40Ar/39Ar-ages of phlogopite in mantle xenoliths from South African kimberlites: Evidence for metasomatic mantle impregnation during the Kibaran orogenic cycle

We applied the 40Ar/39Ar dating method to an extensive suite of phlogopites from kimberlite-hosted mantle xenoliths (dominantly garnet bearing) from the mines of Bultfontein (South Africa), Letseng-la-Terae and Liqhobong (Lesotho). Argon extraction was performed by conventional high resolution stepwise heating technique, laser incremental heating technique and laser spot analysis. All age spectra obtained by conventional analysis indicate various degrees of 40Ar loss during kimberlite emplacement, but never resulted in a total reset of the argon system. Most intriguingly, the sample-specific maximum apparent ages cluster between 1.0 and 1.22 Ga for the phlogopites with the least disturbed age spectra. A maximum apparent age of 1.02 Ga was observed during laser heating analysis. Individual grains tend to yield older ages in their cores, with successively younger ages at their rims. The range in age obtained via the laser fusion technique and with conventional stepwise heating technique agrees with each other, as well as with literature data. The often inferred presence of excess 40Ar in those phlogopites cannot explain the coherent age pattern in the large suite of samples. Hence, the age constraint of 1.0-1.25 Ga is regarded as geologically meaningful and assigned to metasomatism of the local cratonic mantle during the advent of Kibaran orogenesis (1.00-1.25 Ga). The major consequences of our findings are: (i) The argon system of phlogopite can remain closed for long time scales, even at ambient temperatures of 800-1200 °C within the mantle, most likely because the solid/solid partitioning behaviour of Ar between phlogopite and other major phases in the mantle strongly favours phlogopite, or because conventionally inferred diffusivity of argon in phlogopite is seriously overestimated. Thus, the 40Ar/39Ar phlogopite system appears to be a valuable tool for deciphering ancient metasomatic events affecting the lithospheric mantle. (ii) The cratonic lithospheric mantle below southern Africa may have been frequently influenced by different episodes of fluid or melt migration during subduction of oceanic crust at active continental margins.