Ph.D. (Chemistry) / The main objective of this study was to develop sample preparation methods based on separation and preconcentration of trace metals in organic matrices (alcohols, diesel and gasoline). The presence of metals in organic matrices is undesirable, unless they are used as additives. Therefore, it is important to control and monitor their concentrations in fuel and petrochemical products as they tend to affect the quality of these products. Solid phase extraction (SPE), solid phase microextraction (SPME) and hollow fiber-liquid phase microextraction (HF-LPME were used for separation and preconcentration of Ag, Al, As, Cd, Co, Cr, Cu, Fe, Mn, Mo, Ni, Pb, Sb, Ti, V and Zn prior to their inductively coupled plasma optical emission/mass spectrometric (ICP OES/-MS) determination. For solid phase extraction, the exchange efficiencies of different commercial ion exchange resins, namely Dowex 50W-x8, Dowex 1-x8, Dowex MAC-3 and Chelex 100, for preconcentration of metal ions in alcohol and fuel samples, were investigated. The results obtained indicated that Dowex 50W-x8 was suitable for simultaneous preconcentration of cations such as Cd, Co, Ni, Cu, Fe, Mn and Zn, among other metals while Dowex 1-x8 was suitable for metal ions that exists in more than one oxidation states, namely, As, Cr, Mo, Sb and V. Chelex-100 and Dowex MAC-3 were found to be selective to a limited number of target metal analytes. For further applications, Dowex 50W-x8 and Dowex 1-x8 were employed. Furthermore, the applicability of synthetic adsorbents such as nanometer-sized alumina and functionalized cellulose nanofibers for preconcentration of trace metals in gasoline samples, was investigated. Nanometer-sized alumina sorbent was found to be suitable for simultaneous separation and preconcentration of Co, Cr, Mn, Ni and Ti. Functionalized cellulose nanofibers on the hand, were suitable for simultaneous preconcentration of Cd, Cu, Fe, Pb and Zn. The optimization of the experimental parameters was achieved by both univariate and multivariate procedure. The second preconcentration technique was solid phase microextraction which was also used for extraction and enrichment of metal ions in diesel samples using two approaches of the SPME method. The first approach was hollow fiber-solid phase microextraction (HF–SPME) method using fiber-supported sol-gel combined with a cation exchange resin (Dowex 50W-x8). This method showed satisfactory results for the preconcentration of Cd, Cu, Fe, Pb and Zn in diesel and gasoline samples. The second approach was based on membrane solid phase microextraction (MSPME) using titania-alumina hollow fiber. The MSPME method applied was used for extraction and preconcentration of trace amounts of Co, Cr, Mo, Ni, Sb and V in liquid fuel samples. Multivariate techniques were used for optimization of experimental parameters for both approaches. The last preconcentration technique that was developed was hollow fiber- liquid phase microextraction (HF-LPME). In this method fuel samples were first digested before being subjected to HF-LPME system. Ammonium pyrrolidine dithiocarbamate (APDC) and [C6MIM][PF6] ionic liquid were both used as chelating agent and acceptor phase, respectively. Two level factorial and central composite designs were used for multivariate optimization of experimental parameters. Satisfactory results were obtained for extraction and preconcentration of Ag, Al, As, Mn and Ti.

Ph.D. (Chemistry) / Materials that are to be used as reference samples are frequently analysed using costly and sophisticated instrumentation, itself calibrated with similar certified standards. Analytical programmes using such instrumentation are only as accurate as the initial calibrations, and their poor results have led to the adoption by the United States National Bureau of Standards of definitive methods of analysis for the determination cf trace elements ...

The Saint-Urbain and Big Island rutile-bearing ilmenite Fe-Ti oxide deposits are located in the composite 450 km² Saint-Urbain anorthosite (1055-1046 Ma, U-Pb zircon) and in the Lac Allard intrusion (1057-1062 Ma, U-Pb zircon) of the 11,000 km² Havre-Saint Pierre anorthosite suite, respectively, in the Grenville Province of Eastern Canada. Slow cooling rates of 3-4°C/m.y. are estimated for both anorthosites, based on combined U-Pb zircon/rutile/apatite and ⁴⁰Ar/³⁹ Ar biotite/plagioclase geochronology, and resulted from emplacement during the active Ottawan Orogeny. Slow cooling facilitated (1) diffusion of Zr from ilmenite and rutile, producing thin (10-100 microns) zircon rims on these minerals, and (2) formation of sapphirine via sub-so lidus reactions of the type: spinel + orthopyroxene + rutile ± corundum → sapphirine + ilmenite. New chemical and analytical methods were developed to determine the trace element concentrations and Hf isotopic compositions of Ti-based oxides. Rutile is a magmatic phase in the deposits with minimum crystallization temperatures of 781°C to 1016°C, calculated by Zr-in rutile thermometry. Ilmenite present in rutile-free samples has higher Xhem (hematite proportion in ilmenite), higher high field strength element concentrations (Xhem = 30-17; Nb = 16.1-30.5 ppm; Ta 1.28-1.70 ppm), and crystallized at higher temperatures than ilmenite with more fractionated compositions (Xhem = 21-11; Nb = 1.36-3.11 ppm; Ta = <0.18 ppm) from rutile-bearing rocks. The oxide deposits formed by density segregation and accumulation at the bottom of magma reservoirs, in conditions closed to oxygen, from magmas enriched in Fe and Ti. The initial ¹⁷⁶Hf/¹⁷⁷ Hf of rutile and ilmenite (Saint Urbain [SU] = 0.28219-0.28227, Big Island [BI] = 0.28218-0.28222), and the initial Pb isotopic ratios (e.g.²⁰⁶Pb/²⁰⁴ Pb: SU = 17.134-17.164, BI = 17.012-17.036) and ⁸⁷Sr/⁸⁶ Sr (SU = 0.70399-0.70532, BI = 0.70412-0.70427) of plagioclase from the deposits overlap with the initial isotopic ratios of ilmenite and plagioclase from each host anorthosite, which indicates that they have common parent magmas and sources. The parent magmas were derived from a relatively depleted mantle reservoir that appears to be the primary source of all Grenvillian anorthosite massifs and existed for --600 m.y. along the margin of Laurentia during the Proterozoic. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

Please read the abstract in the section 00front of this document / Dissertation (MSc)--University of Pretoria, 2007. / Geology / Unrestricted

Multi-element stream sediment reconnaissance in the Hess River region of the Eastern Yukon has outlined an extensive area characterized by anomalously high molybdenum values. An accessible region in the Hess Mountains, within the high molybdenum zone, was selected for detailed study of trace element levels in stream sediment, rock, soil and vegetation. In view of the frequently observed relationship between high forage molybdenum concentrations and the incidence of copper deficiency in cattle, molybdenum concentrations in plant species likely to be consumed by caribou and moose were of particular interest. High sediment molybdenum values are characteristic of catchments underlain by dark shales and less commonly dark limestone. These rocks and associated soils are rich in molybdenum. Concentrations in vegetation growing on anomalous shaly soils are characteristically low, while most plants growing on soils derived predominantly from limestone are molybdenum-rich. The Mo-Cu status of vegetation on limey soils is typically within the range associated with molybdenum induced hypocuprosis in cattle. Low molybdenum uptake by plants on soils derived from shales likely reflects the unavailability of the molybdate anion, resulting from its adsorption onto clay minerals and sesquioxides under acidic conditions prevalent in these soils. In neutral to mildly basic environments, typical of dark limestone soils, molybdenum adsorption is greatly decreased, and therefore molybdenum is relatively available to plants. In the detailed study area soil pH values are typically similar to pH levels in associated stream water. Therefore by combining stream sediment molybdenum concentrations with stream pH data, catchments likely to contain molybdenum-rich vegetation can be predicted. Unfortunately stream pH values were not obtained in the regional survey. In view of the apparent rarity of dark limestone throughout the Eastern Yukon, however, molybdenum-rich vegetation is not likely to be particularily widespread. Wildlife in this area, therefore, is probably not significantly affected by molybdenum induced copper deficiency. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate

The distribution of trace elements in bedrock, soils and plants, was studied in twelve areas of British Columbia where mineralization was known to occur below different kinds and depths of overburden. Samples were taken from two soil profiles and the bedrock at each location, and second and third year twigs of the principal vegetation found within a radius of fifty feet of each profile were also collected. The bedrock, soil and plant samples were analysed. The soil samples were used for the determination of pH, organic matter content, percentage of material <80 mesh, cation exchange capacity, exchangeable heavy metals, and content of Cu, Mo, Zn, Pb, As, Co, Ni, Fe and Hg. The same elements were determined in the bedrock and vegetation samples. The results were examined graphically for relationships between elemental contents of the bedrock, soil horizons and vegetation. The data were then statistically analysed.(a) soil horizons and plant relationship with bedrock. (b) inter-elemental relationship of individual horizons of soils and of plants, as well as all horizon relationships. (c) multiple correlation study of cation exchange capacity, percentage of organic material and -80 mesh of element content of individual and all soil horizons. These studies showed that, although most of the soil horizons were developed from transported materials, (glacial, alluvial, etc.), there was a highly significant correlation with B and C horizons and bedrock that confirmed the value of soil sampling in prospecting, since horizon development includes the upwards migration of the elements from bedrock. The secondary dispersion of the halo elements (Mo, Zn, Pb, As, Co, Ni, Hg) proved useful as pathfinders where major economic elements may have been masked during the upward migration process. Secondary dispersion may also be in some degree, helpful in identifying the origin of soils and plants. A great divergence in the affinity of various plants for different elements, and of the same species at different locations, was noted. It was also observed that plants have a closer relationship to the soils than to the bedrock itself, but even so, indicate mineralization. The important relationships between elemental distributions in soil horizons and in plants with bedrock, indicated a logarithmic relationship. The multiple correlation study indicated that some of the major factors of influencing the level of element content in soils developed on transported material-covered areas, are the size of the soil particles and frequently the pH of the soil. In general, the study indicated that the distribution of trace elements is highly complex and that bedrock, soils, and plants, should all be combined into one study; since the study of one of these alone would be incomplete without the others. / Land and Food Systems, Faculty of / Graduate

M.Sc. (Chemistry) / The determination of 17 trace elements (As. Ba. Co. Cr. Cu. Mo. Nb. Ni. Pb. Rb. Sr. Th. U. V. Y. Zn and Zr) in uranium ores by X-ray fluorescence spectrometry was investigated in this study. The determination of major elements. however. was also necessary for the calculation of mass absorption coefficients. Major elements were determined on borate melts and trace elements on powder briquettes pressed at 7 t with a binder in liquid form. Initially a method was developed for the determination of the elements of interest in unmineralised rocks The rhodium tube was used for the Group 1 elements (As. Mo. Nb. Pb, Bb, Sr. Th, U. Y and Zr) and the gold tube for the Group 2 elements [Ba, Co. Cr. Cu. Ni. V and Zn). Background and peak overlap corrections were made by means of background and interference factors. Corrections for absorption of radiation by the sample were made by means of mass absorption coefficients. which were calculated from the major element composition or obtained from the relation between the inverse of the mass absorption coefficient and the intensity of the Compton scattering peak. Due to various problems. only the latter method was suitable for uranium ores. The high uranium content in uranium ores mainly affected the Group 1 elements. Because of the high intensity of various UL lines. large overlap corrections were necessary. while only a few completely interference-free background positions were available. Consequently. the Feather and Willis method was used for determining the background intensity at the peak positions as well as for mass absorption coefficients. As a result of the presence of the UL absorption edges both primarx and secondary mass absorption coefficients had to be used for matrix corrections. Furthermore. it was observed that the background intensity in the region of the uranium lines increases with increasing uranium content of the sample instead of the expected decrease due to the increasing mass absorption coefficient. This effect was greater for the LiF(11 0) crystal than for the LiF(100) and was attributed to the scattering of uranium lines in the spectrometer chamber. especially from the crystal. A method was developed to correct the measured intensities for this scattering effect. Calibration lines of the contribution from the scattering of uranium lines to the measured intensity at the different 28 positions versus the uranium peak intensity were plotted by using samples with various uranium concentrations (<2 %) and for which the mass absorption coefficients and concentrations of the various elements were known. The precision of the method was less than 2.5 % at concentrations greater than 50 ppm. With the exception of barium. detection limits varied between 1 and 5 ppm. Accurate results were obtained over large concentration ranges for various unmineralised samples and for uranium ores. The results of the analysis of a number of Karoo uranium ores are given.

Magister Scientiae (Biodiversity and Conservation Biology) / This study aimed to determine the nutrient content (Co, Cr, F, I, Se and V) of various vegetable based food items collected from the Cape Town area in the Western Cape Province of South Africa. This was done to determine which vegetable crops provided the highest concentrations of essential trace elements, and how much they contribute to the daily recommended intake (DRIs) of these trace elements. It also aimed to assess the effects of the addition of the trace elements (Co, Cr, F, I, Se, Si, Sn and V) on seed germination and root growth under controlled conditions in order to calculate their phytotoxicity, and then to biofortify four vegetable crop species, grown in sand culture, with a composite treatment of the trace elements to determine how the addition of these elements will affect the vegetable crops grown under these experimental conditions. From this study, it was shown that trace element content in vegetable crops in the Western Cape Province of South Africa varied between different geographic locations and that certain trace elements were absent from several items collected from some areas. Although some crop species contained sufficient amounts of certain trace elements to satisfy our daily recommended intakes, most of the crops were found to contain insufficient amounts of many of the trace elements to satisfy our needs. Leafy vegetables and tubers were identified as the better vegetable types to biofortify with essential trace elements as they already contain higher concentrations of several of the essential trace elements and should thus be assessed for their effectiveness as crops to be biofortified. When the trace elements were applied directly to cress and lettuce seeds, it was found that all the trace elements, as well as the composite treatments, exerted phytotoxic effects on cress and/or lettuce seeds when applied athighconcentrations. Lettuce was found to be more prone to the effects of these elements. Seed germination was strongly inhibited by fluoride, while several elements affected root growth. When fluoride was left out of the composite treatment, phytotoxicity only occurred at high concentrations. The addition of the trace elements at the high concentrations to already established spinach, cabbage, lettuce and turnip plants were found to affect the uptake of several essential plant nutrients, but the concentrations of the elements affected generally remained higher than the concentrations needed for adequate growth of agricultural crops. Several of the trace elements supplied to the plants were also found to be retained in the roots of the vegetable crops however, as the concentrations supplied to the plants increased, so did the concentrations found in the edible portions of the crops

In 2015, Illinois changed size and harvest limits for catfishes (blue catfish Ictalurus furcatus, flathead catfish Pylodictis olivaris, and channel catfish Ictalurus punctatus) in the Ohio River to match those of neighboring states in order to provide continuity of the regulations and promote a trophy catfish fishery. Regulations imposed a daily limit of one blue catfish or flathead catfish ≥ 35 inches (88.9 cm) and one channel catfish ≥ 28 inches (71.1 cm) per fisher and a 13 inch (33.0 cm) minimum length limit for all species with no bag limit. Although management regulations were implemented, potential efficacy of the implemented regulations and appropriate (i.e. most precise or accurate with fewest samples) monitoring protocols were unknown. Furthermore, there was general lack of understanding of early life movements, natal dispersal timing and principal recruitment sources that aide in determining appropriate spatial scale for monitoring and managing lower Ohio River catfish stocks.To fill these knowledge gaps the following methods were employed: 1) simulation modeling was used to evaluate precision in estimating catch and size distribution metrics for monitoring population trends with increasing sample size (i.e., sampling events), 2) N-mixture modeling was used to estimate size selectivity of multiple gears using detection probability as a robust alternative to size-specific catchability coefficients, 3) otolith microchemistry (Sr:Ca and Ba:Ca) was employed to determine principal recruitment sources, early life movement patterns, and provide fisheries managers with a better understanding of the spatial extent to which management actions should be implemented, 4) Bayesian modeling was used to estimate growth and mortality, 5) Yield-per-recruit modeling was used to estimate and evaluate fishing mortality rates that would result in growth overfishing (FMAX) and yield at FMAX (YPRMAX) for three management scenarios (no regulation, minimum length limit [33.0 cm or greater] and a permissive slot limit [33.0 cm – 88.9 cm; blue catfish and flathead catfish]). The simulation models presented account for the uncertainty associated with heterogeneous selectivity of a gear, and minimize the impact of rare or extreme catch values. Trotlines and low pulse (15-pps) electrofishing generally required the fewer samples to achieve stable values of catch per unit of effort (CPUE), proportional size distribution (quality; PSDQ), and coefficient of variation (CV) than other gears based on simulation modeling. Abundance and detection probabilities were estimated separately for each species of catfish by length category within and across gears, producing a species-gear-size correction for catch bias used in estimating Proportional Size Distribution - Quality (PSDQ). Corrected (i.e., accounting for detection) PSDQ values were lower than uncorrected estimates suggesting a positive bias for larger fish across the entire sampling regime. Managers should use a combination of low pulse electrofishing, trotlines, and high pulse (60-pps) electrofishing in their monitoring efforts for all three species. Based on microchemistry, ictalurid catfishes in the lower Ohio River appear to recruit from multiple sources and make movements across a broad geographic scale. Additionally, some catfish may be originating from outside the portion of the Ohio River that is managed by Illinois (lower 214 km). Fisheries managers should take this into account when implementing management actions. However, most ictalurid catfishes originated from riverine (e.g., Ohio and Mississippi River) natal environments and not from smaller tributaries, and managers should not expect tributaries to compensate for weak year-classes within the river. Based on yield per recruit modeling, catfish stocks are unlikely to benefit from current regulations or a theoretical minimum size limit given the near complete overlap of YPRMAX confidence intervals for all estimable scenarios and the small statistical difference (1 – 3%) based on FMAX between the most permissive and most restrictive scenarios. While statistical differences in FMAX exist, they are likely biologically irrelevant, exceeding the precision of estimation methods for F. While there is some indication that alignment and continuity of management regulations is warranted based on microchemistry, the efficacy of the current permissive slot regulations is questionable based on the models presented and the life history of these fishes. There is no advantage to implementing any of the modeled regulations in terms of increasing FMAX or YPRMAX.

The subsumed work of this dissertation is comprised of three independent but interrelated studies which seek to further the understanding of processes which govern the coprecipitation of trace elements with calcite and aragonite minerals. These studies investigate the effects of: 1) pressure on crystal morphology and trace element incorporation to aragonite; 2) growth rate on uranium partitioning between calcite and fluid; 3) aqueous Mg/Ca on the magnesium partitioning to low-magnesium calcite. The importance of this work is to determine how the environment of formation and growth rate influences the geochemistry of CaCO3 in order to improve existing paleoproxies and develop new ones. In the first study a series of experiments were conducted at 1, 25, 75, 100, and 345 bars of nitrogen – this range covers pressures at the oceanic floor. Aragonite precipitation was induced by the one-time addition of a Na2CO3 solution to an artificial seawater. Results suggest that oceanic floor pressures could affect the crystallization of CaCO3 by altering mineralogical composition and aragonite crystal size. In the second study calcite crystallized from NH4Cl-CaCl2-U solution by diffusion of CO2. The calcite growth rate was monitored by sequential spiking of the calcite-precipitating fluids with REE dopants. The resulting crystals were analyzed using Secondary Ion Mass Spectrometry (SIMS). Results showed that the partitioning of uranium increases with increasing growth rate. Growth entrapment model (GEM) and unified uptake kinetics model (UUKM) explain the obtained data.In the third study CaCO3 precipitated in NaCl solution by continuous addition of CaCl2, MgCl2, and either Na2CO3 or NaHCO3. The Mg/Ca of the fluid was adjusted in an attempt to produce calcite where Mg/Ca would match Mg/Ca in foraminifera shells. It was observed that multiple CaCO3 polymorphs precipitated from fluids at high pH (Na2CO3 doping experiments). This result underscores the potential control of pH and/or supersaturation state on CaCO3 polymorph precipitated from low Mg/Ca solutions. Calcite was the only mineral crystallized at low pH (NaHCO3 doping experiments). It was determined that Mg partition coefficient between calcite and fluid (KMg) negatively correlates with Mg/Ca(Fluid) when it exceeds 0.5 mol/mol; no systematic correlation was observed when 0.05< Mg/Ca(Fluid)<0.5 mol/mol.