Petrology, genesis and geodynamic implication of the Mesoproterozoic–Late Cretaceous Timmasamudram kimberlite cluster, Wajrakarur field, Eastern Dharwar Craton, southern India

New mineralogical and bulk-rock geochemical data for the recently recognised Mesoproterozoic (ca. 1100 Ma) and late Cretaceous (ca. 90 Ma) kimberlites in the Timmasamudram cluster (TKC) of the Wajrakarur kimberlite field (WKF), Eastern Dharwar Craton, southern India, are presented. On the basis of groundmass mineral chemistry (phlogopite, spinel, perovskite and clinopyroxene), bulk-rock chemistry (SiO₂, K₂O, low TiO₂, Ba/Nb and La/Sm), and perovskite Nd isotopic compositions, the TK-1 (macrocrystic variety) and TK-4 (Macrocrystic variety) kimberlites in this cluster are here classified as orangeites (i.e. Group II kimberlites), with geochemical characteristics that are very similar to orangeites previously described from the Bastar Craton in central India, as well as the Kaapvaal Craton in South Africa. The remaining kimberlites (e.g., TK-2, TK-3 and the TK-1 microcrystic variant), are more similar to other 1100 Ma, Group I-type kimberlites of the Eastern Dharwar Craton, as well as the typical Group I kimberlites of the Kaapvaal Craton. Through the application of geochemical modelling, based on published carbonated peridotite/melt trace element partition coefficients, we show that the generation of the TKC kimberlites and the orangeites results from low degrees of partial melting of a metasomatised, carbonated peridotite. Depleted mantle (T_DM) Nd perovskite model ages of the 1100 Ma Timmasamudram kimberlites show that the metasomatic enrichment of their source regions are broadly similar to that of the Mesoproterozoic kimberlites of the EDC. The younger, late Cretaceous (ca. 90 Ma) TK-1 (macrocrystic variant) and TK-4 kimberlites, as well as the orangeites from the Bastar Craton, share similar Nd model ages of 1100 Ma, consistent with a similarity in the timing of source enrichment during the amalgamation of Rodinia supercontinent. The presence of late Cretaceous diamondiferous orangeite activity, presumably related to the location of the Marion hotspot in southern India at the time, suggests that thick lithosphere was preserved, at least locally, up to the late Cretaceous, and was not entirely destroyed during the breakup of Gondwana, as inferred by some recent geophysical models.