School of Earth Sciences - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 4 of 4
  • Item
    Thumbnail Image
    Petrology and geochemistry of mantle xenoliths from the Bultfontein kimberlite (Kimberley, South Africa): new insights into lithospheric mantle fluids
    GIULIANI, ANDREA ( 2013)
    In cratonic areas the lithospheric mantle formed and stabilised during the Archean eon and has been modified through time by multiple episodes of enrichment (i.e. metasomatism) driven by fluids and melts of variable composition. The vast majority of metasomatised mantle rocks are thought to have interacted with basic/ultrabasic and carbonate melts/fluids, variably enriched in alkali and other incompatible elements. The common occurrence of volatile-dominated fluids and brines included in diamonds and mantle silicate minerals suggests that such fluids might be widespread in the Earth’s mantle; however the metasomatising impact of these fluids still require further study. This thesis provides new textural, mineralogical and geochemical data for mantle xenoliths entrained by the Bultfontein kimberlite (Kimberley, South Africa). The studied xenoliths include two mantle polymict breccias, a Ni-mineralised spinel harzburgite and a sulphate-rich MARID (mica-amphibole-rutile-ilmenite-diopside) rock. Mantle polymict breccias are complex mixtures of mantle clasts and minerals cemented together by variable amounts of olivine, phlogopite, ilmenite, rutile, orthopyroxene and sulphides. Groundmass olivine and ilmenite host carbonate-rich inclusions dominated by magnesite, dolomite, alkali-carbonates, phlogopite and kalsilite. These inclusions probably represent an alkali-carbonate melt, which was entrapped during olivine and ilmenite crystallisation in the mantle. This is the first evidence for an alkali-carbonate fluid in the lithospheric mantle above the diamond stability field. The heterogeneous mineralogy and geochemistry of polymict breccias suggest these rocks formed immediately prior to entrainment and transport by kimberlite magmas. Polymict breccias are therefore regarded as failed kimberlite intrusions frozen at depth. The alkali-carbonate melt preserved as inclusions in olivine and ilmenite could be parental to the cementing phases of polymict breccias and could either derive from silicate-carbonate liquid immiscibility of a precursor proto-kimberlite melt or represent a pristine example of primitive kimberlite melt. The Ni-rich spinel harzburgite hosts millimeter-sized mineralised areas that include native nickel, heazlewoodite and Ni-rich silicates (e.g., olivine, phlogopite). The presence of several mineral phases enriched in alkali and volatile species (e.g., phlogopite, phosphates, carbonates, chlorides, djerfisherite) indicates that the transition metal cations were introduced during metasomatism by alkali-rich C–O–H fluids or alkali-carbonate melts. The sulphate-rich MARID sample is traversed by veins dominated by Ba-rich celestine and clinopyroxene, with minor phlogopite, pectolite, sphene, apatite, barite and Sr-Ca carbonates. Celestine hosts the other metasomatic vein phases, but also occurs as inclusions in clinopyroxene, suggesting co-precipitation of these minerals. Celestine was partly replaced by serpentine during alteration by hydrous fluids after kimberlite emplacement in the upper crust. The texture and chemical composition of the metasomatic phases indicate that the MARID rock was infiltrated by a sulphate fluid enriched in Sr, Ba, Na and Ca, with lesser P, Ti, LREE, CO2 and F. A mantle origin for the sulphate fluid is supported by: (i) comparisons between the Sr–S isotopic compositions of celestine, the host kimberlite, crustal and mantle lithologies from the area, and (ii) alteration of celestine by late-stage hydrous fluids. The celestine-bearing veins provide the first evidence for the occurrence of sulphate-dominated fluids in the Earth’s mantle. In summary this thesis provides new insights into the compositions of widespread mantle metasomatic agents, namely alkali-carbonate melts, and documents some of the metasomatic processes occurring in the lithospheric mantle during ascent of primitive or precursor kimberlite magmas. My research also provides unexpected evidence for the occurrence of previously unknown fluids in the Earth’s mantle, including Ni-rich C-O-H fluids and sulphate-dominated fluids.
  • Item
    Thumbnail Image
    Geochemistry and mineralisation of primary and secondary platinum-group elements in the ultramafic "Alaskan-type" Owendale complex and laterites in the Fifield Region, New South Wales, Australia
    Shi, Bielin ( 1995)
    The Owendale Complex belongs to a family of ultramafic-mafic intrusions that is characterised by a zonal, nonstratiform arrangement of the principal ultramafic units. The ultramafic rocks of the Owendale Complex are virtually identical to many of the Alaskan-type intrusions, however the associated gabbroic rocks (wehrlites) are K-rich and Si-undersaturated, in contrast to the tholeiitic gabbroic rocks of the Alaskan examples. The intrusion history of the Owendale Complex is thought to have involved emplacement of a gabbroic intrusion that was invaded by an ultrabasic magma, possibly while the former was still only partly solidified. Emplacement of both magmas probably occurred during Late Devonian tectonism and deformation synchronous with emplacement and crystallisation is necessary to explain the present non-stratiform arrangement of the rock units. The most obvious linkage factor between the two proposed parent magmas (gabbroic and ultrabasic) of the Owendale suites is their mutual affinity with tholeiitic basalt magmas and the similarities of their products with intrusions of alkalic basalt derivation. This suggests the possibility that the Owendale Complex rocks and those of other tholeiitic intrusions of the regions are comagmatic products of an ancestral magma that may have also produced the widespread assemblage of complexes. Viewed from this perspective, the ultramafic rocks of Owendale Complex would thus represent a very minor product of a period of regional magmatic activity. Most alloys, erlichmanite, cooperite and some grains with exclusion texture of Pt-Os-Ir-Pd-Rh are considered to represent a primary high-temperature paragenesis. Concentration of PGE in pegmatoidal units of dunite-wehrlite is explained by the accumulation of platinum-rich alloys that segregated directly from the melt at an early stage in the evolution of the complex. The high-temperature PGM segregate directly from a silicate melt and were not generated by exsolution from spinels or magmatic sulphides. These suggest that fS2 was generally low (subordinate sulphide formation) and, after some influence at the beginning, has given way to rising fO2 (chromite, olivine and Pt-Fe-Cu-Ni alloys formation). After lithification, the ultramafic rocks become subject to "reducing" conditions, i.e., conditions of lower O2 and S2 activities. Ni-Fe alloys, native Fe and Bi formed in cracks which filled the serpentine matrixes. The former PGM (erlichmanite, cooperite and Pt-Fe alloys) were exposed to the reducing conditions via cracks were desulphurated to form porous cooperite with Pt-Fe alloys and multiphase textural Os-Ir-Ni, Pt-Ir aggregates. It is plausible that the veinlets and aggregates of unnamed Rh-Sb-S, (Pt, Ir)2(Fc, Cu)3(S, Sb, AS)3 in the dunites may also have been formed by reduction of Ni-rich sulphides and erlichmanite, Pt-Fe alloys or cooperite. Late PGM are dominated by sperrylite-geversite solid solution resulting from the reaction of early PGM with a fluid phase. A hydrothermal origin is also indicated for native Fe, native Bi and awaruite (NiFe) and the base-metal sulphides (pentlandite, chalcopyrite, sphalerite, arsenopyrite, pyrite, pyrrhotite, and some Ni-Co-Fe sulfide). The cause of the reducing conditions may have been related to H2 production accompanying hydrous alteration of the dunites and clinopyroxenites. The laterites overlying the ultramafic complexes in the Fifield region are exceptionally well-developed and well-preserved weathering profiles. Field, textural and geochemical data all support a chemical weathering origin for the profiles and compatible with meteoric and ground water origins. Meteoric water with intermediate Eh and pH and negligible dissolved species sinks into the laterite where these parameters are modified. The Eh rises and pH decreases to the conditions typical of lateritic soils and the concentration of dissolved species increases. In this state the water is able to take PGE and Au into solution from a finely disseminated form in the bedrock as a part of the process of lateritisation. When the soil solution transports the PGE and Au towards a transitional interface must exist between the ferruginous and saprolite zones with lower Eh, neutral pH and lower concentration of dissolved salts. At this transitional region, deposition of the PGE and Au occurred. The presence of magnetic Pt-Fe-Cu-Ni alloys suggests that hydrothermal solutions play a later role in the Fifield region, and the alloys have grown in situ in a lateritic soil by a process involving laterite water solution in the high Eh, low pH conditions prevalent in such soil, followed by deposition when the conditions become less extreme. Some examples of the Pt-Fe alloys from such an environment become frequently strongly magnetic with larger size. It is assumed that the temperature of the hydrothermal solution is in the range of 300° - 500° C (Bowles, 1990). PGE mineralisation in the primary rocks and laterite in this region has demonstrated a good example of multi-stage process mineralisation including primary high temperature magmatic formation; low temperature postmagmatic hydrothermal alteration and residual lateritic enrichment.
  • Item
    Thumbnail Image
    The geochemistry and petrology of the Enterprise dolerite, Ora Banda, Western Australia
    Gregory, Melissa Joy ( 1998)
    The Enterprise Dolerite was emplaced as an intrusive tholeiitic sill within the Ora Banda Sequence at Ora Banda in the Eastern Goldfields Province of the Yilgarn Craton. The Enterprise Dolerite is now a metamorphic body with modifications in both the mineralogy and geochemistry of the rocks. Careful analysis of petrographic features integrated with geochemical trends have made it possible to interpret the original igneous characteristics of the sill. It is proposed here that the order of crystallisation in the Enterprise Dolerite is plagioclaseolivine- clinopyroxene-quartz. Furthermore, plagioclase and olivine accumulated through crystal settling before a switch to in-situ crystallisation in the remainder of the sill. The bulk chemistry of the Enterprise Dolerite is equivalent to that of the Mt Ellis Sill which occurs at the same stratigraphic position, and it is proposed here that they are continuations of the same intrusive body. This intrusive body is related to the other mafic members of the Ora Banda Sequence, with all members forming a differentiation trend and in which the Big Dick Basalt represents a primary mantle magma. The Enterprise Dolerite/Mt Ellis Sill has evolved in composition along the trend from this primary magma. Finally, the addition and removal of phases has produced a chemically evolving system with differentiation progressing to maxima in silica and iron concentrations which provide very good conditions tor gold deposition. This study proposes that both the Enterprise Dolerite and the Mt Ellis Sill be examined for future potential gold mineralisation.
  • Item
    Thumbnail Image
    Garnet-bearing metabasic rocks at Mount Joel: an investigation into distribution, petrology and equilibrium thermodynamic modelling
    Farrell, Nicole ( 1998)
    Garnet-bearing metabasic rocks at Mount Joel, Yilgarn craton, Western Australia, have been studied to determine their distribution, petrography and mineral equilibria. At depth, the orientation of garnet-bearing rocks is approximately 340°N, dipping 60°-70° to the east and mimics that of chloritoid schist and gold mineralisation. Three mineral assemblages at Mount Joel can contain garnet, including: chloritoid-chlorite-plagioclase-quartz-garnet; chlorite-plagioclase-quartz-garnet; chlorite-hornblende-plagioclase-quartz-garnet. Garnets are manganese-rich, composed of up to 23% spessartine. Bulk rock analysis suggests a correlation between manganese enrichment and the appearance of garnet in mineral assemblages. The chemical relationships are consistent with the garnet-bearing rocks being formed from altered basaltic rocks. Thermodynamic calculations have been undertaken using an internally consistent thermodynamic dataset (Powell and Holland, 1990) and THERMOCALC v2.5. Phase diagrams, including Pressure-Temperature (P-T) Projections, P-T Pseudosections and Temperature-Composition (T-X) Pseudosections, have been used to model the mineral equilibria for FeO-MgO-Al2O3-SiO2-H2O (FMASH), Mn-FeO-MgO-Al2O3-SiO2-H2O (MnFMASH)and CaO-Na2O-MnO-FeO-MgO-Al203-Sí02-H2O (CaNaMnFMASH) systems. Upon addition of manganese to a garnet-free system (FMASH), garnet becomes introduced as a new stable phase. As a result, garnet can be present in low pressure and temperature metabasic rocks, such as those at Mount Joel. The variety of mineral assemblages in garnet-bearing rocks at Mount Joel reflects a range in mineral chemistry of the metabasic rocks, possibly due to a range of alteration processes affecting these rocks. The pressure and temperature conditions of formation of garnet-bearing metabasic rocks at Mount Joel have been constrained to about 510 °C at about 3 kbars.