The alleged kimberlite-carbonatite relationship: evidence from ilmenite and spinel from Premier and Wesselton mines and the Benfontein sill, South Africa

Abstract

Carbonate-rich, SiO_2-poor residua are developed in some kimberlites solidifying as ocelli, layers, or discrete dikes which satisfy petrographic definitions of carbonatite. Arguments that these rocks have mineralogies, antecedents, and comagmatic rocks differing from those of the carbonatites in alkaline rock complexes, including the specific observation that kimberlites and carbonatites contain ilmenites and spinels of different composition, have been used to refute the "alleged kimberlite-carbonatite relationship". New microprobe analyses of ilmenites and spinels from carbonate-rich rocks associated with kimberlites in three South African localities correspond to spinels and ilmenites of carbonatites from alkalic complexes, or have characteristics intermediate between those of carbonatites and kimberlites. The ilmenites are distinguished from kimberlite ilmenites by higher MnO, FeTiO_3, and Nb_2O_5, and by negligible Cr_2O_3. The spinels are distinguished from kimberlite spinels by their Al_2O_3 and Cr_2O_3 contents. There is clearly a genetic relationship between the kimberlites and the carbonate-rich rocks, despite the observation that their ilmenites and spinels are distinctly different, which indicates that the same observation is not a valid argument against a petrogenetic relationship between kimberlites and carbonatites. These rocks are among the diverse products from mantle processes influenced by CO_2, and we believe that the petrogenetic links among them are forged in the upper mantle. We see insufficient justification to deny the name "carbonatite" to carbonate-rich rocks associated with kimberlites if they satisfy the petrographic definition in terms of major mineralogy.

Additional details