Identification of High-Pressure Phases of Rocks and Minerals from Hugoniot Data

Abstract

Crystal structures which have been proposed as possible high-pressure phases of some rocks and minerals are tested by comparing predictions of their pressure—density Hugoniots with shock-wave Hugoniot data. The predictions are based on tentatively established systematic variations of bulk modulus and its pressure derivative with density, composition and crystal field effects. This procedure avoids the ambiguity of extra-polating to zero pressure data which do not clearly define a high-pressure phase. New or more strongly supported interpretations of crystal structure are obtained in seven of the 12 cases considered. More than one phase transition seems to be involved in bronzitite and fayalite. Perovskite is a likely structure for the highest-pressure phase of bronzitite. Electronic spin transitions in Fe2+ and Fe3+ seem to be required to explain the high densities of the magnetite and haematite high pressure phases, and are likely to be involved also in fayalite and Mooihoek dunite. Eclogite data do not clearly define a high-pressure phase, but high-pressure densities approach those of the isochemical mixture of oxides. The high-pressure almandite-pyrope garnet phase seems to have a lower zero-pressure density than previously inferred, but may still be in the ilmenite structure as suggested. The exceptionally high densities attained by the forsterite and enstatite ceramics are reaffirmed. Forsterite seems to achieve a phase about 10 per cent denser than the isochemical oxides mixture, possibly in the K2NiF4 structure. Previous interpretations of Twin Sisters dunite, spinel and rutile high pressure phases are reaffirmed.