The determination of accurate emplacement ages for kimberlites and related rocks has important implications for understanding the processes involved in the genesis and intrusion of kimberlitic magmas. As kimberlitic intrusive events are recorded from Eocene to Proterozoic times, they also provide constraints on mantle and crustal geochemical and geodynamic models through time. In turn, this information is utilized in an economic context to focus diamond exploration efforts. A number of

more »

... ic dating techniques have been applied to kimberlite and related rock geochronology, with varying degrees of success (see reviews by Davis et al., 1996 and Allsopp et al.. 1989) . The most popular methods are Rb-Sr dating of phlogopite and U-Pb dating of zircon and perovskite. As each of these techniques has certain limitations, it is generally recommended that, where possible, more than one method is used. Despite the high potassium contents of kimberlites and related rocks, the K-Ar and 40Ar/39Ar techniques have not been widely applied to dating emplacement events. This is largely a result of the common presence of excess 40Ar and/or the loss of radiogenic argon (cf. Phillips, 1991; Pearson et al., 1997) . Roddick (1984), Smith et al. (1985) and Phillips (1991) have suggested that the most reliable 40Ar/39Ar results are obtained from groundmass micas. However, general application of the method has not been demonstrated. Furthermore, matrix phlogopite is rare in some kimberlites and the separation of pure mineral separates for "conventional" 40Ar/39Ar analyses can be an arduous task, particularly given the fine grain sizes involved (<200pm). In the current study, the above problems are addressed through laser probe 40Ar/39Ar analyses of groundmass phlogopite from a wide variety of kimberlites and related rocks in southern Africa. The improved spatial resolution of the laser probe allows for the analysis of single groundmass phlogopite grains with grain diameters as small as 50p.m. For finer-grained samples, or in cases where phlogopite is rare, grains may be analysed in thin sections. The laser probe obviates the necessity for bulk separates -in most cases, five single phlogopite grains are adequate for determining accurate intrusion ages. The analysis of single grains also serves to identify inherited, xenocrystic, phenocrystic and altered mica grains, which cause problems in bulk analyses. Individual phlogopite grains are step-heated and/or subjected to core-rim laser spot analyses. The reliability of the age data can be assessed through intra-and inter-grain reproducibility of resultsin the current study, most 40Ar/39Ar ages are also compared to results from other dating methods.