A key issue in models of planetary core formation is the interconnectness and potential percolation of iron-sulfide melts in contact with silicates at high temperature and pressure. To address this issue an integrated study of the electrical conductivity-texture-permeability relationships of olivine-sulfide partial-melt samples was performed. This work has application to the interpretation of high conductivity zones in the Earth as revealed by electromagnetic studies and to the origin and development of the Earth's core. The project consisted of three main tasks. (1) Synthesis and characterization of olivine-sulfide partial-melts. (2) Electrical conductivity measurements of the partial-melt and the individual melt and crystalline phases. (3) X-ray microtomographic determination of the 3-D structure and interconnectedness of the melt phase. The results are used to determine a model of permeability of a partially molten solid that incorporates the melt distribution, a goal that has never before been achieved. Material synthesis was accomplished in the piston cylinder apparatus and electrical conductivity measurements were performed at one atmosphere. X-ray computed tomography was performed on recovered samples at the ALS. This work makes use of and further enhances LLNL's strengths in high-pressure material properties, x-ray micro- and nanoscale imaging and development of transport theory.

This report contains highlights of FY 2001 and 2002 technical studies conducted by the Analytical and Nuclear Chemistry Division (ANCD) at Lawrence Livermore National Laboratory (LLNL) in support of the Hydrologic Resources Management Program (HRMP) and the Underground Test Area (UGTA) Project. These programs are administered by the U.S. Department of Energy, National Nuclear Security Administration, Nevada Site Office (NNSA/NSO) through the Defense Programs and Environmental Restoration Divisions, respectively. HRMP-sponsored work emphasizes the Defense Programs goal of responsible management of natural resources at the NTS, while UGTA-funded work focuses on defining the extent of radionuclide contamination in NTS groundwater resulting from underground nuclear testing. The report is organized on a topical basis, and contains eight chapters that reflect the range of technical work performed by LLNL-ANCD in support of HRMP and UGTA. Chapter 1 describes recent hot well sampling efforts at the NTS, and presents the results of chemical and isotopic analyses of groundwater samples from six near-field wells. These include the Cambric (UE-5n), Bilby (U-3cn PS No. 2), Bourbon (UE-7nS), Nash (UE-2ce), Tybo/Benham (ER-20-5 No. 3), and Almendro (U-19v PS No. 1ds) sites. The data generated by the hot well program is vital to the development and validation of contaminant transport models at the NTS. Chapter 2 discusses the results of xenon isotope measurements of groundwater samples from the six near-field wells described in Chapter 1. This work demonstrates that fission xenon is present in the water at levels that are readily measurable and highlights the significant differences in xenon concentrations and isotopic abundances at different sites. These differences provide insight into the early cooling history of nuclear test cavities, and may assist in predicting the distribution of the source term in the near-field environment. Chapter 3 is an investigation of the distribution and abundance of actinides in a nuclear test cavity and chimney. This work demonstrates that early-time processes can widely disperse actinides at low concentrations outside the melt glass, implying that melt glass dissolution may not be the sole mechanism for the release of actinides to groundwater. The study also provides evidence for the isotopic fractionation of plutonium under the extreme conditions accompanying nuclear explosions. In Chapter 4, X-ray absorption spectroscopy measurements were used to determine the redox state of Fe and U in nuclear melt glass samples from the NTS. Both elements were found to occur in mixed valence states (Fe{sup 2+}/Fe{sup 3+} and U{sup 5+}/U{sup 6+}) in all samples. Comparison of the Fe and U redox states with published redox studies of synthetic glasses suggests that plutonium is predominantly in the Pu{sup 4+} oxidation state in the melt glasses. In Chapter 5, alpha autoradiography is used in a NTS field study to investigate the spatial distribution and transport of actinides in soils, and to help identify the size distribution and morphology of the actinide particles. It was found that {alpha}-emitting radionuclides have moved to at least 39 cm depth in the soil profile, far deeper than expected. The methodology that was developed could easily be applied to other field locations where actinides are dispersed in the soil zone. Chapter 6 summarizes the development of a method for measuring environmental levels of {sup 241}Am on the multi-collector inductively coupled plasma mass spectrometer. The method detection limit of 0.017 pCi/L is about two times lower than the best analyses possible by alpha spectrometry. Chapter 7 describes a chlorine-36 study of vertical groundwater transport processes in Frenchman Flat. Mass balance calculations developed from a {sup 36}Cl mixing model at well ER-5-3 No. 2 are used to estimate vertical transport fluxes and average vertical flow velocities through the thick volcanic section underlying the basin. The study also documents the variations in {sup 36}Cl/Cl ratios within the three principal hydrostratigraphic units in Frenchman Flat. Chapter 8 discusses an ongoing stable isotope investigation of precipitation and recharge processes in central Nevada. Precipitation, spring water, and shallow infiltration samples have been collected at four locations on a biannual basis since 1999. The results show that winter precipitation accounts for>90% of the recharge at these sites. Lysimeter data suggest that most of the evaporation occurring during recharge is due to water vapor loss through the soil zone during periods of slow infiltration. In addition to the topical investigations described above, LLNL-ANCD contributed to several other major collaborative technical products during FY 2001 and 2002.

Pressure-volume relationships were measured at room temperature for eight granular materials and one specimen of epoxy foam. The granular materials included hollow ceramic microspheres, spherical molybdenum powder, Ottawa sand, aluminum, copper, titanium and silicon carbide powders and glassy carbon spheres. Measurements were made to 0.9 GPa in a liquid medium press for all of the granular materials and to 3 GPa in a solid medium press for the ceramic microspheres and molybdenum powder. A single specimen of epoxy foam was compressed to 30 MPa in the liquid medium press. Bulk moduli were calculated as a function of pressure for the ceramic microspheres, the molybdenum powder and three other granular materials. The energy expended in compacting the granular materials was determined by numerically integrating pressure-volume curves. More energy was expended per unit volume in compacting the molybdenum powder to 1 GPa than for the other materials, but compaction of the ceramic microspheres required more energy per gram due to their very low initial density. The merge pressure, the pressure at which all porosity is removed, was estimated for each material by plotting porosity against pressure on a semi-log plot. The pressure-volume curves were then extrapolated to the predicted merge pressures and numerically integrated to estimate the energy required to reach full density for each material. The results suggest that the glassy carbon spheres and the ceramic microspheres would require more energy than the other materials to attain full density.

Pressure-volume relationships were measured for unexpanded Expancel microspheres, epoxy foam and one specimen of crushed foam powder. The specimens were jacketed in tin canisters and compressed at ambient temperature and low strain rates to 3 GPa in a solid medium press. Pressures were corrected for friction, and specimen volumes were calculated relative to a nickel standard. The pressure-volume curves for each material show large volume reductions at pressures below 0.1 GPa. The curves stiffen sharply at or near full density. Relatively little volume reduction is observed above 0.1 GPa, and most is recovered on unloading. The energy expended in compressing the materials to 3 GPa and the energy recovered on unloading were determined by numerically integrating the pressure-volume curves. The net energy, which includes absorbed energy, was found to be small. Compressibilities and bulk moduli were determined from the slopes of the pressure-volume curves. The Expancel bulk modulus above 0.1 GPa was found to be similar to that of isopentane. The pressure-volume data were fit to a model from the ceramics literature (Kawakita and Ludde, 1970). The model fits provided estimates of the initial specimen porosities and room pressure bulk moduli.