Journal of the Faculty of Science, Hokkaido University. Series 4, Geology and mineralogy;Vol.XXI, No.3

FONT SIZE:  S M L

Experimental Study of the System Mg3Al2Si3O12-Mg3Cr2Si3O12 at High Pressure and High Temperature

Irifune, Tetsuo

Permalink : http://hdl.handle.net/2115/36737

Abstract

Phase relation between pyrope Mg3Al2Si3O12 and knorringite Mg3Cr2Si3O12 is determined experimentally under high pressure and temperature condition. The solubility limit of the knorringite molecule expands with increasing pressure, and knorringite end member is stable at pressures higher than 105 kbar at 1200℃. A low pressure phase assemblage of garnet s.s. + pyroxene s.s. + spinel s.s. + quartz (coesite) appears on the pyrope-rich side, while this assemblage is replaced by that of garnet s.s. + pyroxene s.s + corundum s.s. on the knorringite-rich side at the pressures above ca. 67 kbar at 1200℃. Measurement of chemical compositions of these phases is carried out, and the distribution of Cr3+ among the coexisting phases is determined. As a result, it is confirmed that the Cr content of the garnet, as well as that of the spinel, is a useful indicator of the equilibrium pressure between spinel and garnet lherzolites. Melting experiment is made for the pyrope end member and also for the garnet solid solution between pyrope and knorringite at high pressures. It is observed that Cr3+ strongly prefers the garnet to the melt at relatively low pressures. However, such a preference is almost absent at the higher pressures around 90 kbar. On the other hand, an anomalous change of the quenched phase from the garnet melt is observed at pressures a round 70 kbar; the quench crystal changes from aluminous pyroxene with garnet stoichiometry to garnet at these pressures. Further, notable decrease in slope of the melting curve is observed for the pyrope end member at around this pressure. It is strongly suggested that these changes would be caused by a pressure-induced structural change, from pyroxene-like structure with tetrahedrally coordinated Al to garnet-like one in which every Al is in an octahedral site, in the pyrope melt at the pressures about 70 kbar. This interpretation is confirmed by the X-ray emission spectroscopic measurement, which shows structural similarity between pyrope glass and the quench crystal. Origin of Cr-rich pyrope and distribution of Cr in the upper mantle are discussed on the basis of the present experimental results. The Cr-rich pyrope, which commonly occurs as an inclusion in diamond, is estimated to be formed as a residual crystal caused by partial melting of the mantle material at pressures corresponding to 120-240 km in depth. It is also suggested that the garnet would be almost a unique phase which accommodates Cr3+ in the upper mantle from ca. 600 km to ca. 100 km depth, above which Cr3+ would be accommodated mainly in spinel and pyroxene. However, the notably high concentration of Cr3+ observed in the garnet inclusions (Cr/Cr+Al=0.2-0.5) is considered to be a local phenomenon limited in the shallower horizons above ca. 240 km.

FULL TEXT:PDF