Hf Isotope Systematics Research Articles

Decoupling of the Lu–Hf, Sm–Nd, and Rb–Sr isotope systems in eclogites and a garnetite from the Sulu ultra-high pressure metamorphic terrane: Causes and implications

The whole-rock Hf, Sr and Nd isotope data of five high-Fe–Ti eclogites, nine high-Al eclogites, and a garnetite from the Sulu ultrahigh pressure (UHP) metamorphic terrane at eastern China were analyzed to resolve the causes for the decoupling of the Lu–Hf, Sm–Nd, and Rb–Sr isotope systems in these UHP rocks and to infer their protolith characteristics. Seven of the nine high-Al eclogites define an 87Rb/86Sr–87Sr/86Sr errorchron age of 192±43Ma (MSWD=2.8), which is within the time span of retrograde metamorphism despite the large uncertainty. The high-Fe–Ti eclogites and garnetite, however, have low 87Rb/86Sr ratios of <0.031 with scattered 87Sr/86Sr ratios of 0.7042–0.7058. Accordingly, it is inferred that the Rb–Sr isotope system in the samples reflects the effects of processes postdating the UHP metamorphism. Despite having different constituent mineral assemblages and whole rock geochemistry, the samples, however, define a 147Sm/144Nd–143Nd/144Nd errorchron age of 232±36Ma (MSWD=4.6). Although the uncertainty of ±36Ma implies incomplete Nd isotope equilibrium among the samples, the large overlap between this errorchron age span and the mineral isochron ages of 245–210Ma for the UHP metamorphism indicates the control of peak metamorphism on the Sm–Nd isotope system. The incomplete Nd isotope re-equilibration was accompanied by metamorphic modification on the Sm/Nd ratios as indicated by the U-shaped LREE patterns. The initial εNd(780) values of the protolith rocks calculated from the Sm/Nd ratios of the samples deviate from the igneous initial εNd(t)–εHf(t) trend to significantly lower values, consistent with the metamorphic increase in the Sm/Nd ratios. In contrast, the Lu/Hf ratios are generally within the range for basalts and do not vary systematically with the 176Hf/177Hf ratios. The protolith εHf(780) values calculated from the Lu/Hf ratios of the samples are nearly identical to the initial εHf(t) values of the ~780Ma magmatic zircon cores from the Yangtze craton, suggesting the dominance of protolith characteristics on the Lu–Hf isotope system. Being controlled by different processes, the Rb–Sr, Sm–Nd, and Lu–Hf isotope systems of the investigated samples are therefore decoupled. The εNd(220)–εHf(220) compositions of the samples deviate from the εNd(t)–εHf(t) trend of intraplate lavas to high εHf(220) at a given εNd(220) and are within the arc lava field. The protolith εNd(780)–εHf(780) compositions calculated from metamorphic initials using igneous Sm/Nd and Lu/Hf ratios show a similar distribution pattern, strengthening the arc affinity of the eclogites. These arc signatures together with the felsic–mafic bimodal geochemical features of the UHP rocks from the Sulu terrane are explained as the characteristics of protoliths generated by backarc rifting.

Age, nature, and origin of Ordovician Zhibenshan granite from the Baoshan terrane in the Sanjiang region and its significance for understanding Proto-Tethys evolution

This contribution reports on a better understanding of the Proto-Tethys evolution in the Sanjiang Tethyan region of China. The manuscript presents laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) zircon ages, Sr–Nd–Hf isotope systematics, and whole-rock major and trace element of Ordovician magmatic rocks from the calc-alkaline Zhibenshan granite, which formed along the northern margin of East Gondwana. The U–Pb zircon dating of monzogranite from the Zhibenshan granite yields crystallization ages of 466–457 Ma. The monzogranite has Cross–Iddings–Pirsson–Washington normative corundum (3.34%) and is peraluminous with Al2O3/(Na2O+ K2O+CaO) molar ratio of 1.26, similar to S-type granites. All samples are enriched in large ion lithophile elements (LILEs, such as Rb, K, U, and Th) and depleted in high field strength elements (HFSEs, e.g. Nb and Ti). These granites are enriched in light rare earth elements (REEs) and depleted in heavy REEs with strongly negative Eu anomalies (δEu = 0.19–0.24). The initial 87Sr/86Sr ratios range from 0.7118 to 0.7176 and εNd(t) values from −11.3 to −10.3 with Nd model ages of 2114–2037 Ma. Magmatic zircons with early Palaeozoic dates have εHf(t) values ranging from −13.3 to −1.8 and Hf model ages from 2258 to 1537 Ma. These geochemical and isotopic features suggest that the Zhibenshan granite originated from an ancient crustal source. Ordovician granites in the Baoshan terrane represent the southward continuation of the early Palaeozoic granitic belt that extended along the northern margin of East Gondwana, providing important evidence for the evolution of the Proto-Tethyan Ocean.

The affinity of Archean crust on the Yilgarn—Albany–Fraser Orogen boundary: Implications for gold mineralisation in the Tropicana Zone

Craton margins can be subject to a wide array of gold genesis and redistribution processes, although high-grade terrains on craton margins are frequently viewed as less prospective than lower-grade counterparts. In contrast to this, the high-grade Tropicana Zone, a newly defined Archean crustal component on the eastern margin of the Yilgarn Craton within the Albany–Fraser Orogen (AFO), contains a significant Proterozoic gold deposit. This deposit and zone comprise mid-amphibolite to granulite-facies gneissic rocks with evidence of partial melting and granite injection. The Tropicana Zone contains significant low-Si, LILE-enriched, granites classed as sanukitoids. Along with the distinctive compositions, the rarity of these rocks within any Archean craton suggests that the granitoid protoliths represent a single suite, emplaced during one event.Due to the intense granulite-facies overprinting of the Tropicana Zone rocks, determination of the magmatic protolith age for these sanukitoids is challenging. Nonetheless, the best age estimate for magmatism is 2692±16 Ma, based on the youngest zircons preserving textural evidence of growth within a viscous silicate melt. This age is older than compositionally similar magmatism found within the Yilgarn Craton, although a sanukitoid in the Northern Foreland of the AFO has a similar age. Furthermore, the granulite-facies metamorphic zircon growth in the Tropicana Zone at 2718–2554 Ma was prolonged compared to that in the Yilgarn Craton. Nonetheless, the Hf isotopic signature of the Tropicana Zone zircon shares strong similarity to that from the Eastern Goldfields Superterrane of the Yilgarn Craton. This implies that the Tropicana Zone reflects a deeper crustal level of the Yilgarn Craton, exhumed and thrust NW an unknown distance over the craton edge. In addition, we observe that the granulite-facies zircons have a less radiogenic Hf-isotope signature than the preserved pre-metamorphic zircon cores. Based on correlations with alpha dose, U and Th content and 176Hf/177Hf we suggest this reflects the preferential destruction and release of unradiogenic Hf from inherited zircon whereas the protolith sanukitoid zircon, with lower U and Th content, was more resistant to mobilisation during high-grade metamorphism. We note this situation may be a more general response of the Hf isotopic system, in which zircon grown in a more mafic melt is less likely to contribute to the metamorphic Hf reservoir than its felsic counterpart.Juvenile granitic veins dated at c. 1780 Ma intruded into the Tropicana Zone indicate that the Tropicana Zone was structurally emplaced at or before c. 1780 Ma, given similar Proterozoic magmatic events are well documented from (para)autochthonous adjacent units. Re–Os dating of pyrite coeval with one generation of gold in these rocks indicates model ages of c. 2100 Ma, supportive of a Palaeoproterozoic age of mineralisation. This mineralisation event is distinct from major Proterozoic tectonothermal events elsewhere in the AFO. Sanukitoid magmas are well-known for gold fertility and were likely the original source of gold in the Tropicana Zone, which was subsequently concentrated into brittle structures during several episodes. Gold mineralisation post-dated peak metamorphic conditions and is significantly younger than gold mineralisation within other parts of the adjoining Yilgarn Craton.

Magma mixing and crust–mantle interaction in the Triassic monzogranites of Bikou Terrane, central China: Constraints from petrology, geochemistry, and zircon U–Pb–Hf isotopic systematics

Mafic microgranular enclaves (MMEs) are common in the monzogranites from the Yangba pluton in the Bikou Terrane of central China. Zircon LA-ICP-MS U–Pb dating yields ages of 208.7±0.7Ma and 209.3±0.9Ma for the host monzogranites, and 211.9±0.8Ma for the MMEs, indicating formation from coeval magmas. The field occurrence and textures including spheroidal shapes, transitional contacts, igneous mineral assemblages, acicular apatites, and oscillatory zoning with repeated resorption surfaces in plagioclase, indicate that the MMEs crystallized from mafic magma that was injected into and mingled with the host felsic magma. The host monzogranites are intermediate-felsic, metaluminous, and lack typical peraluminous minerals or alkaline mafic minerals, suggesting their I-type affinity. The monzogranites display depletion in high field strength elements, with Nb and Ta anomalies, and enrichment in large ion lithophile elements and light rare earth elements (LREE), with slightly negative Eu anomalies (Eu/Eu∗=0.72–0.93). The MMEs from the Yangba pluton are intermediate, metaluminous, and have higher K2O, Al2O3, Fe2O3T, MgO, Ni, and Cr contents than the pluton. The MMEs also display much higher total rare earth element (REE) concentrations and REE patterns that are sub-parallel to those of the host rocks, enrichments in Rb, Th, U, K, and LREE, and depletions in Ba, Nb, Ta, and Ti. These features are similar to those of the host rocks, indicating crust–mantle interaction during their petrogenesis.Zircons from the MMEs have highly variable εHf(t) values (−5.5 to +8.7) with corresponding two-stage Hf model ages (TDM2) of 1.14 to 1.42Ga, indicating they were derived from depleted mantle with crustal contamination. In contrast, zircons from host monzogranite show εHf(t) values ranging from −1.7 to +2.7, with TDM2 of 1.13 to 1.21Ga, corresponding to that of the basement rocks in the Bikou Terrane, which indicates that these granitic magmas were probably derived from the Neoproterozoic juvenile lower crust with minor contribution of Mesoproterozoic ancient lower crust.The integrated petrology and elemental and isotopic geochemistry suggest that the MMEs and host monzogranites were generated by the interaction of a granitic magma and a mafic magma in the lower crust. We envisage that Neoproterozoic subcontinental lithosphere mantle-derived mafic magma intruded the lower continental crust during Late Triassic, following the Late Permian to Early Triassic collision between the South China and North China Blocks.

Eocene magmatism in the Tethyan Himalaya, southern Tibet

Abstract Recent studies in the Yardoi gneiss dome of the Tethyan Himalaya, southern Tibet demonstrated that the Himalayan orogenic belt experienced a major episode of crustal melting in the Mid-Eocene ( c. 44–40 Ma), represented by the genesis of a suite of high Sr/Y two-mica granites (TMGs) accompanied by leucogranites. Geochemical and U–Pb age data indicate that (1) both the leucogranite and the porphyritic leucogranite formed at c. 44–41 Ma are characterized by similar Sr–Nd–Hf isotope systematics to those in the less evolved TMGs; (2) these three suites of rocks might represent a Mid-Eocene igneous complex derived from partial melting of amphibolite under crustal thickened conditions; and (3) in contrast with the TMGs, leucogranites experienced hydrothermal-magma interactions as well as plagioclase fractional crystallization as revealed from pronounced negative Eu anomalies, much lower Sr concentrations, and elevated common Pb and U concentrations in the latest magmatic zircon overgrowth. Our data suggest that magmatic differentiation of more primitive magmas could have played a critical role in the generation of leucogranites. Documentation of partial melting of deep crustal rocks during the earliest phase of the tectonic evolution of the Himalayan orogen implies that the large-scale collisional orogenic belt could experience intensive partial melting under contraction and thickening conditions, which could profoundly affect the geophysical and geochemical properties as well as rheological properties and facilitate the initiation of movement or flow of deep crustal rocks.

Uncoupled O and Hf isotopic systems in zircon from the contrasting granite suites of the New England Orogen, eastern Australia: Implications for studies of Phanerozoic magma genesis

The Permo-Triassic granites of the New England Orogen, eastern Australia, were emplaced into a volcanic arc complex accreted to the eastern Gondwana margin in the Late Devonian or Early Carboniferous. Zircon U–Pb dating shows that the S-type Hillgrove (∼297Ma) and Bundarra (∼287Ma) Supersuites predated intrusion of the I-type Moonbi Supersuite (∼250Ma) by up to 50Ma. The high δ18Ozrn of the S-type granites (10.0–11.5‰), and range of U–Pb ages (∼370–300Ma) and δ18Ozrn (∼5–10‰) of their inherited zircon cores, show that their source rocks were predominantly weathered Carboniferous volcaniclastics, the youngest deposited<25Ma before the granites were emplaced. In contrast, the lower δ18Ozrn (6.9–7.8‰) and lack of inheritance in the I-type granites is consistent with a zircon poor, more juvenile source, probably a mafic igneous underplate mixed with a small amount of volcanogenic and/or oceanic sediment. Despite the differences in source materials, the εHf(t) values of all granites, both S- and I-type, are similar (+5.0±0.5 cf. +5.9±0.5), consistent with both the mafic and sedimentary components in the granite sources being relatively young and similar in Hf isotopic composition at the time of granite genesis. In young, isotopically juvenile orogens, the O isotopic composition of well-dated igneous and inherited zircon can be a much more sensitive indicator of petrogenetic processes than the zircon Hf isotopic compositions alone.

Timing, duration and inversion of prograde Barrovian metamorphism constrained by high resolution Lu–Hf garnet dating: A case study from the Sikkim Himalaya, NE India

We investigated Lu–Hf isotopic systematics in garnets from gradually higher grade metamorphic rocks from the first appearance of garnet at c. 500 °C to biotite dehydration melting at c. 800 °C in the Sikkim Himalaya, India. Exceptionally precise Lu–Hf ages obtained for the Barrovian metasedimentary sequence in the Lesser Himalaya (LH) correspond to the time of early garnet formation on a prograde path and show remarkable correlation with increasing metamorphic grade and decreasing structural depth. The youngest age is reported for the garnet zone (10.6±0.2 Ma) and then the ages become progressively older in the staurolite (12.8±0.3 Ma), kyanite (13.7±0.2 Ma) and sillimanite (14.6±0.1 Ma) zones. The oldest age of 16.8±0.1 Ma was recorded at the top of the sequence in the zone marking the onset of muscovite dehydration melting, directly below the Main Central Thrust (MCT). These ages provide a tight constraint on the timing and duration of the Barrovian sequence formation which lasted about 6 Ma. The age pattern is clearly inverted with respect to structural depth but shows “normal” correlation with the metamorphic grade, i.e. earlier garnet growth in higher grade rocks.The timing of the peak metamorphic conditions estimated for the garnet and kyanite zones suggests that thermal climax occurred in all zones within a relatively short time span about 10–13 Ma ago. The metamorphic inversion in the Sikkim Himalaya must therefore have occurred after the metamorphic peak was attained in all grades.Lu–Hf garnet ages in the overthrust Higher Himalaya (HH) continue to be older but do not show a clear progression. There is about 6 Ma jump to 22.6±0.1 Ma within the MCT zone and even older ages of 26–28 Ma were obtained for migmatites from the lower part of the HH. The Lu–Hf system in garnets from the HH records the time of melting at or near the thermal peak rather than dates initial garnet growth as in the LH.The age pattern obtained in this study provides precise time constraints on the continental collision models which must take into account early and prolonged anatexis in the HH, progressive delay in initial garnet growth in different Barrow zones in the LH, nearly contemporaneous metamorphic peak in all grades and post peak inversion of the coherent LH block during exhumation.

Role of crustal contribution in the early stage of the Damara Orogen, Namibia: New constraints from combined U–Pb and Lu–Hf isotopes from the Goas Magmatic Complex

The tholeiitic to calc-alkaline Goas Complex represents the earliest magmatic activity of the inland branch of the Damara Orogen, and reflects the Pan-African convergence between the Congo and Kalahari Cratons. New results from laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) zircon U–Pb geochronology coupled with single-zircon Lu–Hf isotopic data constrain the age of emplacement and allow the identification of a crustal contribution to the magma sources. No temporal gap is registered between the first tholeiitic intrusion at ca. 575Ma and the main diorite to granite episodes. Invariably sub-chondritic εHft values (−3.8 to −34.4) attest to a significant crustal residence time with long-term reworking of multiple and mixed crustal components. No significant juvenile magma was involved in the magmatogenesis. Hf model ages coupled with an exhaustive dataset of U–Pb geochronological data suggest that the crustal components which fed the Goas sources might have formed during specific orogenic events with a major contribution of material derived from the central-western African Paleoproterozoic Eburnean Orogen. Geochemistry, U–Pb geochronology and Hf isotope systematics however, do not unequivocally clarify the long debate regarding whether or not the Pan-African orogenic cycle included ocean closure and subduction, as the involvement of crustal melting processes during the early stage of the Damaran event seems to have played a major role.

Maturing Arc Signatures Monitored by Trace Element and Hf Isotope Systematics in the Early Cretaceous Zacatecas Volcanic Field, Mexico

Mesozoic growth of continental crust along the southwestern margin of North America and its southern extension in Mexico has been partly explained by the accretion of terranes. These terranes have been considered to be fragments of exotic, intraoceanic island arcs that approached mainland Mexico after the Early Cretaceous. Trace elements and Lu-Hf isotopic systematics for primitive arc successions of the Zacatecas Volcanic Field indicate a close relationship with parts of the northern Guerrero superterrane. Major and trace element systematics of lava flows and dioritic rocks from laccoliths suggest a cogenetic origin of the Zacatecas Formation and Las Pilas Complex rocks, here combined in the Zacatecas Group. This group represents a single arc succession that evolves from a primitive to mature arc. Initial 176Hf/177Hf (age corrected to 130 Ma) ranges from 0.28296 to 0.28307, corresponding to εHf(t) = +9.3 to +13.4, indicating a source related to a depleted mantle wedge with a superimposed subducted sediment contribution. Based on combined field and geochemical evidence, we propose an arc model and suggest a spatial extension of paleoarc spreading north–south from Baja California beyond the present-day Trans-Mexican Volcanic Belt in the Early Cretaceous.

U–Pb age and Hf isotopic data of detrital zircons from the Devonian and Carboniferous sandstones in Yimin area, NE China: New evidences to the collision timing between the Xing’an and Erguna blocks in eastern segment of Central Asian Orogenic Belt

New U–Pb age data and Hf isotope systematics on zircons from Middle Devonian and Lower Carboniferous units from NE China provide evidence for the depositional age of these poorly constrained units as well as for the timing of the collision between the Erguna (Ergun in some references) and Xing’an blocks in the eastern end of Central Asian Orogenic Belt (CAOB). The detrital zircons from the Devonian and Carboniferous sandstones yielded similar age spectra in probability diagrams showing three distinct peaks at ∼450Ma, ∼500Ma and ∼540Ma, respectively, along with Proterozoic and Archean age groups, except for a relatively young age population at ∼360Ma detected for zircons from the Hongshuiquan Fm.The dating results suggest that the depositional time of the Niqiuhe Fm. should be younger than Late Silurian, as indicated by the youngest 422-Ma detrital zircon from this Fm., and that of the Hongshuiguan Fm. with a youngest detrital age of ca. 338Ma, is younger than Middle Mississippian (Visean), which is in accordance with those inferred from biostratigraphic data.These dominant age groups are widely found in the Erguna and Xing’an blocks, it is suggested that the sediments of the studied sandstones from Niuqiuhe and Hongshuiquan Fms come mainly from the Erguna and Xing’an blocks. The εHf (t) value of dated zircons from our sandstones has a wide range (−21 to 16), which shows a mixed signature of that from Paleozoic granitoids in Xing’an block (6–17) and Erguna block (−2 to 6), indicating that Erguna and Xing’an blocks had connected before the Middle Devonian deposition of the Niqiuhe Fm. (422±5Ma). Combined with the age data, we suggest that the Erguna and Xing’an blocks started to accrete at ∼540Ma, reached the peak of collision at ∼500Ma, and went into the post-collision stage from ∼490 to ∼450Ma. The study provides more evidences and constraints for the early Paleozoic evolutions of the inter-continental block amalgamations and accretions in CAOB.

Compositional and Sr–Nd–Hf isotopic variations of Baijingsi eclogites from the North Qilian orogen, China: Causes, protolith origins, and tectonic implications

Eclogites from the Baijingsi area in the North Qilian orogen at NW China were analyzed for major and trace element abundances as well as Sr, Nd, and Hf isotope ratios to evaluate the impacts from subduction processes on these compositional parameters and to reveal the complexity in protolith characteristics. The major element compositions of the Baijingsi eclogites are within the ranges for basalts. However, the absence of systematic variations among major oxides indicates varying extents of metamorphic modifications. Seven samples are characterized by profound Nb and Ta depletions. Six of them form coherent trends in the La versus Th, HFSE, and LREE plots. They are classified as the Group 1 samples. Other eight samples, referred to as the Group 2 samples, define distinct La–Zr and La–Sm trends and have flat variation patterns with slight Th–Nb–Ta–LREE depletions in the primitive mantle-normalized multiple-element diagram. Group 1 dominates the 87Sr/86Sr–87Rb/86Sr, 143Nd/144Nd–147Sm/144Nd and 176Hf/177Hf–176Lu/177Hf trends for relatively larger variations in the isotope and abundance ratios. Based on the errorchron ages from the samples, closure temperatures of the isotope systems, and isotope ratio versus 1/X plots (e.g., 87Sr/86Sr–1/Sr), it is inferred that the Rb–Sr isotope system was subjected to metamorphic modification, whereas the Sm–Nd and Lu–Hf isotope systems mainly reflect protolith characteristics. In addition to the Nb and Ta depletions, the Group 1 samples generally have εHf(510) values (protolith initial at 510Ma) higher than those of MORB at a given εNd(510), reflecting derivation from relatively high Lu/Hf ratio sources. They also plot within the fields for arc lavas in the Ta/Yb–Th/Yb and Th–Ta–Hf/3 diagrams. All these features are consistent with derivation from protoliths of an arc origin. In contrast, the Group 2 samples have εNd(510) values slightly higher than those of the Group 1 samples at a given εHf(510), plotting on the low εNd(510) margin of the MORB field. They also deviate from the Ta/Yb–Th/Yb trend defined by MORB and OIB toward higher Th/Yb values within the fields of back-arc basin basalts. Therefore, it is inferred that the protoliths of the Group 2 samples were generated as back-arc basin basalts possibly associated with the protoliths of the Group 1 samples in a common supra-subduction zone. The εNd(510)–εHf(510) relationship of the Baijingsi eclogites shows an affinity to the Indian ocean type mantle, implying protolith generation at the southern hemisphere on the margin or a fragment of the Gondwana continent, most likely in the Qilian Block. Based on the current tectonic configuration, a southward subduction is required for protolith generation and might eclogitize the arc protoliths by subduction erosion. Alternatively, the northward subduction in a bidirectional subduction system might become dominant during the amalgamation of the Qilian Block and the North China Craton to eclogitize the arc protoliths in the Qilian Block. Resolving these two possibilities requires more petrological, geochemical, and structural evidence from the North Qilian orogen.

Preservation of Re‐Os isotope signatures in pyrite throughout low‐T, high‐P eclogite facies metamorphism

AbstractPyrite in LT–HP eclogites from the western Tianshan orogenic belt yields a Re‐Os age of 378.1 ± 8.9 Ma, which is 30–70 Ma older than ages previously obtained for the same rocks using the Rb–Sr, Sm–Nd, Ar–Ar, U–Pb, and Lu–Hf isotope systems. The Tianshan LT–HP eclogite experienced temperatures of up to ~570 °C combined with pressures of up to 2.1 GPa during metamorphism. These conditions are below the transition of pyrite to pyrrhotite, which defines both pyrite stability and possibly its closure temperature for Re‐Os. Pyrite can preserve Re‐Os signatures through eclogite facies peak metamorphic conditions, and thus allow determination of the formation age of pyrite in the protolith.

Linking continental deep subduction with destruction of a cratonic margin: strongly reworked North China SCLM intruded in the Triassic Sulu UHP belt

The collision between the North and South China cratons in Middle Triassic time (240–225 Ma) created the world’s largest belt of ultrahigh-pressure (UHP) metamorphism. U–Pb ages, Hf isotope systematics and trace element compositions of zircons from the Xugou, Yangkou and Hujialing peridotites in the Sulu UHP terrane mainly record a ~470 Ma tectonothermal event, coeval with the Early Paleozoic kimberlite eruptions within the North China craton. This event is interpreted as the result of metasomatism by fluids/melts derived from multiple sources including a subducting continental slab. The peridotites also contain zircons with ages of ~3.1 Ga, and Hf isotope data imply a component ≥3.2 Ga old. Most zircon Hf depleted mantle model ages are ~1.3 Ga, suggesting that the deep subcontinental lithospheric mantle beneath the southeastern margin of the North China craton experienced a intense mid-Mesoproterozoic metasomatism by asthenospheric components, similar to the case for the eastern part of this craton. Integrating data from peridotites along the southern margin of the craton, we argue that the deep lithosphere of the cratonic margin (≥3.2 Ga old), from which the Xugou, Yangkou and Hujialing peridotites were derived, experienced Proterozoic metasomatic modification, followed by a strong Early Paleozoic (~470 Ma) tectonothermal event and the Early Mesozoic (~230 Ma) collision and northward subduction of the Yangtze craton. The Phanerozoic decratonization of the eastern North China craton, especially along its southern margin, was not earlier than the Triassic continental collision. This work also demonstrates that although zircons are rare in peridotitic rocks, they can be used to unravel the history of specific lithospheric domains and thus contribute to our understanding of the evolution of continental cratons and their margins.

A young solidification age for the lunar magma ocean

The time at which the lunar magma ocean solidified can be determined from the Lu–Hf isotope systematics of lunar rocks derived from magma sources that formed during crystallization of the lunar magma ocean. The final magma ocean crystallization product, termed urKREEP, is enriched in incompatible elements including K, REE and P. We have determined the initial Hf isotopic compositions of four samples, two KREEP basalts and two Mg-suite norites. The incompatible trace element compositions of these samples are controlled by an urKREEP component, and therefore the initial Hf isotopic compositions of these samples represent the Hf isotopic evolution of urKREEP. In order to correct the effects of neutron irradiation on the Hf isotopic compositions of these samples, we have developed a model that uses the stable Hf and Sm isotopic compositions measured on an irradiated sample to determine and correct for the thermal and epithermal neutron fluence that has modified the Hf isotopic composition of the sample. We use our corrected results to calculate a 176Lu–176Hf urKREEP model age of 4353±37Ma and the 176Lu/177Hf of urKREEP to be 0.0153±0.0033. The Lu–Hf model age is concordant with the re-calculated Sm–Nd urKREEP model age of 4389±45Ma, and we take the average of these ages, 4368±29Ma, to represent the time at which urKREEP formed. This age is concordant with the age of the most reliably dated ferroan noritic anorthosite as well as 142Nd model ages for the formation or re-equilibration of mare basalt sources. Taken together, these ages indicate that the Moon experienced a widespread, large-scale magmatic event around 4370Ma, most plausibly attributed to solidification of the lunar magma ocean.

HFSE (High Field Strength Elements)-transport and U–Pb–Hf isotope homogenization mediated by Ca-bearing aqueous fluids at 2.04 Ga: Constraints from zircon, monazite, and garnet of the Venetia Klippe, Limpopo Belt, South Africa

Results from laser-ablation inductively-coupled-plasma mass spectrometry (LA-ICP-MS) and isotope dilution (ID) analyses of minerals and rocks from a single outcrop of the Venetia Klippe of the Limpopo Belt indicate that the U–Pb and Hf isotope system homogenized on the decimetre scale under amphibolite-facies conditions of ⩽645±25°C and ⩽7.0±1.1kbar, i.e. in the presence of an aqueous fluid phase. For a metabasite sample, homogenization is supported by isotope analyses of metamorphic zircon, garnet, and whole rock, which yield a six-point Lu–Hf isochron age of 2039.7±3.4Ma, with initial 176Hf/177Hf of 0.28126±0.00001, and a U–Pb zircon age of 2042±10Ma. The occurrence of a few inherited magmatic zircon cores with ages up to 2705Ma, and with significantly lower initial 176Hf/177Hf of 0.28112, however, indicate that homogenization was incomplete. For a chlorite–biotite–garnet schist isotope homogenization is reflected by within error identical zircon and monazite U–Pb ages of 2045±10Ma and 2041±8Ma, respectively, and by a zircon-garnet-whole rock Lu–Hf isochron age of 2083±63Ma, with an initial 176Hf/177Hf of 0.28140±0.00003. Contemporaneous formation of metamorphic zircon, monazite and garnet in the chlorite schist is not only supported by the isotope data, but also by chlorite inclusions in all three minerals, and by inclusions of metamorphic zircon in garnet. The inclusion textures and the identical initial 176Hf/177Hf support the conclusion that metamorphic zircon grains precipitated from an aqueous fluid phase, after dissolution of zirconium-bearing phases elsewhere, followed by a major HFSE transport, and Hf isotope homogenization. This fluid perhaps was Ca-bearing, as is suggested by the fact that garnet in the schist sample is the only Ca-bearing phase, and that metamorphic monazite, dating the metamorphic peak, is partially replacement by apatite. The fact that the metamorphic zircon rims in the metabasite sample have significantly lower initial 176Hf/177Hf than expected from the Lu–Hf isotope analyses of relic zircon cores and whole rock additionally hint that Ca-HFSE-bearing fluids infiltrated from surrounding quartzo-feldspathic gneisses, which had much lower 176Hf/177Hf at the time of metamorphism.

Lu–Hf isotope systematics of the Hadean–Eoarchean Acasta Gneiss Complex (Northwest Territories, Canada)

The Acasta Gneiss Complex (AGC) is a remnant Hadean–Eoarchean terrane composed of strongly deformed polyphase mafic to felsic gneisses which preserve a multi-stage history of magmatic emplacement, inheritance, and subsequent tectono-thermal modifications. The complexities encountered in such an old terrane fragment have been documented in previous geochronological studies of the AGC (e.g. zircon U–Pb, 147Sm–143Nd), and are evident also in its Lu–Hf isotope systematics. Here, we report new Lu–Hf isotope whole-rock measurements which show that some AGC gneisses were severely disturbed by migmatization and associated mineral segregation, while others preserve their Lu–Hf isotope systematics relatively intact with mostly near- to sub-chondritic initial 176Hf/177Hf ratios. Results reveal identifiable Eoarchean and later (Paleoarchean) magmatic events at around 3960Ma and again at 3600Ma, with a major metamorphism of the complex at 3750Ma. The oldest and least disturbed gneisses have a Lu–Hf regression age of 3946±87Ma, in good agreement with U–Pb zircon geochronology. A role of yet older crust (4000–4200Ma) in the formation of the AGC is also evident, but seems not to have influenced to first order the Lu–Hf isotope systematics of the 3960Ma group. The ca. 3960Ma group is proposed to be representative of its mantle source based on the absence of correlation between εHf(t) and Ce/Pb. It is further suggested that these two parameters show that the ca. 3600Ma gneisses were sourced in part from a mafic lithology belonging to the 3960Ma group, and that multiple sources (mantle and crust) were involved in AGC formation. The identification of preserved Lu–Hf isotope systematics in AGC gneisses means that complementary geochemical and isotopic studies bearing on the petrogenesis of pre-3900Ma rocks are possible. Despite its history of strong deformation and alteration, carefully selected domains within the AGC carry surviving information about the evolution of the mantle–crust system at the Eoarchean–Hadean boundary.

Mantle eclogites and garnet pyroxenites – the meaning of two-point isochrons, Sm–Nd and Lu–Hf closure temperatures and the cooling of the subcratonic mantle

The Earthʼs mantle is a huge metamorphic complex which undergoes permanent changes ruled by plate tectonics. It also has enclaves underneath Archean crust which are exempt from the convecting mantle since at least 2.5 Ga. Since then, this mantle may cool slowly as a result of diminishing heat input from the asthenosphere, of declining heat production from radioactive decay and of denudation of the crust. Under such circumstances, two point garnet–clinopyroxene isochrons from mantle xenoliths (here garnet pyroxenites and eclogites) reflect either cooling ages or eruption ages depending on whether the mantle portion under consideration was below or above the closure temperature of a radiogenic system. Available literature data from the Slave and Kaapvaal craton for the Sm–Nd and Lu–Hf isotope systems and our own new data from Bellsbank (Kaapvaal) provide a whole range of two-point isochron ages from younger than the kimberlite eruption age to early Proterozoic. The meaning of ages other than the kimberlite eruption age is unclear. We use here a compilation of Sm–Nd and Lu–Hf two-point isochron age data from this study and the literature, to assess the meaning of such isochrons. This is achieved by plotting the temperature of last equilibration as derived from the Fe–Mg exchange between garnet and clinopyroxene versus the two-point isochron age. There is a low temperature alignment for both systems of increasing age with decreasing temperature and an alignment around the kimberlite eruption ages at high temperatures. We interpret the intersect between the low temperature limb and the kimberlite eruption age as closure temperature which gives about 920 °C for the Lu–Hf system and about 850 °C for the Sm–Nd system.The differences of the cooling ages between the two isotope systems for individual samples combined with the closure temperatures from this study are used to deduce the cooling of the Slave and Kaapvaal subcratonic mantle from the early Proterozoic until today. We deduce that the subcratonic mantle cooled since 2 Ga from a geothermal gradient of 46 mW/m2 to 38 mW/m2 at 120 Ma ago with a rate of around 0.1 °C/Ma. This value is in agreement with previous estimates which were based on various geophysical and geological constraints, heat production by radioactive elements and diffusivity of radiogenic elements.

Hafnium isotopes and Zr/Hf of rutile and zircon from lower crustal metapelites (Ivrea–Verbano Zone, Italy): Implications for chemical differentiation of the crust

We have used granulite facies metapelites from the classic Ivrea–Verbano Zone (IVZ) lower crustal section, northern Italy, to investigate the behaviour of the hafnium (Hf) isotope system in rutile and zircon under conditions of high temperature (700–1050 °C) metamorphism and partial melting. The Hf isotope composition of rutile is shown to be robustly preserved from these conditions, and is also unaffected by retrogression, making it a powerful new tool for metamorphic petrology. Rutiles from our suite of IVZ metapelite samples have crustal Hf isotope signatures (εHf(i)=−6±1 to −13±1 at 288 Ma), with the most strongly crustal signature (εHf(i)=−9.2±0.3 to −12.9±1.0) recorded by narrow slivers of metapelite incorporated within a gabbroic body. The Hf isotope composition of rutile from these metapelite slivers is the direct result of dissolution of ancient detrital zircon with strongly unradiogenic 176Hf/177Hf at ultra-high temperatures associated with the emplacement of the gabbro. Rutile is an important host of Zr and Hf in high temperature metapelites, containing up to 11 000 ppm Zr and 400 ppm Hf. It has subchondritic Zr/Hf (13–32) that is significantly lower than the Zr/Hf of both zircon (35–52) and the starting bulk rock composition (36–40). Dissolution of zircon into extracted melts is expected to lead to a lowering of bulk rock Zr/Hf towards that of rutile. In the IVZ samples this effect is balanced by the strongly superchondritic Zr/Hf of garnet (40–70). The IVZ case study demonstrates the strong control that dissolution of detrital or inherited zircon can exert on the available Hf isotope budget, whether in a metamorphic or magmatic setting. The link between dissolution of accessory phases into melt and changes in trace element and Hf isotope chemistry points to important differences between high- and ultra-high temperature metamorphism in terms of chemical and isotopic differentiation of the crust.

Zircon U–Pb ages and Lu–Hf isotope systematics from late-tectonic granites, Geita Greenstone Belt: Implications for crustal growth of the Tanzania Craton

Granite plutons that intruded south of lake Victoria and north of the Geita Greenstone Belt have geochemical characteristics similar to high-K granites. When compared to the late high-K granites from the Musoma–Mara region, they have lower SiO2 content and higher TiO2, Al2O3, CaO, Na2O, P2O5, MgO and FeOt. They also show higher values for V, Sr, Zr, Ba and Hf concentrations. All samples display high total REE abundances (∑REE 139–393) and weak to moderate Eu depletion (Eu/Eu*: 0.25–0.77). Their chondrite-normalized pattern indicates that the light REE are moderately fractionated (3.03<Lan/Smn<6.24), whereas the heavy REE are weakly fractionated (1.63<Gdn/Ybn<2.80). The granite plutons are generally undeformed with 207Pb/206Pb ages between 2660 and 2620Ma, and intruded after the main ductile deformation that affected the Geita Greenstone Belt. The 176Hf/177Hf ratios for zircons from all samples are identical within error suggesting evolution from an isotopically uniform reservoir. The ɛHf values plot along the CHUR evolution line indicating that the granite represents juvenile crustal melts derived from greenstone material.Available zircon ages from the Tanzania Craton suggest that crustal growth occurred in three distinct periods, 2850–2800Ma, 2770–2730Ma and 2700–2620Ma, with the dominant period of crustal growth around 2700Ma. Zircon ages from the Sukumaland and Musoma–Mara Greenstone Belts indicate that the greenstone sequences formed during the same three periods of crustal growth.

Trace element budgets and (re-)distribution during subduction-zone ultrahigh pressure metamorphism: Evidence from Western Tianshan, China

We have conducted an LA-ICP-MS in situ trace element study of garnet, epidote group minerals, phengitic muscovite and paragonite in rocks of basaltic and sedimentary protolith from an ultrahigh pressure metamorphic belt along Western Tianshan, China. The data are used to evaluate the capacity of these minerals for hosting incompatible elements in response to subduction-zone metamorphism (SZM). The results confirm existing studies in that the presence and stability of these minerals largely control the geochemical behaviors of elements during SZM.We found that redistribution of rare earth elements (REEs), Th and U into newly-formed minerals during progressive SZM precludes the release of these elements from the down-going ocean crust, which contradicts the common perception in models of slab-dehydration and flux-melting. This suggests that additional processes, such as the involvement of supercritical fluids or hydrous melts formed at depth are required to supply these elements to the mantle wedge for arc magmatism. In addition, the ready release of large ion lithophile elements (LILEs) by different minerals, and the high immobility of REEs in rocks of basaltic protolith indicate that the contribution of altered ocean crust after SZM may not be responsible for the correlated Sr–Nd (Hf) isotope systematics observed in oceanic basalts. That is, subducted ocean crust that has gone through SZM cannot be the major source material for ocean island basalts.