Experimental Actinide Element Partitioning Between Whitlockite, Apatite, Diopsidic Clinopyroxene, and Anhydrous Melt at One Bar and 20 Kilobars Pressure

Citation

Benjamin, Timothy Miller (1980) Experimental Actinide Element Partitioning Between Whitlockite, Apatite, Diopsidic Clinopyroxene, and Anhydrous Melt at One Bar and 20 Kilobars Pressure. Dissertation (Ph.D.), California Institute of Technology. https://resolver.caltech.edu/CaltechTHESIS:03202024-175813071

Abstract

Fission and alpha track radiography techniques have been used to measure partition coefficients (D) at trace (ppm) concentration levels for the actinide elements Th, U, and Pu between synthetic diopsidic clinopyroxene, whitlockite (β-Ca₃(PO₄)₂), apatite, and coexisting "haplobasaltic" silicate liquid at 1 bar and 20 kilobars pressure at oxygen fugacities from 10^(-8.6) to 10^(-0.7) bars. Equilibrium (Rayleigh) partitioning at the crystal-liquid interface is assumed and corrections for actinide zoning and relative alpha and fission fragment ranges have been applied to the measured D values. Reproducibility for both actinide and minor element D values is care­ fully examined as a criterion for crystal-liquid interface equilibrium. The data are mostly compatible with interface equilibrium except at high cooling rates (≥30 deg/hr).

Pu is much more readily incorporated into crystalline phases than is U or Th, under reducing conditions (fO₂=10^(=8.6), because Pu is most likely trivalent whereas U and Th are tetravalent. The effect of changing pressure and liquidus temperature is small where direct comparisons can be made. Definitive valence state assignments cannot be made, but our best estimates of corrected partition coefficients for Pu⁺³, Pu⁺⁴, Th⁺⁴, U⁺⁴, and U⁺⁶ are, for whitlockite: 3.6 / 0.58 / 1.2 / 0.5 / 0.002; for chloro-oxy-apatites: -- / -- / 1.22 / 1.69 / --; and for diopsidic clinopyroxene: 0.06 / -- / 0.002 /0.002 / -- respectively in P-bearing systems.

Strong compositional effects are seen on the addition of P and, to a lesser extent, U. The D_(cpx)for Th and U are a factor of 10, and for Pu⁺³ a factor of three higher in P-free systems relative to P-bearing systems. This is interpreted as a result of actinide stabilization in the melt, possibly by complex formation with PO⁻³₄ groups.

Actinide substitution in these phases is most likely into the Ca site as this site is the largest in each phase and the magnitude of the relative D values, between phases, corresponds to the relative range of Ca site sizes. However experiments run with percent concentrations of UO₂ added to the whitlockite-producing compositions suggest that substitution of UO⁻⁴₄ for PO⁻³₄ is a possibility.

Preliminary results on Sm partitioning at 1 bar pressure and published lanthanide partition coefficients place the partitioning behavior of Pu⁺³ amongst the light rare earth elements with Nd having a factor of two greater D_(whit) than Pu⁺³. These results support the use of Pu/Nd chronology for meteorites provided chemical fractionation effects, which could result in a 60 m.y. error, are assessed.

These results have application in evaluating actinide-actinide and actinide-lanthanide fractionations in natural materials utilized in addressing problems in geochronology, cosmochronology, and petrogenesis. Additionally, these data define a 'bracketing theorem' for selection of unfractionated materials for use in the aforementioned disciplines. This bracketing theorem, based on the sequence of relative crystal/liquid partition coefficients: D_(Nd)>D_(Pu)>D_(Th)≥;D_U, predicts that a sample with unfractionated (relative to solar-system abundances) lanthanide and Th-U abundances will also have a solar-system Pu abundance. Further use of this type of experimentation has clear application to the current problem of radioactive waste disposal.

Thesis Files