MORB Source Research Articles

Redox-variability and controls in subduction zones from an iron-isotope perspective

An ongoing controversy in mantle geochemistry concerns the oxidation state of the sources of island arc lavas (IAL). Three key factors control oxidation–reduction (redox) of IAL sources: (i) metasomatism of the mantle wedge by fluids and/or melts, liberated from the underlying subducted slab; (ii) the oxidation state of the wedge prior to melting and metasomatism; and (iii) the loss of melt from IAL sources. Subsequently, magmatic differentiation by fractional crystallisation, possible crustal contamination and degassing of melts en route to and at the surface can further modify the redox states of IAL. The remote nature of sub-arc processes and the complex interplay between them render direct investigations difficult. However, a possible gauge for redox-controlled, high-temperature pre-eruptive differentiation conditions is variations in stable Fe isotope compositions (expressed here as δ57Fe) in erupting IAL because Fe isotopes can preserve a record of sub-surface mass transfer reactions involving the major element Fe. Here we report Fe isotope compositions of bulk IAL along the active Banda arc, Indonesia, which is well known for a prominent subducted sediment input. In conjunction with other arc rocks, δ57Fe in erupted Banda IAL indicates that fractional crystallisation and possibly crustal contamination primarily control their Fe isotope signatures. When corrected for fractional crystallisation and filtered for contamination, arc magmas that had variable sediment melt contributions in their sources show no resolvable co-variation of δ57Fe with radiogenic isotope tracers. This indicates that crustal recycling in the form of subducted sediment does not alter the Fe isotope character of arc lavas, in agreement with mass balance estimates. Primitive sources of IAL, however, are clearly isotopically lighter than those sourced beneath mid-ocean ridges, indicating either preferential Fe3+-depletion in the mantle wedge by prior, δ57Fe-heavy melt extraction, and/or addition of an isotopically-light slab-derived agent. Based on our findings and previous models of arc redox conditions, we propose a three-stage process to explain the Fe isotope composition of IAL: (i) prior melt depletion lowers Fe3+/ΣFe (Fe3+ over total Fe) in the residues, leaving refractory, δ57Fe-light and possibly reduced mantle wedge protoliths beneath arcs. The oxygen fugacity (fO2) of these refractory protoliths may be up to −2 log10 units reduced relative to the fayalite–magnetite–quartz synthetic oxygen buffer (ΔFMQ); (ii) oxidised, slab-derived fluids, Fe-poor but possibly rich in sulphate (i.e., S6+), trigger melting of depleted protoliths with minimal effect on δ57Fe. Melts derived from this fluid-modified wedge source have high Fe3+/ΣFe, oxidised by the reduction of S6+, but importantly retain the light δ57Fe from their mantle wedge source; (iii) after melt liberation from the mantle wedge, arc magmas initially become progressively oxidised and isotopically heavier in Fe through fractional crystallisation of ferromagnesian silicates. In summary, reduction consequent to Fe3+-rich melt extraction and subsequent oxidation, likely by S6+-rich fluids, results in a “redox yo-yo” in IAL sources. Fractional crystallisation will further oxidise and elevate δ57Fe in erupting IAL. Iron isotope signatures in IAL record a complex magmatic history with no simple relation between δ57Fe and calculated fO2 in erupted lavas. Records of higher fO2 in subduction zones compared to MORB sources are inherited from the subduction component.

Effect of fluoride on Leydig cell viability, cytotoxicity and oxidative damage in vitro

Water contents of the Martian mantle have previously been investigated using Martian meteorites, with several comprehensive studies estimating the water content in the parental melts and mantle source regions of the shergottites and Chassigny. However, no detailed studies have been performed on the Nakhla meteorite. One possible way to determine the water content of a crystallizing melt is to use the water content in nominally anhydrous minerals (NAMs) such as clinopyroxene and olivine. During or after eruption on the surface of a planetary body and during residence in a degassing magma, these minerals may dehydrate. By reversing this process experimentally, original (pre-dehydration) water concentrations can be quantified. In this study, hydrothermal rehydration experiments were performed at 2 kbar and 700 °C on potentially dehydrated Nakhla clinopyroxene crystals. Rehydrated clinopyroxene crystals exhibit water contents of 130 ± 26 (2σ) ppm and are thus similar to values observed in similar phenocrysts from terrestrial basalts. Utilizing clinopyroxene/melt partition coefficients, both the water content of the Nakhla parent melt and mantle source region were estimated. Despite previous assumptions of a relatively dry melt, the basaltic magma crystallizing Nakhla may have had up to 1.42 ± 0.28 (2σ) wt.% H2O. Based on an assumed low degree of partial melting, this estimate can be used to calculate a minimum estimate of the water content for Nakhla’s mantle source region of 72 ± 16 ppm. Combining this value with values determined for other SNC mantle sources, by alternative methods, gives an average mantle value of 102 ± 9 (2σ) ppm H2O for the Martian upper mantle throughout geologic time. This value is lower than the bulk water content of Earth’s upper mantle (∼250 ppm H2O) but similar to Earth’s MORB source (54–330 ppm, average ∼130 ppm H2O).

2.17–2.10 Ga plutonic episodes in the Mineiro belt, São Francisco Craton, Brazil: U-Pb ages, geochemical constraints and tectonics

Geologic information supported by in situ U-Pb zircon ages, and geochemistry coupled with zircon Hf and Nd-Sr isotopic analyses on orthogneisses, metagranitoids and metadiorites unravel the nature and tectonic significance of a calc-alkaline continental arc in the Mineiro belt – southern portion of the São Francisco Craton. The zircon U-Pb ages defined two plutonic episodes: 2174–2149Ma (metaquartz-diorite, metagranitoids and orthogneisses) and 2145–2100Ma (metadiorites, metagranitoids and orthogneisses) whereas several samples contain inherited zircons with concordant 207Pb/206Pb ages as old as 2785Ma. Our rocks show variable SiO2 contents (54.8–77.1wt.%), are metaluminous to peraluminous and mostly magnesian and variably enriched in LREE. The older rocks have La/YbN=2.24 to 11.87 whereas the metadiorites have La/YbN=9.60–13.82 and the youngest rocks in the range 7.79–91.27. Most samples show Nb and Ti negative anomalies that can be attributed to a subducted source whereas some yield marked Pb positive anomalies that highlight the important role of crustal contamination/assimilation. Dy/Dy* and coupled Dy/Yb values (from 0.7 to 0.5 and 1.4 to 2.0, respectively) indicate the metaigneous rocks derived from a LREE enriched MORB source in a subduction zone with mixed Paleoproterozoic and Archean material. This scenario is consistent with the U/Pb data from inherited nuclei and xenocrysts, the range of the zircon ɛHf(t) (from +8 to −10) and ɛNd(t) (from −7 to +0.3) values, as well as with the Lu-Hf TDM ages that vary between 2.3 and 3.2Ga. Our data and regional geologic–tectonic correlations are consistent with the onset of an ultimate collisional event in the Mineiro belt that followed a long-lived accretionary regime active in Proterozoic times, outboard the Archean core of the São Francisco Craton.

Slab and sediment melting during subduction initiation: granitoid dykes from the mantle section of the Oman ophiolite

New geochemical data are presented for a suite of tonalites, granodiorites, trondhjemites and granites intrusive into depleted mantle harzburgites of the Oman–UAE ophiolite. A detailed field, petrological and geochemical examination suggests that these ‘mantle granitoids’ are the product of three processes: (a) the mixing of melts derived from both mafic and metasedimentary sources, (b) interaction with the mantle harzburgite host and (c) the fractional crystallisation of plagioclase, hornblende ± accessory phases. Geochemical data are used to characterise the identity of the protolith(s) by first screening the data for those samples which have experienced fractional crystallisation during emplacement. The resultant ‘reduced’ data set has moderately fractionated REE, with small negative Eu anomalies and fractionated primitive mantle-normalised trace element patterns with high concentrations of fluid mobile elements and lower concentrations of HFS elements and with positive peaks for Rb and Pb and negative troughs for Ba, Nb, Sr and Ti. The character of the protolith was quantified using a melting model based upon a MORB-type basalt similar in composition to the Oman Geotimes lavas and a model using the MUQ (MUd from Queensland) global sediment composition (Kamber et al. Geochim Cosmochim Acta 69:1041–1058, 2005) both with an amphibolite/granulite facies mineralogy. The two compositions bracket the mantle granitoid data set with partial melts of the MORB source yielding trace element compositions lower than the granitoids, whereas melts of the MUQ source yield melts with compositions higher than the granitoids. Mixing of the calculated melt compositions indicates that the measured granitoid compositions represent between 10 and 30 % mixing of a metasedimentary melt into the melt of a mafic source. Current petrological, structural and geochronological data suggest a model for the origin of the Oman ophiolite in which it is formed by spreading above a subduction zone. The results of this study support this model in three ways: (a) trace element data show that the granitoids have interacted with the mantle, implying that they originated below the mantle section of the ophiolite and were emplaced into the mantle wedge from below; (b) a most probable setting for the melting of a sediment–basalt mix is in the upper part of a subducting slab; (c) the high temperature/shallow melting of the sediment and basaltic slab mix can only have been driven by the high temperatures of the overlying mantle wedge; this is most likely during subduction initiation. Thus, sediment melting during subduction initiation represents an important means of refertilising what is otherwise highly depleted mantle.

Water in Hawaiian peridotite minerals: A case for a dry metasomatized oceanic mantle lithosphere

AbstractThe distribution of water concentrations in the oceanic upper mantle has drastic influence on its melting, rheology, and electrical and thermal conductivities and yet is primarily known indirectly from analyses of OIB and MORB. Here, actual mantle samples, eight peridotite xenoliths from Salt Lake Crater (SLC) and one from Pali in Oahu in Hawaii were analyzed by FTIR. Water contents of orthopyroxene, clinopyroxene, and the highest measured in olivine are 116–222, 246–442, and 10–26 ppm weight H2O, respectively. Although pyroxene water contents correlate with indices of partial melting, they are too high to be explained by simple melting modeling. Mantle‐melt interaction modeling reproduces best the SLC data. These peridotites represent depleted oceanic mantle older than the Pacific lithosphere that has been refertilized by nephelinite melts containing <5 weight % H2O. Metasomatism in the Hawaiian peridotites resulted in an apparent decoupling of water and LREE that can be reconciled via assimilation and fractional crystallization. Calculated bulk‐rock water contents for SLC (50–96 ppm H2O) are on the low side of that of the MORB source (50–200 ppm H2O). Preceding metasomatism, the SLC peridotites must have been even drier, with a water content similar to that of the Pali peridotite (45 ppm H2O), a relatively unmetasomatized fragment of the Pacific lithosphere. Moreover, our data show that the oceanic mantle lithosphere above plumes is not necessarily enriched in water. Calculated viscosities using olivine water contents allow to estimate the depth of the lithosphere‐asthenosphere boundary beneath Hawaii at ∼90 km.

The evolution of MORB and plume mantle volatile budgets: Constraints from fission Xe isotopes in Southwest Indian Ridge basalts

We present high-precision measurements of the fission isotopes of xenon (Xe) in basalts from the Southwest Indian Ridge (SWIR) between 16°E and 25°E. Corrections for syn- to post-eruptive atmospheric contamination yield the Xe isotopic compositions of SWIR mantle sources. We solve for the proportions of mantle Xe derived from the primordial mantle Xe budget, recycling of atmospheric Xe, decay of short-lived 129I, fission of extinct 244Pu, and fission of extant 238U. Xe isotope systematics evident in SWIR basalts and other mantle-derived samples provide new insights into the integrated history of mantle source degassing and regassing. We find that recycled atmospheric Xe dominates the Xe inventories of the SWIR Western and Eastern Orthogonal Supersegment mantle sources (∼80–90% of 132Xe is recycled in origin), consistent with results from studies of plume-influenced basalts from Iceland and the Rochambeau Rift. While significant regassing of the mantle is evident, we also find differences in the extent of degassing of the MORB and plume sources. MORB sources are consistently characterized by a lower fraction of fission Xe derived from Pu-fission, indicating a greater extent of degassing relative to the plume source. The prevalence of recycled atmospheric Xe in mantle sources indicates incorporation of depleted recycled material even into mantle sources with primitive He and Ne isotopic compositions. Consequently, depleted lithophile isotopic compositions in mantle sources with primitive He and Ne cannot be interpreted as evidence for a nonchondritic bulk silicate Earth.

Experimental petrology of peridotites, including effects of water and carbon on melting in the Earth’s upper mantle

For over 50 years, the use of high-pressure piston/cylinder apparatus combined with an increasing diversity of microbeam analytical techniques has enabled the study of mantle peridotite compositions and of magmas derived by melting in the upper mantle. The experimental studies have been guided by the petrology and geochemistry of peridotites from diverse settings and by the remarkable range of mantle-derived magma types. Recent experimental study using FTIR spectroscopy to monitor water content of minerals has shown that fertile lherzolite (MORB-source upper mantle) at ~1,000 °C can store ~200 ppm H2O in defect sites in nominally anhydrous minerals (olivine, pyroxenes, garnet and spinel). Water in excess of 200 ppm stabilizes amphibole (pargasite) at P 3 GPa, water in excess of 200 ppm appears as an aqueous vapour phase and this depresses the temperature of the upper mantle solidus. Provided the uppermost mantle (lithosphere) has H2O 500 ppm H2O) overlying the ‘depleted’ MORB source (~200 ppm H2O). From the study of primitive MOR picrites, the modern mantle potential temperature for MORB petrogenesis is ~1,430 °C. The intersection of the 1,430 °C adiabat with the vapour-saturated lherzolite solidus at ~230 km suggests that upwelling beneath mid-ocean ridges begins around this depth. In intraplate volcanism, diapiric upwelling begins from shallower depths and lower temperatures within the asthenosphere and the upwelling lherzolite is enriched in water, carbonate and incompatible elements. Magmas including olivine melilitites, olivine nephelinites, basanites, alkali picrites and tholeiitic picrites are consequences of increasing melt fraction and decreasing pressure at melt segregation. Major element, trace element and isotopic characteristics of island chain or ‘hot-spot’ magmas show that they sample geochemically distinct components in the upper mantle, differing from MORB sources. There is no evidence for higher-temperature ‘hot-spot’ magmas, relative to primitive MORB, but there is evidence for higher water, CO2 and incompatible element contents. The distinctive geochemical signatures of ‘hot-spot’ magmas and their ‘fixed’ position and long-lived activity relative to plate movement are attributed to melt components derived from melting at interfaces between old, oxidised subducted slabs (suspended beneath or within the deeper asthenosphere) and ambient, reduced mantle. In convergent margin volcanism, the inverted temperature gradients inferred for the mantle wedge above the subducting lithosphere introduce further complexity which can be explored by overlaying the phase relations of appropriate mantle and crustal lithologies. Water and carbonate derived from the subducted slab play significant roles, magmas are relatively oxidised, and distinctive primary magmas such as boninites, adakites and island arc ankaramites provide evidence for fluxing of melting in refractory harzburgite to lherzolite by slab-derived hydrous adakitic melt and by wedge-derived carbonatite.

Water in Hawaiian garnet pyroxenites: Implications for water heterogeneity in the mantle

Mapping the compositional variability of the Earth's mantle is fundamental for understanding mantle dynamics, crustal recycling and melt generation. The geochemistry of intraplate oceanic basalts in particular points to a heterogeneous convecting mantle, often explained by variable proportions of fertile (pyroxenite, eclogite) and depleted (peridotitic) domains. A key parameter necessary to constrain the Earth's deep processes is water because it influences melting and affects the physical properties of the Earth's mantle. However, there are only a few direct water determinations on samples of the oceanic mantle. Here we report water concentrations by FTIR on garnet pyroxenite xenoliths from Salt Lake Crater vent, Oahu, Hawaii, in order to constrain the role of lithological variability in the distribution of water in the upper oceanic mantle. The clinopyroxenes have 260 to 576ppm H2O and the orthopyroxenes about half these amounts. Curiously, garnets have water concentrations below detection limits (<0.5ppm H2O). Calculated melts in equilibrium with the pyroxenes have, on average, 2.3±0.4wt.% H2O and 243±83 H2O/Ce. These H2O concentrations are similar to estimates for the H2O concentration of primary magmas of the rejuvenated Hawaiian stage volcanism, and 3 to 5 times higher than primary magmas of the shield stage. This supports earlier conclusions where the pyroxenites are interpreted as high pressure cumulates from alkali magmas similar to the rejuvenated stage magmas within the Pacific lithosphere. The reconstructed bulk pyroxenites have 211 to 467ppm H2O, similar to estimates for the source of Hawaiian magmas and two to four times higher than estimates for the depleted MORB source (~100ppm H2O). Despite their high water concentrations, however, the bulk pyroxenites have much lower H2O/Ce ratios than the MORB source (35–115 vs. 150–210, respectively) or Hawaiian magma sources (>160). The discrepancy between the low H2O/Ce ratios in the bulk pyroxenites and high H2O/Ce in the equilibrium melts is consistent with experimental data that predicts 4 to 5 times higher partition coefficient for Ce than H in these pyroxenes. Therefore, the process of high-pressure crystallization in the oceanic lithosphere will create pyroxene-rich lithologies which are paradoxically, both “wet” (i.e., high H2O concentrations) and “dry” (low H2O/Ce ratios) compared to the source of their parental melts. Phlogopite is present as a trace phase in these rocks (<0.4% modal) with relatively minor contribution on the bulk water contents. The coupled high H2O, low H2O/Ce ratios of the pyroxenites are similar to the inferred source of several Enriched Mantle (EM)-type Ocean Island Basalts (OIB), like the Samoa, Pitcairn, Society, and Kergulen hot spots, as well as the EM-1 type Koolau endmember of the Hawaiian magmas. We suggest that recycling of pyroxenite-bearing oceanic lithosphere can explain the relatively high H2O and low H2O/Ce ratios of some EM-type OIB. Our data suggests a link between lithological variability and heterogeneous water distribution in the upper mantle.

Petrogenesis of basalt–high-Mg andesite–adakite in the Neoarchean Veligallu greenstone terrane: Geochemical evidence for a rifted back-arc crust in the eastern Dharwar craton, India

The ∼2.7Ga Veligallu greenstone terrane in the eastern Dharwar craton is one of the least studied Neoarchean greenstone belts in India. To our knowledge, there is no published dataset that provides a detailed account of the petrogenetic processes, which culminated into the evolution of this greenstone belt. This paper presents a comprehensive account of field, petrography and geochemistry of the basalt–high-Mg andesite–adakitic volcanic rock association in the Veligallu greenstone belt.The basalts are tholeiitic in nature and are characterized by primitive geochemical compositions with ∼49wt% SiO2, ∼9wt% MgO, Mg# 54, Cr∼197ppm, Ni∼122ppm, Y∼27ppm, Yb∼2.7ppm; Zr/Y<3 and Ce/Yb<5. They do not plot on a mixing line between Mesoarchean TTG and N-MORB, therefore, inconsistent with any interaction with the upper continental crust. Accordingly, they are interpreted to have erupted in an oceanic setting. On the basis of Nb/Th ratio the Veligallu basalts are divided onto two groups. The basalts with Nb/Th>8 display systematically flat to light-REE depleted chondrite normalized REE patterns (LaN/YbN=0.76–1.0), in contrast to the samples with Nb/Th<8, which display slight enrichments in light-REE (LaN/YbN=1.0–1.6). The depleted basalts are characterized by primitive mantle normalized Nb concentrations lower than La but equal to Th, in contrast to the enriched basalts, which are characterized by slight depletions of Nb relative to both La and Th. Collectively, as a group, they display Nb/Yb ratio higher than the N-MORB. The enrichment in Nb is attributed to moderate to low degree (∼10%–4%) partial melting of a depleted MORB source. The trace, REE and their interelement ratios in the Veligallu basalts exhibit transitional characteristics between the average N-MORB and the typical Phanerozoic island arc basalts. Overall, the geochemical characteristics of the Veligallu basalts are identical to basalts erupted in the Phanerozoic back-arc basins.The Veligallu high-Mg andesites are orthopyroxene-bearing calc-alkaline andesites and are characterized by moderate SiO2 (∼56wt%), high MgO (∼8wt%), Mg# (61–67), Cr (174–330ppm) and Ni (173–242ppm) contents; low Y (∼13ppm) and Yb (∼1.2ppm) concentrations, relative to typical island arc andesites. They exhibit light-REE enriched (LaN/SmN∼1.7) and heavy-REE depleted (GdN/YbN∼1.9) fractionated REE patterns. On a primitive mantle normalized trace element variation diagram they exhibit negative Nb and Ti, and contrasting positive Zr anomalies relative to the neighboring REE. The La–Yb systematics suggest that these high-Mg andesites were generated by hydrous partial melting of a slab-melt metasomatized mantle wedge in the sub-arc mantle. The felsic volcanic rocks are characterized by high SiO2∼73wt%, Na2O∼4.5wt%, and significantly low concentrations of MgO (∼0.72wt%), Cr∼19ppm, Ni∼23ppm, Y∼5.9, and Yb∼0.44ppm; strongly fractionated chondrite normalized REE patterns (LaN/YbN∼30) with occasionally minor positive Eu anomaly, and primitive mantle normalized negative Nb and Ti, and contrasting positive Zr anomaly. The geochemical attributes of the Veligallu felsic volcanic rocks are analogous to the Cenozoic high silica-type adakites that are produced by partial melting in the sub-arc mantle, and presumably without any significant interaction of the slab-melts with the overlying mantle wedge.The juxtaposed occurrence of basalts with the banded iron formations, and their close spatial association with the high-Mg andesites, in the Veligallu greenstone belt, indicates that the petrogenesis of these volcanic rocks is consistent with their eruption in an intraoceanic setting. The close proximal association of the back-arc type basalts, and the high-Mg andesites and adakites within the Veligallu greenstone belt potentially suggests that this greenstone belt represents a rifted back-arc crust. This event appears to have significantly contributed to the peak of crustal growth activity in the eastern Dharwar craton at 2.7Ga.

Peridotites from the Kamchatsky Mys: evidence of oceanic mantle melting near a hotspot

A suite of mantle peridotites sampled in the Kamchatsky Mys includes spinel lherzolite, clinopyroxene-bearing harzburgite, and harzburgite. Mineral chemistry of olivine, chromian spinel, and clinopyroxene show strongly correlated element patterns typical of peridotite formed by 8% to more than 22% partial melting. Clinopyroxene in the Kamchatka peridotites is compositionally different from that of both abyssal and suprasubduction varieties: Clinopyroxene in lherzolite is depleted in LREE relative to abyssal peridotite and that in harzburgite has very low LREE and Sr unlike the subduction-related counterpart. These composition features indicate that the rocks ultra-depleted in basaltic components originated in the vicinity of a hotspot, possibly, proto-Hawaiian plume, which provided high temperature and melting degree of the MORB source mantle at mid-ocean ridge.

Water in the Moon's interior: Truth and consequences

Geochemical data for H2O and other volatiles, as well as major and trace elements, are reported for 377 samples of lunar volcanic glass from three chemical groups (A15 green, A15 yellow, A17 orange 74 220). These data demonstrate that degassing is a pervasive process that has affected all extrusive lunar rocks. The data are combined with published data to estimate the total composition of the bulk silicate Moon (BSM). The estimated BSM composition for highly volatile elements, constrained by H2O/Ce ratios and S contents in melt inclusions from orange glass sample 74 220, are only moderately depleted compared with the bulk silicate Earth (avg. 0.25X BSE) and essentially overlap the composition of the terrestrial depleted MORB source. In a single giant impact origin for the Moon, the Moon-forming material experiences three stages of evolution characterized by very different timescales. Impact mass ejection and proto-lunar disk evolution both permit system loss of H2O and other volatiles on timescales ranging from days to centuries; the early Moon is likely to have accreted from a thin magma disk of limited volume embedded in, but largely displaced from, the extended distribution of vapor around the Earth. Only the protracted evolution of the lunar magma ocean (LMO) presents a time window sufficiently long (10–200 Ma) for the Moon to gain water during the tail end of accretion. This “hot start” to lunar formation is however not the only model that matches the lunar volatile abundances; a “cold start” in which the proto-lunar disk is largely composed of solid material could result in efficient delivery of terrestrial water to the Moon, while a “warm start” producing a disk of 25% volatile-retentive solids and 75% volatile-depleted magma/vapor is also consistent with the data. At the same time, there exists little evidence that the Moon formed in a singular event, as all detailed planetary accretion models predict several giant impacts in the terrestrial planet region in which the Earth forms. It is thus conceivable that the Moon, like the Earth, experienced a history of heterogeneous accretion.

Constraints from Os-isotope variations on the origin of Lena Trough abyssal peridotites and implications for the composition and evolution of the depleted upper mantle

The Lena Trough is a highly oblique, sparsely magmatic, ultra-slow spreading center located at the smallest distance between North America and Eurasia in the Arctic basin. Competing models suggest that it is either floored by oceanic mantle abyssal peridotites (APs) exposed by lithospheric necking, or by subcontinental mantle exposed in a still juvenile rift. To distinguish between these hypotheses, we have examined mineral major and trace element and whole rock Os-isotope variations in Lena Trough peridotites. Lena Trough peridotites are predominantly LREE-depleted, similar to other AP suites, and have 187Os/188Os ranging from ∼0.118 to 0.130 (Ave.=0.1244). This distribution is nearly identical to that of abyssal peridotites globally. Both the REE patterns and the Os-isotope distribution of the Lena Trough peridotites differ starkly from subcontinental mantle xenoliths sampled at Svalbard adjacent to Lena Trough. This suggests that Lena Trough is a site of oceanic spreading, although mid-ocean ridge volcanism as such has not yet begun.Highly refractory APs from several settings have Os- and Hf-isotope compositions indicating ancient (>1 Ga) melt depletion. Some researchers have proposed that at least some APs do not directly sample the convecting upper mantle source of MORB, but instead sample highly melt-depleted residues either entrained in the convecting mantle or present as a buoyant “slag” floating atop the less-depleted MORB-source mantle. However, ocean island peridotite xenoliths and APs reveal an essentially identical, non-Gaussian distribution of Os-isotopes and also span a similar range in Hf-isotopes. The similar mean and distribution of Os-isotopes between APs and ocean island xenoliths indicate that these two sample types derive from the same heterogeneous mantle reservoir. This similarity is inconsistent with the AP “slag hypothesis” due to the significantly greater depth of origin of ocean island xenoliths with respect to APs.Global correlation between peridotite fertility indices (e.g., whole rock Al2O3, spinel Cr#) and 187Os/188Os in both APs and ocean island xenoliths suggests ancient melting events have played a role in their petrogenesis, with an average age of melt depletion of ∼1.5 Ga. However, APs display a much steeper correlation than ocean island xenoliths, but have a similar y-intercept. The contrasting trends can be explained by recent on-ridge melt extraction superimposed on an ancient melt-depletion trend. The average Os-isotope composition of both APs and ocean island xenoliths is distinctly sub-chondritic, and requires significantly greater average melt depletion of the upper mantle (>10%) than suggested by previous inversions of MORB compositions.

Partial melting control of water contents in the Cenozoic lithospheric mantle of the Cathaysia block of South China

Major and trace element and the H2O contents of minerals in peridotite xenoliths hosted by the Cenozoic basalts of Jiande in the Cathaysia block were evaluated using electron microprobe, laser-ablation ICP-MS and Fourier transform infrared spectroscopy, respectively. The correlations among the major elements of the minerals define a melting trend, and modeling of the Y and Yb contents in the clinopyroxenes indicates that the degree of partial melting ranges from 1% to 15%. Most samples (22 out of 29) show depleted chondrite-normalized rare earth element patterns and a degree of partial melting <4%. The H2O contents (weight in ppm) of the clinopyroxenes, orthopyroxenes and olivines are 390–590ppm, 160–330ppm and ∼0ppm, respectively. Although potential H-loss during xenolith ascent cannot be excluded for olivines, pyroxenes largely preserve the H2O content they have in the mantle prior to sampling by the host basalts, as inferred from (1) the homogenous H2O content within single pyroxene grains, and (2) the equilibrium H2O partitioning between the clinopyroxene and orthopyroxene. Based on the mineral modes and assuming a partition coefficient of 10 for H2O between the clinopyroxene and olivine, the calculated whole-rock H2O contents range from 90 to 220ppm, similar to that of the MORB source. When combined with previously reported data for peridotites hosted by Cenozoic basalts at other localities of the Cathaysia block, the correlation of H2O content with melting index (such as Yb content in cpx, Cr# in spinel) suggests that extent of partial melting is the main factor controlling H2O abundance in these rocks. Nevertheless, the variations in whole-rock H2O contents cannot be perfectly modeled as a simple modal melting process using available experimental partition coefficients of H2O between peridotite and melt. The lack of coherent variations between the H2O contents of the whole rocks and the metasomatic index (La/Yb ratio in clinopyroxene) indicates that mantle metasomatism did not modify the initial H2O contents after the melting event(s). Based on the similarities in major and trace elements, and of H2O contents in the Cathaysia block peridotites and those inferred for the MORB source, we propose that the Cenozoic lithospheric mantle of the Cathaysia block is accreted from the upwelled and cooled asthenospheric mantle. In addition, the H2O contents of the Cenozoic lithospheric mantle of the Cathaysia block are much higher than those of the North China Craton of similar fertility (100ppm vs. 20ppm for whole rock H2O contents), where lithospheric thinning occurred during the Mesozoic. This implies that either the Cathaysia block did not undergo similar lithospheric thinning or that the mechanism of the lithospheric thinning was different.

238U–230Th equilibrium in arc magmas and implications for the time scales of mantle metasomatism

Large excesses of 238U and 226Ra relative to 230Th characterize many arc magmas and are commonly interpreted to represent recent addition of slab-derived fluid to the mantle wedge beneath the arc. A significant proportion of arc magmas are, however, in 238U–230Th radioactive equilibrium. This is generally thought to result from “buffering” of the young slab fluid U-series signal by a sediment component in secular equilibrium. Here we present new 238U–230Th–226Ra and 235U–231Pa measurements for historic andesites from Volcán de Colima, Mexico. In all lavas (230Th/238U) are in equilibrium, whereas (231Pa/235U) and (226Ra/230Th) are significantly greater than one. These data demonstrate that arc magmas with (230Th/238U) equilibrium can have significant 231Pa and 226Ra excesses, precluding ageing of the magmas in the crust as the cause of 230Th/238U equilibrium. Quantitative modeling of metasomatic and melting processes further indicates that addition of sediment melts to a depleted mantle wedge produces significant 230Th excesses and that 238U excesses induced by recent addition of fluids derived from the altered oceanic crust are not sufficient to compensate these 230Th excesses. U-series activity ratios in Colima magmas are best explained by models in which the metasomatised mantle returns to secular equilibrium before melting, implying a time lag ⩾350 kyr, with subsequent production of 231Pa and 226Ra excesses by in-growth during melting rather than by addition of slab fluids. Investigation of a global compilation of U-series data in arc magma indicates that our model proposed for Colima applies to most arc magmas in or near (230Th–238U) equilibrium. The time lag between mantle metasomatism and melting appears to vary between hundreds years to more than 350 kyr in subduction zones. We posit that the absence of U/Th elemental fractionation during melting of arc sources in (230Th/238U) equilibrium reflects a higher fO2 compared to MORB sources that yield magmas with 230Th excesses.

Hafnium–neodymium isotope systematics of the 2.7 Ga Gadwal greenstone terrane, Eastern Dharwar craton, India: Implications for the evolution of the Archean depleted mantle

The Neoarchean Gadwal greenstone belt in the eastern Dharwar craton, India, hosts a well preserved metavolcanic sequence that is dominated by tholeiitic and calc-alkaline basalt-andesite-dacite-rhyolite series, which includes boninitic geochemical varieties. Bulk-rock Lu–Hf and Sm–Nd isotope systematics of these apparently arc-related volcanic rocks yield indistinguishable ages of 2.701±0.024Ga and 2.702±0.026Ga, respectively. On the basis of the close spatial association and identical ages of the different rock types we suggest 2.70±0.03Ga as the age of crystallization of the different rock types within the Gadwal metavolcanic sequence. In contrast, bulk-rock Pb–Pb isotope systematics of the same samples yield a significantly younger and less precise age of 2.466Ga (+0.068/−0.110Ga). We tentatively interpret this younger age to represent a metallogenic and crustal reworking event in the Dharwar craton, which disturbed the U–Pb system but not the Lu–Hf or Sm–Nd systems. The Gadwal metavolcanic rocks have positive initial εHf(2.70Ga)=+1.6 to+8.7 and slightly negative to positive εNd(2.70Ga)=−0.1 to+3.0 values, consistent with an origin from a long term depleted source relative to a chondritic reservoir at ∼2.7Ga. Lack of correlation between initial isotopic compositions and major or trace element indices of fractionation and alteration suggest that the observed isotope variability probably reflects compositional variation in the Gadwal source, similar to that observed in modern day island arcs. Two boninitic samples of the Gadwal sequence have εHf∼8.3 and 8.7, and are more radiogenic than average depleted mantle for the time period 3.2 to 2.5Ga (εHf=4 to 6). Early (perhaps Hadean) differentiation events that led to a depleted and heterogeneous mantle are apparent in the Nd and Hf isotope systematics of 3.7–3.8Ga Isua supracrustal rocks. The radiogenic Hf isotopes of the Gadwal boninites and the Hf, Nd isotope systematics of rocks from other locations in the 3.4 to 2.5Ga time period are consistent with the survival of fragments of an early depleted mantle later in the Archean. From ∼2.0Ga to present, the time-integrated 176Lu/177Hf and 147Sm/144Nd of the depleted mantle appears nearly constant and similar to the present day average MORB source. These data indicate that progressive elimination of early (>4.5Ga) formed heterogeneities in the depleted mantle dominated the history of the Archean mantle, and that portions of early depleted reservoirs survived through the Mesoarchean. These results have implications for the mixing scales for the early terrestrial mantle and the timing of the initiation of present day plate tectonics.

The effects of K2O on the compositions of near-solidus melts of garnet peridotite at 3 GPa and the origin of basalts from enriched mantle

Enrichment in K2O in oceanic island basalts (OIB) is correlated with high SiO2, low CaO/Al2O3, and radiogenic isotopic signatures indicative of enriched mantle sources (EM1 and EM2). These are also chemical characteristics of the petit-spot lavas, which are highly enriched in K2O (3–4 wt%) compared to other primitive oceanic basalts. We present experimentally derived liquids with varying concentrations of K2O in equilibrium with a garnet lherzolite residue at 3 GPa to test the hypothesis that the major element characteristics of EM-type basalts are related to their enrichment in K2O. SiO2 is known to increase with K2O at pressures less than 3 GPa, but it was previously unknown if this effect was significant at the high pressures associated with partial melting at the base of the lithosphere. We find that at 3 GPa for each 1 wt% increase in the K2O content of a garnet lherzolite saturated melt, SiO2 increases by ~0.5 wt% and CaO decreases by ~0.5 wt%. MgO and $$K_{D}^{{{\text{Fe}} - {\text{Mg}}}}$$ each decrease slightly with K2O concentration, as do Na2O and Cr2O3. The effect of K2O alone is not strong enough to account for the SiO2 and CaO signatures associated with high-K2O OIB. The SiO2, CaO, and K2O concentrations of experimentally derived partial melts presented here resemble those of petit-spot lavas, but the Al2O3 concentrations from the experimental melts are greater. Partitioning of K2O between peridotite and melt suggests that petit spots, previously considered to sample ambient asthenosphere, require a source more enriched in K2O than the MORB source.

Pb and Hf isotope variations along the Southeast Indian Ridge and the dynamic distribution of MORB source domains in the upper mantle

Lead and Hf isotope data for a suite of closely spaced (5–10km) basalt glasses, sampled over a distance of ∼1140km along the Southeast Indian Ridge (SEIR) between 88°E and 100°E, confirm the presence of ancient heterogeneous mantle beneath the Indian Ocean. The Pb isotopes show a non-Gaussian, “fat tail” distribution that is bimodal. The paired Hf and Pb isotope data, combined with previously published data, define two populations that reveal the presence of compositional streaks within the upper mantle. The number and density of the streaks follow a Poisson distribution where the characteristic streak thickness is ∼20km. The mantle ‘unit’ giving rise to each MORB sample represents a ‘mixture of mixtures’ with a multi-stage mixing history. There is no correlation of K/Ti or Na8 with Pb or Hf isotope ratios, and both domains encompass both N- and E-MORB sources, ruling out two different regimes of melting in a homogeneous source. Apparently, mantle convection has folded together distinct composite reservoirs of heterogeneous mantle, and stretched them into streaks that remain coherent and discernable units. The Hf and Pb isotope data reveal the presence of two distinct DUPAL mantle components in the upper mantle. The isotope signature of component 1 resembles a continental lithosphere source having Pb isotope composition similar to Karoo picrites (erupted at ∼180Ma). This component likely originated by mixing of Kaapvaal cratonic lithosphere and/or lower crust with the asthenosphere through mantle plume–lithosphere interaction during the early rifting stages of Gondwana. Component 2 likely originated from mixing of recycled oceanic lithosphere plus sediment with the asthenosphere (which had been previously contaminated by component 1) during the upwelling of Indian Ocean mantle plumes over the last 180Ma. The variation in Pb and Hf isotopes along the SEIR reveals that, while the two DUPAL components are spatially distributed in a Poisson fashion, and therefore well-mixed, their relative proportions are spatially variable in the Indian Ocean upper mantle.

English

&nbsp; Sr-Nd-Pb isotopic studies have been carried out on the lavas of Mt. Oku, the central part of the continental sector of the Cameroon volcanic line (CVL). The lavas define a considerable range of Sr-Nd-Pb isotopic compositions characterized by&nbsp;143Nd/144Nd = 0.512626 &ndash; 0.512937,87Sr/86Sr = 0.699411 - 0.704279 and more radiogenic isotopic ratios&nbsp;206Pb/204Pb = 18.3800 &ndash; 19.4480.&nbsp;207Pb/204Pb = 15.4216 &ndash; 15.6448,&nbsp;208Pb/204Pb = 37.1114 &ndash; 39.7507, with radiogenic Pb similar to the HIMU of the continent ocean boundary of the Cameroon volcanic line. These isotopic data display a linear array on the&nbsp;143Nd/144Nd vs&nbsp;87Sr/86Sr diagrams trending from MORB towards bulk earth.&nbsp;143Nd/144Nd and&nbsp;87Sr/86Sr vs.&nbsp;206Pb/204Pb show a continuum of compositions that suggest differing contributions to the parental magmas from three end-members. One end-member corresponds to a mixture of high uranium high lead (HIMU) and depleted mantle (DM) source components while the other end-member corresponds to an enriched mantle source thought to be the MORB source or sub-lithospheric source. The enriched geochemical signatures (EM) of the Oku magmas are unlikely to be the result of crustal assimilation during magma ascent. Rather, they are thought to be derived from a sub-continental lithospheric mantle enriched in incompatible trace elements by ancient metasomatic processes. The HIMU component is thought to have been inherited from the fossil plume that underlay the Equatorial Atlantic during the period 130 to 100 Ma before migrating to its present site. The radiogenic initial&nbsp;206Pb/204Pb ratios of the samples (~18.3800 &ndash; 19.44804) furthermore require the involvement of an ancient HIMU mantle plume in the magmatism and this may have been formed by metasomatism of fluids derived from St. Helena hot plume ca 120 Ma. &nbsp; Key words:&nbsp; Oku, isotopes, metasomatism, HIMU, enriched geochemical signatures (EM), mantle, enrichment.

The origins of volatiles in the terrestrial planets

The origins of volatiles in the terrestrial planets

Mechanisms of magma degassing at mid-oceanic ridges and the local volatile composition (4He–40Ar⁎–CO2) of the mantle by laser ablation analysis of individual MORB vesicles

Individual vesicles in <1cm3 samples of MORB glasses commonly preserve significantly different volatile (He, Ar, CO2) compositions. Five separate MORB glass samples from different ridges were investigated, selectively opening targeted vesicles using a 193nm excimer laser. One sample shows evidence of syn-emplacement selective helium loss from the vesicles in the glass rim close to the crystallised zone of the pillow lava, proving that He contents and low He/Ar ratios do not always reflect mantle and/or magmatic processes. However the composition of the different vesicles of three of these samples covers large ranges in He/Ar and Ar/CO2 with linear variations in plots of ln(He/Ar) vs. ln(Ar/CO2) which are consistent with a Rayleigh distillation at equilibrium and allow the relative abundances of the volatiles in the MORB mantle source to be estimated by correcting for degassing processes on a sample-by-sample basis. This technique presents a new tool for characterising and correcting for volatile fractionation processes that have modified the initial mantle source composition. The results confirm a heterogeneous CO2/3He ratio in the MORB source regions with the Azores mantle source enriched in C/3He compared to that of N-MORBs. The considerable heterogeneity in vesicle compositions within such a small volume (<1cm3) requires injection of less degassed magma at shallow level before eruption. The maximum time interval between magma mixing and quenching on the seafloor is of the order of a few hours.