Jeffery J. Roberts

Partial‐melt electrical conductivity: Influence of melt composition

The electrical conductivity of a partial melt is influenced by many factors, including melt conductivity, crystalline conductivity, and melt fraction, each of which is influenced by temperature. We have performed measurements of bulk conductivity as a function of temperature of an Fo80-basalt partial melt between 684° and 1244°C at controlled oxygen fugacity. Melt fraction and composition variations with temperature calculated using MELTS [Ghiorso and Sack, 1995] indicate that the effect on melt conductivity of changing melt composition is balanced by changes in temperature (T). Thus bulk conductivity as a function of T or melt fraction in this system can be calculated assuming a constant melt conductivity. The bulk conductivity is well modeled by simple parallel calculations, by the Hashin-Shtrikman upper bound, or by Archies law (σpartial melt/σmelt = C1Xmn). We estimate apparent values of the Archies law parameters between 1150° and 1244°C as C1 = 0.73 ± 0.02 and n = 0.98 ±0.01. Estimates of the permeability of the system are obtained by using an electrical conductivity-critical scale length relationship and range from ∼10−14 to 10−18 m2, comparing favorably with previously published values.

Electrical properties of partially saturated Topopah Spring Tuff: Water distribution as a function of saturation

We have measured the dielectric constant and electrical resistivity of Topopah Spring tuff as a function of saturation at 23° and 40°C. Dry and saturated resistivities can differ by more than 4 orders of magnitude. Frequency-dependent impedance measurements indicate that the overall electrical response of the rock arises from at least three conduction mechanisms, which we tentatively identify as conduction through adsorbed water on solid surfaces, conduction through isolated pockets of water, and conduction through continuous (bulk) water. The relative contribution of each mechanism varies with saturation. A continuous (bulk) water film is indicated at saturations as low as 35%. Reasonable agreement was found between predicted permeability based on electrical measurements and microstructural parameters using the Kozeny-Carmen relation and experimentally determined permeabilities. Similarly, good agreement was found between estimates of cation exchange capacity (CEC) based on electrical properties through the Waxman and Smits [1968] equation and published CEC values for tuff.

Anisotropy of electrical conductivity in dry olivine

[1] The electrical conductivity ({sigma}) was measured for a single crystal of San Carlos olivine (Fo{sub 89.1}) for all three principal orientations over oxygen fugacities 10{sup -7} < fO{sub 2} < 10{sup 1} Pa at 1100, 1200, and 1300 C. Fe-doped Pt electrodes were used in conjunction with a conservative range of fO{sub 2}, T, and time to reduce Fe loss resulting in data that is {approx}0.15 log units higher in conductivity than previous studies. At 1200 C and fO{sub 2} = 10{sup -1} Pa, {sigma}{sub [100]} = 10{sup -2.27} S/m, {sigma}{sub [010]} = 10{sup -2.49} S/m, {sigma}{sub [001]} = 10{sup -2.40} S/m. The dependences of {sigma} on T and fO{sub 2} have been simultaneously modeled with undifferentiated mixed conduction of small polarons and Mg vacancies to obtain steady-state fO{sub 2}-independent activation energies: Ea{sub [100]} = 0.32 eV, Ea{sub [010]} = 0.56 eV, Ea{sub [001]} = 0.71 eV. A single crystal of dry olivine would provide a maximum of {approx}10{sup 0.4} S/m azimuthal {sigma} contrast for T < 1500 C. The anisotropic results are combined to create an isotropic model with Ea = 0.53 eV.

On the relationship between microstructure and electrical and hydraulic properties of sand‐clay mixtures

A series of laboratory experiments, including measurements of electrical properties, permeability, and porosity, were performed on saturated sand-clay mixtures. Different mixtures and packing geometries of quartz sand and 0 to 10% Na-montmorillonite clay were investigated using solutions of CaCl2 and deionized water. Two main regions of electrical conduction exist: a region dominated by surface conduction and a region where the ionic strength of the saturating fluid controlled conduction. For low fluid conductivities, the sample geometry was found to greatly affect the magnitude of the surface conductance. The influence of the microstructural properties on the electrical properties was quantified by estimating formation factors, Λ-parameters, and surface conductances. The surface conductances estimated using the theory of Johnson et al. [1986] agreed well with measured values. We suggest that high and low bounds on the expected surface and bulk conductances in a natural system can be derived from the measurements on these artificial geometries.

Carbon‐enhanced electrical conductivity during fracture of rocks

Changes in electrical resistance during rock fracture in the presence of a carbonaceous atmosphere have been investigated using Nugget sandstone and Westerly granite. The experiments were performed in an internally heated, gas-pressure vessel with a load train that produced strain rates between 10−6 and 10−5 s−1. Samples were deformed at temperatures of 354° to 502°C and pressures of 100 to 170 MPa in atmospheres of Ar or mixtures of 95% CO2 with 5% CO or 5% CH4, compositions that are well within the field of graphite stability at the run conditions. In experiments using Nugget sandstone, resistance reached a minimum value when the maximum temperature was achieved and good electrode contact was made. The resistance then increased as the experiment continued, probably due to dry out of the sample, a change in the oxidation state of the Fe-oxide associated with the cement, or destruction of current-bearing pathways. At approximately 200-MPa end load, the rock sample failed. Plots of load and resistance versus time show several interesting features. In one experiment, for example, as the end load reached about 175 MPa, resistance stopped increasing and remained fairly constant for a period of approximately 0.5 hour. During loading, the end load displayed small decreases that were simultaneous with small decreases in resistance; when the end load (and the displacement) indicated rock failure, resistance decreased dramatically, from ∼150 MΩ to 100 MΩ. In a single experiment, the Westerly granite also showed a decrease in resistance during dilatancy. The nature and distribution of carbon in the run products were studied by electron microprobe and time-of-flight secondary-ion mass spectroscopy (TOP-SIMS). Carbon observed by mapping with the former is clearly observed on micro-cracks that, based on the microtexture, are interpreted to have formed during the deformation. The TOF-SIMS data confirm the electron-probe observations that carbon is present on fracture surfaces. These observations and experimental results lead to the hypothesis that as microfractures open in the time leading up to failure along a fracture, carbon is deposited as a continuous film on the new, reactive mineral surfaces, and this produces a decrease in resistance. Subsequent changes in resistance occur as connectivity of the initial fracture network is altered by continued deformation. Such a process may explain some electromagnetic effects associated with earthquakes.

FREQUENCY DEPENDENT ELECTRICAL PROPERTIES OF MINERALS AND PARTIAL-MELTS

The resistance to current flow of minerals and partial-melts is a frequency dependent electrical property. Measurements of the frequency dependent electrical impedance of single crystal olivine, polycrystalline olivine, dunites, metapelites, and partial-melts, between 10−4 and 105 Hz, when plotted in the complex impedance plane, reveal arcs that correspond to different conduction mechanisms in the material being studied. In polycrystalline materials, two impedance arcs related to material properties (as opposed to electrode properties or electrode-sample interactions) are observed. Each impedance arc is activated over a distinct range of frequency, that is, the mechanisms occur in series. Based on experiments comparing single and polycrystalline impedance spectra, experiments on samples with different electrode configurations, and on samples of varying dimension, the mechanisms responsible for these impedance arcs are interpreted as grain interior conduction (σgi), grain boundary conduction (in polycrystalline materials;σgb), and sample-electrode interface effects, from highest to lowest frequency, respectively. Impedance spectra of natural dunitic rocks reveal analogous behavior, that is,σgb andσgi add in series. The grain boundaries do not enhance the conductivity of any of the materials studied (a direct result of the observed series electrical behavior) and, under certain conditions, limit the total conductivity of the grain interior-grain boundary system. By examining the frequency dependence of the electrical properties of partial-melts, it is possible to gain information about microstructure and the distribution of the melt phase and to determine the conditions under which the presence of melt enhances the total conductivity. Impedance spectra of olivine-basalt partial-melts indicate that at least two conduction mechanisms occur in series over the frequency range 10−4-105 Hz, similar to the observed electrical response of melt-absent polycrystalline materials. In a sample containing isolated melt pockets the intermediate frequency grain boundary impedance arc is modified by the presence of melt indicating series conduction behavior. In a sample with an interconnected melt phase the high frequency grain interior impedance arc is modified by the melt phase, indicating the initiation of parallel conduction behavior. Because field EM response versus frequency curves are used to derive conductivity versus depth profiles, it is important to perform laboratory experiments to understand the frequency-dependent electrical behavior of Earth materials. Activation energies determined from studies that measure conductivity at a single frequency may be erroneous because of the shift of the dominant conduction mechanism with frequency as temperature is varied.

Characterizing unsaturated diffusion in porous tuff gravel

latest developments in instrumentation and techniques. Improved understanding of unsaturated diffusion in Evaluation of solute diffusion in unsaturated porous gravel is very gravel will help in the characterization and remediation important for investigations of contaminant transport and remediation, risk assessment, and waste disposal (e.g., the potential high-level effort in gravel deposits at the Hanford Reservation nuclear waste repository at Yucca Mountain, Nevada). For a porous (Washington). It will also help in the invert diffusion aggregate medium such as granular tuff, the total water content is barrier concept for the potential underground high-level comprised of surface water and interior water. The surface water radioactive waste repository at Yucca Mountain, Necomponent (water film around grains and pendular water between vada, where tuff gravel has been considered as an invert the grain contacts) could serve as a predominant diffusion pathway. material (material filling the bottom of a tunnel having a To investigate the extent to which surface water films and contact circular cross-section) to contain radionuclide transport. points affect solute diffusion in unsaturated gravel, we examined the The invert placed between the waste package or drip configuration of water using X-ray computed tomography (CT) in shield and the tuff host rock at Yucca Mountain is an partially saturated gravel and made quantitative measurements of difintegral component of the repository’s performance. If fusion at multiple water contents using two different techniques. In the first, diffusion coefficients of KCl in 2to 4-mm granular tuff at effective, an invert diffusion barrier (caused by slow multiple water contents were calculated from electrical conductivity radionuclide diffusion through the invert) can greatly (EC) measurements using the Nernst–Einstein equation. In the secenhance waste-isolation capacity. Conca and Wright ond, we used laser ablation with inductively coupled plasma–mass (1992) measured effective diffusion coefficients (De) in spectrometry (LA/ICP-MS) to perform microscale mapping, allowing unsaturated soil, gravel, bentonite, and whole rock for the measurement of diffusion coefficients for a mixture of chemical a wide range of volumetric water contents (this free tracers for tuff cubes and tetrahedrons having two contact geometries water content does not include interlayer water in clays (cube–cube and cube–tetrahedron). The X-ray computed tomography or other structural water; Conca and Wright, 2000). images show limited contact between grains, and this could hinder They found that De values in all media were primarily the pathways for diffusive transport. Experimental results show the a function of volumetric water content and not material critical role of surface water in controlling transport pathways and hence the magnitude of diffusion. Even with a bulk volumetric water characteristics. CRWMS M&O (2000b) reported that content of 1.5%, the measured solute diffusion coefficient is as low this diffusion data set was well correlated in terms of a as 1.5 10 14 m2 s 1 for tuff gravel. Currently used diffusion models power-dependence (Archie’s Law type) on the volumetrelating diffusion coefficients to total volumetric water content inaderic water content, and in this study a resultant “univerquately describe unsaturated diffusion behavior in porous gravel at sal” power function was used to represent diffusive very low water contents. transport of radionuclides through the invert. However, in waste emplacement drifts, characterized by a humid environment with or without the presence of liquid S of flow and transport in gravels have rewater, crushed porous rock may provide unique characcently received attention because of the importance teristics that vary greatly from this generic power funcof gravel aquifers, the need to understand contamination (Wang et al., 2001; Hu and Wang, 2003). For examtion characterization and remediation of gravel deposits ple, Conca (1990) placed four different size fractions of in the vadose zone, and the use of gravel as capillary bartuff gravel samples (2–4, 4–6.3, 6.3–9.5, and 15–25.4 mm) riers for waste isolation. As stated in Tokunaga et al. for equilibrium inside a chamber with a nearly 100% (2003), relatively little information is available on the humidity atmosphere. After equilibrating about 70 d, unsaturated hydraulic properties of gravels; this is also all samples of individual grains were observed to be dry, true for transport processes in unsaturated gravel sysdespite the 2.7% intragranular water content. No EC tems. Conca and coworkers published pioneering work could be measured on these samples, resulting in an examining chemical diffusion behavior in porous gravinferred diffusion coefficient below 10 15 m2 s 1, which els, but this work was conducted more than 10 yr ago is the detection limit reported by Conca (1990) using (Conca, 1990; Conca and Wright, 1990, 1992). With the EC for estimating the diffusion coefficient. The inferred improved understanding of water distribution in gravel, low diffusion value of 10 15 m2 s 1 at this water content there is a strong need to investigate the diffusion prodeviates significantly from the “universal” power funccesses in unsaturated porous gravel by employing the tion with a diffusion coefficient (≈2.8 10 12 m2 s 1), which is obtained with continuous fluid introduction. In other words, at the same water content, diffusion in samQ. Hu and J.J. Roberts, 7000 East Ave., MS L-231, Lawrence Livermore National Laboratory, Livermore, CA 94550; T.J. Kneafsey, L. ples prepared using high humidity (without fluid source Tomutsa, and J.S.Y. Wang, 1 Cyclotron Road, MS 90-1116, Lawrence Berkeley National Laboratory, Berkeley, CA 94720. Received 12 Nov. Abbreviations: CT, computed tomography; EC, electrical conductiv2003. Original Research Paper. *Corresponding author (hu7@llnl.gov). ity; ICP-MS, inductively coupled plasma-mass spectrometry; LA/ICPMS, laser ablation with inductively coupled plasma-mass spectromePublished in Vadose Zone Journal 3:1425–1438 (2004).

Electrical properties of microporous rock as a function of saturation and temperature

We measured the electrical resistivity of densely welded tuff (Topopah Spring tuff) as a function of saturation by water between 23 and 145 °C. Measurements at 23–95 °C were performed at ambient pressure and measurements up to 145 °C were performed at pressures up to 5.5 MPa in an externally heated pressure vessel. Pore and confining pressures were controlled independently, allowing electrical measurements as the sample was subjected to boiling conditions. Resistivity changes associated with liquid water flashing to steam were observed. Incremental increases in resistivity with decreasing pore pressure are attributed to vapor pressure lowering caused by capillary tension in the nanometer-size pores. This result has important implications for models of reservoir engineering, geologic nuclear-waste disposal, and geophysical sensing techniques.

In situ high-pressure and high-temperature X-ray microtomographic imaging during large deformation: A new technique for studying mechanical behavior of multiphase composites

We have examined the microstructural evolution of a two-phase composite (olivine + Fe-Ni-S) during large shear deformation, using a newly developed high-pressure X-ray tomography microscope. Two samples were examined: a load-bearing framework–type texture, where the alloy phase (Fe-Ni-S) was present as isolated spherical inclusions, and an interconnected network–type texture, where the alloy phase was concentrated along the silicate grain boundaries and tended to form an interconnected network. The samples, both containing ∼10 vol% alloy inclusions, were compressed to 6 GPa, followed by shear deformation at temperatures up to 800 K. Shear strains were introduced by twisting the samples at high pressure and high temperature. At each imposed shear strain, samples were cooled to ambient temperature and tomographic images collected. The three-dimensional tomographic images were analyzed for textural evolution. We found that in both samples, Fe-Ni-S, which is the weaker phase in the composite, underwent significant deformation. The resulting lens-shaped alloy phase is subparallel to the shear plane and has a laminated, highly anisotropic interconnected weak layer texture. Scanning electron microscopy showed that many alloy inclusions became film-like, with thicknesses <1 μm, suggesting that Fe-Ni-S was highly mobile under nonhydrostatic stress, migrated into silicate grain boundaries, and propagated in a manner similar to melt inclusions in a deforming solid matrix. The grain size of the silicate matrix was significantly reduced under large strain deformation. The strong shape-preferred orientation thus developed can profoundly influence a composites bulk elastic and rheological properties. High-pressure–high temperature tomography not only provides quantitative observations on textural evolution, but also can be compared with simulation results to derive more rigorous models of the mechanical properties of composite materials relevant to Earths deep mantle.

Transient electrical response of San Quintin Dunite as a function of oxygen fugacity changes: Information about charge carriers

The electrical conductivity (σ) of San Quintin dunite (SQD) measured between 950 and 1150°C at controlled oxygen fugacity (fO2) within the olivine stability field shows transients in response to changes in fO2. Such behavior has not been reported previously for measurements made under similar conditions on either olivine single crystals or polycrystalline samples (dunites and lherzolites) in which olivine is the major phase. In general for olivine, an increase in fO2 results in an increase in σ. The transient is manifested as a change in σ from a stable equilibrium value at a specific fO2 to a quickly established subsequent value, the direction of which is opposite that of the final value that will be attained for the fO2 change. This transient may be caused by a changing population of electrons produced by oxygen vacancies, the grain boundaries, or a short-lived defect. We postulate that the transient is observed in this particular dunite because of its large surface area to volume ratio. The transient is more pronounced at relatively high fO23 and temperatures between 950 and 1100°C. At relatively low fO2s and temperatures higher than 1100°C, the effect is diminished. The Seebeck coefficient (S), at 1200°C, is slightly smaller than that reported for single-crystal olivine, and at 1100°C is similar to that reported for single crystal olivine.