A case study: the nepheline basanite UT from Bow Hill in Tasmania, Australia Previous work includes: An experimental study of liquidus phase equilibria to determine conditions of garnet lherzolite saturation (~2.7 GPa and 1200 °C with 4.5 % of dissolved H 2 O and 2 % of CO 2 ) A LAM-ICP-MS and electron micro-probe study of minor and trace element partitioning between peridotite minerals and the hydrous basanite melt A single-crystal X-ray and site refinement study of experimentally produced peridotite phases
Objectives To produce a comprehensive and self- consistent set of peridotite/melt partition coefficients that include volatile as well as non-volatile elements To apply these data to the problem of intraplate magma genesis
Materials and methods Piston-cylinder experiments at GPa and °C using conventional methods H 2 O, C (as CO 2 ), Cl, F, P and S by secondary ion mass spectrometry (SIMS) at the Carnegie Institution of Washington. Major, minor and trace elements by electron microprobe and LAM- ICP-MS at Macquarie University, Australia. Basanite UT (an intraplate basalt & potential near-solidus melt of garnet lherzolite) SiO plag TiO or Al 2 O ne Cr 2 O di Fe 2 O ol FeO*10.05ilm 4.50 MnO 0.20mag 3.83 NiO 0.05ap 3.31 MgO CaO 9.46 SrO 0.18 Na 2 O 4.24 K 2 O 2.09 P 2 O Total x Mg/(Mg + Fe +2 ) = 68.2
Run R °C 1.0 GPa Run R °C 3.0 GPa glass quenched melt (porous crystallite matrix) olivine + cpx graphite inner capsule Pt outer capsule pyroxenes
1:1
H 2 O ppmF ppm Clinopyroxene Orthopyroxene Olivine Garnet Pargasite Phlogopite Cl, S and C have negligible concentrations in nominally anhydrous silicate minerals Cl and S have small to moderate solubilities in amphibole and mica ~ 0.5 wt. CO 2 is soluble in the melt phase at 1-2 GPa
clinopyroxene/melt partitioning of H 2 O
orthopyroxene/melt partitioning of H 2 O
olivine/melt partitioning of H 2 O
Garnet/melt partitioning
Controls on crystal/melt partitioning 1.crystal-chemical effects Tetraherally co-ordinated Al in pyroxenes 2.melt-activity relations Burnham’s (1975) solution model for H 2 O in silicate melts Silver et al.’s (1990) solution model for H 2 O in silicate melts
Effect of iv Al +3 in charge-balancing the addition of H + to the pyroxene lattice [pyx] H + + [iv] Al +3 + [melt] Si +4 [melt] H + + [melt] Al +3 + [iv] Si +4
Burnham’s speciation model for hydrous melts mixing between OH - and 8-oxygen melt units with K D =
Silver et al.’s (1990) speciation model for hydrous melts mixing between OH -, molecular H 2 O and O 2- with D OH cpx/melt = 0.05 molecular H 2 O OH -
Melt activity relations as well as mineral composition play a role in the determination of mineral/melt (Nernst) partition coefficients for H 2 O
Comparisons with non-volatile elements There is a constancy of some volatile to non-volatile element ratios in intraplate (ocean island) and mid-ocean- ridge magmas H 2 O/Ce = 200 ± 50 (Michael 1995) o Higher ratios in Atlantic than Pacific, in some cases the correlation is better for La and/or varies with 87 Sr/ 86 Sr F/Nd Cl/Ba, Cl/K CO 2 /Nb, CO 2 /Ba Similar mineral/melt partition coefficients during intra-mantle fractionation involving the migration of small-degree melts from local MORB sources?
H2OH2O F Cl Cl and F partitioning data from Dalou et al. (2012) Garnet lherzolite/melt Partition coefficients
Crystal-chemical controls on D H2O /D Ce Must also consider the contributing influence of H 2 O concentrations in melts (significant for subduction zones)
Conclusions During peridotite melting D H2O /D Ce increases with increasing pressure and temperature (and therefore depth in the mantle), but decreases with increasing melt H 2 O Therefore - although coupled volatile and non-volatile element enrichments in OIB are consistent with peridotite/melt partitioning, this may require particular circumstances
Thank You
av. OIB normalized to av. MORB D Z peridotite/melt
The influence of iv Al on F in pyroxenes
The influence of Ti and melt H 2 O concentrations on OH in amphibole O3 sites [melt] Ti [melt] O 2- + [M1,3] Mg [O3] OH - [melt] Mg [melt] OH - + [M1,3] Ti + 2 O3 O 2-