Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - EFOX (The formation and evolution of monomineralic oxide layers in mafic intrusions)

Teaser

Layered mafic intrusions represent a major source of precious metals such as platinum group elements (PGE), Ti, V, and Cr. These precious metals are hosted in particular in chromitite (mainly composed of mineral chromite, FeCr2O4) and magnetitite (mainly composed of mineral...

Summary

Layered mafic intrusions represent a major source of precious metals such as platinum group elements (PGE), Ti, V, and Cr. These precious metals are hosted in particular in chromitite (mainly composed of mineral chromite, FeCr2O4) and magnetitite (mainly composed of mineral magnetite (Fig. 1). Fe3O4) layers. The EFOX project focused on untangling the big unknown as for how these thick oxide layers formed and understanding their sub-solidus evolution (i.e. events that occurred after crystallisation of main forming minerals). I proposed an innovative approach by applying detailed microstructural analyses coupled with mineral geochemistry, to study oxide layers of the largest layered intrusion in the world, the Bushveld Complex, South Africa. The Bushveld Complex is divided into 5 zones, but the two most important are the chromitite hosting Critical zone (CZ) and magnetitite hosting Upper zone (UZ). The main objectives are:
A. Characterisation of the original microstructure in the oxide layers;
B. Determination of the extent to which the original microstructure has been modified;
C. Constraining the physical and chemical effects oxide layer formation had on the under- and overlying layers;
D. Determination of the mechanisms by which oxide layers form in layered mafic intrusions, using the predicted criteria listed above.

Work performed

The microstructural analysis was conducted by Electron backscatter direction analysis (EBSD), and it was coupled with elemental mapping (QEMSCAN) and mineral chemistry (electron probe micro-analyser - EMPA). All the analyses were performed at the University of Cambridge. Twenty samples of oxide layers and their surrounding rocks have been analysed by EBSD. The EBSD data produced crystal orientation maps, crystallographic pole figure data, and maps showing intra-crystalline deformation. Additional grain size analaysis have been done on the chromite grains from the chromitite layers. Data was used to interprete the depositional enviroments for the two oxide layers and their surrounding rocks. Samples that have been studied from the magnetitite layers of the Upper Zone of the Bushveld Complex are coming from the Bierkaal borehole drilled in the Western Limb.

The Magnetitite Layers and the surrounding anorthosites
The EBSD data showed that the amount of deformation intragrain microstructures (e.g. microstructure recorded within the single crystal) in the plagioclase crystals is the same in anorthosites that are located below and above the magnentite layers. Anorthosites are also characterised by two populations of plagioclase grains, large >1000 µm in diameter in small grains < 1000 µm (Fig 2). The small population of the plagioclase grain surrounds the large plagioclase grains (Fig. 2). These small grains are not a result of recrystallisation of the large grains but represent a generation of plagioclase that crystallised from the interstitial melt.
Plagioclase crystals within the magnetitite layers show three types of appearance: (1) isolated round grains; (2) single elongate, grains that record evidence of minor plastic deformation with only minor marginal recrystallisation; and (3) grains that preserve their original shape but with more extensive recrystallization along grain boundaries with other plagioclase grains. I proposed that these isolated plagioclase grains moved within this liquid-rich slurry, but in case of plagioclase aggregates, strain localised along the contacts of aggregated grains. The numerous small, rounded plagioclase grains may have originally formed part of recrystallised laths that subsequently disintegrated during the flow of the liquid-rich mush.

The Chromitite Layer and the surrounding anorthosites

The grain size analysis of the chromite crystals from the UG2 chromitite layer showed that there is no significant grain size variation within the main chromitite stream (Fig. 3a). Due to the uniformity of the crystal size, gravitational sorting is unlikely to have a strong effect in the deposition of chromite crystals. Crystallographic data show that chromite grains show no crystallographic preferred orientation and that chromite cluster shows no evidence of crystallographic control of individual crystal within a cluster (Fig. 3d).
The remnant of olivine crystals and pegmatoidal orthopyroxene in underlying harzburgitic rock is likely a product of the metasomatic reaction between olivine-rich rock and late Si-rich melt. Chromite texture is not affected by this metasomatic reaction. The overlying pyroxenite rock on the other side shows strong very magmatic fabric formed by deposition of pyroxene grains from a magmatic current. The pyroxene grains are mantled by fine chromite crystals. Like in the case of magnetitite layers, there is no strong microstructural evidence, that compaction had a significant role in the formation of these rocks.

Final results

The main outcome of the project is that I was able to set two scenarios for the formation of the oxides layers. The textural and geochemical work done on the magnetitite layers of the Upper Zone of the Bushveld Complex showed that magnetitites and the surrounding anorthosites emplaced in the Upper Zone as mobile crystal mush. It also showed that even these rocks are characterised with very high-density rocks, they do not show evidence for viscous compaction, but that deformation microstructure is recorded during long-lasting subsidence of the Bushveld complex.
The textural study on the chromitite layers has been performed on the UG2 layer from the Khuseleka mine site in the Western Limb of the Bushveld Complex. The textural and geochemical study of the chromitite layers and their surrounding rocks revealed that these have deposited in a slowly moving magma. The rocks that underly the thick chromitite layer (UG2), the feldspathic harzburgites have experienced subsequent alteration by late magmatic fluids. The project also showed that fine chromite crystals are very poorly sorted and that they show very uniform grain size variation within the layer.
The outcomes of the Fellowship brought new insights for understanding the fluid dynamics of crystal-rich systems, but also it brought up to attention that it is necessary to re-investigate alternative mechanisms for melt extraction from a cumulate pile. Melt extraction from a cumulate pile is considered an important process that is crucial for understanding the onsets of volcanic eruptions. It was believed this process is governed by compaction of cumulates, and my research has proven otherwise.

At the onset of the MSCA fellowship, I was unfamiliar with the analytical machinery needed to produce and process QEMSCAN elemental maps. I can now set and process the data obtained by ESPRIT post-processing software in order to produce the high-resolution elemental maps. I strengthen up my overall geochemical skills that are complementary to my expertise in microstructural geology. I also extended my knowledge in understanding the fluid dynamics of a particle-rich system. This has greatly broadened up my skills and perfect my understanding of magmatic processes. Gaining these additional skills would not be possible without the independence of Marie Sklodowska-Curie Individual Fellowship.