A thesis submitted to the Faculty of Medicine University of the Witwatersrand, Johannesburg for the Degree of Doctor of Philosophy. / The investigations which make up the body of this thesis touch on various aspects of iron absorption. Although many of the factors governing iron absorption have been elucidated specific problems remain to be defined in greater detail. The aim of these investigations was therefore to define more fully some of the quantitative and qualitative aspects of dietary iron absorption / IT2018

A thesis presented in partial fulfilment of the requirements for the degree of Doctor of Medicine at the University of the Witwatersrand, Johannesburg. / The quantity of iron in the body is regulated largely by the amount absorbed from the gut as the body's capacity to excrete it seems to be very limited. However three conditions have been described in which enormous amounts of iron may accumulate. Such a finding is characteristic of idiopathic haemochromatosis, it is present in a proportion of malnourished South African and is seen also in cases of refractory anaemia treated over long periods with blood transfusions. / IT2018

This project was a joint venture between Loughborough University, the academic investigator, and ICI, the industrial sponsor. The aim was to develop and validate filter cycle scale-up methodologies, based on laboratory experiments at Loughborough and full-scale plant work at several ICI sites. Two ICI products, which experience processing problems during filtration, were chosen for assessment. They were a catalyst material (predominately iron oxide), processed by Synetix using rotary vacuum filters (RVFs), and titanium dioxide (TiO2) processed by Tioxide using a diaphragm filter press. A number of objectives were originally highlighted: (1) to develop and validate scale-up methodologies for a pressure filter and vacuum filter; (2) to develop models required to provide accurate predictions of filter performance; (3) to produce PC simulations of RVF and diaphragm filter press; (4) to produce reliable filter cycle data for process feeds of interest to ICI.

Iron β-diketiminate complexes have great potential as catalysts. Previous work into the coordination chemistry of complexes bearing the β-diketiminate ancillary ligand (Chapter 1) attest to the useful properties of these complexes in catalysis. A handful of literature reports on catalytic systems hint that this could be further extended. Hydrophosphination is a growing field that continues to generate a lot of interest from industry and academia alike. The aims of this project are to investigate hydrophosphination reactions with iron β-diketiminate complexes, to achieve high degrees of regioselectivity from these sterically encumbered complexes and to investigate iron catalysed dehydrocoupling reactions. A combination of synthetic and mechanistic methodologies will be employed in order to achieve definitive insight via NMR spectroscopic analysis, kinetic studies and solid state crystallography. Initial work presented herein (Chapter 2) will focus on the synthesis of iron(II) β-diketiminate complexes. Previously reported literature methods will be explored in order to determine an optimum procedure to use these precatalyst complexes. Initial investigations into hydrophosphination activity of these iron species will then be explored with alkenes. Results of these studies led to serendipitous findings and unexpected results in phosphine dehydrocoupling. The scope of this reactivity was then probed and mechanistic considerations taken into account with findings detailed herein. Radical catalysed reactivity observed will be further discussed. Solvent selectivity will then be discussed with a simple yet highly effective solvent change yielding a complete shift in catalytic activity. Further studies (Chapter 3) highlight the orthogonal reactivity of iron(II) β-diketiminate complexes in hydrophosphination catalysis. Less electronically activated and more atypical substrates have been investigated to determine their activity in hydrophosphination reactions. The synthesis of phosphinoalkenes and phosphinoalkynes for cyclic intramolecular hydrophosphination reactions are detailed along with their catalytic activity. Preliminary mechanistic studies are discussed with radical species again proving crucial to catalytic activity. Selective intermolecular hydrophosphination reactions have been investigated with alkynes. A solvent based switch can be employed wherein the regioselectivity of the reaction is completely altered. Substrate scope, mechanistic considerations and potential future applications are examined in full detail. Dehydrocoupling catalysis can be extended in scope (Chapter 4) from iron catalysed phosphine homocoupling reactions to heterocoupling reactions. Phosphine-silane dehydrocoupling is found to be highly selective for the formation of silaphosphanes, preliminary mechanistic insight and reaction scope is discussed. Analogous amine-silane dehydrocoupling is explored in full. The substrate scope offers insight into reactivity and potential further applications in sequential and tandem catalysis. In depth mechanistic insight is discussed with kinetic analyses. Iron-amido complexes are observed to react in a metathesis mediated cycle via iron hydride species. Finally catalytic alcohol-silane dehydrocoupling is investigated as a synthetic route to protected natural products in organic synthesis. Unsaturated silazanes are potential targets for further dehydrocoupling reactions. Catalytic reactions with pinacolborane led to highly facile desilylation reactions (Chapter 5). Mechanistic considerations hint that the reactions occur via σ-bond metathesis could through iron hydride species. Desilylation activity is then extended to siloxanes and a model developed with potential applications in the depolymerisation of polysilazanes and polysiloxanes.

Lai Chi-hung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1978. / Bibliography: leaves 80-82.

The amine-activated iron-catalysed hydrosilylation of alkenes and alkynes has been developed, and has been found to selectively provide the anti-Markovnikov product. It was established that a range of amines could be used to activate an iron(II) pre-catalyst containing triflate counter ions in situ, with diisopropylethylamine facilitating the highest reactivity (Scheme A1). Pre-catalyst and amine loadings as low as 0.25 mol% and 1 mol%, respectively, could be used for the gram-scale hydrosilylation of 1-octene with phenylsilane, in air. The activator-free iron- and cobalt-catalysed hydrosilylation of alkenes and alkynes was subsequently established, by using pre-catalysts containing tetrafluoroborate counter ions. The methodology was extended to the hydroboration and hydrogenation of olefins. The activation has been proposed to occur via the in situ formation of a hypervalent silicon ‘ate’ complex, formed as a result of the hydrolysis of the BF4 - counter ion. It was found that the iron-catalysed system provided anti-Markovnikov hydrosilylation products, whereas Markovnikov products were formed preferentially when using the cobalt pre-catalyst (Scheme A2). Consequently, the regioselectivity of hydrosilylation could be easily altered.

Ferritins are ubiquitous proteins that serve the dual-function of iron reservoir and sequestering the Fe(II) toxicity. The function of ferritins totally depends on the characteristic spherical structure with a di-iron centre performing the iron oxidation and a hallow cavity enclosing the iron minerals in a bioavailable form. I have characterised the structure, assembly and function of a new member of ferritin superfamily that is natively enclosed within an encapsulin shell. Encapsulin proteins are structurally-related to a virus capsid and form 60-meric or 180-meric icosahedrons. I show that this encapsulin associated ferritin-like protein (EncFtn) possesses two main alpha helices, which assemble in a metal-dependent manner to form a ferroxidase centre at a dimer interface. EncFtn adopts an annular decamer structure in contrast to the 24-meric classical ferritins or 12-meric mini-ferritin (DPS). The resemblance of the dimeric EncFtn and monomeric classical ferritins suggests that it is likely that classical ferritin evolves from EncFtn because of the gene duplication. EncFtn is a catalytically active ferroxidase but with only a limited iron binding ability due to its open structure. The encapsulin itself is not able to oxidise Fe(II), but is able to store about 2200 iron ions. I have demonstrated that the EncFtn must be housed in the encapsulin to achieve a maximum loading of approximately 4200 iron ions. The encapsulin nanocompartments are widely-distributed in both eubacteria and archaeon with distinct life styles and represent a distinct class of iron storage system, where iron oxidation and mineralisation are distributed between two proteins.

The iron-catalysed hydrofunctionalisation of alkenes and alkynes has been developed to give a range of functionalised products with control of regio-, chemo- and stereochemistry. Using a bench-stable iron(II) pre-catalyst, the hydrosilylation, hydroboration, hydrogermylation and hydromagnesiation of alkenes and alkynes has been achieved. Iron-catalysed hydrosilylation, hydroboration and hydrogermylation of terminal, 1,1- and 1,2-disubstituted alkyl and aryl alkenes and alkynes was developed, in which the active iron catalyst was generated in situ (Scheme A1). Alkyl and vinyl silanes and pinacol boronic esters were synthesised in good to excellent yield in the presence of a range of functional groups. Catalyst loadings as low as 0.07 mol% were demonstrated, along with catalyst turn-over frequencies of up to 60 000 mol h−1. The iron-catalysed formal hydrocarboxylation of a range of styrene derivatives has been developed for the synthesis of α-aryl carboxylic acids using carbon dioxide and ethylmagnesium bromide as the stoichiometric hydride source (Scheme A2). Detailed mechanistic studies have shown this reaction proceeds by iron-catalysed hydromagnesiation to give an intermediate benzylic organomagnesium reagent. The nature of the active catalyst and reaction mechanism have been proposed.

Solution density modelling is important in hydrometallurgical processes as accurate predictions of single and mixed electrolytes can be used in the design of equipment and their sizing, heat transfer calculations and choosing of materials for construction. This research project entails modeling of electrolyte solutions by extending the Laliberte and Cooper (compound level) model to ionic level where an electrolyte solution is modeled as a mixture of cations, anions and water molecules. This modeling predicts single and mixed electrolyte density as a function of electrolyte temperature in degrees Celsius; water, cation and anion apparent volumes in cubic centimeters; and their respective concentrations in the electrolyte as mass fractions. The model was developed by fitting single electrolyte density data reported in literature using the least squares method in Microsoft Excel®. The following 26 single electrolyte solutions were used in the fitting exercise: Al2(SO4)3, BaCl2, CaCl2, CdSO4, CoCl2, CuSO4, FeCl3, FeSO4, HCl, HCN, HNO3, K2CO3, LiCl, MgSO4, MnCl2, Na2SO3, NaF, NaI, NaOH, (NH4)2SO4, NiCl2, SrCl2, ZnCl2, ZnBr2, (NH4)2C2O4 and KNO2. The above electrolytes attributed to the following ions: Al3+, Ba2+, Ca2+ Cd2+, Co2+, Cu2+, Fe3+, Fe2+, H+1, K+1, Li+1, Mg2+, Mn+2, Na+1, NH4+1, Ni2+, Sr+2, Zn2+, SO42-, Cl-1, CN-1, NO3-1, CO32-, OH-1, SO32-, Br-1, F-1, I-1, C2O4-2 and NO2-1. This translated to a combination of at least 216 single electrolyte solutions which could be feasibly modeled, and a solution with at most 10 anions for mixed electrolytes, which is comparable with practical hydrometallurgical solutions. A database of volumetric parameters was generated comprising a total of 18 cations and 12 anions. The validation of the developed model was done by predicting densities for both single and mixed electrolytes not used in the fitting exercise. The average density error i.e. the difference between experimental and model density for the single electrolyte solutions was 22.62 kg m-3 with a standard deviation of 39.66 kg m-3. For the mixed electrolytes, the average density error was 12.34 kg m-3 with a standard deviation of 24.48 kg m-3. These calculated errors translated to a maximum percentage average error of less than 4% for single electrolyte solutions and maximum average percentage of less than 3% for mixed electrolyte solutions.

Various concentrations of ferrous and ferric iron solutions were held at room temperature for 60 min before they were assayed for ferrous iron, which may be unstable due to oxidation. The ferrous and ferric solutions (in pH 2 HCl) were maintained as such for 60 min without the use of chelators. There was no significant oxidation of ferrous iron. Also, four different levels of each ferrous and ferric iron were injected into proximal duodenal loops of rat intestine and uptake was determined at four different time intervals. Two iron-replete rats were assigned to each of the treatments. The in situ experiments showed that iron was taken up rapidly from pH 2.0 solutions of ferrous and ferric iron. Maximum amount of iron was taken up in the first 10 min. Uptake of ferrous iron was significantly greater (p < 0.05) than uptake of ferric iron, and there were significant differences in total uptake among the four iron levels used. Uptake, absorption, and adsorption kinetics of both ferrous and ferric iron were determined in situ for both iron-replete and iron-deficient rats. Deficiency caused greater uptake and absorption, confirming a biological adaptation of these processes. Both uptake and absorption were greater for ferrous than for ferric iron and were possibly taken up by different pathways or by a ferrous-ferric pathway with preference for ferrous. Uptake and absorption kinetics were biphasic for both ferrous and ferric iron. The first phase demonstrated saturation kinetics and was followed by a nonsaturable phase at higher concentrations of luminal iron. Iron deficiency altered the uptake and absorption kinetics of ferrous and ferric iron, but not always in a similar manner, suggesting that ferrous and ferric iron were each taken up by a separate pathway. Indications were that enhanced absorption during deficiency was largely due to adaptation of ferric uptake. Iron adsorption was directly proportional to luminal iron concentration, but it was greater for ferric than for ferrous, possibly due to charge interactions. Iron deficiency caused increased adsorption of both ferrous and ferric iron, supporting the notion that adsorption acts to maintain iron in a form available for uptake.