Search USGS⌕ Search

Geology topics

D.S. Burnett

Publications and source records attributed to D.S. Burnett.

3 recordsLinked to original sources

Uranium and minor-element partitioning in Fe-Ti oxides and zircon from partially melted granodiorite, Crater Lake, Oregon

Crystal-liquid partitioning in Fe-Ti oxides and zircon was studied in partially melted granodiorite blocks ejected during the climactic eruption of Mt. Mazama (Crater Lake), Oregon. The blocks, which contain up to 33% rhyolite glass (75 wt% SiO 2 ), are interpreted to be portions of the magma chamber walls that were torn off during eruption. The glass is clear and well homogenized for all measured elements except Zr. Results for Fe-Ti oxides give D U oxide / liq ≈ 0.1. Partitioning of Mg, Mn, Al, Si, V, and Cr in Fe-Ti oxides indicates that grains surrounded by glass are moderately well equilibrated with the melt for many of the minor elements, while those that are inclusions in relict plagioclase are not. Uranium and ytterbium inhomogeneities in zircons indicate that the zircons have only partially equilibrated with the melt and that uranium appears to have been diffusing out of the zircons faster than the zircons were dissolving. Minimum U, Y, and P concentrations in zircons give maximum D U zrc / liq = 13, D Y zrc / liq = 23, and D P zrc / liq = 1, but these are considerably lower than reported by other workers for U and Y. Based on our measurements and given their low abundances in most rocks, Fe-Ti oxides probably do not play a major role in U-Th fractionation during partial melting. The partial melts were undersaturated with zircon and apatite, but both phases are present in our samples. This demonstrates an actual case of non-equilibrium source retention of accessory phases, which in general could be an important trace-element fractionation mechanism. Our results do not support the hypothesis that liquid structure is the dominant factor controlling trace-element partitioning in high-silica rhyolites. Rough calculations based on Zr gradients in the glass indicate that the samples could have been partially molten for 800 to 8000 years.

Oregon↗

Actinide abundances in ordinary chondrites

Measurements of 244 Pu fission Xe, U, Th, and light REE (LREE) abundances, along with modal petrographic determinations of phosphate abundances, were carried out on equilibrated ordinary chondrites in order to define better the solar system Pu abundance and to determine the degree of variation of actinide and LREE abundances. Our data permit comparison of the directly measured Pu/ U ratio with that determined indirectly as ( Pu / Nd ) × ( Nd / U ) assuming that Pu behaves chemically as a LREE. Except for Guareña, and perhaps H chondrites in general, Pu concentrations are similar to that determined previously for St. Séverin, although less precise because of higher trapped Xe contents. Trapped 130 Xe 136 Xe "> 130 Xe 136 Xe ratios appear to vary from meteorite to meteorite, but, relative to AVCC, all are similar in the sense of having less of the interstellar heavy Xe found in carbonaceous chondrite acid residues. The Pu/U and Pu/Nd ratios are consistent with previous data for St. Séverin, but both tend to be slightly higher than those inferred from previous data on Angra dos Reis. Although significant variations exist, the distribution of our Th/U ratios, along with other precise isotope dilution data for ordinary chondrites, is rather symmetric about the CI chondrite value; however, actinide/(LREE) ratios are systematically lower than the CI value. Variations in actinide or LREE absolute and relative abundances are interpreted as reflecting differences in the proportions and/or compositions of more primitive components (chondrules and CAI materials?) incorporated into different regions of the ordinary chondrite parent bodies. The observed variations of Th/U, Nd/U, or Ce/U suggest that measurements of Pu/U on any single equilibrated ordinary chondrite specimen, such as St. Séverin, should statistically be within ±20–30% of the average solar system value, although it is also clear that anomalous samples exist.

Geochimica et Cosmochimica Acta↗