Thomas Stachel

Subducting oceanic crust: The source of deep diamonds

Inclusions of majoritic garnet in diamonds from the Jagersfontein kimberlite formed at unusually great depths of ∼250 to >500 km in the asthenosphere and transition zone. The original host rocks were derived from a much shallower, basaltic (eclogitic) source. The presence of negative Eu anomalies in all majoritic garnets requires a crustal origin, thereby linking these very deep diamond sources to subducting oceanic crust. The carbon isotope values (δ13C) of the host diamonds fall within a narrow range at ∼−20‰, which is fundamentally different from the broad range (−24‰ to −2‰) and bimodal distribution of carbon isotopes of Jagersfontein diamonds that formed in the shallower lithosphere. This indicates that majoritic garnet-bearing diamonds at Jagersfontein inherited their light carbon isotopic composition directly from organic matter contained in a subducting slab. These diamonds were likely formed by direct conversion from graphite, well within the diamond stability field.

New Ca-silicate inclusions in diamonds — tracers from the lower mantle

Abstract Diamonds from the Kankan district of Guinea frequently contain majoritic garnet and ferropericlase mineral inclusions similar to those from Sao Luiz in Brazil [1–3]. Besides these ultra-high pressure phases in the Guinea diamonds, we have now identified, initially by in-situ single crystal X-ray diffraction, a new inclusion paragenesis of Ca-silicates. One diamond contained walstromite-structured CaSiO 3 and three others the mineral assemblage CaSi 2 O 5 (titanite-structured) with larnite (p-Ca 2 SiO 4 ). The first two phases represent new minerals not described before from natural occurrences. From the phase diagram for mantle CaSiO 3 [4], there are indications that primary CaSiO 3 -perovskite underwent successive retrograde phase transformations. Development of equilibrium textures suggests slow exhumation. Ca-silicates are possibly important carriers of strontium, phosphorus and potassium and thereby contain part of the inventory of radioactive elements in the transition zone and the lower mantle. The presence of coesite (formerly stishovite) in two of the diamonds containing Ca-silicates indicates that these diamonds belong to an ‘eclogitic’ suite [5,6], whereas the presence of ferropericlase together with CaSiO 3 and MgSiO 3 in a third diamond may imply a ‘peridotitic’ environment.

Peridotitic diamonds from the Slave and the Kaapvaal cratons—similarities and differences based on a preliminary data set

Abstract A comparison of the diamond productions from Panda (Ekati Mine) and Snap Lake with those from southern Africa shows significant differences: diamonds from the Slave typically are un-resorbed octahedrals or macles, often with opaque coats, and yellow colours are very rare. Diamonds from the Kaapvaal are dominated by resorbed, dodecahedral shapes, coats are absent and yellow colours are common. The first two features suggest exposure to oxidizing fluids/melts during mantle storage and/or transport to the Earths surface, for the Kaapvaal diamond population. Comparing peridotitic inclusions in diamonds from the central and southern Slave (Panda, DO27 and Snap Lake kimberlites) and the Kaapvaal indicates that the diamondiferous mantle lithosphere beneath the Slave is chemically less depleted. Most notable are the almost complete absence of garnet inclusions derived from low-Ca harzburgites and a generally lower Mg-number of Slave inclusions. Geothermobarometric calculations suggest that Slave diamonds originally formed at very similar thermal conditions as observed beneath the Kaapvaal (geothermal gradients corresponding to 40–42 mW/m 2 surface heat flow), but the diamond source regions subsequently cooled by about 100–150 °C to fall on a 37–38 mW/m 2 (surface heat flow) conductive geotherm, as is evidenced from touching (re-equilibrated) inclusions in diamonds, and from xenocrysts and xenoliths. In the Kaapvaal, a similar thermal evolution has previously been recognized for diamonds from the De Beers Pool kimberlites. In part very low aggregation levels of nitrogen impurities in Slave diamonds imply that cooling occurred soon after diamond formation. This may relate elevated temperatures during diamond formation to short-lived magmatic perturbations. Generally high Cr-contents of pyrope garnets (inside and outside of diamonds) indicate that the mantle lithosphere beneath the Slave originally formed as a residue of melt extraction at relatively low pressures (within the stability field of spinelperidotites), possibly during the extraction of oceanic crust. After emplacement of this depleted, oceanic mantle lithosphere into the Slave lithosphere during a subduction event, secondary metasomatic enrichment occurred leading to strong re-enrichment of the deeper (>140 km) lithosphere. Because of the extent of this event and the occurrence of lower mantle diamonds, this may be related to an upwelling plume, but it may equally just reflect a long term evolution with lower mantle diamonds being transported upwards in the course of “normal” mantle convection.

Mineral inclusions in diamonds from the Panda kimberlite, Slave Province, Canada

Ninety diamonds from the Eocene Panda kimberlite (Ekati Mine, Northwest Territories, Canada) were analyzed for the major and trace element compositions of their mineral inclusions using electron microprobe techniques (EPMA) and secondary ion mass spectrometry (SIMS). Additionally, nitrogen aggregation characteristics of the host diamonds were measured using Fourier-transform infrared spectroscopy (FTIRS). Within the cratonic lithosphere, Panda diamonds are principally derived from peridotitic sources (85 %) with a minor content of eclogitic diamonds (10 %). Ferropericlase bearing diamonds (5 %) contain combinations of ferropericlase with either Mg-Al spinel plus olivine or with olivine or with a pure silica phase. The chemical char- acteristics of these inclusions are in accordance with a lithospheric origin from ferropericlase-bearing dunites. Ferropericlase coexisting with CaSiO3 (most likely originally included as Ca-silicate perovskite), however, is regarded as evidence for a lower mantle origin of the host diamond. Major element compositions show that the peridotitic diamonds formed in a moderately depleted environment, indicated by the presence of harzburgitic garnet inclusions with calcium contents generally > 2.5 wt% CaO and olivines with Mg numbers (100*Mg/(Mg+Fe)) of 92-93.5. Rare earth element (REE) concentrations in peridotitic garnets largely follow subdivisions based on major elements with lherzolitic garnets showing middle REE to heavy REE enriched, slightly sinusoidal patterns, whilst harzburgitic garnets have distinctly sinusoidal REEN. Inclusion geothermobarometry indicates formation of peridotitic diamonds in the temperature range 1100-1250°C, following a geothermal gradient of 40-42 mW/m2, in accordance with similar observations world-wide. Touching garnet-olivine and garnet-orthopyroxene inclusion pairs equilibrated at lower temperatures of 1000-1100°C, corresponding to a geothermal gradient around 37 mW/m2. The higher temperatures are considered to be those prevailing during diamond formation. Nitrogen contents in Panda diamonds vary strongly from below detection (< 10 ppm) to 2700 atomic ppm. Nitrogen aggrega- tion ranges from poorly aggregated (Type IaA diamond) to highly aggregated (Type IaB diamond). If all diamonds that show signs of plastic deformation during mantle residence are excluded from the dataset, then a diamond subset becomes apparent with an overall low nitrogen aggregation state of < 30 % B-center. This result may indicate that plastic deformation increases the aggre- gation of nitrogen in Panda diamonds. Taking the Early Archean Re-Os isochron date for sulfide inclusions in Panda diamonds (Westerlund et al., 2003b) at face value, the low aggregation states of undeformed diamonds may indicate mantle residence at rela- tively low temperatures (< 1100°C). If this is the case, the decrease in temperature inferred from the comparison of touching and non-touching inclusion pairs must have occurred soon after diamond formation. Thus diamond formation beneath the central Slave may be restricted to short lived and localized thermal events. An apparent increase in geothermal gradient with depth in the litho- spheric mantle beneath the Central Slave for the time of kimberlite eruptions (Upper Cretaceous to Eocene) may have a similar cause and reflect transient heating of the deep lithosphere during melt infiltration.

Spectroscopic 2D-tomography Residual pressure and strain around mineral inclusions in diamonds

We have studied high-pressure inclusions (Ca-silicates, coesite, graphite) in three large diamonds, one from the Kankan district, Guinea, and the other two from the Panda kimberlite, Ekati diamond mines, Canada. Using the in situ point-by-point mapping technique with a confocal Raman system, the mineralogy of the inclusions, as well as their area distribution pattern ( e.g. , of different Ca-silicate phases) and their order-disorder distribution pattern (shown for graphite/disordered carbon), were determined. Raman mapping of the host diamonds yielded 2D-tomographic pressure and strain distribution patterns and provided information on the residual pressure of the inclusions (∼ 2.3 GPa for a coesite inclusion and ∼ 2.6 GPa for a graphite inclusion). The inclusions are surrounded by haloes of significantly enhanced pressure, several hundred σm across. These haloes exhibit complex pressure relaxation patterns that consist of micro-areas affected by both compressive and dilative strain, with the latter being intensive enough to result in apparent “negative pressures”.

Exhumation of lower mantle inclusions in diamond: ATEM investigation of retrograde phase transitions, reactions and exsolution

A multiphase inclusion (KK-83) in a diamond from the Kankan deposits in Guinea, resulting from complex reactions between primary lower mantle phases, was examined in detail using analytical transmission electron microscopy. The inclusion consists of a large (300 μm) diopside crystal with a small symplectitic intergrowth in one corner, comprising olivine and tetragonal almandine pyrope phase (TAPP). The olivine part of the symplectite contains small exsolutions of Mg-Al-chromite and rare Ca-carbonate. A second inclusion of ferropericlase observed in the same diamond indicates a primary origin within the lower mantle. The clinopyroxene formed through a series of reactions at the expense of touching inclusions of CaSi- and MgSi-perovskite during convective ascent of the diamond through the transition zone. The proposed reaction sequence is: MgSiPvk+CaSiPvk→MgSiIlm+CaSiPvk→ringwoodite+stishovite+CaSiPvk→clinopyroxene+relict ringwoodite. The high Al-content of the bulk symplectite indicates ringwoodite as precursor phase coexisting with clinopyroxene. A slight deficiency of SiO2 during the diopside forming reaction leads to a small amount of relict ringwoodite (<0.1 vol% of the total inclusion is occupied by ringwoodite) coexisting with diopside. The ringwoodite captures the observed high amount of Al. Subsequent breakdown of ringwoodite to wadsleyite+TAPP produces the observed symplectitic intergrowth. During the phase transformation of wadsleyite to olivine, exsolution of Mg-Al-chromite occurs further decreasing the original solubility of Al, Cr and Ti, in accordance with experimental data on the element partitioning between wadsleyite and olivine. Based on these observations, TAPP can form as a retrograde phase within the transition zone of the Earth’s mantle and is not restricted to the upper part of the lower mantle. High Fe3+-contents may favour its formation.

Deep mantle diamonds from South Australia: A record of Pacific subduction at the Gondwanan margin

Diamonds from Jurassic kimberlites at Eurelia, South Australia, contain coexisting inclusions of ferropericlase and MgSi-perovskite that provide evidence for their deep (>670 km) lower mantle origin. Eurelia diamonds formed from mixed carbon sources, likely including subducted carbonate, as indicated by a trend toward isotopically heavy carbon compositions (δ 13 C = 0‰) and low nitrogen concentrations (

A spectroscopic and carbon-isotope study of mixed-habit diamonds: Impurity characteristics and growth environment

Abstract Mixed-habit diamonds have experienced periods of growth where they were bounded by two surface forms at the same time. Such diamonds are relatively rare and therefore under-investigated. Under certain physical and chemical conditions, smooth octahedral faces grow concurrently with rough, hummocky cuboid faces. However, the specific conditions that cause this type of growth are unknown. Here we present a large array of spectroscopic data in an attempt to investigate the impurity and carbon-isotope characteristics, as well as growth conditions, of 13 large (>6 mm diameter) plates cut from mixed-habit diamonds. The diamonds all generally have high nitrogen concentrations (>1400 ppm), with the octahedral sectors enriched by 127-143% compared to their contemporary cuboid sectors. Levels of nitrogen aggregation are generally low (2-23% IaB) with no significant difference between sectors. IR-active hydrogen features are predominantly found in the cuboid sectors with only very small bands in the octahedral sectors. Platelet characteristics are variable; only one sample shows a large B′ band intensity in the octahedral sector, with no platelets occurring in the cuboid sector. Other samples either show a small B′ band in both sectors, or just in the cuboid sector, or none at all. These data support a model that shows the concentration-adjusted aggregation rate of nitrogen to be the same in both sectors, whereas the subsequent platelet development is reduced in the cuboid sectors. This is because the interstitial carbon atoms have interacted with disk-crack-like defects only found in cuboid sectors, which in turn reduces their chances of aggregating to form platelets. These disk-crack-like defects are also thought to be the most likely site for the IR-active hydrogen features and they maybe intrinsic to cuboid growth in mixed-habit diamonds. When they are graphitized, as they are in all of the diamonds in this study, this may reflect a heating event prior to volcanic exhumation. Spectroscopic analysis of the green cathodoluminescence exhibited by all of the diamonds shows nickel centers to be present in only the cuboid sectors. Carbon isotope data, obtained by secondary ion mass spectrometry, show very little variation in seven of the diamonds. The total range of 217 analyses is -7.94 to -9.61 (±0.15)‰, and the largest variation in a single stone is 0.98‰. No fractionation in carbon isotopes is seen between octahedral and cuboid sectors at the same growth horizon. These data suggest that the source fluid chemistry, as well as pressure, temperature, and oxygen fugacity were very stable over time, allowing such large volumes of mixed-habit growth to occur. The high concentration of impurities, namely nitrogen and hydrogen, is probably the critical factor required to cause mixed-habit growth. The impurity and isotopic data fall in line with previous modeling based on diamond growth from reduced carbonates with the loss of a 13C-enriched CO2 component.

Platelet development in cuboid diamonds: insights from micro-FTIR mapping

Fourier transform infrared mapping of diamonds can reveal detailed information on impurities, with a spatial context. We apply this technique, combined with in situ isotopic analysis of carbon, to the study of cuboid diamond growth in a sample that exhibits some mixed-habit growth. While there has been some uncertainty in the literature regarding sectoral differences in nitrogen aggregation and subsequent platelet development, the data from this study appear far more conclusive. We show that despite nitrogen being concentrated in octahedral sectors, there is no detectable difference in the concentration-adjusted rate of nitrogen aggregation within octahedral and cuboid sectors. However, the resultant platelet development is significantly reduced in cuboid sectors compared to contemporaneously formed octahedral sectors. This finding has significant implications for the classification of diamonds using the relationship between their platelet intensity and the absorption caused by B centres. It means that cuboid diamonds naturally fall below the linear relationship that has been termed regular, which would lead to them being incorrectly interpreted as having experienced heating or deformation. The data also support earlier suggestions that large hydrogen concentrations in the diamond-forming fluid may be required for cuboid growth. We further suggest that high nitrogen and hydrogen concentrations are required for mixed-habit diamond growth, which might be the product of specific fluid chemistries that occur in reducing mantle environments.

Carbonatite magmatism and fenitization of the epiclastic caldera-fill at Gross Brukkaros (Namibia)

The Gross Brukkaros inselberg is a dome structure with a crater-shaped central depression within Precambrian/Cambrian country rocks which was active as a depocenter during the Late Cretaceous. The formation of the structure was due to the intrusion and subsequent intermittent depletion of a shallow magma reservoir. Juvenile material has not been recognized hitherto. This is the first account of juvenile lapilli from within the epiclastic fill of the caldera structure. The lapilli are calciocarbonatites and magnesiocarbonatites in composition, but are characteristically low in elements such as P, Nb, Ba and Sr, otherwise typical of carbonatites. This signature, however, is also characteristic of carbonatites from surrounding volcanic centers and necks. The Brukkaros sediments suffered strong metasomatic-hydrothermal alteration, which introduced in a first stage fluids rich in Fe, Ti, Na, Nb, V, K (Ca?, CO2?), and in a second stage the Brukkaros sediments were silicified on a large scale and locally enriched in P, Th and Cr. Si is derived from desilication of the wall rocks (basement?, Nama sediments) of the magma reservoir. Cr was probably mobilized during alteration of the abundant doleritic detritus within the Brukkaros depocenter.