The origin and development of the Earths mantle in the Archean; subcalcic garnets and eclogites as oldest time markers

点击翻译按钮获取中文摘要

The timing of growth and amalgamation of Archean continental nuclei, and the temporal and spatial relationships between processes recorded in the present day subcratonic mantle and its overlying crust are fundamental questions concerning stabilization of long-lived cratonic blocks by buoyant underlying lithospheric mantle (lithospheric keel). A prevailing model favours the origin of the subcratonic mantle mostly as the restite of partial melting at shallow depths followed by subduction. Eclogites and garnet pyroxenites from the SCLM are accordingly the subducted melt products. We have previously used single grain harzburgitic, subcalcic garnets to obtain information on the timing of partial melting and metasomatism of the subcratonic mantle underneath the Kaapvaal craton. This yields defined time slots in agreement with crustal events rather than a continuum of events as seemingly shown by Re-Os model ages. Eclogites and garnet pyroxenites which were also erupted by kimberlites through Archean crust are very heterogeneous as a result of multiple origins of their protoliths and potential chemical changes before and after eclogitisation. Tantalizing questions are the nature of the protoliths, their age and the age of eclogitisation and the comparison to ages from peridotites. Our studies of eclogites and garnet pyroxenites from the Bellsbank diamond mine on the Kaapvaal craton indicates that a number of them represent almost unmodified members of a subducted oceanic crust (= almost unaltered original melt compositions, plagioclase rich cumulates and clinopyroxene-rich cumulates). Their reconstructed bulk rocks linearly align in an isochron diagram and three samples give an age of 4.12 ± 0.06 Ga with eHfi = 3 (±7) i.e. within error the ratio of the primitive mantle at that time. This very old age is not directly matched by any crustal ages from the Kaapvaal craton or by Re-depletion ages from its lithospheric mantle. Lu-Hf-Model ages from garnet alone are minimum ages but they already give ages up to 3.5 Ga. In order to further explore the detailed picture of the origin and modification of the subcratonic mantle, i.e. the distinction of the melting regimes and the applicability of melting models and the timing of first melting and of reenrichment events we want extend our studies of subcalcic to further localities of the Kaapvaal. A second aim is to verify the 4.1 Ga eclogite isochron with further samples from Bellsbank. If correct, it would be the oldest age ever obtained from a mantle eclogite suite and would have profound implications for the production and storage of Hadean crust in the lithospheric mantle.