M.Sc. / Please refer to full text to view abstract

The Kalahari Goldridge mine is located within the Archaean Kraaipan Greenstone Belt about 60 km SW of Mafikeng in the Northwestern Province, South Africa. Several gold deposits are located within approximately north - south-striking banded iron-formation (BIF). Current opencast mining operations are focused on the largest of these (D Zone). The orebody is stratabound and hosted primarily in the BIF, which consists of alternating chert and magnetite-chloritestilpnomelane-sulphide-carbonate bands ranging from mm to cm scale. The ore body varies in thickness from 15 to 45 m along a strike length of about 1.5 km. The BlF is sandwiched between a sericite-carbonate-chlorite schist at the immediate footwall and carbonaceous meta-pelites in the hanging-wall. Further west in the footwall, the schists are underlain by mafic meta-volcanic amphibolite. Overlying the hanging-wall carbonaceous metapeiites are schist units and meta-greywackes that become increasingly conglomeratic up the stratigraphy. Stilpnomelane-, chlorite- and minnesotaite-bearing assemblages in the BlFs indicate metamorphic temperatures of 300 - 450°C and pressures of less than 5 kbars. The BIF generally strikes approximately 3400 and dips from 60 to 75°E. Brittle-ductile deformation is evidenced by small-scale isoclinal folds, brecciation, extension fractures and boudinaging of cherty BIF units. Fold axial planes are sub-parallel to the foliation orientation with sub-vertical plunges parallel to prominent rodding and mineral lineation in the footwall. Gold mineralization at the Kalahari Goldridge deposit is associated with two generations of subhorizontal quartz-carbonate veins dips approximately 20 to 40°W. The first generation consists of ladder vein sets (Group lIA) preferentially developed in Fe-rich meso bands, whilst the second generation consists of large quartz-carbonate veins (Group lIB), which crosscut the entire ore body extending into the footwall and hanging-wall in places. Major structures that control the ore body are related to meso-scale isoclinal folds with fold axes subparallel to mineral elongation lineations, which plunge approximately 067°E. These linear structures form orthogonal orientation with the plane of the mineralized shallowdipping veins indicating stretching and development of fluid - focusing conduits. A second-order controlling feature corresponds to the intersection of the mineralized veins and foliation planes of host rock, plunging approximately 008°N and trending 341°. G0ld is closely associated with sulphides, mainly pyrite and pyrrhotite and to a lesser extent with bismuth tellurides, and carbonate gangue. The ore fluid responsible for the gold deposition is in the C-O-H system with increased CH₄ contents attributed to localized hydrolysis reaction between interbedded carbonaceous sediment and ore fluid. The fluid is characterized by significant C0₂ contents and low salinities below 7.0 wt % NaCl equivalent (averages of 3.5 and 3.0 wt % NaCl equivalent for the first and second episodes of the mineralization respectively) . Calculated values of f0₂. ranging from 10⁻²⁹·⁹⁸ to 10⁻³²·⁹⁶ bars, bracket the C0₂-CH₄ and pyrite-pyrrhotite-magnetite buffer boundaries and reveal the reducing nature of the ore fluid at deposition. Calculated total sulphur content in the ore fluid (mΣs), ranges from 0.011 to 0.018M and is consistent with the range (10⁻³·⁵ to 10⁻¹M) reported for subamphibolite facies ore fluids. The close association of sulphides with the Au and nature of the fluid also give credence that the Au was carried in solution by the Au(HS)₂ - complex. Extensive epigenetic replacement of magnetite and chlorite in BIF and other meta-pelitic sediments in the deposit by sulphides and carbonates, both on meso scopic and microscopic scales gives evidence of an interaction by a CO₂- and H₂S-bearing fluid with the Fe-rich host rocks in the deposit. This facilitated Au precipitation due to changes in the physico-chemical conditions of the ore fluid such as a decrease in the mΣs and pH leading to the destabilization of the reduced sulphur complexes. Local gradients in f0₂ may account for gold precipitation in places within carbonaceous sediments. The fineness of the gold grams (1000*Au/(Au + Ag) ranges from 823 to 921. This compares favourably with the fineness reported for some Archaean BIFhosced deposits (851 - 970). Mass balance transfer calculations indicate that major chemical changes associated with the hydrothermal alteration of BIF include enrichment of Au, Ag, Bi, Te, volatiles (S and CO₂), MgO, Ba, K and Rb but significant depletion of SiO₂ and minor losses of Fe₂O₃. In addition, anomalous enrichment of Sc (average, 1247%) suggests its possible use as an exploration tool in the ferruginous sediments in the Kraaipan greenstone terrane. Evidence from light stable isotopes and fluid inclusions suggests that the mineralized veins crystallized from a single homogeneous fluid source during the two episodes of mineralization under the similar physicochemical conditions. Deposition occurred at temperatures rangmg from 350 to 400°C and fluid pressures ranging from 0.7 to 2.0kbars. Stable isotope constraints indicate the following range for the hydrothermal fluid; θ¹⁸H₂O = 6.65 to 10.48%0, 8¹³CΣc = -6.0 to -8.0 %0 and 8³⁴SΣs = + 1.69 to + 4.0%0 . These data do not offer conclusive evidence for the source of fluid associated with the mineralization at the Kalahari Goldridge deposit as they overlap the range prescribed for fluid derived from devolatization of deep-seated volcano-sedimentary piles near the brittle-ductile transition in greenstone belts during prograde metamorphism, and magmatic hydrothermal fluids. / KMBT_363 / Adobe Acrobat 9.54 Paper Capture Plug-in

The Makganyene Formation forms the base of the Postmasburg Group in the Transvaal Supergroup in the Griqualand West Basin. It consists of diamictites, sandstones, banded iron-formations (BIFs), shales, siltstones and carbonates. It is generally accepted that the Makganyene Formation rests on an erosive regional unconformity throughout the Northern Cape Province. However this study demonstrates that this stratigraphic relationship is not universal, and conformable contacts have been observed. One of the principal aims of this study is to identify the nature of the Makganyene basal contact throughout the Griqualand West Basin. Intensive fieldwork was carried out from Prieska in the south, to Danielskuil in the north. In the Sishen and Hotazel areas, only borehole material was available to assess the stratigraphy. The Griquatown Fault Zone delimits the boundary between the deep basin and platform facies. The Koegas Subgroup is only present south of the Griquatown Fault Zone, where it pinches out. However, the transition Griquatown BIFs-Koegas Subgroup occurs in lacustrine deposits on the Ghaap platform (Beukes, 1983). The Griquatown Fault Zone represents the edge of the basin, which corresponds to a hinge rather than a fault zone. The Makganyene Formation rests with a conformable contact on the Koegas Subgroup south of the Griquatown Hinge Zone, and north of it the Makganyene Formation lies unconformably on the Asbestos Hills Subgroup. The Makganyene Formation displays lateral facies changes that reflect the paleogeography of the Griqualand West Basin, and the development of ice sheets/shelves. The Ghaap platform is characterised by coarse immature sand interbedded with the diamictites. The clasts in this area contain local Asbestos Hills material and no dropstones are present. Such settings are typical of sediments that are being deposited below a grounded ice mass. At the Griquatown Hinge Zone, the sandstone lenses are smaller, and the clasts consist of chert, of which a great number are striated and faceted. In the Matsap area, the presence of dropstones is strong evidence for the presence of a floating ice shelf that released its material by basal melting. Further south, the Makganyene Formation contains stromatolitic bioherms that only form if clastic contamination is minimal and therefore the ice that transported the detritus to the basin did not extend far into open sea conditions. The base of the Hotazel Formation also contains diamictite levels. Dropstones have been identified, implying a glacial origin. The Hotazel diamictites are interbedded with hyaloclastites and BIFs. The Makganyene glacial event, therefore, was not restricted to the Makganyene Formation, but also included the Ongeluk Formation, through to the base of the Hotazel Formation. Petrographic studies of the Makganyene Formation and the base of the Hotazel Formation reveal mineral assemblages that are diagnostic of early to late diagenetic crystallisation and of low-grade metamorphism not exceeding the very low green-schist facies. The facies identified display the same sense of basin deepening, from shallow high-energy Hotazel area on the Ghaap platform, to the deep basin in the Matsap area. Whole-rock geochemical analyses reveal that the elemental composition of the Makganyene Formation is very similar to that of the Asbestos Hills BIFs, which were the most important source of clastic detritus for the Makganyene Formation. However, minor amounts of carbonates of the Campbellrand Subgroup, as well as a felsic crustal input from the Archean granitoid basement, made contributions. On the Ghaap platform, the Makganyene diamictite is enriched in iron, calcium, and magnesium, while in the deeper parts of the basin the diamictites are enriched in detrital elements, such as titanium and aluminium, which occur in the fine clay component. The Hotazel diamictite displays a distinct mafic volcanic input, related to the extrusion of the Ongeluk basaltic andesites, which was incorporated in the glacial sediments. Sequence stratigraphy is based on the recognition of contacts separating the different systems tracts that compose a depositional sequence. However, because the basal contact of the Makganyene Formation has not been properly identified in previous work, no correct model has been proposed so far. Therefore correlations between the Griqualand West and the Transvaal basins, based on lithostratigraphic similarities and extrapolations of unconformities, have to be reviewed, especially since the publication of new radiometric ages contradict all previously proposed correlations. It is proposed here that the Transvaal Supergroup in the Griqualand West Basin represents a continuous depositional event that lasted about 200 Ma. The Makganyene glacial event occurred during changing conditions in the chemistries of the atmosphere and ocean, and in the continental configuration. A Snowball Earth event has been proposed as the causative process of such paleoenvironmental changes. However, evidence presented here of less dramatic glacial conditions, with areas of ice-free waters, implies an alternative to the Snowball Earth event. The paleoenvironmental changes are thought to represent a transition from an anaerobic to aerobic atmosphere, that was responsible for the global cooling of the surface of the Earth, Such a glacial event may have aided in the large-scale precipitation of iron and manganese in areas of intense upwellings.

In the Middelplaats mine area of the Kalahari manganese field, two drill holes (MP53 and MP54) intersected anomalously high-grade manganese ore sitting stratigraphically just above an igneous body (likely a dike or sill). Manganese ore located within approximate 5 meters of the contact with the underlying igneous rocks has been substantially metasomatically upgraded from 25 percent manganese, to over 40 percent whilst the dominant manganese species within the ore has been altered to hausmannite. This report demonstrates the metasomatic alteration is related to devolatilization (removal and/or remobilization of H₂O, CO₂ and CaO) due to contact metamorphism caused by the underlying igneous rocks. The Middelplaats mine is situated in the southwest corner of the Kalahari manganese field where the paleo basin shallows out and ends. Within the mine area, several stratigraphic units pinch out or are truncated by the side of the basin. This pinching out of lithological formations has led to the underlying Ongeluk Formation being in contact with the much younger units of the Hotazel Formation. Therefore, geochemical investigation into the nature and source of the igneous rocks was also undertaken to see if the rocks from the two drill holes were related to one another and/or the underlying Ongeluk Formation. Results of these geochemical studies have demonstrated that the Middelplaats igneous rocks (dolerites) from the two drill holes (MP53 and MP54) share a co-genetic source region. There is also reasonable geochemical evidence that the source region of the Middelplaats igneous rocks was substantially similar to the source region of the Ongeluk Formation. This may indicate that the source region of the Ongeluk Formation was reactivated at some later stage resulting in the emplacement of doleritic dikes or sills in the Middelplaats mine area. The Middelplaats igneous rocks were also found to have undergone a slight but pervasive potassic alteration; with most of the original plagioclase feldspar showing some level of replacement by a potassium enriched feldspar. Although no source for this potassic fluid was found, the devolatilization reaction within the manganese ore appears to have released some potassium into the surrounding rocks. This additional potassium may be responsible for some localized potassic alteration.

A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy. Johannesburg, 2016. / In current Middle Stone Age research there is interest in understanding whether climatic and environmental factors played a role in behaviours related to subsistence, mobility patterns and material culture production. From a palaeoenvironmental perspective, the southern Cape is recognized as an important study region for exploring the link, if any, between past environmental conditions and key MSA occurrences. The research presented in this thesis aimed to contribute to the existing database of past environments in the southern Cape through the geochemical analysis of speleothems from a previously uninvestigated locality in the De Hoop Nature Reserve. Together, the De Hoop speleothems provide a discontinuous record of environmental change from marine oxygen isotope stage MIS 5a to MIS 3 (and the Holocene). Results from the De Hoop records indicate warm summer rain and C4 vegetation in early MIS 5a (c. 85 ka to 80 ka) but more variability by late MIS 5a (c. 79 ka to 74 ka). At Klasies River main site, also on the southern Cape coast, the upper MSA II is associated with the warm early MIS 5a conditions. At Blombos Cave, another important coastal MSA site, the Still Bay occurring within terminal MIS 5a was linked to warm but more variable late MIS 5a conditions. While early MIS 4 (c. 73 ka to 68 ka) was comparatively cooler, conditions were similar to those in early MIS 5a. From this research, the earlier phase of the Howiesons Poort at Klasies River main site and the Howiesons Poort at Klipdrift Shelter were correlated with the early MIS 4 conditions in De Hoop. By late MIS 4 (c. 67 ka to 60 ka), conditions remained cool, but were seemingly more variable than during the earlier part of this iii stage / GR2016

Rustenburg Platinum Mines, Union Section, is located in the mafic phase of the north-western sector of the Bushveld Complex. This part of the Complex is characterised by transgressions of Upper zone ferrogabbros across the lower sequence of mafic rocks. These transgressions have effectively isolated a roughly triangular segment of Lower, Critical and Main zone rocks. It is in the upper part of the Critical zone that the Merensky Reef, with a strike length of 9,5 km, is found to suboutcrop below a 2-3 m black turf soil cover and it constitutes the orebody being mined at Union Section. Underground mining and development have exposed a stratigraphic succession from the Middle Group chromitites to within the lower part of the Main zone gabbros. A detailed investigation of a 100 m stratigraphic succession was undertaken in the upper part of the Critical zone, which includes the Bastard Reef, Merensky Reef, Pseudo Reef, UG 2 and UG 1 layers. Stratigraphic descriptions are presented for both a normal and potholed succession, as well as a description of pothole characteristics. The variations of Merensky Reef thickness, pothole distribution and structure of the are body are also discussed. Forty seven whole-rock major- and trace-element analyses were carried out on selected samples from both normal and potholed successions, although more emphasis was placed on the former. These samples were taken from just below the UG 1 to above the Bastard Reef. Trace elements determined included Sr, Rb, Y, Nb, Zr, Cr, Co, Ni, Cu and V. The cyclical nature of the layering is clearly defined by breaks in the trends of both the major- and trace element analyses, although some elements are strongly influenced by the modal proportions of the mai n mineral phases. Although major breaks are observed for individual elements , insignificant variation of the Mg/Mg+Fe 2 + ratio is observed throughout. New electron microprobe data are presented for the main silicate phases olivine , orthopyroxene, clinopyroxene and plagioclase , while only a limi ted number of analyses were carried out on the spine l phase. There would appear to be a reversal of the expected fractionation trend from the hanging wal l of the UG 2 to the Pseudo Marker layer. There is an upward increase of the Fo content of the olivine from F0₇₆ , ₇₋₈₁,₆ . A similar trend is observed for the orthopyroxene analyses . The Merensky unit, as a whole, exhibits a gradual decrease in the Mg end member of the Ca-poor pyroxenes from En ₇₈,₅₋₇₄ . was found to be markedly more calcic Cumulus plagioclase (An ₇₂, ₈-₈₁, ₇) than the intercumulus phase (An₅₆,₃₋₇₁, ₉) ' A further feature of the cumulus plagioclase is that strong zonation was observed with the cores of individual crystals being consistently lower in Ca relative to their margins . / KMBT_363 / Adobe Acrobat 9.53 Paper Capture Plug-in

The subject of BIF genesis, particularly their environmental conditions and ocean chemistry at the time of deposition and their evolution through time, has been a subject of much contentiousness, generating a wealth of proposed genetic models and constant refinements thereof over the years. The prevailing paradigm within the various schools of thought, is the widespread and generally agreed upon depositional and diagenetic model(s) which advocate for BIF deposition under anoxic marine conditions. According to the prevailing models, the primary depositional environment would have involved a seawater column whereby soluble Fe²⁺ expelled by hydrothermal activity mixed with free O₂ from the shallow photic zone produced by eukaryotes, forming a high valence iron oxy-hydroxide precursor such as FeOOH or Fe(OH)₃. An alternative biological mechanism producing similar ferric precursors would have been in the form of photo-ferrotrophy, whereby oxidation of ferrous iron to the ferric form took place in the absence of biological O₂ production. Irrespective of the exact mode of primary iron precipitation (which remains contentious to date), the precipitated ferric oxy-hydroxide precursor would have reacted with co-precipitated organic matter, thus acting as a suitable electron acceptor for organic carbon remineralisation through Dissimilatory Iron Reduction (DIR), as also observed in many modern anoxic diagenetic environments. DIR-dominated diagenetic models imply a predominantly diagenetic influence in BIF mineralogy and genesis, and use as key evidence the low δ¹³C values relative to the seawater bicarbonate value of ~0 ‰, which is also thought to have been the dissolved bicarbonate isotope composition in the early Precambrian oceans. The carbon for diagenetic carbonate formation would thus have been sourced through a combination of two end-member sources: pore-fluid bicarbonate at ~0 ‰ and particulate organic carbon at circa -28 ‰, resulting in the intermediate δ¹³C values observed in BIFs today. This study targets 65 drillcore samples of the upper Kuruman and Griquatown BIF from the lower Transvaal Supergroup in the Hotazel area, Northern Cape, South Africa, and sets out to explore key aspects in BIF carbonate petrography and geochemistry that are pertinent to current debates surrounding their interpretation with regard to primary versus diagenetic processes. The focus here rests on applications of carbonate (mainly siderite and ankerite) petrography, mineral chemistry, bulk and mineral-specific carbon isotopes and speciation analyses, with a view to obtaining valuable new insights into BIF carbonates as potential records of ocean chemistry for their bulk carbonate-carbon isotope signature. Evaluation of the present results is done in light of pre-existing, widely accepted diagenetic models against a proposed water-column model for the origin of the carbonate species in BIF. The latter utilises a combination of geochemical attributes of the studied carbonates, including the conspicuous Mn enrichment and stratigraphic variability in Mn/Fe ratio of the Griquatown BIF recorded solely in the carbonate fraction of the rocks. Additionally, the carbon isotope signatures of the Griquatown BIF samples are brought into the discussion and provide insights into the potential causes and mechanisms that may have controlled these signatures in a diagenetic versus primary sedimentary environment. Ultimately, implications of the combined observations, findings and arguments presented in this thesis are presented and discussed with particular respect to the redox evolution and carbon cycle of the ocean system prior to the Great Oxidation Event (GOE). A crucial conclusion reached is that, by contrast to previously-proposed models, diagenesis cannot singularly be the major contributing factor in BIF genesis at least with respect to the carbonate fraction in BIF, as it does not readily explain the carbon isotope and mineral-chemical signatures of carbonates in the Griquatown and uppermost Kuruman BIFs. It is proposed instead that these signatures may well record water-column processes of carbon, manganese and iron cycling, and that carbonate formation in the water column and its subsequent transfer to the precursor BIF sediment constitutes a faithful record of such processes. Corollary to that interpretation is the suggestion that the evidently increasing Mn abundance in the carbonate fraction of the Griquatown BIF up-section would point to a chemically evolving depositional basin with time, from being mainly ferruginous as expressed by Mn-poor BIFs in the lower stratigraphic sections (i.e. Kuruman BF) to more manganiferous as recorded in the upper Griquatown BIF, culminating in the deposition of the abnormally enriched in Mn Hotazel BIF at the stratigraphic top of the Transvaal Supergroup. The Paleoproterozoic ocean must therefore have been characterised by long-term active cycling of organic carbon in the water column in the form of an ancient biological pump, albeit with Fe(III) and subsequently Mn(III,IV) oxy-hydroxides being the key electron acceptors within the water column. The highly reproducible stratigraphic isotope profiles for bulk δ¹³C from similar sections further afield over distances up to 20 km, further corroborate unabatedly that bulk carbonate carbon isotope signatures record water column carbon cycling processes rather than widely-proposed anaerobic diagenetic processes.

MESMEG / Department of Mining and Environmental Geology / Literature indicated that several mineralogical identification studies have been carried out on clays but few have focused on the characterisation of geophagic materials from South Africa. Large quantities of earth materials are consumed daily in Mashau Village, however, their mineral content and geochemical compositions had not been determined. Moreover, though the consumption of geophagic materials is very common in the village, the associated health implications had not been addressed. Thus, the main aim of the research was to mineralogically and geochemically characterise geophagic materials commonly ingested in Mashau Village and infer on possible health implications that could result from their consumption. Questionnaires were administered to geophagists in the study area with the aim of generating data on the prevalence of geophagia and the motivations for the practice. Geophagic soils and their parent rocks (for determination of provenance) were sampled and analysed for mineralogical and geochemical content. Geophagic soil samples were subjected to the following physicochemical analyses: colour, particle size distribution, pH, cation exchange capacity (CEC) and electrical conductivity (EC). An x-ray diffractometer (XRD) was used for mineralogical analysis while major oxides and trace elements abundances were determined using x-ray fluorescence (XRF) spectrometry and laser ablation inductively coupled mass spectrometry (LA-ICP-MS), respectively. Furthermore, provenance of the geophagic materials was determined using data obtained from mineralogical and geochemical analysis. Inferred health implications were based on the physico-chemical, mineralogical and geochemical data obtained. Outcomes of the questionnaire survey revealed craving to be the motivation for geophagia in Mashau Village. Although the practice seemed to be prevalent in females of certain ages, it was certainly not limited to gender, age, educational level or socio-economic status. Out of the 20 geophagic samples, 3 samples were brown, 2 had a strong brown colour and another 2 had a light olive brown colour. Other soil colours were less common, as each colour was only observed in one sample. The sand fraction dominated the samples; the clay content was low, giving the samples a sandy clay loamy texture. The pH of the soil ranged from being slightly acidic (5.4) to being slightly alkaline. The CEC values were very high ranging from 17 t0 109 meq/100 g. vii The EC values were also high (ranging from 11.2 to 245 μS/cm) indicating a high amount of soluble salts. Mineralogical analysis of geophagic soils identified quartz, microcline, plagioclase, hornblende, dolomite, muscovite, kaolinite, smectite, talc, anatase, hematite, ilmenite, chlorite and epidote with quartz and kaolinite being the dominant minerals. Actinolite, augite, chlorite, epidote, forsterite, magnetite, muscovite, plagioclase, quartz, sepiolite and microcline were the minerals identified in rock samples. Geochemical analysis for major oxides content (SiO2, TiO2, Al2O3, Fe2O3, MnO, MgO, CaO, Na2O, K2O, P2O5 and Cr2O3) indicated that both geophagic soils and parent rocks were mainly composed of silica and alumina. Trace elements geochemistry showed a depletion of LREEs and an enrichment of HREEs in geophagic soils. The results also revealed that the REEs were enriched in the bulk fraction than in the clay fraction. Relative to the Upper Continental Crust (UCC) compositions, the concentrations of trace elements in geophagic soils were generally low. Provenance determination results showed that geophagic soils in Mashau were derived from basalts and sandstones. Majority of the samples were formed as a result of intense weathering while some were as a result of intermediate weathering. The negative health implications of the studied materials could include perforation of the colon, damage of the dental enamel and anaemia. However, geophagic materials could also be a good source of mineral nutrients and beneficial for reduction of nausea during pregnancy. / NRF

MENVSC / Department of Ecology and Resource Management / Historically, mining activities have generated vast quantities of abandoned tailings dumps in several regions of South Africa and throughout the world. The management and disposal of huge volumes of tailings dumps has constituted a major challenge to the environment. The current study aims to establish the physicochemical properties and mineralogical characterization of the old copper tailings dump in Musina, to reveal the mobility patterns and attenuation dynamics of potentially toxic or heavy metal species as a function of depth, with a view of assessing their potential environmental impact with respect to surface and ground water systems. This information is crucial in the beneficial utilization of copper tailings in the development of sustainable construction materials as part of reuse approach management system. About twelve tailings samples were collected into polyethylene plastic bags from three established tailings profiles drilled by a hand auger. The collected tailings samples were characterized using standard analytical procedures i.e., X-ray fluorescence (XRF), X-ray diffraction (XRD) and scanning electron microscopy/energy dispersive spectroscopy (SEM-EDS). The transfer of potentially toxic or heavy metal species from tailings to water was evaluated using the standardized batch leaching test (EN 12457) and speciation-equilibrium calculations on the aqueous extracts performed by MINTEQA2. The leachate concentration of cations in the collected tailings samples was determined by inductively coupled mass spectrometry (ICP-MS) and the leachate concentration of anions was determined by ion chromatography (IC). A modified sequential extraction scheme was applied on the selected tailings samples of the drilled tailings profiles to further understand the mode of occurrence, the geochemical partitioning and distribution, real mobility, and environmental bioavailability of potentially toxic or heavy metal species in the tailings and tailings-soil interface. The extracted fractions or phases from sequential scheme were as follows: (F1) water-soluble fraction, (F2) exchangeable fraction, (F3) carbonate fraction, (F4) iron and manganese hydroxide associated fraction, (F5) organic matter and secondary sulphide associated fraction, (F6) primary sulphide bound fraction, and (F7) residual or silicate fraction. The results obtained from the seven steps sequential extraction scheme were validated by the determination vi of percentage recoveries from pseudo-total digestion or total metal content of the original sample. The distribution of major elements and potentially toxic or heavy metal species in different leachate fractions obtained after each step of sequential extraction of the selected tailings samples was determined by inductively coupled plasma mass spectrometry (ICP-MS). The appraised data was used to reveal the impact of atmospheric oxygen and infiltrating rain-water on the chemistry of copper tailings dump by depth profiles. Macroscopic properties revealed that the abandoned Musina copper tailings are fine to medium coarse grained, and range in color from light/dark gray at the upper or shallow depth of the tailings, to dark reddish-brown at the deeper zone where the tailings are mixed with the underlying soil or soil-interface. The drilled respective tailings profiles were uniform and slightly varied in both mineralogical and bulk chemical compositions with tailings depth. Mineralogical analysis showed the following order of mineralogical composition within the respective tailings profiles: quartz> epidote> chlorite> muscovite> calcite> hematite. Chalcopyrite was the only sulphide mineral observed by optical microscopy, although not identified or quantified by XRD and SEM-EDS analysis. The observed discrete chalcopyrite grains were attributed to the primary mined ore (i.e., chalcopyrite, chalcocite and bornite) during past copper mining activities in Musina. The tailings profiles were characterized by a medium alkaline pH (7.97-8.37) that corresponds very well with the tailings leachates or pore-water pH (8.36-8.46). This pH was constant and slightly varied with tailings depth in the respective tailings profiles. The high abundance of alumino-silicate minerals and traces of carbonates as calcite coupled with low sulphide mineral content, suggested a high neutralization capacity of the tailings which was in common agreement with an alkaline nature of the copper tailings dump. The chemical composition of major elements within the respective tailings profiles followed the order: Si>Al>Fe>Ca>Mg>K>Na, and corresponds very well with the mineralogical composition of the tailings, whereby alumino-silicates were the most abundant minerals in the tailings samples. Nevertheless, the solid-phase concentration of metals decreases with increasing tailings depth as Cu>Sr>Zr>Ni>Zn and was incongruent with the mineralogical composition within the respective tailings profiles. The main secondary minerals were calcite and hematite, and their proportion increased with increasing tailings vii depth. In addition, hematite formed coatings on the rims and corners of chlorite as observed from optical microscopy, and retained relatively high amounts of potentially toxic or heavy metals (up to 862 ppm of Cu, up to 36 ppm of Ni, and up to 25 ppm of Zn) at the upper and shallow depth of the respective tailings profiles, where bulk density was high and low porosity. Based on batch leaching tests, the amounts of potentially toxic or heavy metal species released into solution were low (0.27-0.34 μg/L Pb, 0.54-0.72 μg/L Ni, 0.88-1.80 μg/L Zn, and 20.21-47.9 μg/L Cu) and decreases with increasing tailings depth, indicating that, presently, the tailings have a minor impact on heavy metals load transported to the receiving surface and groundwater systems. The low concentration of potentially toxic or heavy metal species in solution is primarily due to their retention by secondary Fe oxide phases (i.e., hematite) and the prevailing medium alkaline pH condition of the tailings leachate or pore-water. The observations are consistent with MINTEQA2 speciation calculations, which predicted the precipitation of secondary phase cuprite (Cu2O) as the main solubility-controlling mineral phase for Cu, Zn, and Ni. Primary factors influencing aqueous chemistry at the site are neutralization and dissolution reactions as a function of pH, precipitation, and sorption into hydrous oxides (hematite and cuprite). Based on sequential extraction results, the leachable concentration of potentially toxic or heavy metal species in the water-soluble, exchangeable and carbonate fractions of the respective tailings profiles was relatively low, except for Cu and Mn. For instance, the leachable concentration of Cu and Mn reached 10.84 mg/kg and 321.7 mg/kg at the tailings-soil interface (3 m) in tailings profile C, respectively. The low concentration of potentially toxic or heavy metal species (Cr, Co, Ni, Zn, Cd, and Pb) in these fractions could be due to the low solubility of minerals bearing these trace elements caused by variations in pore-water pH in the respective tailings profiles. The high concentration of Cu and Mn in these fractions suggests their high mobility and therefore most available for uptake in the environment. Except for Cu>Mn>Cr, the contents of potentially toxic or heavy metal species in the Fe and Mn oxides and organic matter or sulphides bound fractions was low, due to the low viii quantity of these fractions in the tailings, despite their high affinity and sorption capacity for potentially toxic or heavy metal species. Likewise, the residual fraction of the respective tailings profiles contained the highest proportion of potentially toxic or heavy metal species. Although the highest potentially toxic or heavy metal species content was in fractions with limited mobility, care must be taken since any geochemical change or shift in the tailings pH or acidic conditions may cause them to be displaced to more mobile fractions, thereby increasing their mobility and environmental bioavailability. Therefore, physicochemical properties of the tailings including pH and mineralogical composition of the tailings samples were the main substrate controlling the geochemical partitioning and distribution, potential mobility, and environmental bioavailability of potentially toxic or heavy metal species by tailings depth. The knowledge of mobility and eco-toxicological significance of tailings is needed when considering tailings dump disposal or reuse in the environment. The addition of copper tailings at 3 and 28 days successfully improved the compressive strength of cement mortar mixtures incorporating tailings at C5 (5%) and C10 (10%) respectively, although with small margin relative to the control mixture (C0). The maximum strength was 31.15 Mpa attained after 28 curing days, and slightly varied when compared with other compressive strength on copper blended cement mortars mixtures in other countries, used for the development of sustainable construction materials. The chemical composition, physical properties and improved compressive strength on cement mortars mixtures incorporating copper tailings, implies that copper tailings are suitable for the development of sustainable construction materials, thereby ensuring job creation, availability of land for development usage, and the reduction of environmental pollution induced by the abandoned copper tailings dumps.

PhDENV / Department of Ecology and Resource Management / Kaolin dominated soils are common in the tropical and subtropical regions. People depend on kaolin-rich soils for agricultural production of food and fiber. The most popular of all South African soils is the Hutton form which accounts for the marvelous redness of the landscape across the Country. The apedal (structureless) soils in the group are characterised by a relatively low CEC (< 11 cmolc kg-1) reflecting oxidic mineralogy with predominantly kaolinitic assemblage. The geochemical and mineralogical composition of soil kaolin has significant implications on soil fertility, geochemical exploration and engineering properties. Despite the dominance of kaolin in these soils, little is known of their properties in the medium. The nature of kaolin minerals in soils varies with parent material, degree of weathering and pedogenic environment. Most studies conducted in South Africa on kaolins are limited to reference kaolins with little or no publication on soil kaolins, hence, this study. This research involved the evaluation of mineralogical and geochemical characteristics of oxidic soils and soil kaolins developed from four (4) selected parent rocks which were basalt, granite, arkosic sandstone, and gneiss. Soils developed from quartzite were selected as control. Representative soil samples collected from profiles developed from the different parent rocks were analysed for physico-chemical, mineralogical, and geochemical data. The mineralogical and geochemical data obtained by x-ray diffractometry (XRD), x-ray fluorescence (XRF), and laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) were used in unraveling the influence of the provenance and degree of weathering on the soil characteristics. The mineralogical and geochemical data for soil kaolins were determined through XRD, scanning electron microscopy (SEM), Fourier transform infrared (FTIR) spectroscopy, thermogravimetric analysis and differential scanning calorimetry, XRF, and LA-ICPMS to establish their mineralogical and geochemical properties with respect to their parent rocks. Comparison between the soil kaolins and selected reference kaolins were also conducted. The phosphorus (P) sorption data acquired photometrically were employed to evaluate the relationship between the P sorption capacities of the soils and soil kaolins. The influence of soil properties on the fertility of the soils were assessed based on the physico-chemical (pH, particle size distribution, and electrical conductivity (EC)) and chemical (organic matter (OM) content, cation exchange capacity (CEC), available P, exchangeable cations (Ca, K, Mg, Na, and Al), and P sorption) data. The mineralogical and geochemical data for the parent rocks were obtained by XRD, optical microscopy, XRF, and LA-ICPMS. Multivariate statistical analyses were also conducted. Results showed that the dominant colour in the studied bulk soils was dusky red (31 %) followed by brown (23 %), reddish brown, yellowish red, and yellowish brown (23 %) as well as strong brown, dark brown, reddish grey, very dark greyish brown, and dark red. Soil textures were clayey to sandy loamy with OM contents between 0.41 and 4.76 %. The pH, EC, CEC, exchangeable cations, and available P values generally ranged from 5.22 to 8.38, 10.25 to 114.40 μS/cm, 2.93 to 18.30 cmol/kg, 0.03 to 13.92 cmol/kg, and <0.01 to 54.99 mg/kg, respectively. Kaolinite and quartz were the dominant phases for soils developed from basalt whereas, quartz and plagioclase were the dominant mineral phases in soils developed from granite, arkosic sandstone, and gneiss, respectively. Other minerals present in the soils were microcline, muscovite, hematite, goethite, montmorrillonite, anatase, gibbsite, chlorite, and actinolite. Geochemical compositions of the bulk soils show relative enrichment of Fe2O3, TiO2, CaO, K2O, MgO, MnO, and Na2O (except for CaO, K2O, MgO, MnO, and Na2O in soils developed from basalt). Chemical index of alteration (CIA), chemical index of weathering (CIW), and plagioclase index of alteration (PIA) values varied between 54.92 and 99.81 % which suggest low to high degree of chemical weathering. The ACN-K and A-CNK-FM diagrams for the different soils also support these observations. Trace elements were generally enriched in soils developed from basalt and gneiss (except for Rb, Sr, and Ba in soils developed from basalt), but were depleted in soils developed from granite and arkosic sandstone (except for Cr and Ta). The principal factors responsible for the mineralogical and geochemical characteristics of the soils were the parent rocks and degree of weathering. In the soil kaolins, the dominant clay mineral was kaolinite accounting for 23 to 85 wt % followed by montmorrillonite, chlorite, and gibbsite. The non-clay minerals like quartz, plagioclase, muscovite, microcline, anatase, goethite, hematite, and actinolite accounted for the remaining percentages. The soil kaolins were characterised by thin platy kaolinite particles with partially to poorly-ordered structural order. The platy kaolinite crystals have their longest dimension sizes between 0.06 and 0.25 μm. The dehydroxylation temperatures for the studied soil kaolins ranged from 425 to 475 ˚C. The SiO2/Al2O3 ratio was lowest in soil kaolins developed from basalt and higher in soils developed from granite, arkosic sandstone, and gneiss which is consistent with their mineralogy since the former have more kaolinite. Higher Fe2O3 and CEC values were obtained relative to reference kaolins which could be attributed to the presence of more structural iron in the soil kaolins as well as their smaller crystal sizes. The presence of weatherable and accessory minerals accounted for the enrichment of Co, Ni, Cu, Zn, and Pb in the soil kaolins. The kaolinite in the soils were formed by leaching and desilication of the primary minerals in the parent rocks under suboxic conditions. H-type P adsorption isotherms obtained for both the soils and soil kaolins indicated their high affinity for phosphorus by chemisorption. The average maximum P adsorption values were in decreasing order of soils developed from basalt > granite > arkosic sandstone > quartzite (control) > gneiss, respectively whereas, for soil kaolins is basalt > granite > quartzite (control) > arkosic sandstone > gneiss, respectively. Relative to other soils developed from different parent rocks, soils developed from basalt (with more clay content) had higher capacity and buffer power for P adsorption. The standard P requirements for the soils ranged from 7.78 to 92.91 mgP/kg and were classified as low based on the Langmuir model. Significant correlation between the P adsorption parameters for the soils and soil kaolins indicated that the later could be taken as a good predictor for P sorption dynamics in the soils. Electrical conductivity of the soils were taken to be negligible in interfering with plant growth. The available P values were generally below the critical level of 12 – 15 mg/kg for soils developed from basalt, gneiss, and quartzite (control) but higher in soils developed from granite and arkosic sandstone. All the soil evaluation factor (SEF) average values estimated were greater than five indicating that they are not of poor soil fertility. The correlation results between the soil properties and P sorption parameters suggest that several variables can influence the P sorption dynamics of the soil. Regression analyses further indicated that CEC, pH, OM, and clay content in the soils account for 99 % bounding P energy variation whereas, Fe2O3 accounts for 76 % P sorption maximum variation in the soils. In addition, variations in Fe2O3 and sand contents in the soils account for 96 % and 95 % maximum buffering capacity and external P requirement (EPR) variations, respectively. Models to advance the interplay between the various soil properties and P sorption parameters in the soils were developed. Mineralogical and geochemical characteristics of the soils were principally controlled by the parent rocks and degree of weathering. The soil kaolins displayed significant differences relative to reference kaolins. Langmuir model is most suited for describing P sorption in soils and soil kaolins developed from different parent rocks within the studied area. P sorption parameters for the soils can readily be obtained from the P sorption parameters of the kaolins present in them. EPR obtained and models for predicting P sorption parameters from selected soil properties developed for the various soils will improve the efficiency of routine P fertilizer applications. Iron oxide (Fe2O3) played the most crucial role in explaining the P sorption dynamics of the soils. The major contributions from this study have been: better understanding of the influence of parent rock characteristics and degree of weathering on the soil characteristics, the nature of soil kaolins and its influence on soil properties as well as P sorption dynamics in soils have been better established, and improvement of the understanding on the relationship between soil properties and P sorption dynamics in the soils. / NRF