Ben Harte

Determination of partition coefficients between apatite, clinopyroxene, amphibole, and melt in natural spinel lherzolites from Yemen: Implications for wet melting of the lithospheric mantle

Plio-Quaternary basaltic volcanic fields along the south coast of Yemen contain mantle xenoliths of anhydrous and metasomatized amphibole-(± apatite) bearing spinel lherzolites. Partial melting of clinopyroxene and/or amphibole in a closed system produced silicate glass from which new clinopyroxene, olivine, and spinel have crystallized. Partition coefficients between mineral phases in the matrix and particularly those (apatite, clinopyroxene, amphibole) formed during the metasomatic event, and between glass and the new clinopyroxene crystallized from it have been calculated by analyzing trace element composition with an ion microprobe. Partition coefficients between apatite and clinopyroxene or amphibole are large (e.g., ≫ 1) for the REEs, Ba, Sr, Y, and Hf and confirm that apatite, when present in the mantle, is an important repository for these elements. On the other hand, partitioning between amphibole and clinopyroxene is very close to unity for most of the analyzed trace elements, except for Ba, Nb, Zr, and Hf. Trace element ratios such as Zr/Nb, Ba/Zr, or Ba/Nb can thus be used in basaltic rocks to assess the presence of amphibole in their mantle sources. Partition coefficients between clinopyroxene and melt are within the range of published values for most of the analyzed elements, with some low values for Zr (e.g., 0.030–0.353) and near unity for the HREE. Using partition coefficients between amphibole and clinopyroxene and Depx/melt from the literature, we calculate the partitioning of trace elements between amphibole and a basaltic melt. The Damp/melt obtained are lower than most of the published values, but are in good agreement with recent values measured in experiments at 1.5 GPa and 1100°C. The partition coefficient for Nb between amphibole and melt is always low (<0.20), which indicates that residual amphibole during partial melting in hydrated mantle (amphibole-bearing peridotite) cannot account for Nb depletion in arc magmas.

Diamond formation in the deep mantle: the record of mineral inclusions and their distribution in relation to mantle dehydration zones

Abstract Studies of the inclusions contained in natural diamonds have shown the occurrence of minerals which must have formed at depths below the lithosphere and which may be closely matched with the silicate mineral assemblages determined by high pressure and temperature experimental studies for depths of 300 to 800 km in the Earth’s mantle. The inclusions come principally from two main depth zones: (1) the lower asthenosphere and upper transition zone; (2) the Upper Mantle/Lower Mantle (UM/LM) boundary region and the uppermost LM. The inclusions from zone 1 are very largely majoritic garnets (with or without clinopyroxene) which indicate bulk compositions of eclogitic/metabasic affinity. The minerals from zone 2 include Ca-Si and Mg-Si perovskites and ferropericlase and are dominantly of metaperidotitic bulk composition, but include some possible metabasite assemblages. In many of these natural assemblages, the tetragonal almandine pyrope phase occurs rather than the garnet found in experiments. As natural diamonds are believed to crystallize in fluids/melts, the hypothesis is developed that the restriction of diamonds and inclusions of particular compositions to the above two depth intervals is because they are controlled by loci of fluid/melt occurrence. Attention is focused on subduction zones because both suites of inclusions show some evidence of subducted protoliths. The lower zone (600−800 km) coincides with the region where dehydration may be expected for hydrous ringwoodite and dense hydrous Mg-silicates formed in subducted peridotites. The dehydration of lawsonite in subducted metabasites provides a particular location for melt formation and the inclusion of the shallower (~300 km) majoritic inclusions. For the deeper majoritic inclusions in the region of the upper transition zone, melt development may occur as a consequence of the hydrous wadsleyite-to-olivine transformation, and such melt may then interact with the upper crustal portion of a subducting slab. These suggestions offer an explanation of the depth restrictions and the compositional restrictions of the inclusions. The differences in δ13C values in the host diamonds for the two suites of inclusions may also be explained on this basis.

PARTITIONING OF TRACE ELEMENTS BETWEEN CLINOPYROXENE AND GARNET : DATA FROM MANTLE ECLOGITES

Concentrations of REE, Ba, Sr, Y and Zr have been measured by ion microprobe (SIMS) for garnet and clinopyroxene in a suite of eclogite xenoliths from the Roberts Victor kimberlite pipe in South Africa. The xenoliths have a restricted set of temperatures of formation estimated at 1100 ± 100°C at 5 GPa, but span a wide range of compositions; garnet Ca/(R2+) ratios range from 0.08 to 0.50, whilst the associated clinopyroxenes also show increasing jadeite solid solution as their Ca/(R2+) ratios increase. The cpx/grt partition coefficients (Di) for REE, Sr and Y decrease progressively as Ca/(R2+ in the minerals increase; in the LREE this change amounts to approximately three orders of magnitude. Evidence for such large changes in partition coefficients as a function of Ca content is also seen in other data for mantle eclogites from South Africa and Siberia. Cpx and grit with low Ca/(R2+ ratios have (Di) for HREE which are similar to those implied by mineral/liquid data for basic-ultrabasic melts. Poor correlations of Ba, La and Lu partition coefficients with cpx-grt Ca contents in the present dataset, are the result of late-stage partial alteration and relatively large errors of measurement associated with low concentrations. Zr partition coefficients show only a weak correlation with mineral Ca/(R2+). Strong correlations are demonstrated between cpx/grt Di (calculated on a ppm or molar basis) and molar cpx/grt DCa for most REE, Sr and Y. These may be expressed in the form: In Di∗cpxgrt =A ln DCa∗cpxgrt + BB (Eq. (1)), where A and B are empirical constants for each element and vary with ionic radius in the REE. The partitioning of a given trace element progressively favours garnet rather than pyroxene as increasing amounts of Ca are more strongly partitioned into garnet than clinopyroxene; and the magnitude of this effect increases with ionic radius in the REE. The empirical linear calibrations (Eq. (1)) given for REE, Y and Sr allow calculation of the (Di) of a cpx-grt pair from their Ca contents (for 1100 ± 100°C, ∼ 5 GPa). Using the well-documented R3+ Al[4] Ca−1Si−1 (YAG-type) substitution relationship for REE and Y in garnets as the basis of an exchange reaction, it is suggested that the A term in empirical Eq. (1) may be related to variations in activity coefficients as the major element (Ca) composition changes, whilst term B is largely dependent on the standard state free energy.

Pelite facies series and the temperatures and pressures of Dalradian metamorphism in E Scotland

Summary An attempt is made to synthesise data on the regional metamorphism of pelites in the eastern Scottish Dalradian. Zonal sequences (facies series) of mineral assemblages are presented and variants of the traditionally recognised Barrovian and Buchan metamorphism are separated. A Schreinemakers net and petrogenetic grid for the system KFMASH is presented and fixed in P-T space using experimental data. The P-T data for the various facies series are used in conjunction with other constraints to develop a model of isotherm and isobar distribution in the region, and the effects of post metamorphic folding of the isopleths are noted. P-T gradients normal to isobars suggest overall ‘geothermal’ gradients for most of the area to be convex to the T-axis, and it is suggested that they result from magma intrusion at depth. The region adjacent to the Highland Boundary fault shows strong horizontal and vertical temperature gradients which are tentatively interpreted to indicate the presence of a synmetamorphic tectonic boundary to the Dalradian metamorphic belt.

Extreme crustal oxygen isotope signatures preserved in coesite in diamond

The anomalously high and low oxygen isotope values observed in eclogite xenoliths from the upper mantle beneath cratons have been interpreted as indicating that the parent rock of the eclogites experienced alteration on the ancient sea floor. Recognition of this genetic lineage has provided the foundation for a model of the evolution of the continents whereby imbricated slabs of oceanic lithosphere underpin and promote stabilization of early cratons. Early crustal growth is thought to have been enhanced by the addition of slab-derived magmas, leaving an eclogite residuum in the upper mantle beneath the cratons. But the oxygen isotope anomalies observed in eclogite xenoliths are small relative to those in altered ocean-floor basalt and intermediate-stage subduction-zone eclogites, and this has hindered acceptance of the hypothesis that the eclogite xenoliths represent subducted and metamorphosed ocean-floor basalts. We present here the oxygen isotope composition of eclogitic mineral inclusions, analysed in situ in diamonds using an ion microprobe/secondary ion mass spectrometer. The oxygen isotope values of coesite (a polymorph of SiO2) inclusions are substantially higher than previously reported for xenoliths from the subcratonic mantle, but are typical of subduction-zone meta-basalts, and accordingly provide strong support for the link between altered ocean-floor basalts and mantle eclogite xenoliths.

Petrography and geological history of upper mantle xenoliths from the matsoku kimberlite pipe

Abstract Upper mantle (garnet-peridotite facies) xenoliths from the Matsoku kimberlite pipe consist predominantly of olivine, orthopyroxene, clinopyroxene and garnet, with a wide range of modal values. In some xenoliths metasomatism, under the ambient mantle P.T. conditions indicated by all xenoliths, has added the primary-metasomatic minerals: phlogopite, ilmenite, rutile, pyrrhotite, pentlandite and chalcopyrite. This metasomatism has probably occurred within the kimberlite magma envelope at depth, and been caused by fluid derived from the kimberlite magma. Omitting metasomatic aspects, the xenoliths show a wide and broadly continuous range of textural varieties, which may be referred to three major groups: coarse-grained (showing restricted deformation and recrystallization), flaser (relatively strongly deformed and showing extensive recrystallization) and even-textured (extensively annealed following deformation). These form a progressive sequence of deformation and annealing, the latter involving recovery, recrystallization and grain growth. It is evident from the textural features, in conjunction with aspects of mineral chemistry, that the xenoliths cannot be in any simple sense cognate with the kimberlite magma, and that a wide variety of textural types occurs within a restricted location in the upper mantle. The deformation and annealing responsible for the textural variety are probably caused both by widespread mantle creep and more localized phenomena occurring within the kimberlite magma envelope at depth. Petrographic and chemical characteristics suggest that the banding seen in some xenoliths is dominantly of cumulate igneous origin.

Carbon isotope ratios and nitrogen abundances in relation to cathodoluminescence characteristics for some diamonds from the Kaapvaal Province, S. Africa

Abstract Secondary ion mass spectrometry (SIMS) techniques have been used to study the variation of C isotope ratio and N abundance within selected diamonds in relation to their crystal growth zones. The growth zones are seen in cathodoluminescence (CL), and include both octahedral and cuboid zones within typical diamonds of external octahedral morphology. Compositions were determined by use of a primary 133Cs+ ion beam and measurement of 12C− , 13C− , and 12C14N− secondary ions at high mass resolution on a Cameca ims-4f ion microprobe at Edinburgh University. In each of the diamonds, different growth zones have marked differences in N abundance, which are as great as 0−1400 ppm within one diamond. Changes of several hundred ppm N are common across both octahedral and cuboid growth zones, and appear sharp and abrupt at the boundaries of the growth zones. In general for the common blue CL, luminescence increases with N abundance. The changes in N abundance across fine scale (~100 μm) growth zones show that the total N contents determined by IR spectroscopy may show great variations of abundance. In contrast, within detection limits, δ13C appears constant across many growth zone boundaries. Thus the factors controlling uptake of N from the fluid/melt reservoir in which natural diamonds grow often do not influence δ13C. No evidence of progressive variation or fractionation of C isotopes during growth was found. Some original variation in C isotope composition may have been eliminated by diffusion of C atoms subsequent to growth, because of the storage of natural diamonds over millions of years in the Earth’s mantle at temperatures of 950−1250°C. Such atomic mobility does not homogenize N distribution because of the proven tendency of N to form aggregates of atoms. A survey of experimental estimates of single atom (C) diffusion parameters, suggests that diffusion distances of ~100 μm are likely at high temperatures (~1100°C) over long time periods (~1.0 Ga). Therefore, with refinement of the diffusion parameters and measurements, the extent of C isotope homogenization in natural diamonds, as well as N aggregation state, might provide quantitative evidence of their time-temperature history under mantle conditions.

SILICATE GLASSES IN SPINEL LHERZOLITES FROM YEMEN : ORIGIN AND CHEMICAL COMPOSITION

Abstract Anhydrous and amphibole-bearing spinel lherzolites found in alkali basalts from Southern Yemen contain melt-pockets with silicate glasses from which have crystallized euhedral clinopyroxenes, olivines and spinels. These glasses have variable major element compositions related to variations in the composition of the melting phases. In the hydrous lherzolites, the variations in the major element composition of the glasses are directly correlated to the variations in the composition of residual amphiboles in the melt-pockets. The same observations can be made for the composition of clinopyroxene and glass in the anhydrous lherzolites. The trace element compositions of the residual phases and the glasses are consistent in the different samples and all the observations indicate that the glasses originated by in-situ melting of amphibole or clinopyroxene in the lherzolites. If infiltration of a metasomatic melt or fluid caused the melting event, these data are indicative of very low fluid/rock ratios and attainment of conditions close to equilibrium between the fluid and the solid phases. Surprisingly, Cl-rich apatite, when present, was not melted in these samples, which indicates that apatite can be a residual phase during partial melting at mantle conditions. Comparison of the glasses in the Yemen lherzolites with published chemical compositions for mantle glasses shows a dichotomy between glasses produced by (a) in-situ melting of amphibole or clinopyroxene (± phlogopite), either in closed-system or during reaction with a metasomatic fluid at low fluid/rock ratios, and (b) glasses resulting from interaction between the peridotite and a metasomatic melt or fluid at higher fluid/rock ratios.

Distribution of trace elements between amphibole and clinopyroxene from mantle peridotites of the Eifel (western Germany): An ion-microprobe study

Abstract The pargasite/diopside partition coefficients ( D ) for the rare-earth elements (REE), Sr, Y, Zr and Hf, determined with ion-microprobe analysis on low-temperature West Eifel mantle peridotites, have values of between 1 and 2. In contrast, Nb and Ba are strongly enriched in amphibole with partition coefficients of D Nb = 100–200 and D Ba > 1000. Small sample-to-sample variations in the REE partition coefficients in the West Eifel samples appear to be related to compositional factors, specifically the Na content of clinopyroxene. The REE partition coefficients of edenitic-pargasitic hornblende/diopside pairs from East Eifel xenoliths have higher values ( D = 3–8) than those from the West Eifel and these differences may be attributable to the lower Na and K contents of the East Eifel xenoliths. The partition coefficients for amphibole/clinopyroxene pairs of natural peridotites differ from the literature values for those of high-temperature melts. Compared to the West Eifel xenoliths most of the melt-related partition coefficients are higher for LREE ( D = 2–4), but decrease and become similar for HREE ( D = 1). It is unclear to what extent these differences are the result of temperature, pressure and composition, or the result of lack of equilibrium.

SIMS stable isotope measurement: counting statistics and analytical precision

Abstract Analytical precision is vital in the interpretation of stable isotope data collected by secondary ion mass spectrometry (SIMS) given the small analysis volumes and the small magnitude of natural isotopic variations. The observed precision of a set of measurements is represented by the standard deviation (precision of an individual measurement) or the standard error of the mean (precision of the mean value). The SIMS data show both systematic variations with time and random Poisson variability, but the former largely cancel out when data for two different isotopes are expressed as a ratio. The precision of a SIMS isotope ratio routinely matches that predicted by Poisson counting statistics and can approach that of conventional bulk analysis techniques for counting times of several hours. All sample analyses must be calibrated for instrumental mass fractionation using SIMS analyses of a standard material. There is often a gradual drift in the mass fractionation with time, but this can be modelled by least-squares regression of the standard isotope ratios. Drift in the sample analyses is eliminated by using the relevant point on this regression line to calibrate each sample. The final precision of a corrected isotope ratio must take into account the scatter in both the sample and the standard data.