DI31A-2546
Constraining the Protolith of Large, Macroscopically Layered Kyanite-bearing Eclogite Xenoliths from the Kaapvaal Craton, South Africa

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Chelsea Claire Rebelo, University of Cape Town, Cape Town, South Africa
Abstract:
We describe the geochemistry of a suite of ten layered, kyanite-bearing eclogites from the Roberts Victor kimberlite, Kaapvaal Craton, South Africa. All samples are characterized by clear zonation corresponding to the presence or absence of kyanite and the state of preservation of clinopyroxene. The sampled zones are defined as KF (free of kyanite, with well-preserved garnet and clinopyroxene), KZ (preserved clinopyroxene is absent; garnet occurs with kyanite); and TZ (partially preserved clinopyroxene and more altered garnet; kyanite may or may not be present). We report on results of petrographic observations, mineral chemistry, and trace element and oxygen isotope analyses for the different zones. Results from adjacent zones are presented as geochemical transects perpendicular to the layering for each xenolith. We use the results to infer the geochemical evolution of likely protoliths for the various samples.

Mineral chemistry of garnets across the different zones shows decreasing FeO*, Cr2O3, MgO and MnO and increasing in CaO from KF into KZ. Clinopyroxene shows increasing in Al2O3 and Na2O from KF into KZ. Clinopyroxenes approach a more jadeitic composition towards the KZ and are more diopsidic in KF.

Trace element analyses were conducted with laser ablation ICP-MS on hand-picked mineral separates. Garnets show subchondritic LREE depletion and superchondritic HREE enrichment in both KF and KZ. Positive europium anomalies in garnet are present in all zones, with relatively larger anomalies in garnet grains from KZ. Preserved clinopyroxenes show complementary REE patterns to garnet grains, with superchondritic LREEs and subchondritic HREEs.

For the suite of xenoliths, results from mass spectrometry on oxygen extracted by laser fluorination on mineral separates show that δ18O values are above the accepted mantle value and fall within the range of hydrothermally altered oceanic crustal material. At this stage of the research, the bulk protolith is interpreted to be of crustal origin.