Bina, C. R., and B. J. Wood, The olivine-spinel transitions: Experimental and thermodynamic constraints and implications for the nature of the 400 km seismic discontinuity, Journal of Geophysical Research, 92, 4853-4866, doi:10.1029/JB092iB06p04853, 1987.
Abstract. The sequence of high-pressure phase transitions alpha -> beta -> gamma in olivine is traditionally used as a model for seismic velocity variations in the 200- to 650-km-depth interval in a mantle of peridotitic bulk composition. It has been proposed that the observed seismic velocity increase at 400-km depth is too sharp and of too small a magnitude to be attributable to the alpha -> beta phase change and that the upper mantle must thus be chemically stratified, with the 400-km discontinuity being due either to a combination of phase changes in a layer of pyroxene/garnet-rich ``piclogite'' composition or to a chemical boundary between such a piclogite layer and an overlying peridotitic layer. Using available calorimetric, thermoelastic, and synthesis data (and their associated experimental uncertainties), we have derived internally consistent high-pressure phase relations for the Mg2SiO4-Fe2SiO4 join. We find that the divariant transition alpha -> alpha + beta -> beta, which is generally regarded as occurring over a broad depth interval for mantle olivine compositions, is, in fact, extremely sharp. The seismic discontinuity corresponding to the alpha -> alpha + beta -> beta transition in (Mg0.9Fe0.1)2SiO4 should occur over a depth interval (isothermal) of about 6 km at a depth of approximately 400 km; the sharpness of this transition is quite insensitive to uncertainties in the constraining calorimetric, thermoelastic, and synthesis data. In addition, we have computed seismic velocity profiles for a model mantle consisting of pure olivine of (Mg0.9Fe0.1)2SiO4 composition. Comparison of these computed profiles to those derived from recent seismic studies indicates that the magnitude of the observed velocity increase at 400-km depth is consistent with a mantle transition zone composed of about 60-70% olivine. We conclude that there is no need to infer the existence of pyroxene/garnet-rich compositions, such as eclogite or ``piclogite,'' in the transition zone, since an upper mantle of homogeneous olivine-rich peridotitic composition is consistent with the available seismic velocity data.Copyright © 1987 American Geophysical Union