Abstract
Hydrogen as an atomic impurity in mantle minerals is recurrently proposed as a key element impacting significantly on many mantle properties and processes such as melting temperature and mechanical strength. Nevertheless, interpretation based on the natural samples remains weak as we do not have yet a robust world-wild database for hydrogen concentrations in mantle minerals and rocks. Here, we report the first hydrogen concentrations in nominally anhydrous minerals from a rare selection of ultramafic rocks and minerals embedded in Mesoproterozoic Wajrakarur kimberlites (Eastern Dharwar craton, India). Based on key chemical elements, we demonstrate that olivine, pyroxenes and garnet from the Dharwar craton are of mantle origin. We quantify the hydrogen concentrations using Fourier transform infrared spectroscopy (FTIR) and mineral-specific FTIR calibrations. Calculated hydrogen concentrations are, in average, 18 ppm wt H2O in olivine, 70 ppm wt H2O in orthopyroxene and 207 ppm wt H2O in clinopyroxene. Garnet has highly variable hydrogen concentration ranging from 0 to 258 ppm wt H2O, probably influenced by nano-scale inclusions. The average of clean garnet spectra yields 14.5 ppm wt H2O. The reconstructed hydrogen bulk concentrations of Dharwar peridotites yields 40 #$ %&' ppm wt H2O. This value is two to five times lower than the estimated hydrogen concentration in the lithospheric mantle, and agree well with the lower range of hydrogen bulk concentration from the current data base for the upper mantle minerals transported by kimberlites from other cratons (e.g., South Africa, Siberia). The low hydrogen concentration in mantle minerals, together with petrological and geochemical evidence of carbonated silicate melt metasomatism in Dharwar cratonic lithospheric mantle, suggest that these xenoliths are possibly related to proto-kimberlite melts with low water activity prior to being transported to the surface by the Mesoproterozoic Wajrakarur kimberlites. These observations, valid to a depth of ~165-km, suggest that cratonic lithosphere beneath the Dharwar craton may not be particularly indicative of an abnormal hydrogen-rich southern Indian lithosphere in the late Archean and that hydroxylic weakening in olivine would induced a negligible effect on the mantle viscosity of Indian subcontinent.