Highly Siderophile Elements in H Chondrites GEOL394

Highly Siderophile Elements in H Chondrites GEOL394 Jonathan Tino Dr. Richard Walker Dr. Katherine Bermingham Greg Archer 1

Abstract Chondritic meteorites are undifferentiated solar system bodies that have not melted since the beginning of solar system formation. This study examines a subset of these meteorites that are known as the H chondrites due to their high iron metal content. Isotope dilution of highly siderophile elements (HSEs) Re, Os, Ir, Ru, Pt, and Pd has been applied to H chondrite samples of three metamorphic grades to determine the extent of equilibration of these trace elements between metal and silicate. Equilibrium is assessed by comparing the measured ratios of the concentration of the HSEs to established equilibrium concentration ratios, or D values, of metals relative to silicates (O Neill et. al. 1995). The study also assesses whether fine metal grains (<150 µm) contain higher concentrations than coarser grained metals (>150 µm). Thermal Ionization Mass Spectrometry (TIMS) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS) analyses have been used to quantify isotopic ratios and concentrations of HSEs in fine grained metals, coarse grained metals, and silicate material within the Avanhandava (H4), Richardton (H5), and ALHA 78115 (H6) meteorites. Metamorphic grades 4-6 (in increasing degree of metamorphism, with 4 being the least and 6 being the most) have been chosen in order to assess the possibility of a positive correlation between increasing metamorphic grade and increasing concentration ratio. Similar methods have not previously been applied to H5 or H6 chondrites. After obtaining HSE data for fine grained metal, coarse grained metal, and silicates, I conclude that the HSE in metals and silicates are not in chemical equilibrium for any petrologic type. This is based on the observation that the concentration ratios of Richardton H5 fine metal concentration relative to Richardton H5 silicate concentration reaches a factor of 3833/1. This is the highest degree of separation for any element between any two samples, and it falls well short of the previously established concentration ratio indicative of HSE in chemical equilibrium between these two reservoirs. There is a possibility that open system behavior or metal contamination in the silicate fractions gave rise to this result. It is also possible that the complete chemical equilibrium of these highly refractory HSE does not occur until the sample reaches much higher peak metamorphic temperatures, at which point partial melting would begin and the rock would become an H7 chondrite. 2

Introduction Background: Chondritic Meteorites Chondritic meteorites are among the oldest rocks in our solar system. Their components can include a stony matrix, chondrules, metal grains of various size, and refractory inclusions (McSween, 1987). Chondrules, the namesake of chondritic meteorites, are millimeter-scale spherical grains that cooled from the rock s initially molten state prior to becoming incorporated in the chondrite s matrix that are composed of silicate minerals as well as metal grains. Refractory inclusions are calcium-aluminum inclusions (CAIs) and amoeboid olivine aggregates (AOAs). CAIs are thought to be the first substances to condense from the primordial dust and gas cloud known as the solar nebula. These inclusions can constitute anywhere from 0.01-10% of the volume of a chondrite (Scott and Krott, 2003). CAIs are made up of spinel, melilite (Ca,Na)2(Al,Mg,Fe 2+ ) [(Al,Si)SiO7], hibonite ((Ca,Ce)(Al,Ti,Mg)12O19), Al-Ti diopside, and perovskite, all of which are absent in other chondrite components. Amoeboid olivine aggregates contain some Fe-Ni metal, granular massive olivine [(Fe,Mg)2SiO4], spinel [(Mg,Fe) (Al,Cr)2O4], anorthite (CaAl2Si2O8), Al-diopside, and rarely melilite. Metal grains, primarily comprised of Fe-Ni, can be up to 70% of a chondrite by volume (Scott and Krott, 2003). Matrix material, primarily silicates and fine grained Fe-Ni metal, is volatile rich, can contain sulfides, oxides, organic material, and even rare presolar grains, which are created by stars that existed before ours formed (Scott and Krott, 2003). The chondrites commonly comprise up to 80% matrix. The origin of each component is still disputed, but the background knowledge of the solar system suggests that they formed in different areas of the solar nebula (McSween, 1987). Figure 1: Comparison of elemental abundances in C1 carbonaceous chondrites and the solar photosphere demonstrates an approximate 1:1 correlation. Solar atmosphere elemental abundances are derived using spectroscopy. Red arrow indicates a 1:1 correlation line. 3

Additionally, each component is likely to have condensed at a different time in the formation of the solar system. Solar system cooling and varying volatilities of the chemical elements provide good evidence for the oldest materials being highly refractory. The cloud was originally hotter, and so the first solids to form must have comprised the least volatile elements. Calcium and aluminum form some of the most refractory compounds, and thus they condense at much higher temperatures relative to most elements. The refractory nature of the calcium and aluminum compounds is the reason that CAIs are presumed to be the oldest components both in chondrites and possibly the solar system. The approximate chemical abundance of elements in the solar system can be determined by analyzing C1 carbonaceous chondrites. These are a class of meteorites that contain high amounts of water (3-22% by volume) and organic material, as well as silicates, oxides, and sulfides (Norton, 2002). Data from Ahrens et. al. (1989) suggest that there is an approximate 1:1 correlation between relative abundances, normalized to 10 6 Si atoms, of the solar atmosphere and C1 carbonaceous chondrites (Fig. 1). The largest variations from the solar abundance are caused by volatility of the elements in question. Abundances of highly refractory elements are closer to a 1:1 ratio between carbonaceous chondrites and the solar atmosphere, and abundances of volatile elements are not as close to a 1:1 ratio. Due to intense stellar winds during the sun s T-Tauri phase, most inner solar system terrestrial bodies are depleted in volatiles relative both to outer solar system gaseous bodies and the sun (White, 2013). After formation, temperatures in chondritic parent bodies are not thought to have reached the temperatures necessary in order to melt. Therefore, any diffusion of elements between minerals occured through grain boundary diffusion in metamorphic regimes. I have examined the group known as the ordinary chondrites. These chondrites comprise the majority of the undifferentiated meteorite group, and they are the majority of meteorites found on Earth (Scott and Krott, 2003). Specifically, this project s focus is examining H chondrites. This subgroup of ordinary chondrites is high in iron metal and total iron content, which makes comparisons of trace elements in metals and silicates easier due to their high concentrations compared to other types of chondrites. The chondrites are split into categories based on their petrologic type, which is rated on a scale from 1-6 and is based on the level of metamorphism the meteorite experienced. Levels 1-2 are reserved for aqueous alteration, which is not observed in H chondrites, so H chondrites begin their classification with grade 3. Grades 3-6 are thermal metamorphic grades, with 6 being the highest grade, which are common in H chondrites. H3 chondrites, which are relatively unaltered, are metamorphosed between 400 C and 600 C, H4s are metamorphosed between 600 C and 700 C, H5s form are metamorphosed between 700 C and 750 C, and finally H6s are metamorphosed between 750 C and 950 C (McSween, 1987). One theory, called the Onion Shell Model, suggests that most H chondrites originated from one parent body, or that they were one proto-planetary mass initially before impacts dislodged the pieces that eventually found their way to 4

Earth (Wood, 2003). Heat production within the parent body is thought to come from planetary accretion and radiogenic heat produced by the decay of 26 Al=> 26 Mg+ + B, a positron decay reaction (McSween, 1987). The model shown in figure 2 illustrates the perceived source of each metamorphic grade among the H chondrites in this proposed parent body. This thermal model is based on the assumption that heat transfers slowly in silicate bodies, and the heat is therefore stratified into distinct temperature layers. However, Kessel et. al. (2007) used the Fe-Mg exchange between the olivine and spinel that constitute H chondrites to determine that the peak metamorphic temperatures for H4 through H6 chondrites are all in excess of ~730 C. Kessel s study also concludes that the onion shell model is not consistent with such a narrow range of peak metamorphic temperatures for grades that are supposed to occur in layers of distinct heat ranges. Figure 2: From Wood (2003), this illustration shows the higher metamorphic grade H chondrites as having originated at a deeper (hotter) location within the parent body than H3 and H4 chondrites. Temperature and pressure are greater within the parent body than they are near the surface. Thin sections demonstrate the high degree of recrystallization in H6 relative to H4. Highly Siderophile Elements The trace elements examined in this study are the highly siderophile elements (HSEs). These elements are notable for partitioning strongly into iron metal relative to silicate phases. The elements utilized in this study are rhenium, osmium, iridium, ruthenium, platinum, and palladium, ordered by increasing volatility from Re to Pd. First, they partition strongly into metals relative to silicates when the system is in equilibrium, with partition coefficients of >10 4 (O Neill et. al, 1995). These partition coefficients, or D values, indicate the concentration ratio of an element in one reservoir to another, in this case, metal and silicate. Therefore, there are roughly 10 4 atoms of the element in the metal grain for 5

each 1 in the silicate material. Next, the potential for multiple oxidation states and their varying volatilities enables them to record nebular condensation, evaporation, and parent body processes such as aqueous alteration (Horan et. al, 2009). The solar system-forming process began at a higher temperature and cooled as the heat dissipated over time. Therefore, the more refractory elements, which condense at higher temperatures, will condense earlier in the solar system s history. Third, they can be used to determine metal melting and crystallization processes provided those values are well defined for the HSE (Horan et. al, 2009). The presence of sulfur and phosphorous also can have an impact on how HSEs partition when both liquid and solid iron phases are present, but this study is concerned only with metal grains that have been solid since the beginning of solar system formation. The fourth use for HSEs in this study is the 187 Re- 187 Os radiogenic isotope systematics (Horan et. al, 2009). This system can be used to determine whether or not the chondrite underwent any open system behavior after its components initially coalesced. The isotope ratios for metals and silicates can be plotted on a primordial isochron, and if they do not correlate with the 4.57 Ga age isochron, one can infer that some open system behavior occurred. A likely explanation is shock metamorphism from an impact, which can also potentially be dated using the 187 Re- 187 Os systematics. Overall, the two uses for HSEs most important to this study are their ability to constrain ages and assess open system behavior in chondrites using 187 Re- 187 Os systematics and their ability to constrain metal-silicate equilibration. Figure 3: Data from Horan et. al (2003) representing concentrations of HSEs in H chondrites from 4-6. Forest Vale (H4), Forest City (H5), Ochansk (H4), Zag (H3-6), Rose City (H5) are represented. These data are bulk rock analyses normalized to a CI chondrite value. 6

The primary basis of this study is the work done by Horan et. al. (2003, 2009). In Figure 2, data from chondrites in the same range of metamorphic grades as the ones in this study demonstrate bulk rock HSE abundance (Horan et. al. 2003). It was concluded that shock metamorphism during impacts does not fractionate HSEs significantly, but that it can reset Re-Os ratios partially. The higher grade Zag (H3- H6) and Rose City (H5) samples have higher abundances relative to the lower H4 samples. The Zag and Rose City also experienced relatively high degrees shock metamorphism, which is noted to cause both high and low HSE concentration anomalies. Horan (2003) deduced that Re and Ru may be more easily oxidized in these conditions, leading them to be more mobile than other HSE. Additionally, because these are bulk analyses, the components are not sieved and filtered into separate analyses. Sample inhomogeneity can lead to uncertainty in measurements, as Horan discusses in her study. At the time, there were relatively few data available to compare to these results, and what was available also had large variations (Horan et. al 2003). In addition, chondrites with the same metamorphic designation may not have the same abundances of HSEs. Compositional variations within each parent body as well as the dubious nature of classifying these rocks can account for this difference. For example, classification of the Faucett meteorite has changed from H4 to H5 and back again since its discovery (The Meteoritical Society, Lunar and Planetary Institute). Figure 4: Abundances of HSEs in the H4 chondrite Ochansk. Each fraction was isolated by magnetic separation. Additional sieving is done in order to separate fine metal (<25 micron) and medium (125-25 micron) metal from coarse (>250 micron) metal. 7

A follow up to the 2003 study in 2009 by Horan et. al. focused on separating the components of H chondrites into compositionally distinct fractions. Although it was not the main goal, this study took steps towards mitigating the uncertainty from sample inhomogeneity by determining specific abundances for distinct components of each chondrite. Measurements of isotopic abundances from nonmagnetic, slightly magnetic, coarse grained metals, and fine grained metals were recorded. This study focused on assessing HSE equilibration in the H chondrites based on their D value (>10 4 ) of elements in metals compared to silicates. The nonmagnetic fractions are the silicates, and the slightly magnetic fractions are poorly separated metals and silicates. This study only included H4 chondrites, and the data, represented in figure 3, show metal silicate concentration ratios to be less than 10 4. By definition, this means that the H4 chondrites analyzed in the study did not show HSEs in equilibrium conditions. The fine grained metal fractions had a higher concentration for all HSEs than coarse grained metal fractions. Previously, the H4-6 chondrites were known as the equilibrated H chondrites (Horan et. al, 2003), but this conclusion now seems to be in question. The 2009 study shows that the H4 chondrite does not contain equilibrated HSEs, because the concentration ratio for most of the metals relative to silicates in Figure 4 is much closer to 10 2 than 10 4. A study using similar methods to those used by Horan et. al. (2009) has not been done for grades H5 and H6. Hypothesis My study focused on measuring HSE abundances in the Avanhandava (H4), Richardton (H5), and ALHA 78115 (H6) meteorites. These have been separated into fractions based on size and composition. I have completed isotope dilution, Thermal Ionization Mass Spectrometry (TIMS), and Inductively Coupled Plasma Mass Spectrometry (ICP-MS) to evaluate the following hypotheses: A) Increasing thermal metamorphic grade in H chondrites will correlate with increasing concentration ratios in metal relative to silicate approaching or exceeding the D value of 10 4 that is expected for HSEs in chemical equilibrium. I predict that H4 chondrites will not be equilibrated, (metal/silicate<10 4 ) and H6 chondrites will be equilibrated (metal/silicate>10 4 ). If the concentration ratio is greater than 10 4, then I can say that the HSEs are in chemical equilibrium for that grade of H chondrite. I can demonstrate an increasing degree of equilibration of HSEs with increasing thermal metamorphic grade by measuring how close to 10 4 (the D value of partitioning) the concentration ratio in metal relative to silicate is for each sample. An increase in concentration ratio between metal and silicate from H4 to H6 would verify statement A. B) Fine grained metals (<150µm) will have a higher concentration of HSEs than coarse grained metals (>150µm) relative to silicates. This phenomenon has 8

been observed in previous studies on metal grains in chondrites [Rambaldi et. al (1997) and Campbell and Humayun (2003)]. Hypothesis B is simple to test, but more difficult to explain. In the discussion, I will compare the data for coarse and fine metal across all three meteorites in order to find significant differences. A difference in HSE concentrations between fine and coarse fractions could be explained by small-scale grain boundary diffusion. When a smaller metal grain has the same initial concentration of an HSE as a larger grain, and the concentration of the HSE in the silicate matrix is the same surrounding both, it follows that a metamorphic event will cause the same amount to enter each grain. However, due to the smaller initial volume, the smaller grain will have a much higher concentration. Additionally, Rambaldi et. al. (1977) suggested that the difference in concentration is due to different mineralogy in grain sizes. They found that coarser grains had a higher kamacite concentration, and that finer grains had a higher taenite concentration. Both are Fe-Ni solid metals, but taenite is more abundant in Ni. A later study by Campbell and Humayun (2003) found that kamacite/taenite partitioning occurred at ~1020K, and cooled rapidly to prevent further equilibration of metals. However, they determined that the highly refractory siderophile elements, including Re, Os, Ru, Ir, and Pt, had localized variations from grain to grain that are unrelated to the kamacite/taenite partitioning. There is no dominant theory (that I am currently aware of) to explain this phenomenon, so it will be interesting to investigate. Methods Samples of Avanhandava (H4), Richardton (H5), and ALHA 78115 (H6) meteorites were obtained from Dr. Terry Blackburn of U.C. Santa Cruz. They were already sorted and processed such that silicates/non magnetics, coarse grained metals, and fine grained metals for each sample were separated. However, for the analysis to be truly representative, it was necessary to purify each sample further by means of mortar and pestle crushing of the metal grains in order to remove any small silicate material. This was followed by magnetic separation with a powerful magnet. They were also sieved in addition to the sieving done by Dr. Blackburn in order to separate metals into those >150 microns (coarse grained) and those <150 microns (fine grained). Each sample was then crushed using a mortar and pestle, washed into a watch glass, and purified with a magnet up to 4 times, or until no silicate dust remained within the water of the watch glass. For silicate fractions, the goal was to remove metal from the watch glass and dry down the remaining silicate. Each sample was subsequently dried with a heat lamp, and then placed into glass vials with the appropriate label. When a sample reaches the point where it is significantly rid of silicates, (or metals, depending on 9

the analysis) it is ready to be digested in a carius tube. This allows us to begin the next process, called isotope dilution. Isotope Dilution For coarse metals, I weighed out aliquots of roughly 5 mg for each of the three meteorites. Next, I weighed out spikes for each sample in 4 Teflon beakers, 1 for each meteorite and a total analytical blank. My spike and sample weights are represented in table 1. I used two different high precision balances; one that has up to six decimal precision (for sample weight) and another that has up to five (for spike weights). Adding spikes of known concentration and mass is crucial for calculating sample concentration data, and this is the basis of isotope dilution. These data are important in reducing the TIMS measurements, which will be discussed later. Next, I took the samples and add them to labeled carius tubes in an ice bath using weighing paper and glass funnels. It is important to keep these cool, as the osmium will oxidize once exposed to the acid, aqua regia and become volatile (easily turned into a gas). Keeping temperatures down will reduce the likelihood of losing sample. The previously mentioned digestion acid, known as aqua regia, is a mixture of 2:1 concentrated nitric acid (HNO3) and hydrochloric acid (HCl). Next, weigh ~1 ml of the HCL and add it to the appropriate beakers which contain the spikes. It is important to add the HCl first and expose the mixture to cold in order to mitigate osmium loss. Then, ~2ml of HNO3 may be added to each appropriate carius tube. This was allowed to cool, and then I rinsed the funnels with a small amount of Milli- Q purified water (the Milli-Q system s filters ensure high degrees of purity in H2O) to clean out any sample or acid that may remain. Once the tubes are on ice, I sealed them and began digestion. Next, the tubes were sealed with a blowtorch and allowed to cool. At this stage, it is not necessary to freeze them because nothing should be able to escape if the seal was done correctly. The tubes were placed in steel jackets and heated in an oven to 240 C for 24 hours. Once the digestion is complete, the osmium was extracted using 3 ml of Carbon Tetrachloride (CCl4) in a centrifuge tube for each sample. One also must prepare 4 ml of hydrobromic acid (HBr) in each of the Teflon beakers that once held the spikes for each sample. These must be cleaned thoroughly with Milli-Q purified H2O. Because the aqua regia oxidizes the osmium in the samples, the tubes are exposed to water ice or liquid nitrogen to prevent loss of OsO4, then opened, again using the torch, and the contents are thawed and quickly added to the 10

centrifuge tube. The samples are shaken vigorously to enable the OsO4 to be captured by the CCl4. The initial 3mL of CCl4 are then extracted once the mixture has settled out, as CCl4 is denser than the acidic phase containing the rest of the highly siderophile elements. For each sample, use a different, clean pipette to transfer the CCl4 to its appropriate Teflon beaker containing the HBr. This process is repeated 2 more times to ensure that nearly all of the OsO4 is extracted. In the metal analysis, there was no significant silicate residue, so it was not necessary to use the centrifuge to separate the sample from the residue. Significant centrifuging is required for silicate samples, where residual silicate material is certain to be an issue. Silicate matter that ends up in your sample will make eluting solutions through resin columns problematic. I centrifuged the samples after each time I shook them. Otherwise, I would have undone the separation. Once the extraction was completed, I sealed each beaker and place it on a hot plate set to 85 C for 2-3 hours. This transfers the osmium from the Tetrachloride to the HBr acid through a reduction reaction, removing any oxygen from OsO4 and isolating the osmium atoms. When the osmium reduction is complete, the sample is dried under a heat lamp until it is a droplet that is approximately 40 µl in volume. It is then transferred to the inside of the cap of a clean conical bottom Teflon beaker. Subsequently, the sample is dried completely under a heat lamp. Now, it is ready for micro distillation, the last step before the sample can be loaded onto a platinum filament for TIMS analysis. Micro Distillation Mix the dry sample with 20-30µL of dichromate (H2SO4-CrO3) on the inner cap, and then use a pipette to place 15µL of HBr in the tip of the conical bottom. The hot plate is set to ~85 C, and the beaker is placed upside down and wrapped with tin foil, allowing for the HBr tip to be uncovered, on the hot plate. The function of HBr is to reduce the osmium again, but you must let the solution heat for roughly 3 hours to ensure complete reduction. If the dichromate becomes red and turns yellow when Milli-Q purified water is added, it indicates that the osmium was properly oxidized by the dichromate and has collected in the HBr. Then, the sample may be dried until it is powder using a heat lamp. The isotope dilution chemistry for osmium is now complete, and the samples are ready to be loaded onto a platinum filament for TIMS analysis. All chemistry is done following the IGL HSE Manual (2012). Thermal Ionization Mass Spectrometer Analysis Samples are loaded onto platinum filaments by dissolving the dried product of osmium micro distillation into 2 µl of HBr. A filament loading platform runs current through your metal filaments, heating the drop of HBr until it dries 11

completely. Each load will contain 1 µl of osmium bearing HBr, and will be coated in Barium Hydroxide (Ba(OH)2), which enables the samples to ionize properly. Now, the samples can be placed onto the sample wheel and into the TIMS itself. The heating element will slowly heat the filaments up to the temperature required for analysis (750-850C for Os). The source functions as a way to focus your cloud of ions into a beam that can be sent through the analyzer gate and into the detector array itself. The TIMS will take several hours using Faraday detector cups to analyze for osmium. These detectors are designed to attract ions and measure the resulting currents to determine the number of ions making contact with the cup (Brown and Tautfest, 1956). It is important to note that TIMS takes place in near vacuum conditions in order to avoid sample contamination. The ionized sample is focused into a beam and run through the analyzer gate, where a magnet will deflect beams of ions of varying isotopic masses into the appropriate Faraday cup. The TIMS data collection software will measure counts of different isotopes of osmium in each cup, which is calibrated to detect osmium isotopes of different masses. The currents generated by the ions making contact with the Faraday cups are converted into measurements of concentration and relative abundance ratios between one osmium isotope and another. For example, measurements of 187 Os/ 188 Os osmium are useful in determining the age of a sample due to the radiogenic nature of 187 Os. Each sample I measured represented an average of 80 ratios, and each standard was an average of ~60 ratios. The data are corrected and reduced by the program that collected it, but needed to be further reduced. Because small amounts of O2 bled into the near-vacuum mass spectrometer tube, the data need to be corrected to account for this effect. The oxidized osmium increases signal strength, and it improves the quality of the data. An excel macro designed to correct for the oxygen present in the osmium further reduces the data in order to obtain true ratios and concentrations. Additionally, blank correction needs to be done to account for background contamination in the isotope dilution chemistry. This will be discussed further in the results section. Inductively Coupled Plasma-Mass Spectrometry In order to complete ICP-MS on the Nu Plasma mass spectrometer, I used anion resin separation column chemistry to distinguish between the dissolved solids in our HSE sample cuts. For each sample, I labeled a column and filled it with anion resin, calibrated and cleaned the resin by allowing a series of acids and water to filter through it, and collected Pt and Ir, Re and Ru, and Pd in three separate beakers for each sample. The columns are calibrated to release these elements in certain concentrated acids after other acids have cleaned out contaminating elements. A secondary column is used to clean up the Re and Ru cuts, which uses similar techniques to the primary column. The details of this column chemistry can be found in the IGL HSE Manual. Once the concentrated acids have been dried, they are dissolved again in a miniscule amount of aqua regia and dried again in order to 12

oxidize any organic contaminants. They are then re-dissolved in 5% HNO3 and placed into small centrifuge tubes for the Nu Plasma ICP-MS analysis. The Nu Plasma does not use filaments, and requires a sample to be dissolved in dilute acid in order to run. Otherwise, it shares many of the same aspects as the TIMS regarding the software and collector array. The loading mechanism is referred to as an ion pump, and this allows the maintenance of vacuum and solves the issues certain HSE have with electroplating the insides of the TIMS. The machine is multi-collector, with 16 Faraday cups and 5 ion counters (The Nu Plasma II MC-ICP-MS: Neodymium and Hafnium Isotope Ratio Measurements). The former are used for low concentration samples, where the latter can handle higher concentration samples. The sample is recorded by the strength of the resulting signal, so it is important not to run samples on IC that may overload them and break the instrument. Additionally, the Pd cut must be shaken up in its centrifuge tube and only added to the tube immediately before the analysis, as it tends to adhere to plastic if left to settle for too long. For Re and Ru, I mixed a W standard in with my samples in order for the software to correct for fractionation. I needed an element with an isotopic mass of 185 to correct our Re 187, so W 185 was used. The ability to run multiple isotopes of elements at the same time in order to correct for their fractionation is one of the larger draws to using this method for this suite of five HSE. The data collected with this analysis are put into the same blank correction spreadsheet as the osmium data. 13

Results Table 1: Weights, HSE concentrations, and isotopic ratios for all samples along with absolute 2sigma standard error. These isotopic ratio data are used in the isochron plot, and the concentration values are used in the main results plot, Figure 6. In figure six, the concentrations are all normalized to the average Orgueil data for each element. Normalization to a C1 carbonaceous chondrite is a standard practice that allows for plots of data to be more easily comparable to one another though their values may differ by orders of magnitude. Figure 5 : coarse metal HSE concentration data from H4, H5, and H6 chondrites with coarse metal HSE data from Horan et. al. 2009. All data are normalized to Orgueil, a carbonaceous chondrite. X-axis shows the suite of HSE in order of increasing volatility from left to right. 14

Concentration Relative to Orgueil The concentrations of HSE measured in coarse metal HSE, displayed in Table 1, are consistent with trends of increasing concentration with increasing metamorphic grade. However, this was not the case for the silicate data for osmium, which demonstrated no correlation between petrologic type and increasing concentration. Similarly, the osmium concentrations for fine grained metal fractions did not display increasing concentration with increasing petrologic type. In addition, the degree of difference between the silicate concentrations and both metal fractions was around two orders of magnitude at most for osmium. Concentration data for Ochansk (H4) and Djala (H3.8) from Horan et. al (2003) compare closely to the data I measured in figure 5. In figure 6, all data gathered are shown normalized to a C1 chondrite, Orgueil. The other elements follow the same concentration trends as Os for each meteorite. The largest degree of fractionation occurs between Pd concentrations in H5 silicate and H5 fine metal. H5 has the highest overall concentrations in fine grained metals, and the fine grained metal concentrations overall are the highest out of the three main fractions analyzed. 100 Comparison of HSE data for H Chondrites 10 1 0.1 0.01 0.001 Re Os Ir Ru Pt Pd Avanhandava H4 coarse Richardton H5 coarse ALHA 78115 H6 coarse Avanhandava H4 Silicate Richardton H5 Silicate ALHA 78115 H6 Silicate Avanhandava H4 FM Richardton H5 FM ALHA 78115 H6 FM Figure 6 : HSE data I have collected compared on the same graph. The Y-axis shows concentrations relative to Orgueil, a carbonaceous chondrite. The X-axis represents the HSE in order of increasing volatility from left to right. Measuring Re on the Nu Plasma enabled me to evaluate the Re-Os isotope systematics for those three samples. I plotted the 187 Re/ 188 Os to 187 Os/ 188 Os on a 15

primordial 4.57 Ga isochron, and the results indicate plenty of deviation from that line for silicates and fine metals. The coarse metals show the least amount of deviation. In some cases, isochrons are used to date rock samples using the radiogenic decay rate to measure the time passed since the system closed. In this case, I am assessing the possibility of open system behavior after the system initially cooled. If the data plot significantly off the trend of the primordial isochron, the possibility that the system lost or gained rhenium or osmium is high. Ordinary chondrites are considered primitive meteorites, and the accepted age is around 4.57 Ga. Therefore, one can assume that any deviation from that isochron indicates the possibility that rhenium and osmium fractionated after an event in the past. Coarse Metal Data vs 4.57 Ga Primordial Isochron 0.14 0.13 187 Os/ 188 Os 0.12 0.11 0.1 0.09 0 0.1 0.2 0.3 0.4 0.5 0.6 187 Re/ 188 Os Avanhandava H4 coarse Richardton H5 coarse ALHA 78115 H6 coarse Avanhandava H4 Silicate Richardton H5 silicate ALHA 78115 Silicate Avanhandava H4 Fine Richardton H5 Fine ALHA 78115 H6 Fine Figure 7 : Re-Os data plotted on a primordial 4.57 Ga Isochron. This comparison enables the assessment of open system behavior. The solar system initial ratio (y intercept of isochron) for 187 Os/ 188 Os is 0.09517. 16

Discussion I have completed the osmium chemistry, TIMS osmium analyses, and ICP-MS analyses for all samples. Firstly, there is a notable difference in the silicate osmium data I measured for the H5 and H6 fractions relative to what Hypothesis A predicted. It follows that with more heat affecting those meteorites, there will be an increasing degree of trace HSE metals diffusing into the iron metal grains. However, there appears to be the opposite effect for all HSE, with ALHA 78115 silicates actually containing the highest concentration of HSE. This implies that the hypothesis A cannot be true, as the trend in the silicate data for all HSE does not comply with the one predicted in A. Additionally, figure 6 demonstrates that there is not a 10 4 :1 concentration ratio between either of the metal fractions and the silicate fraction, so the criteria for chemical equilibrium are not met. The closest sample to meeting this comes in the comparison of Richardton (H5) fine grained metals relative to silicates. This one is not quite three orders of magnitude, which falls well short of the goal ratio of 10 4 for all HSE outlined in O Neill et. al. For H4 samples, this is expected and it fits the hypothesis that they are not equilibrated. Yet, the H5 and H6 samples also do not meet the criterion for chemical equilibration of any HSE. This means that hypothesis A can now be falsified, and the null hypothesis is verified. There is no apparent cause to this variation in the expected trends. The degree of exposure to weathering increases the likelihood of open system behavior in the silicate samples, but this would presumably affect the metals as well. Shock metamorphism from meteorites entering atmosphere and impacting the surface can also create temporary windows for the system to open. Fortunately, one can assess open system behavior of our coarse grained metals by using Re-Os isotopic systematics to create an isochron. Because figure 7 appears to demonstrate negligible scatter for coarse grained metals, it is unlikely that the open system behavior was responsible for the much larger variations in the silicate samples. However, upon plotting the silicate data, it becomes apparent that significant deviation from the isochron is prevalent in all of my meteorites. Shock metamorphism or chemical weathering could have caused this. The open system could have enabled HSE from the metals to move back into the silicates over time, and it would explain the unexpectedly high abundance of HSE in silicates from all three petrologic types of meteorites. Another possibility is that my methods for separating the silicates from metals are flawed, and the resulting inconsistency in the silicate data for osmium relative to what my hypothesis predicts is based in metal contamination. Table 1 shows that the concentration of osmium in fine grained metals is very high relative to silicates. Therefore, if even a small amount was left in the silicates, it could significantly alter the results. I processed the samples very thoroughly, going over each aliquot of silicate powder five or more times with a powerful magnet. Still, it is possible that very weakly magnetic metals remained in the sample, and this would likely impact the results. However, because all of the samples were subjected to the 17

same exacting processing, it would mean that they should all be abnormally high in concentration rather than the H4 and H6 silicates having over 10x higher concentrations than the H5 silicate sample. These results may also indicate that the samples never reached a temperature that would have brought HSE in the two reservoirs into chemical equilibrium, or that it did not stay at that temperature long enough for the elements to diffuse from silicates into metal grains completely. A possible experiment to solve this question would involve artificially heating these samples incrementally, allowing them to cool, and then processing them as normal to assess the temperature at which HSE go into chemical equilibrium in metal relative to silicates. It is already demonstrated by planetary bodies such as the Earth that melting events partition HSE into planetary cores almost completely. It would be important to maintain the solid form of the sample in order to mimic conditions in the H chondrite parent body accurately, but it would be an interesting direction to explore in future studies. Conclusions Overall, my concentration data demonstrates that there is no chemical equilibration in any of the metamorphic grades of H chondrite that I examined in this study. The concentration ratios simply did not meet the criterion of having a metal to silicate concentration ratio (D-value) of 10 4 for all HSE established by O neill et al. Given this evidence, I must reject hypothesis A and accept the null hypothesis as true, which is that there is no trend in increasing D-value with increasing metamorphic grade for H chondrite meteorites. It is unclear what causes these concentrations to deviate from the requirements for chemical equilibration between metal and silicate reservoirs. Shock metamorphism causing an open system would certainly allow HSE to migrate into the silicates and explain the high concentration of HSE in silicate relative to metals. 18

Bibliography Ahrens, et al, 1989 Formation of atmospheres during accretion of the terrestrial planets, in Origin and Evolution of Planetary and Satellite Atmospheres, S.K. Atreya, et al, editors, University of Arizona press, Tucson. Brown, K.L., and G.W. Tautfest. "Faraday Cup Monitors for High Energy Electron Beams." AIP Review of Scientific Instruments 27.9 (1956): Faraday Cup Monitors for High Energy Electron Beams. American Institute of Physics, 29 Dec. 2004. Web. 26 Apr. 2015. Campbell, Andrew J., and Munir Humayun. "Formation of Metal in Grosvenor Mountains 95551 and Comparison to Ordinary Chondrites." Geochimica Et Cosmochimica Acta 67.13 (2003): 2481-495. Horan, M. "Highly Siderophile Elements in Chondrites." Chemical Geology 196.1-4 (2003): 27-42. Horan, M.f., C.m.o D. Alexander, and R.j. Walker. "Highly Siderophile Element Evidence for Early Solar System Processes in Components from Ordinary Chondrites." Geochimica Et Cosmochimica Acta 73.22 (2009): 6984-997. IGL HSE Manual (2012) Kessel, Ronit, John R. Beckett, and Edward M. Stolper. "The Thermal History of Equilibrated Ordinary Chondrites and the Relationship between Textural Maturity and Temperature." Geochimica Et Cosmochimica Acta 71.7 (2007): 1855-881. Web. McSween, Harry Y. Meteorites and Their Parent Planets. Cambridge: Cambridge UP, 1999. Print. "Meteoritical Bulletin: Entry for Faucett." The Meteoritical Society. Lunar and Planetary Institute, n.d. Web. 15 Apr. 2015. Norton, O. Richard (2002). The Cambridge Encyclopedia of Meteorites. Cambridge: Cambridge University Press. pp. 121 124. The Nu Plasma II MC-ICP-MS: Neodymium and Hafnium Isotope Ratio Measurements O'neill, H.st.c., D.b. Dingwell, A. Borisov, B. Spettel, and H. Palme. "Experimental Petrochemistry of Some Highly Siderophile Elements at High Temperatures, and Some Implications for Core Formation and the Mantle's Early History." Chemical Geology 120.3-4 (1995): 255-73. 19

Rambaldi, E. "Trace Element Content of Metals from H- and LL-group Chondrites." Earth and Planetary Science Letters 36.2 (1977): 347-58. Scott, E.R.D., and A.N. Krott. "Chondrites and Their Components." Treatise on Geochemistry. Vol. 1. San Diego :: Pergamon Imprint, 2003. 143-200. White, William M. Geochemistry. 1st ed. West Sussex, UK: Wiley-Blackwell, 2013. Print. Wood, John A. "Planetary Science: Of Asteroids and Onions." Nature 422.6931 (2003): 479-81. Web. 20