Making Cratonic Lithospheric Mantle

Abstract

The origin of cratonic lithospheric mantle has been attributed to either high-pressure (5–7 GPa) melting in hot mantle plumes or low-pressure (<5 GPa) melting in mid-ocean ridges or suprasubduction zones. To resolve this long-standing debate, it is necessary to confirm under what depths the incipient cratonic mantle melted. Compared with most cratonic mantle xenoliths and xenocrysts, which commonly experienced metasomatic modification after melt extraction, diamond inclusions with predominantly harzburgitic compositions can better track the compositional signature of pristine cratonic mantle. This paper presents thermodynamic calculations performed with THERMOCALC software to establish a quantitative relationship between garnet and cratonic peridotite compositions. Along the normal cratonic geotherm (40 mW/m2), the XCa [atomic Ca/(Ca + Mg + Fe2+)] and Cr# [atomic Cr/(Cr + Al) × 100] values in garnet depend mainly on the bulk CaO/Al2O3 and Cr# values, respectively. Furthermore, mantle melting modeling shows that the Cr# of the residue displays a negative correlation with melting pressure at melt fractions of greater than ~15%. Therefore, the high Cr# values (mostly 12–36) of global garnet inclusions in diamond support a low-pressure (<4 GPa) origin for cratonic lithospheric mantle. More importantly, the incipient cratonic mantle calculated from the garnet inclusions exhibits similar bulk CaO/Al2O3 and Cr# compositions to oceanic lithosphere but different values from arc lithosphere with higher CaO/Al2O3 and Cr# values. These results imply that cratonic mantle was initially formed by the extensive melting of hot ambient mantle in the Archean ocean ridge environments. The shallow oceanic lithosphere was subsequently stacked to generate the thick and stable cratonic lithospheric mantle.