Mesozoic lithospheric mantle of the northeastern Siberian craton (evidence from inclusions in kimberlite)

Abstract

Several thousand clinopyroxene, garnet, and phlogopite inclusions of mantle rocks from Jurassic and Triassic kimberlites in the northeastern Siberian craton have been analyzed and compared with their counterparts from Paleozoic kimberlites, including those rich in diamond. The new and published mineral chemistry data make a basis for an updated classification of kimberlite-hosted clinopyroxenes according to peridotitic and mafic (eclogite and pyroxenite) parageneses. The obtained results place constraints on the stability field of high-Na lherzolitic clinopyroxenes, which affect the coexisting garnet and decrease its Ca contents. As follows from analyses of the mantle minerals from Mesozoic kimberlites, the cratonic lithosphere contained more pyroxenite and eclogite in the Mesozoic than in the Paleozoic. It virtually lacked ultradepleted harzburgite-dunite lithologies and contained scarce eclogitic diamonds. On the other hand, both inclusions in diamond and individual eclogitic minerals from Mesozoic kimberlites differ from eclogitic inclusions in diamond from Triassic sediments in the northeastern Siberian craton. Xenocrystic phlogopites from the D’yanga pipe have 40Ar/39Ar ages of 384.6, 432.4, and 563.4 Ma, which record several stages of metasomatic impact on the lithosphere. These phlogopites are younger than most of Paleozoic phlogopites from the central part of the craton (Udachnaya kimberlite). Therefore, hydrous mantle metasomatism acted much later on the craton periphery than in the center. Monomineral clinopyroxene thermobarometry shows that Jurassic kimberlites from the northeastern craton part trapped lithospheric material from different maximum depths (170 km in the D’yanga pipe and mostly <130 km in other pipes). The inferred thermal thickness of cratonic lithosphere decreased progressively from ~260 km in the Devonian-Carboniferous to ~225 km in the Triassic and to ~200 km in the Jurassic, while the heat flux (Hasterok-Chapman model) was 34.9, 36.7, and 39.0 mW/m2, respectively. Dissimilar PT patterns of samples from closely spaced coeval kimberlites suggest different emplacement scenarios, which influenced both the PT variations across the lithosphere and the diamond potential of kimberlites.