The NOBLE project

NOBLE logo

The Origin, Accretion and Differentiation of Extreme Volatiles in Terrestrial Planets

Identifying the mechanisms by which the terrestrial planets acquired, retained and redistributed extreme volatiles and highly incompatible elements remains a fundamental challenge in the Earth and planetary sciences.

The halogens, Cl, Br and I, are highly incompatible, volatile and represent a powerful potential tracer of these processes. Although Cl is readily analysed, the concentrations of Br and I within most samples of interest are below the detection limit of conventional techniques and the halogens as a tracer set has been almost completely ignored.

Argus VI

Dr Lorraine Ruzié working with the Argus VI mass spectrometer

Neutron irradiation of samples converts halogens to noble gas isotopes that can be measured by conventional or laser resonance mass spectrometry. Pioneered by Manchester, this innovation in analytical technique development now provides detection limits that far exceed any other approach, is independent of matrix effects, and links the halogen results to naturally occurring noble gases; a key tracer set that Manchester has a lot of experience in interpreting.

Halogen Phase Partitioning

  • Dr Bastian Joachim
  • Figure 1: a simplified schematic model for oceanic volcanism.

    Figure 1: a simplified schematic model for oceanic volcanism.

    Below Mid-Ocean Ridges, the earth upper mantle melts partially over a range of pressures between ~10-20 kbar, which is a depth of about 30-80 km.
    These partial melts rise to the surface and crystallize as Mid-Ocean-Ridge Basalts (MORB).

    The source for melts, which crystallizes at the surface as Ocean Island Basalts (OIB), is the deep mantle. Their origin are mantle plumes, which are characterized by upwelling solid diapiric pods at the 660-km boundary. Decompression of the solid material during upwelling leads to the generation of partial melts, which reach the surface in intra-plate hotspots like Hawaii. Consequently, MORBs and OIBs are crystallized partial melts of the earth mantle, which can easily be sampled, as they are present at the earth surface.
    Halogens are, as large anions, excluded from most mineral structures, which means that they will preferably partition into the melt. So far, estimated halogen concentrations of the MORB-source mantle vary by orders of magnitude while those of Bromine and Iodine in the OIB-source mantle are non-existent (Pyle and Mather 2009; Aiuppa et al. 2009)

    Figure 2 (modified after Green and Ringwood 1967) shows the mineralogy of a typical mantle pyrolite, consisting of olivine, orthopyroxene (opx), clynopyroxene (cpx), plagioclase (Plg) and some minor phases (mP).

    Figure 2 (modified after Green and Ringwood 1967) shows the mineralogy of a typical mantle pyrolite

    We want to perform experiments that simulate partial melting of the earth mantle, so that we are able to study the mineral melt partitioning behaviour of halogens. By combining these data with halogen concentrations that will be measured in natural MORBs and OIBs, we will be able to determine the abundances of halogens in the mantle source regions.

    Mineralogy
    The green rectangle in Figure 2 (modified after Green and Ringwood 1967) shows the mineralogy of a typical mantle pyrolite, consisting of olivine, orthopyroxene (opx), clynopyroxene (cpx), plagioclase (Plg) and some minor phases (mP). The size of the blocks within the rectangle is proportional to the respective mineral ratio.

    Partial melting of this mantle pyrolite will result in generation of magma (red), coexisting with olivine and orthopyroxene. This melt will rise to the earth surface and crystallize as a basalt, consisting of olivine, orthopyroxene, clynopyroxene, plagioclase and some minor phases.
    We want to simulate experimentally partial melting of the earth mantle, which will give us the opportunity to determine the partitioning behaviour of halogens between a basaltic melt and olivine respectively orthopyroxene.

    Figure 3 (modified after Litasov and Ohtani, 2002): a phase diagram of the model CMAS-Pyrolite composition.

    Figure 3 (modified after Litasov and Ohtani, 2002): a phase diagram of the model CMAS-Pyrolite composition.

    As starting composition, we use in a first step the model primitive mantle composition proposed by Jagoutz et al. (1979) simplified to four components (CMAS) CaO (4.28 wt%), MgO (44.35 wt%), Al2O3 (5.31 wt%) and SiO2 (46.07 wt%) according to the procedure of O´Hara (1968). 0.2 wt% of the respective halogens are added as CaBr2, CaF2, CaCl2 and CaI2. In future experiments, we will stepwise modify the chemical compositions by e.g. adding Fe or varying the Al-content, to determine the effect of these on the partitioning behaviour and get stepwise closer to an understanding of more complex conditions, as they appear in nature.

    The red circle represents the pressure temperature range, which we plan to use for our experiments. A melt (L) is coexisting with olivine (Ol) and orthopyroxene (Opx).

    Figure 4: a schematic sketch of the inner part of a conventional Boyd and England Type Piston Cylinder Apparatus.

    Figure 4: a schematic sketch of the inner part of a conventional Boyd and England Type Piston Cylinder Apparatus.


    Experimental Setup

    Experiments are performed in a conventional Boyd and England Type Piston Cylinder Apparatus. Figure 4 shows a schematic sketch of the inner part.
    The sample is located in the centre within a platinum capsule (length: 1 cm). Al2O3-powder and a Talc-Pyrex assembly serve as pressure medium. The pressure is generated by a cylindrical piston, which condenses the sample by pushing from the bottom. Heat is produced by a graphite furnace and controlled via a W-Re thermocouple. Pressures range between 10 – 25 kbar, temperatures between 1500 – 1720°C.

    Preliminary Experiments
    Figure 5 shows a BSE-image of an experiment that was performed at 10 kbar, first heated to 1720°C and slowly cooled with a rate of 10°C/min to 1500°C. After 5 h the sample was quenched and afterwards polished for further analysis.

    Forsterite grains with a diameter of about 20 up to 150 µm are embedded in a melt with almost MORB-composition, giving us the opportunity to investigate halogen partitioning between forsterite (olivine) and melt at conditions relevant for mantle melting (10 kbar, 1500°C).
    Determination of the halogen contents within crystals and melt will be performed by Microprobe, Ionprobe (SIMS) and for very low concentrations via neutron-irradiation coupled with conventional noble gas mass spectrometry (Ni-NGMS).

    Figure 5: BSE image of polished, quenched experimental sample

    Figure 5: BSE image of polished, quenched experimental sample

    Halogens in Meteorites

  • Dr Patricia Clay
  • BSE image of ALH 77295 enstatite chondrite (EH3).

    BSE image of ALH 77295 enstatite chondrite (EH3).

    The halogens (Cl, Br and I), due to their highly incompatible behavior during melting, makes them excellent tracers of geo- and cosmochemical processing. However, they are present in ultra low abundances (particularly Br and I) in most extraterrestrial materials. This low abundance unfortunately makes them difficult to measure by most traditional methods. However, at the University of Manchester, the ability to measure these elements through Neutron-Irradiation Noble Gas Mass Spectrometry, or NI-NGMS, has been developed. This technique utilizes the noble gases as proxies for the halogens through neutron-irradiation of samples. In this way, the halogens can be measured in many samples of geo- and cosmochemical interest, tracing geochemical processing in terrestrial and extraterrestrial materials to unravel the origin, retention and distribution of the halogens through time.

    MIL07139 chips

    Chips of MIL 07139 meteorite for neutron irradiation and halogen analysis.

    Using NI-NGMS we are able to accurately construct a picture of the halogen reservoirs in the precursors to planets, i.e., materials that accreted to form the planets, such as primitive chondritic metreorites. With this information we can better understand the evolution of volatile budget on our own planet. We are also pursuing a ‘halogen inventory’ of other important bodies in our Solar System, such as the Moon and Mars, which will provide valuable insight into the halogen response to planetary processes (e.g., melting, differentiation, impacts and crust formation). In conjunction with other threads of research (Link to Halogens in MORBS and OIBs) and (Link to Halogen Phase Partitioning) we are constructing a clearer picture of the overall origin, distribution and l behavior of halogens important reservoirs in the Earth (e.g., MORB, OIB) and other planetary bodies through time.

    Halogens in MORBs and OIBs

  • Dr Lorraine Ruzié
  • Because of their incompatibility, relatively high concentrations and distinct elemental compositions in surface reservoirs, the heavy halogens (Cl, Br, I) represent key tracers of volatile transport processes in the Earth. However, as pointed out recently by Aiuppa et al. 2009, their concentrations and distributions of halogens in the mantle remain deep, dark and mysterious. The difficulty to conduct high precision simultaneous data of I with Cl and Br on material of interest (e.g. basalts glass-samples) explains the scarcity of data in the literature (Schilling et al. 1980; Deruelle et al 1992; Jambon et al. 1995). For example in MORB samples, Cl, Br and I may be as low as 1ppm, 10 ppb and 1.0 ppb respectively.

    Thin section of basalt

    Thin section of basalt sample

    The Manchester Isotope Geochemistry and Cosmochemistry group has pioneered an innovative halogen analytical technique involving neutron irradiation of samples to convert halogens to noble gases. Neutron irradiation converts Cl, Br and I into isotopes of Ar, Kr and Xe, respectively (Johnson, 2000; Burgess, 2002). The production of irradiation-derived noble gas isotopes are orders of magnitude higher than their natural abundances and readily measured using conventional noble gas mass spectrometry (NI-NGMS). Current NI-NGMS detection limits are 1×10-12g for I and up to three times better for Cl. In one milligram of sample, detection limits are at 1ppb. These limits will be pushed using multi-collector mass spectrometers (Argus, Helix) offering a factor 10 improvement in precision and detection limits. This provides detection limits unmatched by any other technique.

    For the first time, we are in a position to accurately determine the halogen content and variance in different mantle samples/reservoirs and because of the technique being used, link these to the noble gases. Our aim is now to conduct analyses on MORB and OIB glass-samples in order to reconstruct their respective mantle-source halogen compositions.