An unusually complete suite of eclogite types: Eclogite xenoliths from the Zero kimberlite, South Africa

Abstract

This first investigation of eclogite xenoliths sampled by the Zero kimberlite show that the suite is exceptional concerning kimberlite age and diversity of eclogite types. The Zero kimberlite intruded as part of a sequence of non-diamond bearing kimberlites within the oldest kimberlite province on the Kaapvaal craton (1600 Ma; Northern Cape Province, South Africa) towards the western craton margin. Distinct eclogite parageneses, including bimineralic eclogites, usually rare orthopyroxene-bearing eclogites with and without sanidine in addition to one kyanite eclogite occur in the suite. Garnets exsolved from clinopyroxene host occur in all Zero eclogites, whereas in orthopyroxene bearing eclogites, orthopyroxene and, if present, sanidine are also exsolved from clinopyroxene. That most eclogites occur with quartz, rutile and rarely K-feldspar inclusions in garnets might be taken to imply a crustal precursor. However, this is in conflict with the analyzed oxygen isotopic composition of the eclogites, which yield δ18O values of 5.18-5.47, close to values for the uncontaminated Earth"s mantle. Irrespective of the original source of the eclogite suite, a complex history within the mantle can already be inferred from the exsolved constituents of the sanidine eclogites. Thermobarometrical results indicate the last P/T condition prior to the kimberlite eruption was at 900-950 °C at 3.5 GPa. The previous stage was deduced from the reconstructed K2O-content of clinopyroxene host prior to the exsolution of sanidine. High K2O contents in clinopyroxene, which apparently resulted from the breakdown of dehydrating phlogopite are typical of high pressures and can be used as a crude barometer. The K2O content of 0.2 wt% calculated by phase map analysis requires pressure conditions of 6-8 GPa. Hence the eclogites were carried to 180-210 km depth prior to their exhumation. The eclogite suite from the ~1.6 Ga Zero kimberlite is mineralogically unusually diverse, bears evidence for multiple P-T-t-stages and is characterized by mantle-like oxygen isotopic compositions. In-situ trace element analyses of minerals show that sanidine bearing, orthopyroxene bearing, bimineralic and kyanite eclogites, have REE abundances close to, or even more depleted than, primitive mantle, demonstrating that they originated from a fractionated igneous source. Most eclogites show enrichment in LREE to different extents caused by a metasomatic imprint at a late stage. The kyanite eclogite and one bimineralic eclogite occur with slight positive Eu-anomalies, whereas in the remaining eclogites no indication of plagioclase in the protolith is present. High Ni and Cr contents of some orthopyroxene bearing eclogites indicative that olivine and chromium-spinel were present in the protolith. Low trace element abundance including low Nb, Ta, Zr and Hf appears to be characteristic for the eclogite suite. According to trace element modeling combined with modeled crystal accumulation from potential parental magmas we can show that the protolith of the eclogites were generated by igneous fractionation from picritic magma generated at 1.8 GPa. The best model for the Zero eclogites resulted from crystallization of the picrite at 1-1.5 GPa, indicating development of the eclogite protolith in the upper mantle, suggesting a formation within the sub-continental lithosphere. That the resulted cumulate consisted of spinel and clinopyroxene in the presence of plagioclase has significant implications for the genesis of the Zero eclogite. Although the presence of plagioclase in the protolith is not indicated in the trace element pattern by Eu- anomalies in most Zero eclogites, the modeled trace element patterns show that small amounts of plagioclase were present in the protolith, but in insufficient amounts to lead to noticeable positive Eu anomalies. The requirement for plagioclase is in contrast to the hypothesis for crystallization at mantle pressures. The metamorphic history of the Zero eclogites can be used to constrain geodynamic processes at the western margin of the Kaapvaal craton during the early Proterozoic. The eclogitic protoliths were formed at about 30-45 km, and then were brought to depths of at least 180-240 km (6-8 GPa), witnessing that sub-continental lithosphere was displaced at least 150 km deeper into the mantle. Construction of the P-T-t-path draws largely on sanidine bearing eclogites, which originated from metasomatic veins consisting of phlogopite and clinopyroxene pervading cumulates prior to metamorphism. The eclogitization of the metasomatised protolith was caused by the collision of the Kheis terrane (2 Ga) with the Kaapvaal craton (2.7 Ga) at the western margin of the craton. As the eclogite protolith was metasomatised prior to the eclogitization, it is likely that the protolith formed in a magma chamber situated underneath the Kheis belt which was later underplated prior to the collision with the Kaapvaal craton. Hence, the geological situation argues that the Zero eclogites should not be notably older than the Kheis belt, which is about 2 Ga.