Subduction zones, such as the Andean convergent margin, are the sites at which new continental crust is generated, and where subducting material is either recycled to the crust via arc magmatism or transferred to the deep mantle. The composition of arc magmas and associated new continental crust reflects variable contributions from mantle, crustal and subducted reservoirs. Insights into crustal growth and recycling processes in the southern Central Andes, specifically in the Pampean flat-slab segment, have been gained by utilising a range of petrological, geochronological and geochemical techniques. These techniques have been applied to a suite of Late Cretaceous (~73 Ma) to Late Miocene (~6 Ma) intrusive (granitoids) and extrusive (basalts to rhyolites) arc rocks collected from an east - west transect across the Andean Cordillera. The oxygen and hafnium isotopic composition of the accessory mineral zircon allows mantle-derived melts contaminated with older, upper continental crustal to be identified. Boron isotopic compositions of melt inclusions, combined with concentrations of certain incompatible trace elements, can be used to assess the source and influence of fluids derived from subducting material on the melt source region. The southern Central Andes provides a particularly interesting area to study these processes as the thickness of the continental crust has increased significantly over the course of the Cenozoic (from ~35 km to >50 km) and the angle of the subducting Nazca plate has shallowed since ~18 Ma, causing the position of the volcanic arc to migrate to the east. In order to unravel the complexities involved with constraining the contributions to arc magmas at an active continental margin, a range of geochronological, geochemical, and geothermobarometric techniques, including high resolution, micro-analysis of mineral phases and melt inclusions, have been applied. High resolution, U-Pb dating of magmatic zircon has improved regional stratigraphy in the Pampean flat-slab segment (between ~29 and 32 °S) and provided an accurate temporal constraint for geochemical and geothermobarometric data. The results of in-situ O and Lu-Hf isotope analysis of zircon show both distinct temporal and spatial variations across the Andean arc. The observed isotopic variability is attributed to variable contamination of mantle-derived melts with distinct Andean basement terranes, which vary east – west in composition and age. ‘Mantle-like’ δ18O(zircon) values, juvenile initial ƐHf(zircon) values and a lack of inherited, xenocrystic zircon cores, suggests the Late Cretaceous (~73 Ma) to Eocene (~39 Ma) plutons located in the Principal Cordillera of Chile, experienced very little interaction with the upper continental crust. Amphibole – plagioclase geothermobarometry indicates these calc-alkaline granitoids, which form extensive north – south trending belts, were emplaced at shallow depths in the crust (~4 – 5 km). Therefore the Late Cretaceous to Late Eocene is interpreted as a period of significant upper crustal growth. The isotopic variability in the Late Oligocene (~26 Ma) to Late Miocene (~6 Ma) arc magmatic rocks demonstrates that during thickening of the continental crust and migration of the Andean arc to the east, arc magmas assimilated Late Paleozoic to Early Mesozoic basement. In addition, arc magmas erupted/emplaced in the Argentinean Precordillera (i.e. farthest east from the trench) assimilated a Grenville-aged (~ 1330 – 1030 Ma) basement. The youngest arc magmas (~6 Ma) erupted in the Frontal Cordillera also show evidence for the assimilation of this ancient basement terrane, potentially signalling under-thrusting beneath the Frontal Cordillera. Overall, the later part of the Cenozoic represents a period of crustal reworking. Boron concentrations and isotope ratios measured in pyroxene hosted melt inclusions and for the first time in zircon hosted melt inclusions, are higher than the values expected for the mantle wedge and show significant variations with time. The source of the Paleocene (~61 Ma) arc magmas were influenced by fluids primarily derived from altered oceanic crust. Lower δ11B values and boron concentrations obtained for Oligocene (25 – 23 Ma) arc magmatic rocks reflects a diminished influence of slab-derived fluids reflecting a greater depth to the top of the slab. Fluids derived from serpentinite influenced the source of the arc magmas after ~19.5 Ma. This has been linked with the intersection of the Juan Fernández Ridge, a volcanic seamount chain associated with hydrated and serpentinised oceanic lithosphere.

Oxidized hydrous intermediate composition magmas are responsible for porphyry copper (Cu ±Mo ±Au) deposits and epithermal Au ore deposits formed globally in the shallow crust (Sillitoe, 2010; Seedorff et al., 2005). Recently, zircon geochemistry has been used to characterize both productive and barren intrusions associated with porphyry Cu-Au ore deposits. Zircon composition differs slightly between the two intrusive groups, and researchers have proposed that zircon in productive intrusions has crystallized from a relatively more oxidized melt compared to barren intrusions (Ballard et al., 2002; Muñoz et al., 2012). Zircon rare earth elements record anomalies in Ce and Eu contents that allow estimation of the ratio of oxidized versus reduced species, i.e. Ce⁴⁺/Ce³⁺ (Ce[superscript IV]/Ce[superscript III]) and Eu³⁺/Eu²⁺ (Eu/Eu*)[subscript CN]. This study focuses on understanding the compositions of Eocene magmas associated with sediment hosted Carlin gold deposits and the gold-copper ores of the Battle Mountain porphyry Cu-Au-skarn district in northern Nevada. Zircon trace element composition was analyzed using LA-ICP-MS and SHRIMP-RG to determine differences between mineralizing and non-mineralizing intrusions in northern Nevada and to compare these compositions with known porphyry Cu-Au type magmas. These zircon and rock compositional data was then used to test the hypothesis of a magmatic origin of the Carlin type gold deposits (Muntean et al., 2011). Zircon U-Pb ages were calculated using multiple SHRIMP-RG spot analyses of each sample for two Carlin biotite porphyry dikes, two Battle Mountain porphyry dikes and the granodiorite of the Copper Canyon stock. The new U-Pb age dates for Carlin porphyry dikes are 38.7 ± 0.5 Ma and 38.8 ± 0.4 Ma. The age of the Copper Canyon stock is 38.0 ± 0.7 Ma, and the age of the Battle Mountain porphyry dikes are 40.2 ± 0.4 Ma and 41.3 ± 0.4 Ma. The Carlin dike ages are the same age, within uncertainty, with previous studies conducted (Mortensesn et al., 2000). The productive porphyry dikes from the Battle Mountain district have Ce(IV)/Ce(III) ratios of 500 to 10000 and a wide range of (Eu/Eu*)[subscript CN] values between 0.3 and 0.7 respectively. Carlin porphyry dikes have Ce(IV)/Ce(III) values between 100 and 1000, and a more limited (Eu/Eu*)[subscript CN] range of 0.5 to 0.7. Barren Eocene intrusions at Harrison Pass and Caetano have much lower Ce(IV)/Ce(III) ratios that range from 20 to 500, and have a very large span of (Eu/Eu*)[subscript CN] from 0.03 to 0.6. Calculated Ce(IV)/Ce(III) and (Eu/Eu*)[subscript CN] of zircon of this study illustrate a distinction between productive and barren intrusions in northern Nevada, and demonstrate a geochemical link between porphyry type magmas and dikes associated with Carlin type gold deposits. These ratios may provide a useful means of evaluating potentially economic geologic terranes and serving as a method to infer relative oxidation state of zircon bearing intrusive rocks. / Graduation date: 2013