The Geochemistry of Mafic and Ultramafic Rocks from the Archaean Greenstone Belts of Sierra Leone

H. R. Rollinson
Department of Geography and Geology, The College of St Paul and St Mary, The Park, Cheltenham, Glos. GL50 2RH, UK

Abstract: The Archaean (c. 2800 Ma) ultramafic rocks in eastern Sierra Leone cut basalt lavas and are mostly olivine-rich cumulates either iron-rich (Fo85–86) and derived from a basaltic or picritic parent, or more magnesian (Fo92–93) derived from an ultramafic melt with c. 18–25 wt. % MgO. In central Sierra Leone the ultramafic rocks are lavas predating tholeiitic basalts.

The basalts show a wide variation in Zr/Y, suggesting that garnet was present in the source region of some of these rocks but not others. This implies that melting took place at different depths in the mantle. The REE evidence for basaltic rocks in the upper part of the Nimini belt succession suggests that they were derived from a mantle source region which had already suffered melt extraction. Ti/Zr ratios in the basaltic rocks are also variable and individual belts define different trends on a Ti vs. Zr plot implying that the basaltic rocks evolved in geographically distinct magma chambers. It is likely that the basaltic rocks evolved from a parental liquid with Ti/Zr = 90 via shallow level crystal fractionation. The source region for these rocks therefore had a lower than chondritic Ti/Zr.

There are two possible models for the basaltic and ultramafic magmas in the Sierra Leone greenstone belts. First that the ultramafic and basaltic liquids were derived from mantle diapirs of differing size, but originating in the same region of the mantle. Ultramafic liquids were produced in small diapirs, which store large melt fractions, and basaltic liquids in larger diapirs which segregate larger melt fractions. A second model is based upon the double diffusion process suggested for magma chambers at mid-ocean ridges and involves a transient magma chamber from which basalts, derived from parental ultramafic liquids, are erupted, with ultramafic liquids rising directly to the surface when the magma chamber is frozen. The available data does not discriminate between these two models.

Mineralogical Magazine; September 1983 v. 47; no. 344; p. 267-280; DOI: 10.1180/minmag.1983.047.344.01
© 1983, The Mineralogical Society
Mineralogical Society (www.minersoc.org)