The Insoluble Organic Matter (IOM) found in meteorites, particularly carbonaceous chondrites, is an abundant reservoir for carbon material. Although the chemical structure of this IOM has been analyzed, questions still remain as to what molecule(s) and chemical reactions have produced it. The carbonaceous chondrite, Murchison, was analyzed with NMR spectroscopy revealing an abundance of furan and aromatic carbons in its chemical structure. With the formose reaction as a guideline, formose products were created using formaldehyde and glycolaldehyde in order to create products that could potentially be structurally similar to the IOM found in carbonaceous chondrites. Using NMR spectroscopy to analyze the chemical structure of these products, they were found to contain many of the same functional groups as the IOM from Murchison. The main difference was the increased amount of methine carbon present in the formose products, which also led to a lower amount of aromatic carbon than the Murchison. A possible solution to decrease the amount of methine is to find a way to dehydrogenate the formose products; therefore, increasing the amount of aromatic carbons due to creation of double bonds from the dehydrogenation mechanism. Overall, the formose reaction can still be considered to be a possible reaction pathway for the synthesis of primitive IOM. Further studies into how these organics evolved through chemical reactions will be able to yield more insight into some of the most primitive chemistry taking place in our galaxy.

The low-metamorphosed state of Neoarchean banded iron formations (BIFs) from Abitibi, Quebec allows for the study of BIFs with minimal post-depositional alteration. Biological oxidation of ferrous iron has been suggested as a possible process by which BIFs form and if correct, organic remains of these microbial communities may be preserved within sediments. An investigation of the carbonaceous material in the BIFs from this region was performed to determine the occurrence of carbonaceous material and if it is associated with apatite, which is a mineral that can form from the diagenetic maturation of organic remains. Mass spectrometry, laser Raman spectroscopy, scanning electron microscopy and optical microscopy were used to provide information about the origin and mode of occurrence of carbonaceous material. Optical microscopy and SEM analyses of thin sections show that apatite exists in association with carbonaceous material but also with carbonate, magnetite and hematite. Laser Raman spectra confirm that apatite is associated with carbonaceous material, which is characterized by D- and G-band peak intensities consistent with the metamorphic history of these rocks. Micro-drilled areas with concentrations of apatite and carbonaceous material were analyzed for carbon isotopes and revealed δ¹³C values between -15 to -45 permil, and organic contents generally around 0.3%. Several large negative δ¹³C values indicate significant carbon isotope fractionation similar to previously published data. The significance and interpretation of these results will be discussed.

The surface adsorption behavior of 12 amino acids (L-alanine, L-serine, L-aspartic acid, L-proline, L-phenylalanine, L-valine, L-arginine, δ-amino valeric acid, glycine, L-lysine, L-isoleucine, and β-alanine) on 21 minerals (quartz, calcite, enstatite, illite, olivine, pyrrhotite, pyrite, alkali basalt, albite, analcime, chlorite, barite, hydroxyl apatite, hematite, magnetite, aluminum hydroxide, kaolin, silica gel, corundum, rutile, and montmorillonite) was determined via batch adsorption experiments. Absorption was determined for concentrations between 10^-4 and 10^-6M in the presence of 0.1M NaCl, and between pH values of 3 and 9 at 25°C. The equilibrated solutions were centrifuged, filtered, derivatized using a fluorescent amino group tag (dansyl-chloride) and analyzed by HPLC. Adsorption was standardized using BET surface area measurements for each mineral to give the number of mols of each amino acid adsorbed per square meter for each mineral.

The results indicate an enormous difference in the adsorption of amino acids between minerals, along with major differences in the adsorption of individual amino acids on the same mineral surface. There is also a change in the absorbance of amino acids as the pH changes. Many previous studies of amino acid concentration and catalysis by minerals have used clay minerals because of their high surface areas, however, this data suggests that the surfaces of minerals such as calcite, quartz and pyrite have even higher affinities for amino acids. The results suggest mineral surfaces that could be optimal locations for the polymerization of molecules linked to the origin of life.

Mineral surfaces may have facilitated the concentration and polymerization of simple biomolecules into macromolecules while promoting the development of homochirality. In this study, rutile and aspartic acid (Asp) were investigated as a possible system in this scenario. Batch adsorption experiments were performed to examine the adsorption of Asp as a function of pH and total concentration. A constant background electrolyte of 0.1 M NaCl was applied to the system and all solutions were purged with Ar gas to eliminate carbon dioxide contamination. Asp adsorbed onto rutile to the highest extent over the pH range 3-5.5 suggesting that an acidic environment is required for the adsorption between Asp and rutile to occur in significant amounts. This pH range of maximum adsorption is constrained between the isoelectric point of Asp and the point of zero charge of rutile, which indicates that electrostatic effects are influencing Asp adsorption. Both the L- and D- enantiomers of Asp were individually adsorbed onto the rutile surface to determine the potential of the system for chiral selection. Preliminary results indicate that D-Asp adsorbs in greater amounts than L-Asp at higher Asp concentrations. This may provide insight on the emergence of chiral selection in macromolecules, although a more thorough study is required to investigate this difference in adsorption.

Using a pulsed laser heating system, in conjunction with a diamond anvil cell, thermal conductivity and thermal diffusivity for MgSiO₃ perovskite has been measured for the temperature ranges of 1800 K to 4200 K at a constant pressure of 125 GPa. At 4000 K and 125 GPa the thermal conductivity has been measured to be 12.5 (W/m*K) and thermal diffusivity has been calculated to be 12.2 * 10^-6 (m²/s). In situ x-ray diffraction tests of the sample performed at 125GPa and 2500 K show that the original enstatite (Mg_0.9Fe_0.1)SiO₃ has compositionally changed into two separate assemblages of perovskite and postperovskite. This mixture of the two phases as well as the diffusivity results give insight into the composition of the lower mantle at the D” layer along with information about the geodynamics of the coremantle boundary.

The partitioning behavior of sulfur and oxygen between metal and silicate was observed at 2 GPa and 8 GPa for temperatures of 2273 K. Experiments were carried out using a piston cylinder apparatus and multi-anvil press. Three different starting materials, all of CI composition were used, along with two different capsules. FeO concentration was adjusted to vary oxygen fugacity relative to the iron-wüstite buffer. Sulfur was introduced in small amounts as FeS metal. Samples were recovered and analyzed using an electron microprobe. Data is available for 2 runs at a pressure of 2 GPa. In the Fe-C-S-O system, it is known that at low pressures the C presence produces two immiscible liquids. Indeed, at 2 GPa, using a Ccapsule, two immiscible liquids were observed in the metal, one sulfur rich alloy and one sulfur poor alloy. Using weight percent data obtained from the microprobe analysis, least square mass balance calculations were used to find the composition of the total metal in the system.

The partition coefficient of sulfur was calculated and the behavior of S was compared to the oxygen fugacity of our system. Also, the weight percent of O, as reported from microprobe data, was studied as a function of the oxygen fugacity of the system. Results from comparing the partition coefficient of sulfur and weight percent of O in the metal as a function of oxygen fugacity, demonstrate that as the system is more oxidized, S and O increase in the metal. This result is consistent with previous work. Additionally, in the Fe-C-S-O system our results suggest that the partition coefficient of oxygen is independent of the C and S content in the metal. Future analysis of runs using MgO capsules will provide a comparison for the Fe-S-O system. Analysis of 8 GPa will also provide the pressure effect on the Fe-C-SO system.

Shear wave splitting is a valuable tool which can be used to gain information about seismic anisotropy. In turn, seismic anisotropy can help us learn about mantle flow and deformation. Despite advances in our understanding of the structure of subducting slabs, mantle flow in subduction zones is still poorly understood. Data from fifty-four broadband seismic stations in Japan’s F-net array were examined for evidence of shear wave splitting using the SplitLab software. Japan was chosen for this study because of its dense seismic instrumentation and location near two subduction zones. Direct S waves from local earthquakes in the subducting slabs were analyzed for all fifty-four stations, and teleseismic SKS waves were examined at four stations in the Izu-Bonin region.

Results for local events in Tohoku and Hokkaido showed evidence of trench parallel fast directions close to the trench with a flip to trench erpendicular farther away from the trench. This is consistent with previous observations as well as with the presence of B-type olivine in the shallow corner of the mantle wedge and subduction parallel flow in the deeper mantle wedge. There was a significant amount of complexity in the splitting patterns from the outhwestern Japan region. This is consistent with a complex flow field due to the interaction between the Pacific and Philippine plates. The Izu-Bonin region yielded results which were consistent with subduction parallel flow. Events detected at these stations were likely too deep and too far from the trench to sample any B-type olivine. The SKS waves in the Izu-Bonin region showed significant complexity which may be due to multiple anisotropic layers. Delay times from all regions besides Tohoku ranged from 0.85 to 2.2s with a regional average of ~1.4 s. Although there was a limited data set, delay times from the Tohoku region were ~0.3 s with a bandpass filter of 0.125-0.5 Hz. This is longer than the ~0.1 s delay times obtained by Nakajima and Hasegawa (2004) at 2-8 Hz. This provides strong evidence that delay times are frequency dependent and this must be taken into account for future studies in Japan and elsewhere.

This study established a dataset of local shear wave splitting results for fifty-four broadband stations which had not been previously analyzed. The work presented here complements previous work on teleseismic and local splitting in Japan and provides a more detailed picture of anisotropy in both the mantle wedge and the sub-slab mantle. Further analysis and interpretation of these measurements should enhance our understanding of mantle flow and slab-mantle interaction in subduction zones both in Japan and on a global scale.

Subduction, the sinking of oceanic plates into the mantle, is a central component of plate tectonic theory. However, the process by which subduction initiates is not well understood. Gurnis et al. (2004) developed a finite element model of subduction initiation using the dynamics of elastic flexure, viscous flow, and plastic failure constrained by temperature, stress, and strain rates. They found that three conditions were necessary for subduction to initiate: the presence of a fracture zone, an offset in the age (and therefore density) of the ocean floor along the fracture zone, and compressive stress (leading to shortening) oriented normally to the fracture zone. To evaluate the Gurnis et al. model, digital seafloor age data and two different global stress models were used to calculate a parameter predicting the location of relatively likely subduction initiation on the ocean floor in the present day or near future.

Principal stresses and the azimuth of the greatest principal stress were calculated from the stress tensor data. No preexisting database of fault locations on the ocean floor was available so as part of this project, a dataset of fracture zones associated with age offsets was compiled in MATLAB by digitizing faults visible on a map of the seafloor age data. Using MATLAB, a “subduction initiation likelihood parameter” at each fault segment was calculated as the magnitude of the stress vector normal to that fault. This parameter was then normalized by the greatest stress value to span from -1 (most extensional) to 1 (most compressional). A value of zero indicates a strike-slip regime with no normal stress.

Both stress models predict regions where subduction initiation is likely. Using the Ghosh (2008) model, high likelihood for subduction initiation is predicted in the Indian Ocean, south of Australia, in the northern Pacific Ocean, and off the east coast of South America. Using the Lithgow-Bertelloni and Guynn (2004) model, high likelihood for subduction initiation is predicted in the Indian Ocean, south of Australia, and in the northern Pacific Ocean. Both stress models predict that subduction initiation is relatively likely in the Indian Ocean basin; however, no subduction has initiated in the Indian Ocean basin since the collision of India with Eurasia ~50 Myr ago. This observation may present a challenge for models of the subduction initiation process, such as that of Gurnis et al. Future work on this project will include looking for geophysical evidence of subduction initiation (e.g. crustal shortening and thickening, deep earthquakes) at sites where we predict it to be likely; we also plan to evaluate the stress regime in very young (<5 Myr) subduction zones. This observation-based approach to evaluating the predictions of numerical models should lead to a better understanding of the subduction initiation process.

A fitting technique called particle swarm optimization (PSO) was used to fit radial velocity curves to radial velocity data of seven stars known to have planetary companions that were also likely to host additional planets. The stars that were analyzed were 47 UMa, HD 142, HD 6434, HD 37124, HD 68988, HD 216437 and HD 213240. Before the fitting was done, an assessment of the parameters of the PSO program was made in order to optimize its efficiency. The assessment revealed that PSO is not overparticular and requires no alteration for radial velocity curve fitting. 120 PSO runs were performed for each star. A different random number seed was used for each run, ensuring that the parameter space was explored in a different way each time. HD 6434, HD 37124, HD 68988, HD 216437 and HD 213240 were not considered further because they did not yield a good χ₂. HD 142 and 47 UMa were examined further. Two, three and four planet fits were attempted for both stars with the 4 planet fits yielding the best χ₂. For 47 UMa, it was found that there is most likely a planet with a period anywhere between 8,000 and 20,000 days. The PSO program also found four short period possibilities at 13, 53, 270 and 320 days. A 5,000 to 8,000 day period was observed around HD 142 as well as two shorter periods at 3.4 and 30 days. The long period planets for both HD 142 and 47 UMa require a few full orbits to confirm and thus a few more decades of data. Additionally the short period planets for both stars cannot be confirmed until more data is acquired.

We have performed density functional electronic structure calculations for oxygen at 350 GPa, using the Quantum Espresso Package with ultrasoft pseudopotentials and PBE exchange and correlation functional. The random search method was used to determine the enthalpies and lattice parameters of monatomic oxygen at 350 GPa. Other chalcogen elements (Po, Te, Se, S) all take a rhombohedral β-Po structure when compressed. Polonium forms its β structure at zero pressure, Te at 11 GPa, Se at 60 GPa, and Sulfur at 153 GPa. As you move up the table, higher pressures are needed to make the element take a β-Po form. Our hypothesis was that at pressures near 350 GPa, oxygen would have a β-Po structure as well.

Within random search, several starting configurations are randomly chosen and relaxed, until the final structures, with the lowest enthalpy, are found several times. First, the crystal lattice is generated by randomly selecting cell-vector lengths between 0.5 and 1.5 (in arbitrary units) and three cell angles between 40° and 140°. The cell vectors are then scaled to match a new volume, which is also chosen randomly between 0.5 and 1.5 of some physically sensible volume. Then, atomic positions are obtained by generating three random numbers between 0 and 1 for each atom, which represents the positions of the atoms in terms of the crystal vectors. First-principles methods are then used to relax the cell towards the closest minimum in enthalpy.

We performed a random search using one oxygen atom per unit cell. Therefore, only the six cell degrees of freedom were necessary. 150 random starting systems were selected and then converged towards equilibrium at 350 GPa. 1000 iterations or more were necessary for each structure to reach hydrostaticity of the stress tensor. The systems’ enthalpies were then calculated and the structures with the lowest enthalpies were analyzed. Eight different values for enthalpy local minima were obtained. The global minimum was obtained with 14 out of 150 samples. This global minimum had an enthalpy that was 50 meV lower than that of the closest local minima and had a C2/m space group. The results show that oxygen forms a monoclinic arrangement of 1-D chains at 350 GPa and zero temperature. For the global minimum, the spacings between oxygens along the chains is 1.44 Å, while the distance from one chain to the closest other is 1.98 Å.