Zirconium-Oxide ceramics stabilized with -8-wt% Yttrium-Oxide can be employed to sense pH in high temperature (>90°C) aqueous environments with an accuracy of ±0.05 pH log units (Lvov et. al., in press), and to sense the fugacity of oxygen (fO2) in low temperature (>230°C) gaseous environments with an accuracy of ±0.2 (fO2) log units. The major components, m the two commercially available yttria-stabilized ceramics are yttria (-8-wt%) and zirconia (-91-wt%) with minor amounts of Ti, Fe and U. The textural differences in the two ceramics produce significantly different emf vs. 10,000/T responses. Response error can be introduced by: the ionic contribution of the softening glass that may be present, the catalytic action of the Pt sensor components, and the presence of Ti and Fe in the ceramic, which has been shown (Merino et. al., 1996) to alter the oxygen diffusivity of the ceramic. The first type of ceramic contains a 3-dimensionally-continuous Ca-Al-Si feldspathic glass that acts as a sintering aid during manufacturing. The glass, which has a higher ionic conductivity than the zirconia ceramic, reduces the bulk resistivity and induces an error over the temperature ranges higher than the softening point of the glass. The glass also reduces durability of the ceramic. When the glass hydrates it produces zeolites, which grow primarily in the triple-grain-junctions of the ceramic, thus mechanically weakening the ceramic and generating electronic, ionic and mechanical stability problems. The second type of ceramic contains no grain boundary glass, but does contain discrete silicate and oxide phases (such as diopside, wollastonite, periclase, silica, etc.) in the triple-grain-junctions. Because there is no inter-granular glass, the type-two ceramic does have a greater bulk resistivity compared with the type one ceramic. In a gas-sensing configuration, resistivity will affect the minimum temperature of sensor operation. A sensor with a higher bulk resistivity must reach a higher minimum temperature before the sensor will sense oxygen. The work presented in this manuscript suggests that the same is true for the sensor in its aqueous pH configuration. In addition to the mechanical degradation, there are also chemical leaching issues with both ceramics. While zirconia is relatively unleachable in its pure form, the addition of yttria, while creating the necessary lattice defects, increases the vulnerability of the solid solution grains to acidic solutions. This would possibly create ceramic durability problems during long-term down-hole operation. The ceramics do function well as a sensor and can produce highly accurate results (with calibration) and if the durability issues are considered during sensor assembly, the ceramic sensor could be highly desirable for many high temperature geologic and industrial applications. (Manna et al., 2002 and Manna et al., 2001). / Earth and Environmental Science / Accompanied by one .pdf file: 1) Manna-Supplemental-2002.pdf

A 1993, airborne geophysical survey found numerous localized "bull's­-eye" magnetic anomalies, apparently of natural origin, associated with dolines on the Oak Ridge Reservation (ORR), Oak Ridge, Tennessee. Susceptibility measurements made on a soil core extracted from one such doline located on Copper Ridge (Anomaly B) showed the soil to be magnetic throughout the soil profile (average susceptibility >200 x 10^-5 S.I. units), while a soil core removed from outside the doline showed an average susceptibility of less than 20 x 10^-5 S.I. units. Pedologic characterization of the soil cores both inside and outside of the doline showed a series of stacked soil profiles, all in the early stage of pedogenesis, with A-horizons lying directly on C-horizons that were transforming to Bw-horizons, indicating periodic burial by fresh colluvium or alluvium. Based on mottling seen throughout the soil cores, which is evidence of hydromorphy, I identified of the soils as aquepts using the USDA soil classification. In the majority of the soil profiles the magnetic susceptibility peaked at the base of the A-horizons. I investigated nine different mechanisms discussed in the literature for the formation of magnetic soil. I used scanning electron microscopy (SEM), energy dispersive X-ray (EDX), and X-ray diffractometry (XRD), to supplement pedologic characterization. Based on the XRD results, the source of the soil magnetism was identified as the ferromagnetic mineral, maghemite. SEM measurements showed that the magnetic particles extracted from the soil cores were too large and rounded to be biogenic maghemite. Apparently the maghemite formed by anaerobic microbial iron reduction followed by the formation of maghemite, or by abiological weathering and reduction of an iron-bearing mineral followed by oxidation. Maghemite occurs in the soil both inside and outside of the magnetic anomaly, but was found in higher concentrations inside the anomaly. EDX analysis showed that the soil core extracted from inside the magnetic anomaly contained more iron than the core extracted outside the anomaly, suggesting a different parent material for the soil inside the doline. ORR reports and soil maps indicate that the parent material is an ancient alluvium. Ancient alluviums found on the ORR have more iron as a result of the weathering of iron rich sediment deposited by an ancient river. Through topographic inversion these ancient alluviums are now found on top of Copper Ridge. I found evidence to support an alluvial source for the soil inside the doline, including rounded metaquartzite and illite (weathered biotite). I conclude that soils derived from ancient alluviums have accumulated in dolines on the ORR and that hydromorphy within the dolines has increased the maghemite concentration. The magnetic anomalies detected by the airborne geophysical survey resulted from a combination of increased soil thickness in the dolines, and the higher maghemite concentration that developed in these iron rich soils. / Earth and Environmental Science / Accompanied by one .pdf file: 1) Rivers-Supplemental-2002.pdf

Marked facies changes occur at and just above 3rd order boundaries in Lower Cretaceous carbonate platform deposits of the French Jura and the Dorset Coast. In contrast, the boundaries of cycles just below 3rd order boundaries show smaller facies changes and the facies are very restricted. Two studied intervals (lower and upper) in these carbonate platform deposits are near the lower and upper boundaries of a 3rd order sequence (Anderson, 2000; 2001a & b). These intervals are described in terms of facies and their cyclic stacking pattern and are correlated within each region. The studied intervals are well exposed at four localities in the French Jura: Saleve, Chapeau de Gendarme, Crozet, and Yenne, and two on the Dorset coast: Durlston Bay and Stair Hole. Correlation involves using unconformities interpreted from sequence stratigraphy techniques, biostratigraphy from previous literature, distinct marker beds such as paleosols or freshwater beds, and the interpretation and matching of the hierarchic cyclic structure at each locality. The interpretation of sequences and the stacking pattern at each locality is based on the Milankovitch orbital forcing model. The degree of facies change within a sequence and at sequence boundaries is a function of the magnitude of the sea-­level rise. Cycles and sequences usually are characterized by shallowing upward facies with more massive limestone, usually calcarenite, beds near the base and marl, shale, or paleosols toward the top. This pattern of facies change is seen throughout each locality and in cycles of every scale in the Milankovitch hierarchy. In each region cycles near the basin margin are missing or highly truncated, whereas in the most complete sections (type sections) located more basinward a more complete Milankovitch stacking pattern is preserved. Paleosols tended to form near major cycle boundaries and as a consequences are associated with formational and member boundaries. Paleosols are classified using USDA (1998) and Mack et al. (1993) standards by examining macro- and microanalysis, molecular weathering ratios determined from ICP-MS data, and clay mineralogy from X­ray diffraction. Four paleosols were sampled. Two of these paleosols are from the study interval in the French Jura and two are sampled from the study interval in the Dorset coast. The analysis of the paleosols in England (samples D3-5 and Dl-6) indicate moderate development and are classified as Inceptisols. Whereas the paleosols in the French Jura (samples R-2 and S-19) are poorly developed and are classified as Entisols. / Earth and Environmental Science / Accompanied by one .pdf file: 1) Seier-Supplemental-2002.pdf

Six million hectares at about 1500 sites in the U.S. are contaminated with unexploded ordnance (UXO) with the cost of cleanup estimated at $500 billion. The development and implementation of ultra-low (<10 m) altitude airborne methods to detect UXO has the potential to cut this cost by half. Unfortunately, magnetic data acquired at these low altitudes are severely contaminated with flight-height related noise which arises because of the strong dependence of the anomaly strength, T, on the detection distance, r, Toc1 / r^3. Survey altitudes commonly fluctuate by more than 25% even in ideal conditions. Data acquired above the mean flight height are attenuated with respect to the data acquired at the mean height, while data acquired below the mean height are magnified. This causes large corrugations (easily 100% of the background magnitude) in the data, inducing error in the calculation of target depth and position, and creation of false positives. The noise may be removed by migration of the survey measurements to a common altitude using upward and downward continuation. Three published algorithms (Taylor-series, Chessboard, and equivalent source) were tested on four synthetic data sets and on data acquired at a UXO test site on the Badlands Bombing Range in South Dakota. Corrections were applied to both the profile and the gridded data. The results indicate that the grid-based equivalent source method, combined with a preferential low-pass filter to stabilize downward continuation, is the most accurate and versatile of the methods. Application to field data showed an eightfold increase in the signal-to-noise ratio, which was characteristic of the synthetic data as well. These results are promising for the future of ultra-low magnetic surveys to detect UXO. / Earth and Environmental Science / Accompanied by one .pdf file: 1) Weikel-Supplemental-2003.pdf

Dissolution kinetics and ion-release rates were measured for a halogenated, trioctahedral biotite isolated from stream alluvium down-gradient from an acidic metal­enriched plume in Pinal Creek Basin, Arizona. Experiments were performed using flow-­through columns in which unground 150-250 µm biotite and bulk alluvial sediment from Pinal Creek basin were reacted with a solution of dilute sulfuric acid (pH 3.4 and 2.9) for five months under flowing argon gas at 25°C. The dissolution rate of the halogenated biotite was 2.16 x 10^-13 mol m^-2 s^-1 at pH 3. After 1400 hours, mineral dissolution was constant and congruent. The order of the biotite dissolution reaction with respect to hydrogen ion activity (n) was 0.52 for output pH values between 3.1 and 3.5. Dissolution rates for this study were slightly slower than rates previously measured for biotite and phlogopite in the same pH ranges, possibly due to sample composition, preparation and/or reactor design. This study is one of the few where Cl release rates were measured and averaged 0.08 picomol m^-2 s^-1. Inverse geochemical modeling of ground-water evolution between wells in Pinal Creek basin by Glynn and Brown (1996) suggested that an additional chloride source was necessary in some model simulations to achieve chloride mass balance. A model calculation of Cl input from biotite dissolution along the flow path modeled by Glynn and Brown (1996) using the biotite dissolution rate and the rate of Cl released determined in this study determined that Cl contribution from the dissolution of halogenated biotite can be as much as 10 mg/L. / Earth and Environmental Science

The goal of this thesis was to evaluate the relationship between the measurement of potentials using SP electrodes and redox zones in a series of laboratory experiments. The experiments tested three hypotheses from the literature to explain field detection of redox zones around ore bodies and contaminant plumes. The first hypothesis is that SP electrodes detect current from ion flow driven by redox reactions. The second hypothesis is that SP electrodes detect Eh differences directly. The third hypothesis is that the electric field detected by SP is created by charge separation across the boundary between two different redox zones. Experiments were performed in beakers, glass vessels with flitted plate separating two solutions, and in a sand tank with Pt-Pt, SP-Pt and SP-SP electrode pairs. Solutions with various Eh values were used in the three set ups in different combinations to create contrasting Eh. Small-scale beaker experiments were used to design the sand tank experiments. Results from the two solutions separated by flitted glass tested the hypotheses that electrodes could not measure Eh differences between solutions. Control tank measurements were used as a reference for a tank with a buried plume. The tank with the buried plume tested the third hypothesis. An experiment was done in a tank containing sand saturated with salt solution and battery-induced voltage of 6000 mV to test the first hypothesis. The beaker experiment showed that SP electrodes do not respond to the Eh of the surrounding solution. Voltages between electrode pairs in a single solution should be zero because there is no Eh contrast. However, when a combination of SP-Pt electrodes was used, voltages in the hundreds of m V were measured. Thus, Pt and SP responded differently to the single solution. Results in the fritted glass vessel with two solutions of contrasting Eh confirmed that the SP electrode does not respond to the Eh of the surrounding solution. When Pt and SP electrodes were placed in KMnO4/FeCl2 and KMnO4/H2O2 solution pairs respectively, the recorded voltages were around four hundred m Vs. Similar values were obtained in two experiments with the SP electrode placed in FeCl2 first and H2O2 afterwards. Thus, the SP-Pt pair measured the Eh of the solution where the Pt electrode was inserted but not an Eh difference. Both the Eh and SP-Pt electrode pairs were able to read potential variations in the tank with a buried plume, but the SP-SP pair did not. In the tank with a buried plume, the Eh was 20-40 mV higher near the plume, and SP-Pt read 25-35 mV higher near the plume. SP-SP readings were 5-7 mV throughout this tank showing no difference near the plume. Voltages read by SP-SP in the tank with a battery varied from 1100 mV to 82 mV. The voltage readings decreased inversely proportional to the distances of the wires from the SP electrodes, and confirmed that the SP electrodes could detect current in the sand tank. In all experiments, SP-SP measured neither the Eh of independent solutions nor Eh differences between solutions, as proposed in the literature. The lack of response from the SP electrode may be due to Eh buffering of the copper wire suspended in the copper sulfate solution within the SP electrode. A new explanation must be found for the relationship between the SP and redox zones. / Earth and Environmental Science

Shells materials of mollusks, Mercenaria mercenria, Haliotis fulgens, Haliotis corrugate, and Perna canaliculus were employed in this thesis to develop feasible techniques to detect the causes of color in costly pearls. Samples of shell materials were prepared in the form of powders, extractions, and fragments for different analytical methods. The results of Infrared (KBr disc), X-ray Diffractometer (XRD) and Scanning Electron Microscopy (SEM) with Energy Dispersive X-Ray Spectroscopic (EDS) analyses indicate that the major chemical composition of nacre of shells is CaCO3, mainly in the form of aragonite. The XRD results also indicate that there are calcite and dolomite present in the nacre layer of M. mercenaria shell. The EDS and Inductively Coupled Plasma-Mass Spectrometry (ICPMS) results indicate that the purple nacre layer of M. mercenaria contains lower concentration of calcium carbonate than the white layer. The Infrared and UV-Visible spectra of extracted solutions of purple and white M. mercenaria indicate that the compositions of extractions are similar except some varieties of amide molecules. But no specific molecular structures related to common organic pigments can be inferred from the spectra. The results of ICPMS and EDS analyses show that concentrations of transition metal and rare earth trace elements in the purple and white M. mercenaria have no significant differences within the limits of detection (ppm). The SEM images (magnifications up to 50000X) of the ridge and valley areas of M. mercenaria shell indicate that different sizes of grains make up the purple and white layers of M. mercenaria, and the sizes of particles in purple layers are two times larger than the sizes of particles in white layers (about 2.5 µm). The heat testing (L.O.I.,~2.5 Wt%) and the different shrinkage and expanding phenomena of the sample fabric of purple and white M. mercenaria under the various water vapor pressure in environmental SEM support the idea that the purple nacre layers in the M. mercenaria have 0.58 Wt% more organic content than the white ones. The different colors in the shells are not caused by the different concentrations of certain metal elements or the organic pigments. Optical effects caused by the structure may be the cause for the different colors in shells, but whether the different optical effects are caused by the crystal sizes of calcite and aragonite or the different content of organic proteins in the shells has not been established. / Earth and Environmental Science

Aquifer pumping tests conducted by the U.S. Geological Survey (USGS) at the John Evans Site (North Penn 6 Superfund Site, Lansdale, PA) produced anomalous water level changes in observations wells. In some observations wells, changes were not radially symmetric, and more distant wells had greater drawdown than proximal wells. Triassic sandstones and shales at the site are tilted, and contain a large number of fractures, which may affect groundwater movement. Therefore, the computer code FRAC3DVS, which allows simulation of fractures, was used to develop a conceptual model of these anomalies, which was generally consistent with hydrogeologic properties. Both steady state and transient models were constructed. Hydrologic properties and the number, location, orientation, and aperture of fractures were varied in order to simulate the results of the aquifer pumping test. The steady state modeling showed that the range of influence of a few connected fractures created deviations in the flow field on the scale of 10s of meters, similar to the site scale. Thus a simplified network of fractures was used to model the flow anomalies. Five transient simulations from the site-specific model are presented to demonstrate the calibration process and goodness of fit. The best fit, having a root mean square error of 0.07 m, was obtained using a model with three continuous dipping fracture sets with an aperture of 1.0E-4 m and a matrix hydraulic conductivity of 1.2E-9 m/s. The transient modeling showed that anomalies were produced when tilted fractures that only intersect some wells are used and when the contrast between the fracture and matrix hydraulic conductivity is high. / Earth and Environmental Science / Accompanied by one .pdf file: 1) Elsea-Supplemental-2006.pdf

In this study, quantitative modeling techniques were used to examine groundwater flow and saltwater intrusion in Cape May County, New Jersey. Increased pumping of freshwater from coastal aquifers following urbanization has triggered invasion of seawater and forced many municipalities to abandon and seal formerly productive supply wells. Previous studies predicted a sharp interface between saltwater and freshwater based on simple advective models. Such models do not account for hydrodynamic dispersion and influences of geologic heterogeneity on solute transport, nor do they account for buoyancy-driven flow. To simulate saltwater intrusion in Cape May, a numerical model was used that couples equations of flow and solute transport by advection and hydrodynamic dispersion, accounts for permeability variation in space and considers flow due to lateral density gradients. In addition, unlike previous studies which examined only the physical mixing of seawater and freshwater, chemical reactions associated with saltwater intrusion were examined. The hydrologic models suggest two major flow regimes. A surficial flow system, controlled by local topographic highs and lows, predominates in the uppermost Holly Beach aquifer. In the deeper Cohansey Formation, groundwater migrates across the peninsula primarily from the northeast to the southwest, reflecting the regional slope. The extent of saltwater intrusion is controlled mainly by the vertical permeability of clay confining units between aquifers. High vertical permeabilities, to some extent, prevent saltwater intrusion by allowing surface water to discharge into confined aquifers. The results of geochemical modeling suggest that chemical processes may occur during intrusion as a result of mixing between saltwater and freshwater. Na+ exhibits conservative behavior at low salinities and non-conservative removal at high salinities. Ca2+ displays non-conservative addition over the entire range of groundwater compositions, the amount of enrichment increasing with chloride concentration. H+ exhibits non-conservative removal primarily at low chloride concentrations. At low salinities, mineral dissolution may control non-conservative behavior, which is supported by the conservative behavior of Na+ (suggesting little ion-exchange), the considerable depletion of H+, the undersaturation of calcite in Cohansey groundwater, and the relatively high ratio of Ca2+ to Na+. However, at high salinities ion-exchange predominates as suggested by 2:1 removal of Na+ relative to enrichment of Ca2+, the lower amounts of H+ depletion, and the higher saturation state with respect to calcite. In addition, ion-exchange, overall, may be more important in liberating Ca2+ than mineral dissolution. / Earth and Environmental Science

The purpose of this study was to determine whether the pathways through which groundwater travels in a karst aquifer change between seasons and during storm events. A specific goal was to determine how these pathways affect the water chemistry and sediment in the discharge from a karst spring. From July 2002 through December 2003, storm and continuous monitoring data were collected from Nolte Spring, a former community water supply in southeastern Pennsylvania. Storm samples were collected using an automatic sampler. Additional samples were collected by hand during site visits performed on a bi-weekly to monthly basis. All samples were analyzed for suspended sediment and ions. Datalogging sensors were installed in the spring for the entire study period to record stage, conductivity and temperature. A conceptual model for the study spring was devised to describe storm response and seasonal changes in terms of recharge and groundwater flow pathways. The stage and conductivity at baseflow varied during the study period, as characterized by two end­members. The drought end-member was characterized by low stage and high conductivity. Frequent snowmelt and rain led to high stage and low conductivity during the flood end-member. The relatively high conductivity observed at times during the study period is indicative of slower groundwater flow paths to the spring. Storm responses were determined by different pathways based on recharge type, which was classified as point or diffuse recharge, based on the mean storm intensity. Point recharge events are caused by high intensity storms (> or equal to 1.5 cm/day) and the storm response was an initial increase in conductivity followed by a decrease. Diffuse recharge events are caused by low intensity storms(< or equal to 1.5 cm rain/day) and snowmelt, and the response was a decrease in conductivity. Geochemistry and sediment mineralogy were also affected by the flow pathways. During the drought, the P-CO2 was the lowest of the monitoring period, indicating outgassing from dewatered conduits. The Sic was slightly above saturation, and calcite was observed in the sediment samples, indicating possible calcite precipitation. In the growing season of 2003, P-CO2 was the highest of the monitoring period, due to increased recharge and plant activity. Calcite was also observed in the sediment collected during this time. Various calcite morphologies suggest dislodging of calcite particles from the wall rock and transport of precipitated calcite from another area of the karst aquifer or the soil-bedrock interface as possible reasons for the existence of calcite in the spring. During Winter and early Spring 2003 the suspended sediment mineralogy was mostly siliciclastics such as clays and silt. Sediment concentrations in Nolte Spring were lower than expected, possibly due to the absence of stagnant conduits in the karst aquifer where sediment would accumulate and be flushed during storms. The karst system feeding Nolte Spring has characteristics of both conduit and matrix/fracture flow. Specifically, calcite undersaturation points to conduit flow, whereas stable temperature and low coefficient of variation of hardness (CV) points to matrix/fracture flow. The results of this study also reveal the importance of long-term continuous monitoring, which shows the changeable nature of the karst aquifer in terms of storm water pathway and spring response. / Earth and Environmental Science / Accompanied by one .pdf file: 1) Tancredi-Supplemental-2004.pdf