School of Earth Sciences - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 6 of 6
  • Item
    Thumbnail Image
    Gabbro magmatism in the Lachlan Orogen, southeastern Australia: implications for mafic–felsic associations and granitoid petrogenesis
    Whelan, Joanne Amy ( 2016)
    The nature and role of mafic endmembers in granitoid petrogenesis is poorly constrained in the Lachlan Orogen of southeastern Australia. Most previous studies have focussed on intermediate–felsic magmas and attempted to model the nature of their primitive precursors. Ordovician–Devonian magmatic rocks of gabbroic composition are exposed as rare volumetrically minor intrusions throughout the Lachlan Orogen, and are temporally and spatially associated with more voluminous granitoids. A detailed study of tholeiitic and alkali gabbros exposed in the Kuark Zone of eastern Victoria provides new insights into gabbro petrogenesis and the source regions of such magmas which in turn have implications for the generation of granitoids, particularly I- and A-type magmas, throughout the Lachlan Orogen. The Arte Igneous Complex and Scrubby Flat Gabbro are poorly exposed mafic intrusions spatially associated with A-, I- and S-type magmas. Although major- and trace-element variations do not always show clear evidence for a geochemical link between these units, variations in Sr, Nd and Hf isotope compositions indicate a shared source for at least some of the magmas. Alkali gabbros in particular preserve considerable Sr-Nd isotopic heterogeneity ranging greater than 10 epsilon Nd units within a small geographic area interpreted to represent the root of the Arte intrusion. It is proposed herein that magmatic differentiation occurred via fractional crystallisation and cumulate processes; however, it is argued that much of the Sr-Nd isotopic variation was inherited from a heterogeneous source region. A model involving a small degree (<20%) of partial melting of greenstone basement can explain the variation within the alkali gabbro of the Arte Igneous Complex. Subsequent higher degrees of partial melting (>30%) can explain the more voluminous, less heterogeneous tholeiitic gabbros. Spatially associated A- I- and S-type granitoids and the gabbros is more cryptic, some geochemical and Sr-Nd isotope links are apparent. Importantly, the I-type intrusions are interpreted to have been derived magmas fromed by partial melting of more intermediate compositions within greenstone basement (c.f. ultramafic to mafic for the gabbros), thus they share a similar heterogeneous source region with the gabbro rocks. In contrast, the S-types intrusions have a more complex link to the gabbros and are interpreted to have been derived via partial melting and assimilation of Ordovician turbidites by tonalite magmas of the Arte Igneous Complex. Comparison of the magmatic rocks of the Kuark Zone with other twelve of the 20 known Lachlan Orogen gabbros reveals similar isotopic heterogeneity. This requires that the source heterogeneity is present on a local- and regional-scale. Cambrian greenstone basement exposed in rare fault-bounded belts throughout the Lachlan Orogen have the same isotopic heterogeneity. Moreover, the same heterogeneity is observed in I-type granitoids of the Lachlan Orogen. The implication is that I-type magmas may also be generated by partial melting of greenstone basement rocks, thus both gabbroic and I-type magmas image their source region. There is a correlation between the age of gabbros and Sr-Nd isotope values, with younger gabbros characterised by on average more isotopically juvenile compositions. The ca. 380 Ma Bingie Bingie Suite that approachs compositions of depleted mantle. A number of the gabbros exhibit arc-like trace-element characteristics, however, given that the greenstones were generated in a Cambrian arc environment, these signatures may be inherited from their source (greenstones) rather than the gabbro magmas themselves being generated in a subduction zone setting. The chemical characteristics of most of the gabbros are consistent with the partial melting of greenstone basement in a back arc basin setting under extension. Influx of new mantle-derived magma is only likely to have occurred to produce the youngest mafic rocks (e.g., Mount Buller Igneous Complex). The results of this study provide new insights into the source regions for mafic intrusions of gabbro/diorite composition in the Lachlan Orogen. In light of this new information, these insights present an opportunity to re-examine the petrogenetic models for I- and A-type granitoids in particular.
  • Item
    Thumbnail Image
    The petrogenesis of Brazilian kimberlites and kamafugites intruded along the 125° lineament: improved geochemical and geochronological constraints on magmatism in Rondonia and the Alto Paranaiba Igneous Province
    Felgate, Matthew Richard ( 2014)
    Despite a number of studies, the origin of ultrapotassic magmatism in Brazil remains contentious. In particular, problems relate to the timing of Cretaceous kimberlite and kamafugite magmatism within the Alto Paranaiba Igneous Province (APIP), the origin and relationship between the Cretaceous kimberlites and kamafugites, and the geochemical nature of the Permo-Triassic kimberlite magmatism. Many of these uncertainties arise as a consequence of either the implementation of now superseded analytical techniques or the use of highly altered samples. This study seeks to address these issues by performing state-of-the-art geochemical and geochronological analyses on a number of fresh kimberlite and kamafugite drill core samples situated along the 125° lineament. The first part of the thesis provides a comprehensive geochronological study of Brazilian kimberlites and kamafugites. This is built around a combination of the most recent U/Pb perovskite, Rb-Sr phlogopite, and Ar-Ar phlogopite dating techniques that together provide 33 age estimates, representing 22 separate intrusions, of which 11 are kimberlites and 11 kamafugites. The results are in broad agreement with previously published data, identifying a Permo-Triassic age for the Amazonian kimberlites, and kimberlites and kamafugites of Cretaceous age in the APIP. The new data from this study also support an age decrease from NW to SE within the APIP. However, data from this study suggest that this is not a continuous younging, instead indicating that two distinct magmatic episodes may exist. The data also reveal the presence of an “older” mid Cretaceous cluster (~88 Ma) encompassing Goias and the NW APIP (Southern Goiás and NW Minas Gerais) and a slightly younger Cretaceous cluster (~80 Ma) in the Central/Southern APIP. The importance of these new findings is discussed with reference to previously proposed petrogenetic models and casts doubt on the applicability of the plume-related hypothesis for ultrapossic magmatism in the region. In the second part of the thesis, an in-depth geochemical study of 15 kimberlites and 16 kamafugites from the APIP (27 samples) and Rondonia/Mato Grosso (4 samples) is presented. Bulk rock major and trace element data, together with Nd and Hf isotope determinations are combined with perovskite trace element, Sr and Nd isotope analyses in order to provide a comprehensive geochemical dataset. The elemental and isotopic similarity of the APIP kimberlite to South African transitional kimberlites, the APIP kamafugites to Ugandan and Chinese kamafugites and, for the first time, the Amazonian kimberlites to a mixture of South African Group I and transitional kimberlites is established. The overlapping range in Sr isotopic signatures identified between the APIP kimberlite and kamafugite magmas during previous bulk rock studies is confirmed by in situ perovskite analysis, with APIP kimberlites ranging between 70496 and 0.70596 and APIP kamafugites ranging between 0.70499 and 0.70574. Modelling suggests that this range can not result from contamination and must represent a heterogeneous source component, common in the formation of both rock types. It is argued that the combination of geochemical results is best explained by derivation of both rock types from the heterogeneous subcontinental lithospheric mantle. Kimberlites are believed to have been sourced from carbonated garnet lherzolite whereas the kamafugites were generated from phlogopite and clinopyroxene rich vein assemblages, with small volume MARID type metasomes indicated by the kamafugite Hf signatures. Finally, the geochronological and geochemical information is combined into a petrogenetic model: it is argued that the Cretaceous magmatism is likely not the result of plume heating but is instead a consequence of adiabatic decompression melting of the heterogeneous SCLM caused by extension. In parallel with the study of Brazilian ultrapotassic magmatism the thesis also investigates technical aspects of the U-Pb perovskite dating technique. Using a subset of Indian and Brazilian kimberlite samples, in situ perovskite dating was performed and combined with isotope dilution analysis of the co-magmatic, low uranium phase, titanomagnetite, in an attempt to improve precision of the Tera-Wasserburg regression. The inclusion of the titanomagnetite phase improved precision in all cases and enabled ages to be determined for the Indian samples without recourse to an assumed Stacey-Kramers common Pb value. The investigation also highlights some of the shortcomings associated with using an assumed Stacey-Kramers Pb value for young unradiogenic perovskite samples, whilst also highlighting the circularity of the commonly employed 207Pb correction method.
  • Item
    Thumbnail Image
    Geochronology and geochemistry of Cenozoic volcanism in relation to epithermal gold mineralisation in western Java, Indonesia
    Titisari, Anastasia Dewi ( 2014)
    Western Java in Indonesia is a well known gold deposit district that contains a number of world class ore deposits. The district hosts mainly low-sulphidation epithermal gold deposits and the most important gold deposits occur in the Pongkor, Cibaliung, Cikotok, and Papandayan districts. Although Java has a long record of volcanic activity, little is known of its pre-Pleistocene arc history, particularly in relation to the timing and geochemical evolution of the volcanic successions associated with ore mineralisation. Moreover, much of the available geochronological data for volcanic rocks in the region are based on imprecise K-Ar results. Therefore a combination of 40Ar/39Ar dating techniques and elemental geochemical methods (major and trace element analyses) have been utilised in the current project. Most of volcanic samples from the current study are characterised by enriched LILE and LREE compositions, which are characteristic of calc-alkaline arcs. However, Papandayan basaltic samples exhibit depleted LREE contents, typical of island arc tholeiites. The more enriched LILE and LREE compositions present in some Pongkor samples and Papandayan intrusive rocks, are indicative of high-K calc-alkaline and shoshonite arcs. Trends in Nb/Y, Th/Nb, Ce/Yb, and Ce/La(N) ratios reflect temporal evolution of the arcs, from: i) a primitive arc (low Nb/Y, Th/Nb, Ce/Yb and high Ce/La(N) ratios) characterised by the tholeiite basaltic samples; ii) an evolved arc (high Nb/Y, Th/Nb, Ce/Yb and low Ce/La(N) values) typified by the high K – shoshonite volcanic samples; and iii) a mature arc (with intermediate Nb/Y, Th/Nb, Ce/Yb and Ce/La(N) values) represented by the calc-alkaline volcanic samples. Trace element signatures of La and Ce suggest a broadly similar magma source for all rocks in the region. 40Ar/39Ar dating of volcanic rocks that host gold mineralisation are representative of the western Java magmatic arc, give the oldest age of ca. 18 Ma for the Papandayan district; ages ranging from ~11 Ma to ~9.5 Ma for the Cibaliung district, from ~18 Ma to ~4.5 Ma for the Cikotok district, and the youngest ages from 2.74 ± 0.03 Ma to ca. 2 Ma for the Pongkor district. Adularia crystallisation that is associated with western Java gold mineralisation shows 40Ar/39Ar ages from the oldest to most recent: ca. 18 Ma for the Papandayan district; from 12.44 ± 0.19 Ma to 9.39 ± 0.75 Ma for the Cibaliung district; 5.36 ± 0.46 Ma and 3.43 ± 0.04 Ma for the Cikotok district; and from 2.02 ± 0.03 Ma to 1.80 ± 0.03 Ma for the Pongkor district. Some host volcanic rocks have been affected by hydrothermal alteration, which has resulted in partial resetting of ages towards the time of gold mineralisation. Variation of the mineralisation ages suggests multiple generation of adularia growth in the mineralised veins. The dating results suggest that the magmatic arc across western Java are most likely linked to Southeast Asia tectonic evolution, from Early Miocene counter clock wise rotation of Kalimantan to Late Miocene – Pliocene subduction of the Eurasian continental plate beneath the Indo-Australian oceanic plate. The new and existing age data allow for a reconstruction of the western Java magmatic arc, with three main events identified: an Early Miocene primitive tholeiite arc (20 – 18 Ma), a Middle Miocene mature calc-alkaline arc (13 – 9 Ma) and a Late Miocene – Pliocene evolved high-K calc-alkaline and shoshonitic arc (7 – 2 Ma). The reconstruction indicates that the Early Miocene Papandayan low sulphidation epithermal system (with some indications of high sulphidation activity) is related to a basement comprising thinned island arc crust. In contrast the Miocene – Pleistocene low-sulphidation epithermal mineralisation system of the Cibaliung, Cikotok and Pongkor districts, which is associated with a calc-alkaline arc, was constructed on Sundaland continental crust.
  • Item
    Thumbnail Image
    The geochemistry and geochronology of Tanzanian kimberlites
    Chesler, Rachel ( 2012)
    The Tanzanian Craton is host to numerous kimberlite intrusions, although few have been studied in any detail. Due to its proximity to the East African rift zone, this region also provides a unique opportunity to investigate the effect of rifting on kimberlite volcanism. To date, however, no geochemical studies have been conducted on Tanzanian kimberlites, largely due to their altered and weathered nature, and only four intrusions have published age constraints. This thesis seeks to redress these deficiencies by the application of modern high precision analytical techniques to the study of kimberlites from this previously neglected region. The study focused on 30 separate kimberlite intrusions located on the Tanzanian Craton. Due to their highly altered and weathered character, perovskite was used as a proxy to obtain geochemical and geochronological information on the host kimberlites. Perovskite is a common accessory mineral in kimberlites and contains substantial U, Sr and Nd and lesser amounts of Pb and Hf. Consequently, perovskite provides the means to determine both age constraints and initial isotopic signatures for the host kimberlites. Perovskite extracted from the Tanzanian kimberlites was utilised for U-Pb geochronology and analysed for trace element and Sr-Nd-Hf isotopic compositions. Similar to kimberlites worldwide, these samples show extreme enrichment in large ion lithophiles and LREE, whereas they are heavy REE depleted. They are characterised by moderately radiogenic 87Sr/86Sri compositions (0.70268 – 0.70476) and ɛSr values (-14.56 to 4.46), restricted 143Nd/144Ndi compositions (0.51138 to 0.51277) and ɛNd values (1.8 to 3.8), and very restricted 176Hf/177Hfi compositions (0.282058 to 0.282966) and ɛHf values (0.8 to 8.2). All samples plot within the range previously defined for Group I kimberlites. This study provides compelling evidence for the utility of perovskite analysis in highly altered kimberlitic samples that would otherwise be intractable to study. The ages obtained for these samples define two periods of kimberlite volcanism for the Tanzanian Craton. Early Mesoproterozoic kimberlites include the Itanana kimberlite (1083 Ma) and the previously dated Bubiki kimberlite (1097Ma). All other kimberlites for which ages were obtained are significantly younger, with ages ranging from 44-80 Ma. The latter group includes some of the youngest reliably dated kimberlites known; the Kikhomango (44 Ma) and X073 (47 Ma) intrusions. Ages from both the older and younger Tanzanian kimberlites coincide with periods of known kimberlite activity worldwide. Previous studies of the Tanzanian kimberlites have focused on the thermobarometry of garnet macrocrysts derived from disaggregated mantle xenoliths. Garnet compositions vary with proximity to the East African Rift, suggesting that the SCLM in this region was affected by rift related heating prior to kimberlite emplacement. If the source of these kimberlites was within the SCLM, systematic geographic variation in the isotopic and trace element compositions in relation to their proximity to the rift would be expected. This has not been observed in this study, with the entire region displaying no significant geographic variation in geochemical composition. Therefore the current results suggest a deeper source for the Tanzanian kimberlite magmas, below the influence of rifting in the region.
  • Item
    Thumbnail Image
    The timing and origin of orogenic gold mineralisation in the western Lachlan Orogen, southeast Australia: constraints from 40Ar/39Ar dating and halogen and noble gas geochemistry
    Fairmaid, Alison Maree ( 2012)
    The Ballarat East gold deposit (408t) is the second largest orogenic gold deposit in the Western Lachlan Orogen, southeast Australia. The western Lachlan Orogen is characterised by a thick package of Ordovician turbiditic sedimentary rocks overlying Cambrian oceanic volcanic sequences. The region was variably affected by multiple major deformation/metamorphism and magmatism events during the Cambrian to Devonian. The Ballarat East gold deposit is located in the Bendigo structural zone of the Western Lachlan Orogen and is hosted in Ordovician sediments of the Castlemaine Supergroup. Gold mineralisation in the Ballarat East deposit is sited in quartz and quartz-carbonate veins within goldfield-scale, west-dipping reverse faults. Two major lode types are present: 1) lode type ‘1’ is characterised by arsenopyrite-dominated quartz veins associated with early movement on reverse faults, whereas 2) lode type ‘2’ is related to structurally later, shallow east-dipping, pyrite-sphalerite-galena-white-mica dominated veins, emanating from reverse faults. Previous studies have suggested that gold mineralisation in the Western Lachlan Orogen occurred at ~440Ma, as a result of metamorphic devolatilisation reactions in the lower crust. However the age of mineralisation at the Ballarat East deposit is only broadly constrained to a period between 460 and 370 Ma, and the source of the gold-bearing fluids could include metamorphosed volcanic rocks, sedimentary rocks and/or granites. In order to provide a more robust chronological framework for gold mineralisation at the Ballarat East deposit, several samples of detrital and hydrothermal potassium-rich minerals were collected and analysed by 40Ar/39Ar dating. In addition, fluid inclusions in portions of quartz and quartz-carbonate veins were characterised by micro-thermometry and halogen/noble gas isotopic tracer methods to further constrain the source(s) of the gold mineralising fluids. The 40Ar/39Ar data obtained from detrital muscovite grains yield ages between 530 – 460 Ma and are concordant with previously published detrital ages. The vein muscovite/sericite ages fall into three age groupings as follows: 445 – 435 Ma (lode type ‘1’), 420 – 415 Ma (lode type ‘2a’) and 380 – 370 Ma (lode type ‘2b’). The gold-bearing quartz veins (from both lode types) contain low salinity (average 4 wt.% NaCl eq.) aqueous H2O inclusions and mixed H2O-CO2 fluid inclusions. Fluid inclusion 40Ar/36Ar values range from 322 (close to Air Saturated Water; ~296) up to a maximum of 4503, and 40Ar/36Ar is strongly correlated with Cl/36Ar. Fluid inclusions have variable Br/Cl values between 1.66 10-3 and 2.91 × 10-3 and I/Cl values between 153 × 10-6 and 501 × 10-6, with a strong correlation between Br/Cl and I/Cl. The fluid inclusion 84Kr/36Ar and 129Xe/36Ar values are variable but show a systematic enrichment in the heavier noble gases. The 40Ar/39Ar ages suggest gold mineralisation at the Ballarat East deposit occurred in three main episodes at ca. 445 Ma, ca. 420 Ma and ca. 380 – 370 Ma. All episodes of mineralisation are associated with fluid inclusions of similar composition. This fluid is suggested to reflect a deeply sourced fluid, possibly originating by devolatilisation of altered volcanic rocks (e.g. basalts). In this scenario, the fluid would have acquired additional noble gases and organic Br plus I by interaction with sedimentary rocks, including organic-rich shales that are found beneath and surrounding the deposit. The data are compatible with genetic models for orogenic Au in which gold mineralisation was initiated by metamorphic devolatilisation in the lower crust, linked to Lachlan Orogenesis at ca. 440 Ma.
  • Item
    Thumbnail Image
    Geochronology of Timor-Leste and seismo-tectonics of the southern Banda Arc
    Ely, Kim Susan ( 2009)
    Arc–continent collision is a significant plate boundary process that results in crustal growth. Since the early stages of evolution are often obscured in mature orogens, more complete understanding of the processes involved in arc–continent collision require study of young, active collision settings. The Banda Arc presents an exceptional opportunity to study a young arc–continent collision zone. This thesis presents aspects of the geology and geochronology of Ataúro and the Aileu Complex of Timor-Leste, and the tectonics of the Banda Arc. U–Pb dating of detrital zircons from the Aileu Complex by LA-ICPMS show major age modes at 270–440 Ma, 860–1240 Ma and 1460–1870 Ma. The youngest zircon populations indicate a maximum depositional age of 270 Ma. The detrital zircon age populations and evidence for juvenile sediments within the sequence favours a synorogenic setting of deposition of sediments sourced from an East Malaya – Indochina terrane. Previous uncertainty in aspects of the cooling history for the Aileu Complex is resolved with 39Ar/40Ar geochronology of hornblende. Cooling ages of 6–10 Ma are established, with the highest metamorphic grade parts of the Complex yielding the older ages. Cooling ages of 10 Ma imply that metamorphism of the Aileu Complex must have commenced by at least ~12 Ma. Metamorphism at this time is attributed to an arc setting rather than the direct result of collision of the Australian continent with the Banda Arc, an interpretation consistent with the new provenance data. Geological mapping of Ataúro, an island in the volcanic Banda Arc north of Timor, reveals a volcanic history of bi-modal subaqueous volcanism. 39Ar/40Ar geochronology of hornblende from dacitic lavas confirms that volcanism ceased by ~3 Ma. Following the cessation of volcanism, coral reef marine terraces have been uplifted to elevations of 700 m above sea level. Continuity of the terraces at constant elevations around the island reflects regional-scale uplift most likely linked to sublithospheric processes such as slab detachment. North of Timor, the near complete absence of intermediate depth seismicity beneath the inactive segment of the arc is attributed to a slab window that has opened in the collision zone and extends to 350 km below the surface. Differences in seismic moment release around this slab window indicate asymmetric rupture, propagating to the east at a much faster rate than to the west. If the lower boundary of this seismic gap signifies the original slab rupture then the slab window represents ~4 m.y. of subsequent subduction and implies that collision preceded the end of volcanism by at least 1 m.y. Variations in seismic moment release and stress state across the transition from subduction of oceanic crust to arc–continent collision in the Banda Arc are investigated using earthquake catalogues. It is shown that the slab under the western Savu Sea is unusual in that intermediate depth (70–300 km) events indicate that the slab is largely in down-dip compression at this depth range, beneath a region of the arc that has the closest spacing of volcanoes in the Sunda–Banda arc system. This unusual state of stress is attributed to subduction of a northern extension of the Scott Plateau. Present day deformation in the Savu Sea region may be analogous with the earliest stages of collision north of Timor.