A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science. Johannesburg, 2017. / Massif-type anorthosites formed in the Proterozoic Eon are the most voluminous anorthosite occurrences on Earth, reaching tens of thousands of square kilometers in aerial extent. While they formed throughout the Proterozoic, most formed during a 700 Ma period between 1800 and 1100 Ma. The rocks are dominated by plagioclase (typically 70 – 95 volume %) of intermediate composition (An40-65). Olivine, orthopyroxene, clinopyroxene and Fe-Ti oxides make up the minor mafic proportion. While most researchers agree that the anorthosites formed from a high-alumina basaltic parental magma, there are disparate views on how that parental magma was generated. Whether the parental magma formed by partial melting of the lower crust, or by mantle melting, is a topic of much debate. The anorthosites commonly have crust-like isotopic signatures, but this could be produced by melting of the lower crust, or by crustal contamination of mantle-derived magmas. Many Proterozoic anorthosite complexes consist of both olivine-bearing and orthopyroxene-bearing anorthosites. This has been attributed to variable amounts of crustal contamination of mantle-derived magmas, based on evidence from isotopes and field relations. While geochemical and petrologic evidence for crustal contamination is plentiful, existing experimental work shows that a thermal divide exists for high-alumina basalts fractionating at lower crustal depths, casting doubts on whether fractionation of a mantle melt could produce anorthosite. Here I use high-pressure experiments to test whether the fractionation of high-alumina basalt can form anorthosites, and to what extent crustal contamination affects the fractionation sequence. The results are compared to new geochemical and petrologic data from the Kunene Anorthosite Complex (KAC), in Angola and Namibia.
The KAC is one of the largest anorthosite complexes in the world, with an area of ~18 000 km2. The KAC (1438 – 1319 Ma) has an elongate shape and intruded into Palaeoproterozoic to Mesoproterozoic country rocks (~2200 to 1635 Ma) at the southern margin of the Congo craton. It is associated with a suite of granitoid rocks of variable composition, which are akin to the granitoids associated with nearly all Proterozoic anorthosites. The granitoids have been shown to be coeval with the anorthosites, but were from a chemically independent magma series. The most distinctive granitoids in the KAC are the Red Granites, which outcrop around the southern margins of the complex, and also cross-cut the complex in a NE-SW linear belt, dividing the complex roughly into northern and southern domains. The rocks of the KAC are highly variable in terms of mode, mineral chemistry, and texture, but there is a general trend of more olivine-bearing anorthosites north of the granite belt, and orthopyroxene-bearing anorthosites to the south. The olivine-bearing rocks (or
leucotroctolites) typically contain plagioclase and cumulus and/or intercumulus olivine, with lesser interstitial orthopyroxene and/or clinopyroxene, Fe-Ti oxides, and biotite. The orthopyroxene-bearing anorthosites (or leuconorites) contain cumulus plagioclase ± cumulus orthopyroxene, and interstitial orthopyroxene, clinopyroxene, oxides and biotite. The leucotroctolites are characterized by more calcic plagioclase (An56-75), while the leuconorites contain more intermediate plagioclase (An48-56). The variability of the rocks across the complex suggests that the KAC consists of several coalesced plutons with different histories. The petrologic data and field observations in this study are consistent with the leuconorites of the complex being derived from a mantle-derived magma that experienced contamination by silica-rich rocks, crystallizing orthopyroxene rather than olivine, and less calcic plagioclase. The leucotroctolites experienced less or no contamination.
To test whether the mineral dichotomy and the variations in plagioclase chemistry observed in Proterozoic anorthosites are due to variably contaminated mantle-derived magma, piston cylinder experiments were conducted on a synthetic high-alumina basalt (HAB) composition, as well as a mixture of this HAB with 30% of a Red Granite composition. Experiments were conducted at 10 kbar, to simulate the depth at which anorthosite differentiation most likely begins (based on Al-in-orthopyroxene geobarometry of highly aluminous orthopyroxene megacrysts that occur in many massifs). The uncontaminated experiments produced olivine as the first liquidus phase, followed by plagioclase (An65-68), and then by clinopyroxene, pigeonite and ilmenite at progressively lower temperatures. Residual liquids evolve towards more silica-rich compositions with decreasing temperature. The contamination experiments produced liquidus orthopyroxene, followed by plagioclase (An51-56), and then by pigeonite at lower temperatures. The experiments show that contamination of a primitive HAB magma by granitic material, most likely produced by partial melting of the lower crust during anorthosite formation, can shift the mineral assemblages of the crystallizing anorthosite from olivinebearing to orthopyroxene-bearing, and produce less calcic plagioclase than the uncontaminated HAB magma. This could explain the observation of olivine-bearing and orthopyroxene-bearing anorthosites in the KAC and many other Proterozoic anorthosites.
Previous high-pressure experimental studies, using a slightly more evolved HAB composition, indicated the presence of a thermal divide, which causes liquids to evolve to more Si-poor compositions. The experimental results presented in this study however, do not show a thermal divide, indicating that small variations in experimental starting composition can cause large differences in the liquid line of descent. The results of this study indicate that partial melting of the mantle can produce anorthosite parental magmas, and that the range in
mineral assemblages of the anorthosites can be accounted for by crustal contamination of a mantle-derived magma.
Fractionation of the experimental starting compositions was also modeled using the MELTS algorithm. These calculations produce a close match to the experimental liquid trends. This allows for modeling of a variety of compositional and environmental variables. The MELTS modeling shows that as little as 10% contamination of HAB magma with a granitic composition may position the magma in the orthopyroxene stability field, forming orthopyroxene-bearing anorthosites. The modeling also shows that a variety of silica-rich contaminants, including granites, granodiorites and tonalities, produce similar results and liquid evolution trends, so a range of granitoid compositions may successfully produce the shift in mineral assemblages of the anorthosites. This suggests that crustal contamination of mantle-derived HAB could be a widespread process and the primary mechanism that produces the distinctive crust-like signatures in Proterozoic anorthosites.
In summary, the mineralogical and chemical diversity observed in Proterozoic anorthosites can be produced by variable amounts of crustal contamination of mantle-derived, highalumina basaltic magma. The experimental results in this study combined with field observations, and geochemical and isotopic data, provide evidence for a model of massif-type anorthosite petrogenesis. Orthopyroxene-bearing rocks formed from an originally highalumina basaltic magma that experienced contamination by granitic partial melts of the lower crust, during ponding of the magma at the Moho. This process preconditioned the surrounding crust and possibly prevented further anatexis. Following emplacement of orthopyroxene-bearing anorthosites, subsequent magma pulses ponded at the Moho did not assimilate any/as much granitic material, as they were interacting with preconditioned crust, and formed olivine-bearing anorthosites. With better constraints on the parental magma composition, magma source, and crustal contamination processes, addressing aspects such as the tectonic setting and emplacement mechanisms of these massive intrusions should be prioritized. Understanding these enigmatic aspects of anorthosite petrogenesis is leading the anorthosite community towards answering the ultimate questions of why massif-type anorthosites are restricted to the Proterozoic. / XL2018
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/23584 |
Date | January 2017 |
Creators | Hill, Catherine Mary |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Thesis |
Format | Online resource ([xv], 150 leaves), application/pdf |
Page generated in 0.0021 seconds