The European Kupferschiefer Province contains multiple sediment-hosted stratiform copper (SSC) deposits and has been mined for many centuries. The mineralized rocks of the Kupferschiefer are stratabound and hosted by a stratigraphic succession of Late Permian terrestrial sandstones (Rotliegend Formation) and transgressive marine mudstones and limestones (Zechstein Formation). The formation of the Kupferschiefer deposits has been described by various genetic models, which primarily differ in terms of the timing of ore stage sulfide formation relative to host rock deposition (i.e. syn-genetic vs. epigenetic). In this study, samples from two drill cores in the Spremberg-Graustein Kupferschiefer deposit (eastern Germany) are described. Reflected light and scanning electron microscope (SEM) petrography has been used to determine key paragenetic relationships and in situ sulfur isotope values of pyrite and chalcopyrite (secondary ion mass spectrometry; SIMS) have been generated to determine pathways of sulfide formation. The extensive replacement of carbonate and feldspar by ore stage sulfides provides evidence that hydrothermal activity post-dated the formation of diagenetic phases in all units. The highly negative δ34S values of pyrite (–41.9‰ to –35.7‰) and chalcopyrite (–38.9‰ to –34.5‰) indicate that reduced sulfur was generated via open system organoclastic sulfate reduction (OSR). The indistinguishable δ34Schalcopyrite values preserved in the Rotliegend sandstones and Zechstein mudstones suggest a common origin of sulfides in distinct lithologies. To reconcile the petrographic evidence of host rock replacement with the isotopic evidence of open system sulfate reduction requires an external source of bacteriogenic sulfur, most likely in the form of a low temperature sulfur rich brine. The infiltration of low temperature brines transporting bacteriogenic sulfur is a key feature of genetic models in other sediment hosted systems (e.g., Irish Zn Ore Field). If this model is applied to the Kupferschiefer district, exploration programs should target subbasins with evidence of low temperature brine circulation from marginal sedimentary facies.

The German Research Centre for Geosciences (GFZ) in Potsdam hosts a CAMECA 1280-HR large geometry secondary ion mass spectrometer (SIMS) with a web-based user node at the University of the Witwatersrand, South Africa. A major theme of our facility is high-precision, high-accuracy, high-spatial resolution analyses of light isotope ratios in a variety of natural and experimental materials.The latest analytical developments from the GFZ SIMS laboratory focus on the development, assessment and use of new reference materials for stable isotope analysis. Particularly for oxygen, our repeatability from 15-µm diameter domains is now typically better than ±0.15‰ (1s). However, the total uncertainty on such analyses is commonly larger because of significant differences (in some cases more than one ‰) among the isotope ratios of reference materials reported by multiple, highly regarded gas source mass spectrometry laboratories. This issue of interlaboratory bias during reference material characterization inevitably impacts all in situ data employing such materials and must be duly considered.

Abstract The origin of Italian kamafugites and lamproites is a matter of debate, not least due to their “crustal signature” displayed by trace element compositions and isotopic ratios, but also due to puzzling geodynamic significance. We combine in situ EMPA and LA‐ICP‐MS analyses with in situ analyses of oxygen isotopes (SIMS) on olivine from the Pleistocene San Venanzo kamafugites and Torre Alfina lamproites. Lamproitic olivine shows extremely high Mg# and Ni concentrations whereas Ca and Mn concentrations are low. Their δ18OV‐SMOW values are very high up to +11.5 ‰. In kamafugites we recognize three genetically different olivine groups: (a) phenocrystic one with high Mg#, very low Ni, high Ca and Mn. Values of δ18OV‐SMOW are up to +10.9 ‰; (b) melt‐related xenocrystic grains that compositionally resemble lamproitic olivine; (c) skarn‐related almost pure forsterite of extreme δ18OV‐SMOW ∼27 ‰, with negligible amounts of minor and trace elements. The melting and crystallization conditions of Italian kamafugites and lamproites indicate compositionally heterogeneous mantle sources on very small scales. Distinct geochemical features of the olivine macrocryst populations observed in kamafugite point to a range of processes occurring both within the magma storage and transport system. We suggest that the diversity of metasomatic agents was involved in mantle processes on local scales, coupled with magma mixing and/or the uptake of xenocrysts during magma ascend.

Using olivine composition, we observe small-scale mantle heterogeneity resulting from the recycling of the different sediment compositions

Here we report on the oxygen isotope compositions of four proposed apatite reference materials (chlorapatite MGMH#133648 and fluorapatite specimens MGMH#128441A, MZ‐TH and ES‐MM). The samples were initially screened for 18O/16O homogeneity using secondary ion mass spectrometry (SIMS) followed by δ18O determinations in six gas source isotope ratio mass spectrometry laboratories (GS‐IRMS) using a variety of analytical protocols for determining either phosphate‐bonded or “bulk” oxygen compositions. We also report preliminary δ17O and Δ’17O data, major and trace element compositions collected using EPMA, as well as CO32− and OH− contents in the apatite structure assessed using thermogravimetric analysis and infrared spectroscopy. The repeatability of our SIMS measurements was better than ± 0.25 ‰ (1s) for all four materials that cover a wide range of 103δ18O values between +5.8 and +21.7. The GS‐IRMS results show, however, a significant offset of 103δ18O values between the “phosphate” and “bulk” analyses that could not be correlated with chemical characteristics of the studied samples. Therefore, we provide two sets of working values specific to these two classes of analytical methodologies as well as current working values for SIMS data calibration.

The Singhbhum Craton, one of five major Archean cratons in the Indian subcontinent, contains abundant well-preserved Paleoarchean supracrustals and granitoids. This study presents zircon U-Pb ages and whole rock geochemistry of tonalite-trondhjemite-granodiorites (TTGs) and granites from the Bonai Granite Complex (BGC) and Older Metamorphic Tonalite Gneiss (OMTG), which are separated from each other by the Jamda-KoiraNoamundi Iron Ore Group (IOG) supracrustals. Emplacement ages obtained in this study indicate that a major episode of TTG magmatism took place in BGC around 3368 +/- 8 Ma (1 sigma), followed by granitic magmatism around 3331 +/- 33 Ma (1 sigma). In contrast, a TTG from the Deo Nala area representing OMTG yielded crystallization age of 3312 +/- 8 Ma (1 sigma). The emplacement and evolution of the BGC were coeval with granitoid magmatism from the central part of the Singhbhum Craton. Whole rock geochemical data identify both high- and low-HREE TTGs in both the BGC and OMTG to the west and east of the IOG basin, respectively. The trace element systematics of high-HREE Bonai TTG are similar to those of Icelandic dacites, suggestive of their derivation from a garnet free, plagioclase rich amphibolite. The low-HREE TTGs of the BGC and OMTG were derived from an amphibolite source with varying amounts of garnet. The potassic granites of the BGC were sourced from the older TTGs which had undergone partial melting at a shallow depth. The evolution of the BGC and OMTG can be attributed to the partial melting under a thickened mafic crust. Dome and keel structures and emplacement ages of granitoids from the west and east of the Jamda-Koira-Noamundi IOG basin, support the origin of these Paleoarchean granitoids in a stagnant lid regime. High geothermal gradients induced by heat supplied by mantle upwelling appear to have induced the melting of the thickened crust, to form the TTG. Delamination induced mafic-ultramafic underplating resulted in melting of early formed TTGs, to form the younger potassic granites of the BGC at similar to 3.33 Ga.

Because of their robustness against resetting, in situU-Pb ages of zircons in lunar impactites have the potential to provide constraints on the lunar bombardment history that may complement the more common K-Ar ages. Most previous work has focused on relatively large zircons that show growth zoning and ages were mostly interpreted as early igneous crystallization ages. Here we combine highresolution mineralogical imaging and in situU-Pb dating by ion microprobe to identify, characterize and date <20 mu m size zircons in thin sections of lunar impact breccias. Several tens of grains of zircons of this size range were identified in thin sections of impactites from the Apollo 15 and 16 landing sites. Small zircons are more abundant in both noritic and evolved clinopyroxene, SiO2 or K-feldspar bearing lithologies compared to anorthositic bulk compositions. Both granular zircon aggregates and overgrowth on existing zircon or baddeleyite (in breccias 15455 and 67915) are interpreted to reflect high-temperature recrystallization of zircons or its high-temperature-pressure precursor phases, following shock heating events by impact. In contrast, conchoidal or poikilitic zircons < 10 mu m in Fe-Ni metal bearing noritic clasts or matrix (67915, 67955) crystallized in situfrom impact melt. Most U-Pb ages of the 24 analyzed grains are either concordant or reverse discordant with Pb-207-Pb-206 ages ranging from 4.15 to 4.25 Ga. The small age range, combined with a large textural spectrum and the frequent presence of Fe-Ni metal suggest zircon crystallization from impact melt and recrystallization of preexisting zirconium-bearing minerals by impact heating. Such `impact' zircons with 4.2 Ga ages have now been reported from most Apollo landing sites, suggesting widespread formation and modification of zircons by basin-forming impacts at this time. The contrast between U-Pb zircon (predominantly 4.2 Ga) and K-Ar feldspar ages (predominantly 3.9 Ga) likely reflects resetting of the latter chronometer by impact heating. (C) 2021 The Author(s). Published by Elsevier B.V.

The Lindero gold deposit is located in the Southern Puna plateau, northwest Argentina. The deposit is centered in a cluster of six subvolcanic intrusions emplaced at the margin of the Arizaro Basin. Three alteration types were recognized: (i) Ca-Na silicate (clinopyroxene + magnetite +K-feldspar + quartz + calcite plagioclase), (ii) K silicate (K-feldspar + quartz + magnetite biotite +/- anhydrite) and (iii) chlorite-calcite alteration. The highest ore grades are linked to the K-silicate alteration. The proven plus probable reserves of Lindero are 84,226 t with average grades of 0.63 g/t Au and 0.11% Cu. A previous study assigned Lindero to the iron oxide copper-gold (IOCG) deposit type but many features of Lindero suggest that it is a porphyry gold deposit, including (i) the temporal and spatial link between alteration and the intrusive bodies, (ii) the alteration distribution pattern, particularly the small volume of rock affected by Ca-Na silicate alteration, (iii) the Au-rich and Cu-poor mineralization style. The magmatic complex at Lindero comprises an early-mineral unit (FPD), four inter-mineral units (CPD1, PBFD, CPD2 and DDP) and one post-mineral unit (PMI). In-situ U/Pb SIMS dating of the oldest (FPD), an intermediate (PBFD) and the youngest (DDP) intrusive units, confirms a middle Miocene age. The weighted mean ages of the oldest and youngest units are indistinguishable, with 15.36 +/- 0.13 Ma (n = 21) and 15.47 +/- 0.11 Ma (n = 16), respectively. Individual ages from each unit range by similar to 1 m.y. and the overall spread of zircon ages is 15.92 +/- 0.23 to 14.44 +/- 0.33 Ma. We suggest that emplacement of the subvolcanic stocks took place within this span time, likely at the lower end of this range (15.0-14.4 Ma). Two 40Ar/39Ar ages of hydrothermal biotites from the K-silicate alteration (14.99 +/- 0.16 Ma and 14.93 +/- 0.12 Ma) indicate that hydrothermal alteration began practically simultaneously with the emplacement of the porphyry units. All of the intrusive units are similar compositionally. They show a fineto medium-grained porphyric texture (1-4 mm) comprising plagioclase, amphibole, clinopyroxene and scarce quartz phenocrysts (40-55 vol % of phenocrysts) in a K-feldspar +/- quartz microcrystalline (0.02-0.07 mm) groundmass, except in the post-mineral unit which has a cryptocrystalline groundmass. Whole-rock analyses reveal a narrow range of dioritic composition (58.6-61.9 wt % SiO2) and high-K calc-alkaline character for all units. Trace element features (low Ba/Nb ratios, high Nb) of the Lindero magmas indicate a back-arc affinity, similar to those from the Southern Puna and distinct from the Central Volcanic Zone (CVZ) frontal arc. The Sr and Nd isotope ratios (Sr-87/Sr-86 = 0.706042 to 0.706607; Nd-143/Nd-144 = 0.512501 to 0.512582) from Lindero intrusives are also similar to Southern Puna back arc volcanic rocks. The Pb isotope ratios of Lindero (Pb-206/Pb-204 = 18.79 to 18.83; Pb-207/Pb-204 =15.60 to 15.63; Pb-208/Pb-204 = 38.66 to 38.74) overlap with both back-arc and arc magmas in the CVZ. The narrow age range, spatial association and uniform chemical and isotopic composition of Lindero porphyry units suggest that were derived from a common magma source, which underwent fractionation and/or crustal assimilation before emplacement as suggest by the low concentrations of Mg, Cr, Ni and Sr. The Lindero porphyry units show chemical and isotopic similarities with those from porphyry gold deposits in the Maricunga belt, Chile, and with the porphyry copper deposits of Argentina located in a back-arc setting; however, they differ from porphyry copper deposits in the frontal-arc setting of Chile, notably by the lack of an adakite-like signature (high Sr/Y ratio).

Instrumental mass fractionation (IMF) associated with Secondary Ion Mass Spectrometry (SIMS) measurements of oxygen isotope compositions in silicate glasses was studied using a set of 27 synthesized glasses spanning a compositionally broad range of six major oxides: SiO2, TiO2, Al2O3, total FeO (FeOt), MgO and CaO. The impact of chemical composition on the IMF values was investigated using a Cameca 1280-HR during a single SIMS analytical session operated under constant instrumental conditions. The data measured were compared with the 818O values obtained by laser fluorination gas source mass-spectrometry (LF). The offset between the 818O(LF) and 818O(SIMS) was found to reach up to 5%o. Our data document that SIMS oxygen isotope matrix effects in silicate glasses strongly depend on the chemical composition of silicate glasses, here the cation-oxygen bond strength was found to have a strong influence on the IMF value. We tested a variety of models based on single oxide contents and various composition-dependent parameters, but none were fully satisfactory for predicting IMF. Neither mean atomic mass nor NBO/T (the ratio of non-bridging oxygens per tetrahedrally coordinated cation) show a strong correlation with the IMF values (R2 of 0.45 and 0.46, respectively). Among the single oxides, only the model based on the SiO2 content is useful for prediction of the IMF in silicate glasses, but this model has a large standard error (1a = +/- 0.90%o) and was also found to break down for glasses with high Na and K contents. We propose an empirical model based on the correlation of six major element oxides that shows a strong correlation with IMF (R2 = 0.98, 1a = +/- 0.40%o). This model describes the experimental data with uncertainties that are roughly a factor of two better than the correction methods proposed in earlier studies. We also investigated using the correlation between IMF and isotope I-18O index, which describes the correlation between atomic bond strength and relative oxygen isotope enrichment in silicate substances (R2 = 0.87). Although our efforts provide refinements to the SIMS determination of 818O in natural silicate glasses, truly accurate IMF corrections will need further refinements related to the impact of alkali elements.

Abstract The Raman spectra of five [4]B-bearing tourmalines of different composition synthesized at 700 °C/4.0 GPa (including first-time synthesis of Na-Li-[4]B-tourmaline, Ca-Li-[4]B-tourmaline, and Ca-bearing ◻-[4]B-tourmaline) reveal a strong correlation between the tetrahedral boron content and the summed relative intensity of all OH-stretching bands between 3300–3430 cm–1. The band shift to low wavenumbers is explained by strong O3-H···O5 hydrogen bridge bonding. Applying the regression equation to natural [4]B-bearing tourmaline from the Koralpe (Austria) reproduces the EMPA-derived value perfectly [EMPA: 0.67(12) [4]B pfu vs. Raman: 0.66(13) [4]B pfu]. This demonstrates that Raman spectroscopy provides a fast and easy-to-use tool for the quantification of tetrahedral boron in tourmaline. The knowledge of the amount of tetrahedral boron in tourmaline has important implications for the better understanding and modeling of B-isotope fractionation between tourmaline and fluid/melt, widely used as a tracer of mass transfer processes.

This study presents new secondary ion mass spectrometry (SIMS) reference materials (RMs) for measuring water contents in nominally anhydrous orthopyroxenes from upper mantle peridotites. The enstatitic reference orthopyroxenes from spinel peridotite xenoliths have Mg#s between 0.83 and 0.86, Al 2 O 3 ranges between 4.02 and 5.56 wt%, and Cr 2 O 3 ranges between 0.21 and 0.69 wt%. Based on Fourier-transform infrared spectroscopy (FTIR) characterizations, the water contents of the eleven reference orthopyroxenes vary from dry to 249 ± 6 µg/g H 2 O. Using these reference grains, a set of orthopyroxene samples obtained from variably altered abyssal spinel peridotites from the Atlantic and Arctic Ridges as well as from the Izu-Bonin-Mariana forearc region was analyzed by SIMS and FTIR regarding their incorporation of water. The major element composition of the sample orthopyroxenes is typical of spinel peridotites from the upper mantle, characterized by Mg#s between 0.90 and 0.92, Al 2 O 3 between 1.66 and 5.34 wt%, and Cr 2 O 3 between 0.62 and 0.96 wt%. Water contents as measured by SIMS range from 68 ± 7 to 261 ± 11 µg/g H 2 O and correlate well with Al 2 O 3 contents (r = 0.80) and Cr#s (r . = -0.89). We also describe in detail an optimized strategy, employing both SIMS and FTIR, for quantifying structural water in highly altered samples such as abyssal peridotite. This approach first analyzes individual oriented grains by polarized FTIR, which provides an overview of alteration. Subsequently, the same grain along with others of the same sample is measured using SIMS, thereby gaining information about homogeneity at the hand sample scale, which is key for understanding the geological history of these rocks.

Crustal recycling into Earth's deep mantle has been inferred from both seismic tomography and geochemical observations. As a further line of evidence, we report on zircons having a wide range of ages that were recovered from the Aladag chromitites, providing direct evidence for crustal recycling. Mesozoic zircons represent the earliest stages of Neotethyan seafloor spreading magmatism, whereas Neoproterozoic and Mesoproterozoic-Archean zircons record recycled old crustal material entrained in the ophiolitic mantle melt sources. The incorporation of such old, crustal zircons into the Neotethyan mantle might have followed from previous subduction events, from lithospheric delamination, and/or from the dismantling of West Gondwana in the Permo-Triassic. Subduction-affiliated ophiolitic melts may have picked up these recycled zircons and integrated them into chromitites, which precipitated from peridotite-melt interactions in a mantle wedge beneath the Inner-Tauride (a Neotethyan seaway) seafloor spreading system. Common in ophiolitic chromitites within the Neotethyan realm, such unusually old zircons present unique archives for tracking crustal recycling and mantle processes during rift-drift, seafloor spreading and subduction zone evolution of the oceanic mantle. We conclude that ophiolitic mantle peridotite and chromitite may be an important archive for preserving the recycling history of crustal material back into Earth's deep mantle.

Our most important Zn resources occur within clastic-dominated (CD-type) deposits, which are located in a small number of Proterozoic and Phanerozoic sedimentary basins. The most common model for CD-type mineralisation involves sedimentary exhalative (SEDEX) processes, i.e. the venting of metal bearing fluids into a restricted, anoxic H2S-bearing (euxinic) water column. In the Carpentaria Zn Province (Australia), multiple world class deposits are hosted in Proterozoic (1.6 Ca) stratigraphy, where models of the ancient sulfur cycle have also been developed. Focusing on the most recent discovery - the leena deposit - we report bulk rock and isotopic data (delta S-34(pyrite) values) that provide information on the sulfur cycle during the diagenetic and hydrothermal evolution of the Teena sub-basin. in contrast to the SEDEX model, intervals containing abundant pyrite with highly positive delta S-34 values (>25 parts per thousand) correspond with euxinic conditions that developed due to high organic loading (i.e. productivity) and not basin restriction. This basin wide feature, which can also be mistaken as a hydrothermal pyrite halo, is genetically unrelated to the subsequent hydrothermal mineralisation that formed beneath the palaeo-seafloor. The formation of CD-type deposits in the Proterozoic does not, therefore, require euxinic conditions.

The Carpentaria province (McArthur basin and Mount Isa inlier) in northern Australia is one of the most important districts for clastic-dominated (CD-type) massive sulfide deposits. The George Fisher Zn-Pb-Ag deposit, located in this province, is hosted by the carbonaceous Urquhart Shale Formation (ca. 1654 Ma) in a region that has an active history of metamorphism and tectonism. In this study, paragenetically constrained pyrite in samples from the George Fisher deposit and unmineralized Urquhart Shale have been analyzed in situ using secondary ion mass spectrometry (SIMS) of sulfur isotopes (delta S-34 values). Samples were taken from four drill cores through the main orebodies at George Fisher and one drill core through correlative, unmineralized Urquhart Shale (Shovel Flats area). Five generations of pyrite were identified at George Fisher and record a protracted history of sulfate reduction under diagenetic and subsequent hydrothermal conditions: (1) fine-grained, subhedral-spheroidal pyrite (Py-0), (2) coarse-grained, anhedral pyrite (Py-1) associated with ore-stage 1 sphalerite and galena, (3) coarse-grained, euhedral pyrite (Py-2) associated with ore-stage 2 sphalerite, galena, and pyrrhotite, (4) massive subhedral to euhedral pyrite (Py-3) associated with ore-stage 3 chalcopyrite, pyrrhotite, galena, and sphalerite, and (5) coarse-grained euhedral pyrite (Py-euh), which occurs only in unmineralized rocks. In the unmineralized Shovel Flats drill core, only Py-0 and Py-euh are present. Whereas pre-ore pyrite (Py-0) preserves negative delta S-34 values (-8.1 to 11.8 parts per thousand, the ore-stage pyrites (Py-1, Py-2, and Py-3) have higher delta S-34 values (7.8-33.3, 1.9-12.7, and 23.4-28.2 parts per thousand, respectively). The highest delta S-34 values (7.2-33.9 parts per thousand) are preserved in Py-euh. In combination with petrographic observations, the 6 34 5 values of pyrite provide evidence of three different processes responsible for the reduction of sulfate at George Fisher. Reduced sulfur in fine-grained pyrite (Py-0) formed via microbial sulfate reduction (MSR) under open-system conditions prior to the first generation of hydrothermal pyrite (Py-1) in ore-stage 1, which most likely formed via thermochemical sulfate reduction (TSR). During deformation, previously formed sulfide phases were then recycled and replaced during a second hydrothermal event (ore-stage 2), resulting in intermediate sulfur isotope values. Another syndeformational hydrothermal Cu event, involving a sulfate-bearing fluid, formed ore-stage 3 via TSR. This study demonstrates that the fine-grained pyrite formed pre-ore under conditions open to sulfate and outlines the role of multiple stages of sulfide formation in producing high-grade Zn-Pb-Ag orebodies in the Mount Isa inlier.

The Huxingshan tungsten deposit (similar to 0.21 Mt at 0.3 wt% WO3) is located in the Jiangnan Massif tungsten ore belt, South China. Here, the W mineralization is restricted to quartz and muscovite-quartz veins intruding variable country rocks of the Lower Cambrian strata. Rb-Sr isochron ages derived from fluid inclusions trapped in muscovite-quartz veins (ca. 134 +/- 2 Ma) agree with the zircon U-Pb crystallization age of the associated Huxingshan granite (137.8 +/- 0.5 Ma) and may thus suggest a close petrogenetic relationship between both rock types. Zircon epsilon Hf (t) values of the Huxingshan granite (-16.2 to +6.6) overlap with those of the specially related metasedimentary Banxi and Lengjiaxi Groups, consistent with melting of Neoproterozoic Yangtze lower crustal materials with mantle melts input to the source. We suggest that the highly differentiated signatures of tungsten and beryllium granite might be inherited from the origin chemistry of source rocks and further modified by highly fractional crystallization rather than by fluid-rock interaction processes. During this process, tungsten and beryllium was enriched in the residual melts/fluids, which finally separated from the solidifying melt body and were introduced into the county rock to precipitate scheelite and beryl along skarn and greisen horizons by large-scale fluid movements. The outlined processes are likely to be coupled to large-scale melting of continental crust and associated granitic magmatism under a regime of subduction of the paleo-Pacific Plate beneath South China and subsequent tearing of the slab. The newly discovered Huxingshan deposit underlines the huge prospecting potential for the northwestern Jiangnan Massif.

Coupled oxygen-hafnium isotope and trace element geochemical data were obtained using thirty eight previously dated zircon grains extracted from five mafic to intermediate crustal xenoliths of the Wyoming Craton (Montana, USA). Xenoliths include mid to lower crustal (642-817 degrees C and 3.5-9.4 kbar) mafic granulites and amphibolites with dominantly Mesoproterozoic (1772-1874 Ma) and minor Paleoproterozoic to Late Archean (2004-2534 Ma) Pb-207/Pb-206 zircon ages. Zircon oxygen isotope data indicate derivation from melts in equilibrium with a mantle source that interacted with limited supracrustal material (delta O-18= 4.4-5.7 parts per thousand), as well as the incorporation of supracrustal fluids or melts into mantle source regions (delta O-18= 6.0-8.1 parts per thousand). The small withinsample isotopic variability suggests that primary zircon did not exchange with isotopically distinct fluids or melts after initial formation. Initial zircon Hf isotopic values are highly variable across all xenoliths (epsilon(Hf) - +3.7 to -17.6), consistent with protolith derivation from mantle sources that incorporated evolved, unradiogenic material or were modified by subduction-related fluids. Within a single granulite xenolith, two zircon types are recognized based on CL imagery, Hf isotopes and U-Pb ages (Type I and Type II). Type I magmatic zircons show dispersed ages (ca. 1700-2534 Ma) and unradiogenic initial Hf (epsilon(Hf) = -17.6 to -1.5, Hf-176/Hf-177= 0.281074-0.281232). The spread in ages and initial epsilon(Hf), but narrow range in initial Hf-176/Hf-177, points to variable Pb loss in response to dissolution-recrystallization of pre-existing zircon. Type II metamorphic zircon yields a younger Proterozoic population (ca. 1700-2155 Ma) with more radiogenic initial Hf relative to Type I zircon (epsilon(Hf) = -7.9 to +1.4, Hf-176/Hf-177 = 0.281427-0.281578); this type represents newly grown metamorphic zircon that formed in the solid-state and incorporated Zr and Hf from pre-existing zircon and silicate matrix/metamorphic phases. REE patterns from all xenoliths are steep and positively sloping without discernible HREE depletion relative to LREE, implying zircon crystallization/recrystallization in the absence of garnet. Negative Eu anomalies signify simultaneous zircon and feldspar crystallization. Solid-state recrystallization may have lead to variations in LREE, Eu and Ce in certain xenoliths. Xenoliths containing magmatic zircon (1834 +/- 19 Ma) with mantle-like delta O-18 (4.4-5.5 parts per thousand) and radiogenic initial epsilon(Hf) (-2.3 to +3.7) likely formed through crystallization of melts derived from a mantle source that incorporated minor amounts of subducted sedimentary/supracrustal material. Proterozoic (1874 +/- 8 Ma) xenoliths with elevated delta O-18 (6.0-7.0%) and unradiogenic initial epsilon(Hf) (-8.2 to -9.6) within magmatic zircon represent melt products of subduction-induced melting and metasomatism of the overlying mantle wedge in the vicinity of the northern GFTZ. Older (ca. 2534 Ma) xenoliths containing zircons with elevated delta O-18 (6.4-7.2 parts per thousand) and unradiogenic epsilon(Hf) (up to -17.6) represent crystallization of protolith magmas extracted from a mantle source metasomatized by subduction-derived fluids and melts in the Late Archean or earlier. Zircon geochronology and isotope systematics within Mesoproterozoic xenoliths support a model of ocean-closure and subsequent continental collision between the Medicine Hat Block and Wyoming Craton, resulting in the formation of subduction-related melts at ca. 1834-1874 Ma, followed by ca. 1770 Ma collision-related metamorphism thereafter.