Crystal Field Effects in Mantle Minerals

Citation

Gaffney, Edward Stowell (1973) Crystal Field Effects in Mantle Minerals. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/3DS0-XR25. https://resolver.caltech.edu/CaltechTHESIS:07202018-110502938

Abstract

The behavior of Fe²⁺ in the lower mantle will depend on the effects of crystal fields. A point charge model, scaled to fit observed spectra at low pressures is developed to predict these effects. Two of the three parameters needed to predict spin-pairing transitions can only be determined from spin-forbidden electronic transitions. The spectra of garnet, gillespite and peridot are examined and found to have such absorption features. Assignment of these spectra leads to values of the Racah parameters, B and C, as well as the crystal field parameter Dq.

A new experimental technique, which allows the measurement of optical absorption spectra of solids in the visible region during shock loading, is described. Results are discussed for periclase and ruby. The ruby data indicate that the point charge model is good to at least 15 percent (volume) compression.

The effects of low-spin Fe²⁺ in the earth's lower mantle are investigated in considerable detail. The existence of low-spin Fe²⁺ permits the formation of a separate phase since Mg²⁺ and low-spin Fe²⁺ may not form solid solutions. The bulk elastic behavior of such phases is predicted from volume-bulk modulus systematics and compared with available shock wave data. It is likely that the high pressure phases of several ferrous iron compounds involve low-spin Fe²⁺ Iron will be spin-paired in the mantle below 1200 km and likely at higher levels as well. The observed density and bulk modulus in the lower mantle are inconsistent with any combination of phases in a pyrolite bulk composition but can be fit quite well by a model with all Fe²⁺ spin-paired below 630 km and nearly olivine composition at that depth, with MgO decreasing to almost a pyroxene composition at the core.

An origin of the upper mantle from the lower mantle by chemical fractionation is proposed. The spin-pairing of Fe²⁺ provides an excellent mechanism for both iron and silicon enrichment in the lower mantle by partial melting yielding a pyrolite upper mantle, and hence, a chemically inhomogeneous mantle. This removes the motivation for reducing FeO and SiO₂ in the mantle to supply Fe and Si for the core.

Thesis Files