Search

This work was supported by NSF grants EAR-1447404 and EAR-1650033, and NERC grant NE/R002134/1. PS would also like to cite the support of a Carnegie Trust Research Incentive Grant, which helped the setup of various isotope techniques in the St Andrews Isotope Geochemistry (STAiG) laboratories. FM thanks the ERC under the European Community’s H2020 framework program/ERC grant agreement # 637503 (Pristine) and for the UnivEarthS Labex program (no. ANR-10-LABX-0023 and ANR-11-IDEX-0005-02). Parts of this work were supported by IPGP multidisciplinary program PARI, and by Region île-de-France SESAME Grant (no. 12015908). The silicon isotope composition of detrital quartz and zircon have the potential to inform us about secular changes to the silica cycle and weathering reactions on Earth. However, inferring source melt Si isotope composition from out-of-context minerals is hampered by the fact that, to-date, there is limited Si isotope equilibrium fractionation data for minerals. Here, we report experimental data to constrain Si isotope equilibrium fractionation between zircon and quartz, using two fundamentally different strategies, but with the same experimental design. First, zircon and quartz were hydrothermally synthesized from Zr(OH)4 and SiO2 at 1.5 GPa and temperatures of 725, 800, and 900 oC. The second experimental strategy utilized the three-isotope method; the starting materials consisted of natural zircon and isotopically-labelled SiO2. Three sets of hydrothermal time-series experiments were conducted at the same pressure and temperatures as the direct synthesis experiments. For all experiments, quartz and zircon were separated and 30Si/28Si and 29Si/28Si ratios were measured by solution multi-collector inductively coupled plasma mass spectrometry. The three-isotope method, which provides the best indicator of equilibrium fractionations, yields the following relationship: Δ30Si(qtz-zrc) = (0.53±0.14) × 106 /T2 where Δ30Si(qtz-zrc) is the relative difference in 30Si/28Si between quartz and zircon in permil, T is temperature in K, and the error is 2 s.e. This relationship can be used to calculate the fractionation between zircon and other phases, and to estimate the Si isotope composition of the melt from which a zircon crystallized. The results may be used to assess equilibrium-disequilibrium isotope fractionations between quartz and zircon and co-existing phases in igneous rocks. These data can also be applied to out-of-context zircon (and quartz) to estimate the isotope composition of the host rock. Zircons crystallizing from a melt derived from purely igneous sources – i.e., without the involvement of “weathered” material – are expected to display a δ30SiNBS-28 (permil deviation of the 30Si/28Si from the NBS-28 standard) range from -0.7 to -0.35‰. Deviations from this range indicate assimilation of non-igneous (i.e., sedimentary) material in the melt source. Postprint

FM acknowledge funding from ERC grant agreement No. 101001282 (METAL), the UnivEarthS Labex program (numbers: ANR-10-LABX-0023 and ANR-11-IDEX-0005-02), the IPGP multidisciplinary program PARI, the Region île-de-France SESAME Grants no. 12015908, EX047016, and the IdEx Université de Paris grant, ANR-18-IDEX-0001 and the DIM ACAV+. The source of and timing of delivery of the volatile elements to Earth is a question that is fundamental to understanding how our planet evolved. Here, we show that primitive meteorites have resolved mass-independent Zn isotope anomalies from the terrestrial reservoir. Carbonaceous chondrites (CC), likely originating from the outer Solar System are distinct from non-CC, and Earth is intermediate between these two components. Modelling based on these data indicates that around 30% of Earth's budget of Zn and other moderately volatile material derives from the participation of 6% of CC-like materials during Earth's accretion, with the remaining coming from NC meteorites. This implies that, despite the relatively minor mass of Earth thought to derive from CC-like material, the CC component of Earth was relatively and significantly volatile-enriched; this is in line with the observation that the terrestrial elemental abundance pattern of moderately volatile elements could be explained by a carbonaceous source, and with the carbonaceous chondrite-like isotopic budget of more volatile-rich material accreted later in Earth's accretion history (e.g. Hg, Se, N, noble gases). Publisher PDF

Silicon (Si) is the defining element of silicate reservoirs yet, despite its dominance in major Earth processes, there is still no clear understanding of how much is hosted in Earth's core, how the enriched continental crust forms or even if the crust is isotopically different from the mantle because of a long history of weathering, erosion and subduction. With the advent of multiple collector inductively coupled plasma mass spectrometry it has become relatively straightforward to explore small (100 ppm level) mass dependent variations in Si isotopic composition resulting from high temperature fractionation and to develop new isotopic fingerprints of magmatic processes and source regions. This paper reviews the technique developments, the new data and the veracity of current interpretations. Only a small Si isotopic effect is associated with basalt formation via mantle melting. However, there now is compelling evidence, based on a considerable number of samples (> 100), that the silicate Earth is isotopically fractionated by 100–200 ppm per amu to a heavy composition relative to that of chondrites and also all differentiated stony meteorites. This could reflect variability in the circumstellar disc, but this is not well supported by data for enstatite chondrites which are isotopically light. The most plausible current explanation is that Si is a light element in Earth's core and that differences in the bond stiffness between silicate- and metal-hosted Si resulted in substantial fractionation. Interestingly, the Moon has the same Si isotope composition as Earth's mantle, which is hard to explain unless the Moon's atoms were mainly derived from Earth. Differentiated magmatic sequences such as those of Hekla, Iceland display a systematic relationship between isotopic composition and Si content. More complex magmatic suites, such as I- and S-type granites, reveal a range of isotopic compositions not well correlated with chemical composition. Similar effects are found in lower crustal granulite facies xenoliths. Nonetheless the overall composition of the continental crust is only slightly heavy relative to the mantle; in fact the effect is barely resolvable. Therefore, the amounts of surface dissolved heavy Si removed to the mantle via weathering over time have been small or these losses have been balanced by subduction of isotopically light clay.

International audience; Core formation is the main differentiation event in the history of a planet. However, the chemical composition of planetary cores and the physicochemical conditions prevailing during core formation remain poorly understood. The asteroid 4-Vesta is the smallest extant planetary body known to have differentiated a metallic core. Howardite, Eucrite, Diogenite (HED) meteorites, which are thought to sample 4-Vesta, provide us with an opportunity to study core formation in planetary embryos. Partitioning of elements between the core and mantle of a planet fractionates their isotopes according to formation conditions. One such element, silicon, shows large isotopic fractionation between metal and silicate, and its partitioning into a metallic core is only possible under very distinctive conditions of pressure, oxygen fugacity and temperature. Therefore, the silicon isotope system is a powerful tracer with which to study core formation in planetary bodies. Here we show through high-precision measurement of Si stable isotopes that HED meteorites are significantly enriched in the heavier isotopes compared to chondrites. This is consistent with the core of 4-Vesta containing at least 1 wt% of Si, which in turn suggests that 4-Vesta's differentiation occurred under more reducing conditions (ΔIW∼−4) than those previously suggested from analysis of the distribution of moderately siderophile elements in HEDs.

The upper continental crust (UCC) is the major source of silicon (Si) to the oceans and yet its isotopic composition is not well constrained. In an effort to investigate the degree of heterogeneity and provide a robust estimate for the average Si isotopic composition of the UCC, a representative selection of well-characterised, continentally-derived clastic sediments have been analysed using high-precision MC-ICPMS. Analyses of loess samples define a narrow range of Si isotopic compositions (δ30Si = −0.28‰ to −0.15‰). This is thought to reflect the primary igneous mineralogy and predominance of mechanical weathering in the formation of such samples. The average loess δ30Si is −0.22 ± 0.07‰ (2 s.d.), identical to average granite and felsic igneous compositions. Therefore, minor chemical weathering does not resolvably affect bulk rock δ30Si, and loess is a good proxy for the Si isotopic composition of unweathered, crystalline, continental crust. The Si isotopic compositions of shales display much more variability (δ30Si = −0.82‰ to 0.00‰). Shale Si isotope compositions do not correlate well with canonical proxies for chemical weathering, such as CIA values, but do correlate negatively with insoluble element concentrations and Al/Si ratios. This implies that more intensive or prolonged chemical weathering of a sedimentary source, with attendant desilicification, is required before resolvable negative Si isotopic fractionation occurs. Shale δ30Si values that are more positive than those of felsic igneous rocks most likely indicate the presence of marine-derived silica in such samples. Using the data gathered in this study, combined with already published granite Si isotope analyses, a weighted average composition of δ30Si = −0.25 ± 0.16‰ (2 s.d.) for the UCC has been calculated.

The zinc stable isotope system has been successfully applied to many and varied fields in geochemistry, but to date it is still not completely clear how this isotope system is affected by igneous processes. In order to evaluate the potential application of Zn isotopes as a proxy for planetary differentiation and volatile history, it is important to constrain the magnitude of Zn isotopic fractionation induced by magmatic differentiation. In this study we present high-precision Zn isotope analyses of two sets of chemically diverse, cogenetic samples from Kilauea Iki lava lake, Hawaii, and Hekla volcano, Iceland, which both show clear evidence of having undergone variable and significant degrees of magmatic differentiation. The Kilauea Iki samples display small but resolvable variations in Zn isotope composition (0.26‰

The silicon (Si) cycle is of great current interest but the isotopic composition of the continental crust has not been determined. Magmatic differentiation generates liquids with heavier Si and the lower crust, thought to be dominated by cumulates and restites, is predicted to have a light isotopic composition. This is borne out by the composition of many types of granite, which appear to have relative light Si for their silica content. Here we report the Si isotopic compositions of two granulite facies xenolith suites, from the Chudleigh and McBride volcanic provinces, Australia, providing new constraints on deep crustal processes and the average composition of the deep continental crust. The xenoliths display a range of isotopic compositions (δ30Si=−0.43‰ to −0.15‰) comparable to that measured previously for igneous rocks. The isotopic compositions of the McBride xenoliths reflect assimilation and fractional crystallisation (AFC) and/or partial melting processes. Silicon and O isotopes are correlated in the McBride suite and can be explained by AFC of various evolved parent melts. In contrast, the Chudleigh xenoliths have Si isotope compositions predominantly controlled by the specific mineralogy of individual cumulates. Using the xenolith data and a number of weighting methods, the Si isotope compositions of the lower and middle crust are calculated to be δ30Si=−0.29±0.04‰ (95% s.e.) and −0.23±0.04‰ (95% s.e.) respectively. These values are almost identical to the composition of the Bulk Silicate Earth, implying minimal isotope fractionation associated with continent formation and no light lower crustal reservoir.

Of the primitive meteorite groups, the enstatite chondrites are among the most chemically dissimilar to terrestrial, which should preclude them as major components of the proto-Earth. However, for many isotope systems (most notably oxygen), enstatite chondrites show very little variation away from terrestrial, hinting at a common origin. One isotope system which appears to differ from this trend is silicon, but no satisfactory explanation has been proposed as to why these meteorites should be significantly lighter than both the silicate Earth and other primitive meteorite groups. This study presents a comprehensive investigation into the Si isotope composition of enstatite chondrites and their differentiated counterparts, the aubrites, and confirms that these meteorites are, with respect to Si isotopes, the lightest macroscale solar system objects so far analysed. Crucially, the results show that EH chondrites are significantly lighter (δ30Si=−0.77±0.08‰) than EL chondrites (δ30Si=−0.59±0.09‰) and aubrites (δ30Si=−0.60±0.11‰). Silicon isotope analyses of the metal-free components of EH and EL reveal that these are identical, within error, to each other and with carbonaceous/ordinary chondrite bulk measurements (viz., δ30Si ∼−0.47), which is taken as evidence that Si isotope variation in the nebular gas is not the cause of the light Si isotope enrichment in enstatite chondrites. From this, one can infer that silicates condensing from the solar nebula over a wide range of compositions and, presumably, heliocentric distances have very similar Si isotope compositions. A statistically significant negative correlation between bulk δ30Si and Si content in enstatite chondrite kamacite indicates that the presence of isotopically light Si in the metal phase (as the result of formation under reducing conditions) is the principal cause of the bulk enrichment in lighter Si isotopes. Based on the Si isotope offset between enstatite meteorites and the silicate Earth, the extant enstatite meteorites cannot represent a significant proportion of the material that accreted to form the proto-Earth, as unfeasibly large amounts of Si would have to enter the core to satisfy mass balance equations. However, we posit that material enriched in refractory lithophile elements which condensed in the same region as E-chondrites, could still be an important component.

Because weathering results in preferential incorporation of the lighter Si isotopes into clay minerals, the Si in magmas derived from (meta)sedimentary lithologies enriched in such phases should also be isotopically light, relative to melts derived from igneous sources. With the advent of new high precision techniques it should therefore be possible to resolve such differences. Hence, Si isotopes have to potential to distinguish between, and characterise, different magmatic source regions.To explore this, here we report the Si isotopic compositions of a suite of Phanerozoic I-, S- and A-type granites and mineral separates, as well as samples of associated country rock. The whole-rock data fall between δ 30Si=-0.40 and -0.11±0.04‰ (95%s.e.). This range is more limited than that defined by previous low precision measurements of granites, but is significantly broader than that displayed by more recent, high precision, measurements of high-Si extrusive igneous rocks. As predicted, S-type granites are, on average, isotopically lighter than I- and A-type samples, and negative correlations with alumina saturation index (ASI) values and initial 87Sr/ 86Sr provide evidence that Si isotope variations in granites can be explained by variable incorporation of an enriched sedimentary component. However, the relatively small isotopic perturbations and apparent degree of overlap between granite affinities suggest that Si isotopes are not as sensitive to sedimentary input, relative to more established proxies such as, for example, O isotopes. The average Si isotope value of all granites analysed in this study is δ 30Si=-0.23±0.15‰ (2s.d.), which is entirely consistent with a recently published estimate for the continental crust, based on an andesitic bulk composition and the well defined relationship between Si isotopes and magmatic fractionation. © 2012 Elsevier Ltd.