussianum and other Eriosema plants that are used for erectile dysfunction and secondly, to develop synthetic methods for kraussianone 1 and structurally related compounds. A reversed-phase HPLC-PDA method was developed for the analysis of the extracts of plants from different sources, two of which were authentic E. kraussianum collected from the Drakensberg and Pietermaritzburg. The roots of other Eriosema species called ubangalala and uqonsi in Isizulu were also analysed. These plants were bought from the local herbal traders. The extracts of the two E. kraussianum plants and one uqonsi sample showed a similar chemical profile, even though there were variations in the relative concentrations of the metabolites within each plant. In these three plants, kraussianone 1, the most active metabolite of E. kraussianum, occurred in relatively low quantities, whereas kraussianone 2 was one of the major constituents. The other commercial plants that were analysed contained different compounds from those found in E. kraussianum. The HPLC method developed herein facilitates rapid identification and relative quantification of metabolites from E. kraussianum. Strategies based on semi-synthesis and total synthesis were employed for the preparation of kraussianone 1. The semi-synthetic route was based on the transformation of the prenyl side chain of kraussianone 2 into a linear dimethylpyran scaffold fused to the A-ring. Two routes were investigated for the semi-synthesis of kraussianone 1 from kraussianone 2. In the first route, the dimethylchromene ring was to be prepared by the acid-catalysed cyclisation of the prenyl group of kraussianone 2, followed by dehydrogenation of the resulting dimethylchroman chromophore. This route was abandoned due to poor regioselectivity of the cyclisation reaction and the difficulty of oxidising the dimethylchroman scaffold on the phloroglucinol moiety into a dimethylchromene. The second strategy involved selective protection of the OH-2', followed by DDQ-mediated oxidative cyclisation of the prenyl group to OH-7. This was the most viable route and kraussianone 1 was prepared in an overall yield of 54% from kraussianone 2. The total synthesis of kraussianone 1, on the other hand, employed the Suzuki-Miyaura reaction for the construction of the isoflavone nucleus and the regioselective introduction of the dimethylpyran scaffolds to the A- and B-rings. The key precursors in this synthesis were 3-iodo-5,7-dimethoxymethoxychromone and a boronic acid coupling partner, 7- benzyloxy-2,2-dimethylchromene-6-boronic acid, already bearing the prerequisite chromene scaffold attached to the B-ring. The isoflavones genistein, 2-hydroxygenistein, eriosemaone D and a geranyl analogue of kraussianone 1 were prepared via the route developed for the total synthesis of kraussianone 1 by structural modifications of rings A and B. Furthermore, this synthetic approach was expanded to the synthesis of the coumarochromones lupinalbin A and lupinalbin H. The development of the feasible semi-synthetic and total synthetic routes described herein for kraussianone 1 is of importance for the production of material for an in depth study of the pharmacological activities and the structure-activity relationship studies of kraussianone 1 and related compounds. / Thesis (Ph.D.)-University of KwaZulu-Natal, 2011.

This study has shown that alkythioboranes 1,3,2-dithiaborolane and 1,3,2-dithiaborinane, disproportionated significantly during their synthesis. Their interaction with 1-octene has been investigated, and the rate constants, enthalpies and entropies of the hydroboration process, have been determined. The thermodynamic and kinetic parameters obtained have shown that the hydroboration reaction is sluggish and proceeded via an associative mechanism. / Thesis (Ph.D.) - University of KwaZulu-Natal, Pietermaritzburg, 2008.

The principal goal of this work was to synthesise and fully characterise a range of platinum group metal chelates of bis(pyridine-imine) ligands. These four-nitrogen donor Schiff base ligands are underdeveloped relative to their salen (ONNO donor) counterparts. The purified metal complexes were to be tested for their cytotoxicity against cancer cell lines and their mode of interaction (expected to be intercalation) studied. The syntheses, spectroscopic and structural properties of some novel and some already known bis(pyridine-imine) ligands are described. Furthermore, UV-Vis, IR and NMR spectroscopy, as well as electrospray ionisation mass spectrometry, have been used to characterize the ligands and comparisons were made with relevant literature. Fifteen Schiff base ligands were successfully condensed from a 2-formyl or 2-ketopyridine starting material and a diamine bridging group, while the attempted syntheses of a further three are described. Literature methods or variations thereof were employed in the syntheses of these derivatives, which generally resulted in good yields. X-ray quality crystals were obtained and X-ray structures were determined for four novel ligands and five unexpected cyclised hexahydropyrimidine- and imidazole-containing bidentate ligands, of which three were novel structures. The series of bicationic palladium and platinum complexes synthesised here were analysed by NMR, IR and UV-Vis techniques, as well as X-ray crystallography when possible. The complexes were prepared by reacting the free ligands with platinum group metal salts in refluxing acetonitrile. The complexes exhibit infrared bands for the imine C=N stretch between 1604–1670 cm-1; around 15–40 cm-1 lower than the free ligands. The 1H NMR spectra in the CH=N chemical shift region also display shifts (0.1 to 0.8 ppm for the palladium complexes or 0.4 to 0.9 ppm for the platinum complexes) which are consistent with metallation. X-ray crystal structures were obtained for eight novel metal complexes, which all crystallised in the monoclinic crystal system. The solid state analysis shows changes in the free ligands upon introduction of the metal ions, caused by the coordination process. Metallation of the free ligands led to twisting of the ligands due to size effects and the spatial restrictions of the coordination geometry of the central metal ions. The structures were generally solved by direct methods and refined to R1 = 0.0729 or less. Singlecrystal X-ray diffraction analysis confirmed that the mononuclear complexes exhibit a distorted square planar coordination sphere composed of the four donor nitrogen atoms (two imine and two pyridyl nitrogen atoms) from the Schiff base ligands. The complexes were generally very stable and differed by the type of metal ion (platinum(II) or palladium(II)) and the diamine bridging group (2-carbon or 3-carbon linking chain with various substituted groups). A range of unconventional F···H−C contacts is revealed to play an important role in the overall bonding and crystal packing of many complexes. To better understand our measured data and to separate the intrinsic properties of our molecules from intermolecular interactions, theoretical calculations at the DFT (B3LYP) level were carried out. These calculations predict the structural and spectroscopic properties for the free ligands and their metallated counterparts. DFT simulations were performed for the fifteen synthesised ligands (B3LYP functional, 6-31G** basis set), as well as three projected ligands that could not be synthesised, and for all proposed thirtysix metal complexes (B3LYP functional, SDD basis set). DFT simulations were used to obtain theoretical IR frequency data which was compared to the literature and used to prove the location of a local minimum energy structure in the geometry optimisation rather than a transition state. Our results collectively showed the B3LYP level of theory to be useful in the prediction of IR frequencies for these Schiff base ligands as well as the platinum(II) and palladium(II) complexes. Geometry optimisations performed using density functional theory for the free ligands and metal complexes were compared, where possible, with X-ray data. For the free ligands the main structural differences were observed for the position of the pyridyl-imine "arms" of the ligands with relatively good agreement between the two structures for the bridging groups. On the other hand, the metal complexes showed greater discrepancies for the bridging groups rather than for the pyridine rings. These differences were also affected by the length of the carbon bridge; the longer the carbon bridge, the less variation was observed. The trends in the variation of bond distances and angles with the metal ion identity and ligand structure delineated by DFT simulations were matched by similar trends in the X-ray data. Differences between the gas phase calculated geometries and those determined by Xray diffraction were attributed to packing effects in the solid state and intermolecular interactions not being accounted for during DFT computations. The impressive similarity observed between the structures obtained from X-ray crystallography and computed by DFT shows the applicability of these computations at the B3LYP level of theory. They were able to accurately predict structural and spectroscopic properties for the ligands and complexes presented in this work. The in vitro cytotoxicities of some of the metal complexes were evaluated against two mouse cancer cell lines. The compounds tested had lower than expected cytotoxicity towards the cell lines studied with IC50 values > 100 μM. The interaction of the complexes with calf thymus DNA has been explored by using absorption studies. From the observed changes in absorption possible modes of binding to DNA have been proposed for the metal complexes. Significant changes were observed in the absorption spectra upon interaction of the complexes with DNA, from which large binding constants were determined. The changes were, however, not as intense as expected nor were the spectral variations typical of intercalating drugs. There were also no bathochromic shifts nor any isobestic points observed for most of the metal complexes, except one. For the cationic PGM chelates studied, the lack of shift in the wavelength of the visible range MLCT band maximum and small changes in the band intensity were more indicative of weak electrostatic interactions. Our spectral data were thus interpreted as being consistent with non-intercalative binding; either simple electrostatic adduct formation or major/minor groove binding. The binding constant values (Kb) of the metal complexes estimated from the titration with calf thymus DNA monitored by absorption spectroscopy were all in the range of 105 M-1. This is within the range of those reported for other platinum(II) and palladium(II) metal complexes that have shown intercalation and/or groove binding. With the lack of inhibition of growth of the selected cancer cell lines it is possible that the compounds do not reach nuclear DNA in living cells or that such compounds are possibly substrates for efflux transporters. All the metal complexes in this work were determined to not be pure DNA intercalators as had been expected; however, in some cases partial intercalation cannot be ruled out from the spectral data. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2011.

The base metal and platinum group metal extractive processes for two South African refineries were examined using x-ray powder diffraction. Base metal refinery matte contained nickel sulphide (Ni3S2), copper sulphide (Cu2S), nickel-copper alloy (Ni-Cu), iron oxide (Fe3O4) - nickel iron oxide (NiFe2O4) and nickel oxide (NiO). Nickel iron oxide and cobalt sulphide (CO3S4) were found in matte aeration cavities. Matte fast cooling caused alloy zoning. Alloy stoichiometry and weight percentages of matte constituents were calculated. Magnetic separation of the alloy phase was unfeasible due to high bulk and poor separation from Ni3S2. Sulphuric acid matte leaching produced residues containing copper sulphide (Cul.8S) and nickel sulphides (Ni3S4, NiS). Potassium cyanide residue leaching produced NiS-rich residues, while thiourea / hydrochloric acid leaching produced residues almost free of copper sulphide. PGM concentrates were examined. Base metals occurred as: copper sulphate pentahydrate; copper sulphate hydroxide hydrate (CU4SO4(OH)6'H2O); copper sulphide (CuS); copper platinum (Cu3Pt); nickel oxide (NiO); nickel iron oxide (NiFe2O4). Lead occurred as lead sulphate and silicon as quartz and enstatite ((Mg,Fe,Al)SiO3). Heating the concentrate to 260°C with sulphuric acid converted copper sulphide to copper sulphate. PGMs were mostly metallic. Overdrying concentrates caused pgm sintering. Drying with sulphuric acid reduced sintering. PGM oxidative leach solubility was examined. Platinum sulphide (Pt,Pd)S caused low platinum and palladium solubility. (Ru,Rh,Ir,Pt)AsS caused low rhodium and ruthenium solubility. Platinum-rhodium (Rh0.57Pt0.43), ruthenium-osmium (~Ru16Os), iridosmine (Os,Ir,Ru) also caused low pgm solubility. Silver chloride remained in residues. Concentrate overdrying produced insoluble sintered platinum, palladium monoxide, and ruthenium dioxide. In minor concentrates sintered platinum, rhodium selenide (RhSe2+x), ruthenium dioxide and possibly palladium monoxide lowered pgm solubility. Precipitates and salts were examined. Iron precipitated as iron oxide hydroxide (B-FeOOH); gold as metallic gold; lead as lead chloride. Common salts were: sodium chloride; sodium iron hydroxide sulphate hydrate (Na2Fe(SO4)2(OH)'3H2O); ammonium chloride; sodium carbonate monohydrate. PGM-Iead fusion was examined. Fusion produced the insoluble alloy Pb(Pd,Pt)3. Ruthenium dioxide reacted with lead carbonate forming lead ruthenium oxide (Pb2Ru2O6.5). Nitric acid insoluble residues also contained lead sulphate and iron oxide (Fe3O4). This study demonstrated that solution problems can be understood by identifying the crystalline insoluble phases in intractable residues. / Thesis (Ph.D.)-University of Natal, Durban, 1994.

Abstract available in pdf file.

Many physics and engineering problems are modeled by differential equations. In many instances these equations are nonlinear and exact solutions are difficult to obtain. Numerical schemes are often used to find approximate solutions. However, numerical solutions do not describe the qualitative behaviour of mechanical systems and are insufficient in determining the general properties of certain systems of equations. The need for analytical methods is self-evident and major developments were seen in the 1990’s. With the aid of faster processing equipment today, we are able to compute analytical solutions to highly nonlinear equations that are more accurate than numerical solutions. In this study we discuss solutions to nonlinear partial differential equations with focus on non-perturbation analytical methods. The non-perturbation methods of choice are the homotopy analysis method (HAM) developed by Shijun Liao and the variational iteration method (VIM) developed by Ji-Huan He. The aim is to compare the solutions obtained by these modern day analytical methods against each other focusing on accuracy, convergence and computational efficiency. The methods were applied to three test problems, namely, the heat equation, Burgers equation and the Bratu equation. The solutions were compared against both the exact results as well as solutions generated using the finite difference method, in some cases. The results obtained show that the HAM successfully produces solutions which are accurate, faster converging and requires less computational resources than the VIM. However, the VIM still provides accurate solutions that are also in good agreement with the closed form solutions of the test problems. The FDM also produced good results which were used as a further comparison to the analytical solutions. The findings of this study is in agreement with those published in the literature. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.

Since the discovery of Wilkinson’s catalyst and its usefulness in the homogeneous hydrogenation of olefins many investigations have been carried out on trivalent, tertiary phosphine–rhodium complexes.¹ Studies have shown that N-Heterocyclic carbenes as ligands offer increased stability to the complex and possess similar electronic properties as phosphine ligands.² The applications of the traditional catalyst are limited due to the limited stability of its solutions and its susceptibility to attack from the environment i.e. oxygen and moisture. The hydrogenation of olefins and other unsaturated compound is of great importance for the fine chemical and petroleum industries. The aim is to produce more stable and active versions of the traditional catalyst and also to demonstrate their improved stability and activity in catalytic applications. This study involves the investigation of the effects of ligand modification on Wilkinson type hydrogenation catalysts. Five Rhodium-phosphine complexes 1a: Rh(PPh₃)₃Cl, 1b: Rh(PPh₂Me)₃Cl, 1c: Rh(PPh₂Et)₃Cl, 1d: Rh(PPhMe₂)₃Cl, 1e: Rh(PPhMe₂)₃Cl have been synthesised and characterised by means of melting point,¹H NMR, ¹³C NMR, ³¹P NMR, IR and Mass Spectroscopy. Complexes 1d and 1e have also been characterised by means of elemental analysis and single crystal XRD. Five rhodium-N-heterocyclic carbene complexes 2a: Rh(COD)ImesCl [Imes =1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene] , 2b: Rh(COD)(diisopropylphenyl)₂Cl 2c: Rh(COD)(adamantyl)²Cl, 2d: Rh(COD)(diisopropyl)²Cl 2e: Rh(COD)(ditertbutyl)²Cl have been synthesised and characterised by means of melting point, ¹H NMR, ¹³C NMR, IR and Mass Spectroscopy. Five rhodium-NHC-CO complexes 3a: Rh(CO)₂ImesCl, 3b: Rh(CO)₂(diisopropylphenyl)₂Cl, 3c: Rh(CO)₂(adamantyl)₂Cl , 3d: Rh(CO)₂(diisopropyl)₂Cl, 3e: Rh(CO)₂(ditertbutyl)₂Cl, have been synthesised and characterised by means of ¹H NMR, ¹³C NMR, IR and Mass Spectroscopy. Complexes 1a, 1d, 1e, 2a, 2b, 2c, 2d, 2e were tested in the hydrogenation of simple alkenes under mild conditions. For the rhodium-phosphine complexes the catalyst efficiency based on TOF increases in the following order: 1a > 1d > 1e or RhCl₃(PPhMe₂)₃ > RhCl₃(PPhEt₂)₃ > RhCl(PPh₃)₃. For the rhodium-(COD)-NHC complexes catalyst efficiency based on TOF increases in the following order: 2d > 2b > 2e > 2a > 2c. While rhodium-phosphine complexes are far more active than rhodium-(COD)-NHC complexes, the latter seem to be active for a longer time and hence more stable under mild hydrogenation conditions. / Thesis (M.Sc.)-University of KwaZulu-Natal, Durban, 2010.

This study is focused on the analysis of transformer oil commonly used in South African transformers and the characterisation of the blends of these oils with either synthetic antioxidants or those extracted from natural sources. The virgin naphthenic based uninhibited transformer oil was shown to have no acidic or saponifiable constituents. Although the oil used in this study is naphthenic based it was found to contain bicyclic compounds identified as tetralin (aromatic) and decalin by GC-MS. The oils were void of typical wear metals and only additive metals/non-metals were present. The blending of transformer oils with synthetic antioxidants was successful as confirmed by ¹H NMR, ¹³C NMR and FT-IR data. The blending showed increased antioxidant activity of the oils as well as thermal stabilities (DPPH assay and TGA respectively). Apparent activation energy for the decomposition reaction was determined by means of the Ozawa, Flynn and Wall (OFW) and Kissinger models. The apparent activation energy for BHT-B, DBP-B and VO were found to be 57.0, 63.7 and 69.3 kJ molˉ¹ , respectively for 10% mass conversion obtained from TGA data (OFW model). The dissipation factor for the oil blends was established to be favourably lower than that of unblended oils; however an undesirable decrease in the dielectric breakdown voltage of the blends was observed (16, 26 and 50 kV for BHT-B, DBP-B and VO respectively). The blending of transformer oil with synthetic antioxidants is a common practice, however greener approaches are always favoured. In this study, a greener approach was employed by blending the transformer oil with turmeric plant extract in an effort to improve the oil’s performance. Turmeric is known for its antimicrobial, antioxidant and anticarcinogenic properties. Solvent extraction of turmeric powder and subsequent blending into transformer oils, revealed the dissolution of 5 components from turmeric into the oil, viz., curcumene, sesquiphellandrene, ar-turmerone, α-turmerone and β-turmerone (curlone). In this work, we analysed novel turmeric-oil blends in terms of oxidation and thermal stability in comparison to that of virgin transformer oil. The DPPH assay confirmed that addition of these blends into the oil increased its ability to scavenge the free radical DPPH. Thermogravimetric analysis showed an increase in thermal stability when compared to that of virgin unblended oil. Evidently blending of synthetic antioxidants/turmeric extract has potential to enhance the performance and increase the life-span of transformer oils. / Thesis (M.Sc.)-University of KwaZulu-Natal, Westville, 2012.