Project title: Compositional heterogeneity of the Earth's convecting mantle: Constraints from primitive melts in Continental Flood Basalt provinces
The sublithospheric-mantle has long been thought to be heterogeneous; signals observed in the isotopic and trace-element compositions of mantle-derived melts may relate to underlying lithological heterogeneity, which in turn must derive from geodynamic processes such as subduction and delamination.
Continental flood basalt (CFB) provinces are thought to derive from upwelling hot mantle (plumes) which potentially sample deep material from the lower mantle. Heterogeneity in major and trace elements and isotopic compositions in rare, low-volume primitive magmas in CFB settings relate to deeper mantle processes and are the product of both melting regime and lithological heterogeneity in the mantle source. Of particular interest is a rare primitive magma type known as 'ferropicrite', which has been suggested to derive from high pressure partial melting of pyroxenite, which itself derives from subducted oceanic crust. If correct, this is strong evidence for the deep recycling of ancient oceanic plates and their return to the surface in mantle plume starting heads.
My work examines the origin of CFB primitive melts with a particular forcus on those from the Paraná-Etendeka CFB province, which can reveal the mantle processes which produce the vast CFBs that are associated with major environmental change and mass extinctions. I develop major and trace element models and use geothermometry to understand the origin and early processing of melts formed at elevated temperatures beneath thick lithosphere in continental setting. Predicting the major and trace element chemistry of pyroxenite partial melts and seeing if it really does explain the origin of the enigmatic ferropicrite is of particular interest.
An important source of data for this is melt inclusions in olivine, which are preserved in quasi-closed system and should record the relative timing of the homogenisation of any initial melt heterogeneity in these mantle-derived melts, showing how they recombine and crystallise in deep crustal sills.
Analytical techniques used include EPMA, SIMS, LA-ICPMS and XANES, which provide constraints for modelling. The slow-cooled nature of the samples means that additional steps must be taken to prepare melt inclusions for analysis - they must be re-melted in a gas-mixing furnace.
Ferropicrite sample, Etendeka, Namibia:
Melt inclusions, in various stages of homogenisation: