The information derived from this investigation will be applicable to such diverse problems as the origin of island arc volcanic series and the trace element characteristics of the mantle as well as being applicable to the understanding of the formation and evolution of the majority of volcanic rocks on earth. In addition, this data will be used to calculate mineral- liquid expressions to improve our existing model for the simulation of igneous differentiation. To understand the evolution of natural volcanic magmas, we must understand the chemical relationship between the magmatic liquid and the minerals that crystallize from it. At present, there is not enough data to define this chemical relationship for major and trace elements for several important minerals in basalt magmas. The focus of our proposed two-year project is to determine experimentally the mineral-liquid relationship for magnetite and ilmenite in natural silicate liquids that range in composition from basalt to dacite. Our knowledge of trace element behavior in magmas that crystallize magnetite and ilmenite is largely restricted to a few experiments and to measurements on mechanically separated oxides. The published results of these studies can vary by a factor of 100 for a given element. Our two objectives are to define magnetite and ilmenite-liquid phase equilibria, and to determine the geochemical behavior of several trace elements between magnetite, ilmenite and magmatic liquid. The trace elements to be studied include members of the first transition series (V, Ti, Ni, Sc) and the high field strength elements (Hf, Zr, Nb, Ta, Th, U). We will determine the major and trace element compositions of the magnetite and co-existing glass by electron microprobe. The concentration of the trace elements will be raised above the detection limits of the microprobe by adding pure oxides of those elements to the samples in the O.5-1.O% range. In addition, low concentration dopant (10-200 ppm) experiments will be conducted and analyzed by ion probe. Most experiments will be done at 1- atm. However, to test the effect of pressure, experiments will be conducted at 5, 10, 15, and 25 kb on three of the starting compositions. The range of oxide saturation and oxide composition will be extended by Ti and Fe addition, and by using a range of oxygen fugacities.
