Search

We explore the possibility that the short-lived radionuclides $^{26}$Al, $^{60}$Fe, $^{107}$Pd, and $^{182}$Hf inferred to be present in the proto-solar cloud originated from $3-8 M_{\odot}$ Asymptotic Giant Branch (AGB) stars. Models of AGB stars with initial mass above 5$M_{\odot}$ are prolific producers of $^{26}$Al owing to hot bottom burning (HBB). In contrast, $^{60}$Fe, $^{107}$Pd, and $^{182}$Hf are produced by neutron captures: $^{107}$Pd and $^{182}$Hf in models $\lesssim 5 M_{\odot}$; and $^{60}$Fe in models with higher mass. We mix stellar yields from solar-metallicity AGB models into a cloud of solar mass and composition to investigate if it is possible to explain the abundances of the four radioactive nuclides at the Sun's birth using one single value of the mixing ratio between the AGB yields and the initial cloud material. We find that AGB stars that experience efficient HBB ($\geq 6 M_{\odot}$) cannot provide a solution because they produce too little $^{182}$Hf and $^{107}$Pd relative to $^{26}$Al and $^{60}$Fe. Lower-mass AGB cannot provide a solution because they produce too little $^{26}$Al relative to $^{107}$Pd and $^{182}$Hf. A self-consistent solution may be found for AGB stars with masses in-between ($4-5.5 M_{\odot}$), provided HBB is stronger than in our models and the $^{13}$C($\alpha$, n)$^{16}$O neutron source is mildly activated. If stars of M $< 5.5 M_{\odot}$ are the source of the radioactive nuclides, then some basis for their existence in proto-solar clouds needs to be explored, given that the stellar lifetimes are longer than the molecular cloud lifetimes., Comment: 10 pages; accepted for publication in ApJ

We present a detailed characterization of the mineralogy, petrology, and oxygen isotopic compositions of twelve FUN CAIs, including C1 and EK1-4-1 from Allende (CV), that were previously shown to have large isotopic fractionation patterns for magnesium and oxygen, and large isotopic anomalies of several elements. The other samples show more modest patterns of isotopic fractionation and have smaller but significant isotopic anomalies. All FUN CAIs studied are coarse-grained igneous inclusions: Type B, forsterite-bearing Type B, compact Type A, and hibonite-rich. Some inclusions consist of two mineralogically distinct lithologies, forsterite-rich and forsterite-free/poor. All the CV FUN CAIs experienced postcrystallization open-system iron-alkali-halogen metasomatic alteration resulting in the formation of secondary minerals commonly observed in non-FUN CAIs from CV chondrites. The CR FUN CAI GG#3 shows no evidence for alteration. In all samples, clear evidence of oxygen isotopic fractionation was found. Most samples were initially ^(16)O-rich. On a three-oxygen isotope diagram, various minerals in each FUN CAI (spinel, forsterite, hibonite, dmisteinbergite, most fassaite grains, and melilite (only in GG#3)), define mass-dependent fractionation lines with a similar slope of ∼0.5. The different inclusions have different Δ^(17)O values ranging from ∼−25‰ to ∼−16‰. Melilite and plagioclase in the CV FUN CAIs have ^(16)O-poor compositions (Δ^(17)O ∼−3‰) and plot near the intercept of the Allende CAI line and the terrestrial fractionation line. We infer that mass-dependent fractionation effects of oxygen isotopes in FUN CAI minerals are due to evaporation during melt crystallization. Differences in Δ^(17)O values of mass-dependent fractionation lines defined by minerals in individual FUN CAIs are inferred to reflect differences in Δ^(17)O values of their precursors. Differences in δ^(18)O values of minerals defining the mass-dependent fractionation lines in several FUN CAIs are consistent with their inferred crystallization sequence, suggesting these minerals crystallized during melt evaporation. In other FUN CAIs, no clear correlation between δ^(18)O values of individual minerals and their inferred crystallization sequence is observed, possibly indicating gas-melt back reaction and oxygen-isotope exchange in a ^(16)O-rich gaseous reservoir. After oxygen-isotope fractionation, some FUN CAIs could have experienced partial melting and gas-melt oxygen-isotope exchange in a ^(16)O-poor gaseous reservoir that resulted in crystallization of ^(16)O-depleted fassaite, melilite and plagioclase. The final oxygen isotopic compositions of melilite and plagioclase in the CV FUN CAIs may have been established on the CV parent asteroid as a result of isotope exchange with a ^(16)O-poor fluid during hydrothermal alteration. We conclude that FUN CAIs are part of a general family of refractory inclusions showing various degrees of fractionation effects due to evaporative processes superimposed on sampling of isotopically heterogeneous material. These processes have been experienced both by FUN and non-FUN igneous CAIs. Generally, the inclusions identified as FUN show larger isotope fractionation effects than non-FUN CAIs. There is a wide spread in UN isotopic anomalies in a large number of CAIs not exhibiting large fractionation effects in oxygen, magnesium, and silicon. The question of why some FUN CAIs show more extreme UN isotopic effects is attributed by us to limited sampling and not a special source of isotopically anomalous material. We consider the majority of igneous CAIs to be the result of several stages of thermal processing (evaporation, condensation, and melting) of aggregates of solid precursors composed of incompletely isotopically homogenized materials. The unknown nuclear effects in CAIs are common to both FUN and non-FUN CAIs, and are not a special characteristic of FUN inclusions but represent the spectrum of results from sampling a very heterogeneous medium in the accreting Solar System.

A model for the groundwater transport of naturally occurring U, Th, Ra, and Rn nuclides in the ^(238)U and ^(232)Th decay series is discussed. The model developed here takes into account transport by advection and the physico-chemical processes of weathering, decay, α-recoil, and sorption at the water-rock interface. It describes the evolution along a flowline of the activities of the ^(238)U and ^(232)Th decay series nuclides in groundwater. Simple sets of relationships governing the activities of the various species in solution are derived, and these can be used both to calculate effective retardation factors and to interpret groundwater data. For the activities of each nuclide, a general solution to the transport equation has been obtained, which shows that the activities reach a constant value after a distance ϰ_i, characteristic of each nuclide. Where ϰ_i is much longer than the aquifer length, (for ^(238)U, ^(234)U, and ^(232)Th), the activities grow linearly with distance. Where ϰ_i is short compared to the aquifer length, (for ^(234)Th, ^(230)Th, ^(228)Th, ^(228)Ra, and ^(224)Ra), the activities rapidly reach a constant or quasi-constant activity value. For ^(226)Ra and ^(222)Rn, the limiting activity is reached after 1 km. High δ ^(234)U values (proportional to the ratio ^(ɛ234)Th/^(W238)U) can be obtained through high recoil fraction and/or low weathering rates. The activity ratios ^(230)Th/^(232)Th, ^(228)Ra/^(226)Ra and ^(224)Ra/^(226)Ra have been considered in the cases where either weathering or recoil is the predominant process of input from the mineral grain. Typical values for weathering rates and recoil fractions for a sandy aquifer indicate that recoil is the dominant process for Th isotopic ratios in the water. Measured data for Ra isotope activity ratios indicate that recoil is the process generally controlling the Ra isotopic composition in water. Higher isotopic ratios can be explained by different desorption kinetics of Ra. However, the model does not provide an explanation for ^(228)Ra/^(226)Ra and ^(224)Ra/^(226)Ra activity ratios less than unity. From the model, the highest ^(222)Rn emanation equals 2_ɛ. This is in agreement with the hypothesis that ^(222)Rn activity can be used as a first approximation for input by recoil (Krishnaswamiet al 1982). However, high ^(222)Rn emanation cannot be explained by production from the surface layer as formulated in the model. Other possibilities involve models including surface precipitation, where the surface layer is not in steady-state.

A steady state upper mantle model for the rare gases has been constructed which explains the available observational data of mantle He, Ne, Ar, and Xe isotope compositions and provides specific predictions regarding the rare gas isotopic compositions of the lower mantle, subduction of rare gases, and mantle rare gas concentrations. The model incorporates two mantle reservoirs; an undegassed lower mantle (P) and a highly degassed upper mantle (D). Chemical species are transferred into D within mass flows from P at plumes and from the atmosphere by subduction. Rare gases in D are derived from mixing of these inflows with radiogenic nuclides produced in situ. The upper mantle is degassed at mid‐ocean ridges and hotspots. Flows of each isotope into D are balanced by flows out of D, so that upper mantle concentrations are in steady state. Rare gases with distinct ^3He/^4He, ^(20)Ne/^(22)Ne, ^(129)Xe/^(130)Xe, and ^(136)Xe/^(130)Xe are stored in P and are transferred into D. In P, isotopic shifts are due to decay of U‐ and Th‐ decay series nuclides, ^(40)K, ^(129)I, and ^(244)Pu over 4.5 Ga. Radiogenic ^(136)Xe in P is dominantly from ^(244)Pu. In the well‐outgassed D reservoir, additional isotopic shifts are due to decay of U‐ and Th‐ series nuclides and ^(40)K over a residence time of ∼1.4 Ga. Since ^4He, ^(21)Ne, ^(40)Ar, and ^(136)Xe are produced in proportions fixed by nuclear parameters, the resulting isotopic shifts are correlated. The model predicts that the shift in ^(21)Ne/^(22)Ne in D relative to that in P is the same as that for ^4He/^3He in the respective mantle reservoirs. This is compatible with the available data for MORB and hotspots. The minimum ^(40)Ar/^(36)Ar, ^(129)Xe/^(130)Xe, and ^(136)Xe/^(130)Xe ratios in P are found to be substantially greater than the atmospheric ratios. The range in Ne, Ar, and Xe isotopes measured in MORB are interpreted as reflecting contamination of mantle rare gases by variable proportions of atmospheric rare gases. Subduction is not significant for He and Ne, but may account for a substantial fraction of Ar and Xe in D. The rare gas relative abundances in P are different than that of the atmosphere and are consistent with possible early solar system reservoirs as found in meteorites. The ^3He/^(22)Ne and ^(20)Ne/^(36)Ar ratios of P are within the range for meteorites with ‘solar’ Ne isotope compositions. The ^(130)Xe/^(36)Ar ratio of the lower mantle is greater than that of the atmosphere and may be as high as the ratio found for meteoritic ‘planetary’ rare gases. In the model, atmospheric rare gas isotope compositions are distinct from those of the mantle. If the Earth originally had uniform concentrations of rare gases, degassing of the upper mantle would have provided only a small proportion of the nonradiogenic rare gases presently in the atmosphere. The remainder was derived from late‐accreted material with higher concentrations of rare gases. However, radiogenic ^(129)Xe and ^(136)Xe abundances imply a substantial loss of rare gases up to ∼10^8 years after meteorite formation either from the early Earth or from late‐accreting protoplanetary materials. Rare gases must have been lost during accretion and the moon‐forming impact, so that nonradiogenic rare gases in the atmosphere must have been supplied by subsequently accreted material with nonradiogenic Xe, possibly from comets. Fractionation of atmospheric Xe isotopes relative to other early solar system components occurred either on late‐accreting materials or during loss from the Earth.

The death of Ernst Kunibert Zinner brings great sadness to the community of cosmochemists and astrophysicists. He pioneered the isotopic and chemical study of grains formed in ejecta from very diverse stars that exploded before the Solar System formed. The identification of the chemical compounds, which he found in grains of circumstellar condensates, and their isotopic compositions has permitted great advances in our understanding of stellar nucleosynthesis. Ernst Zinner was Research Professor at Washington University, where he was supported by research grants from NASA and some from the NSF.

There is variability in the Mg isotopic composition that is a reflection of the widespread heterogeneity in the isotopic composition of the elements in the solar system at approximately 100 ppm. Measurements on a single calcium-aluminum-rich inclusion (CAI) gave a good correlation of ^(26)Mg/^(24)Mg with ^(27)Al/^(24)Mg, yielding an isochron corresponding to an initial (^(26)Al/^(27)Al)_o = (5.27 ± 0.18) × 10^(−5) and an initial (^(26)Mg/^(24)Mg)_o = −0.127 ± 0.032‰ relative to the standard. This isochron is parallel to that obtained by Jacobsen et al. (2008), but is distinctively offset. This demonstrates that there are different initial Mg isotopic compositions in different samples with the same ^(26)Al/^(27)Al. No inference about uniformity/heterogeneity of ^(26)Al/^(27)Al on a macro scale can be based on the initial (^(26)Mg/^(24)Mg)_o values. Different values of ^(26)Al/^(27)Al for samples representing the same point in time would prove heterogeneity of ^(26)Al/^(27)Al. The important issue is whether the bulk solar inventory of ^(26)Al/^(27)Al was approximately 5 × 10^(−5) at some point in the early solar system. We discuss ultra refractory phases of solar type oxygen isotope composition with ^(26)Al/^(27)Al from approximately 5 × 10^(−5) to below 0.2 × 10^(−5). We argue that the real issues are: intrinsic heterogeneity in the parent cloud; mechanism and timing for the later production of ^(16)O-poor material; and the relationship to earlier formed ^(16)O-rich material in the disk. ^(26)Al-free refractories can be produced at a later time by late infall, if there is an adequate heat source, or from original heterogeneities in the placental molecular cloud from which the solar system formed.

It is known that there must be some weak form of transport (called cool bottom processing, or CBP) acting in low mass RGB and AGB stars, adding nuclei, newly produced near the hydrogen-burning shell, to the convective envelope. We assume that this extra-mixing originates in a stellar dynamo operated by the differential rotation below the envelope, maintaining toroidal magnetic fields near the hydrogen-burning shell. We use a phenomenological approach to the buoyancy of magnetic flux tubes, assuming that they induce matter circulation as needed by CBP models. This establishes requirements on the fields necessary to transport material from zones where some nuclear burning takes place, through the radiative layer, and into the convective envelope. Magnetic field strengths are determined by the transport rates needed by CBP for the model stellar structure of a star of initially 1.5 solar mass, in both the AGB and RGB phases. The field required for the AGB star in the processing zone is B_0 ~ 5x10^6 G; at the base of the convective envelope this yields an intensity B_E < 10^4 G (approximately). For the RGB case, B_0 ~ 5x10^4 to 4x10^5 G, and the corresponding B_E are ~ 450 to 3500 G. These results are consistent with existing observations on AGB stars. They also hint at the basis for high field sources in some planetary nebulae and the very large fields found in some white dwarfs. It is concluded that transport by magnetic buoyancy should be considered as a possible mechanism for extra mixing through the radiative zone, as is required by both stellar observations and the extensive isotopic data on circumstellar condensates found in meteorites., 26 pages, 4 figures, accepted by Astrophysical Journal

Many bodies of data are examined in this monograph; the complexity of this data is evident, and the difficulties of resolving events is very great. The crustal history of the earth is complex, and our understanding of the mechanisms of both tectonism of the emplacement and mobilization of rocks is poor. Nevertheless, since the ultimate goal of the science is an understanding of crustal history, it is important to speculate and consider various ideas that treat of these processes.

(Abridged) We review abundances of short-lived nuclides in the early solar system (ESS) and the methods used to determine them. We compare them to the inventory for a uniform galactic production model. Within a factor of two, observed abundances of several isotopes are compatible with this model. I-129 is an exception, with an ESS inventory much lower than expected. The isotopes Pd-107, Fe-60, Ca-41, Cl-36, Al-26, and Be-10 require late addition to the solar nebula. Be-10 is the product of particle irradiation of the solar system as probably is Cl-36. Late injection by a supernova (SN) cannot be responsible for most short-lived nuclei without excessively producing Mn-53; it can be the source of Mn-53 and maybe Fe-60. If a late SN is responsible for these two nuclei, it still cannot make Pd-107 and other isotopes. We emphasize an AGB star as a source of nuclei, including Fe-60 and explore this possibility with new stellar models. A dilution factor of about 4e-3 gives reasonable amounts of many nuclei. We discuss the role of irradiation for Al-26, Cl-36 and Ca-41. Conflict between scenarios is emphasized as well as the absence of a global interpretation for the existing data. Abundances of actinides indicate a quiescent interval of about 1e8 years for actinide group production in order to explain the data on Pu-244 and new bounds on Cm-247. This interval is not compatible with Hf-182 data, so a separate type of r-process is needed for at least the actinides, distinct from the two types previously identified. The apparent coincidence of the I-129 and trans-actinide time scales suggests that the last actinide contribution was from an r-process that produced actinides without fission recycling so that the yields at Ba and below were governed by fission., 92 pages, 14 figure files, in press at Nuclear Physics A

Late-formed halogen-rich phases in a refractory inclusion and a chondrule from the Allende meteorite exhibit large S-36 excesses that linearly correlate with the chlorine concentration, providing strong evidence in support of the existence of the short-lived nuclide Cl-36 (mean life of 0.43 Myr) in the early solar system. The inferred Cl-36/Cl-35 ratios at the time when these phases formed are very high (similar to 4 x 10(-6)) and essentially the same for the inclusion and the chondrule and confirm the earlier report of S-36 excess in another meteorite. In addition, the Cl-36 is decoupled from Al-26. The observed and any possible higher levels of Cl-36 cannot be the result of a supernova or AGB stellar source but require a late episode of energetic particle bombardment by the early Sun, in support of the arguments based on the previous discovery of Be-10. It is now clear that a blend of several sources is required to explain the short-lived nuclei when the solar system formed.

Major element concentrations and ^(87)Sr/^(86)Sr ratios were measured in groundwaters and bedrock from the Madison aquifer in western South Dakota. In this region, the Madison aquifer is primarily comprised of dolomite belonging to the Madison Limestone Group. The purpose of the study was to investigate controls on the downgradient evolution of dissolved Sr^(2+) in a carbonate groundwater system that is recharged by waters with high ^(87)Sr/^(86)Sr ratios draining Precambrian basement rocks and to establish the sources of Sr^(2+) added to the waters by reaction with the aquifer lithology. A mass-balance model following previous workers was used to calculate amounts and effective rates of mineral dissolution and precipitation during groundwater transport along a ∼240-km flow path. Both the calculated reaction rates and data for the Sr isotope geochemistry of the reacting phases were then used to develop a self-consistent and quantitative description of the concentration and isotopic composition of dissolved Sr^(2+) in the aquifer waters. The major ion chemistry of Madison aquifer groundwater is known to evolve according to dolomite dissolution, anhydrite dissolution, calcite precipitation, and ion-exchange with clay minerals. Dissolved ^(87)Sr/^(86)Sr ratios in the Madison aquifer decrease downgradient. Input waters draining the igneous Black Hills have ^(87)Sr/^(86)Sr ratios of ∼0.723, while highly evolved waters in the aquifer have ^(87)Sr/^(86)Sr ratios of ∼0.708. Dissolved Sr^(2+) concentrations undergo a concurrent increase from ∼1200 to 66,000 nmol/l. Model results indicate that dolomite dissolution exerts a critical control on the major ion chemistry but is not the primary source of nonradiogenic Sr^(2+), as both the dissolution rate and Sr concentration of dolomite are very low. Anhydrite is readily soluble in water, has a low ^(87)Sr/^(86)Sr ratio (∼0.708), and a very high Sr concentration (∼50,000 nmol/g). Anhydrite is also greatly undersaturated in the groundwaters, and the downgradient evolution of Sr^(2+) accompanies an ∼80-fold increase in dissolved SO_4^(2−). While anhydrite has a very low abundance in the aquifer rocks, the flow model indicates that anhydrite dissolution provides ∼300 times more Sr^(2+) per liter of water relative to dolomite dissolution. These findings suggest that anhydrite dissolution governs the Sr^(2+) geochemistry of Madison aquifer groundwaters, whereas dolomite dissolution and calcite precipitation control the bulk chemistry. Lastly, it is shown that the calculation of net mineral masses transferred to solution requires very high relative proportions of anhydrite and clay to dolomite. Mineral masses obtained by this approach actually represent bulk contributions without consideration of the intrinsic reaction rates of the phases or their modal abundances in the aquifer rocks. However, consideration of these parameters in a transport equation for flow through porous media shows that the high apparent abundances are not a real requirement. This latter approach is consistent with laboratory experiments and indicates that as little as 0.04 wt.% of very soluble anhydrite and 0.3 wt.% of exchangeable clay in the aquifer rock are sufficient to produce the observed groundwater chemistry in most of the flow path.

We report on a dedicated effort to identify and study metal-poor stars strongly enhanced in r-process elements ([r/Fe] > 1 dex; hereafter r-II stars), the Hamburg/ESO R-process Enhanced Star survey (HERES). Moderate-resolution (~2A) follow-up spectroscopy has been obtained for metal-poor giant candidates selected from the Hamburg/ESO objective-prism survey (HES) as well as the HK survey to identify sharp-lined stars with [Fe/H] < -2.5dex. For several hundred confirmed metal-poor giants brighter than B~16.5mag (most of them from the HES), ``snapshot'' spectra (R~20,000; S/N~30 per pixel) are being obtained with VLT/UVES, with the main aim of finding the 2-3% r-II stars expected to be among them. These are studied in detail by means of higher resolution and higher S/N spectra. In this paper we describe a pilot study based on a set of 35 stars, including 23 from the HK survey, 8 from the HES, and 4 comparison stars. We discovered two new r-II stars, CS29497-004 ([Eu/Fe] = 1.64 +/- 0.22) and CS29491-069 ([Eu/Fe] = 1.08 +/- 0.23). A first abundance analysis of CS29497-004 yields that its abundances of Ba to Dy are on average enhanced by 1.5dex with respect to iron and the Sun and match a scaled solar r-process pattern well, while Th is underabundant relative to that pattern by 0.3dex, which we attribute to radioactive decay. That is, CS29497-004 seems not to belong to the class of r-process enhanced stars displaying an ``actinide boost'', like CS31082-001 (Hill et al. 2002), or CS30306-132 (Honda et al. 2004b). The abundance pattern agrees well with predictions of the phenomenological model of Qian & Wasserburg., Accepted for publication in A&A

The isotopic composition of Ca (Δ^(44)Ca/^(40)Ca) in calcite crystals has been determined relative to that in the parent solutions by TIMS using a double spike. Solutions were exposed to an atmosphere of NH_3 and CO_2, provided by the decomposition of (NH_4)_2CO_3, following the procedure developed by previous workers. Alkalinity, pH and concentrations of CO_3^(2−), HCO_3^−, and CO_2 in solution were determined. The procedures permitted us to determine Δ(^(44)Ca/^(40)Ca) over a range of pH conditions, with the associated ranges of alkalinity. Two solutions with greatly different Ca concentrations were used, but, in all cases, the condition [Ca^(2+)]>>[CO_3^(2−)] was met. A wide range in Δ(^(44)Ca/^(40)Ca) was found for the calcite crystals, extending from 0.04 ± 0.13‰ to −1.34 ± 0.15‰, generally anti-correlating with the amount of Ca removed from the solution. The results show that Δ(^(44)Ca/^(40)Ca) is a linear function of the saturation state of the solution with respect to calcite (Ω). The two parameters are very well correlated over a wide range in Ω for each solution with a given [Ca]. The linear correlation extended from Δ(^(44)Ca/^(40)Ca) = −1.34 ± 0.15‰ to 0.04 ± 0.13‰, with the slopes directly dependent on [Ca]. Solutions, which were vigorously stirred, showed a much smaller range in Δ(^(44)Ca/^(40)Ca) and gave values of −0.42 ± 0.14‰, with the largest effect at low Ω. It is concluded that the diffusive flow of CO_3^(2−) into the immediate neighborhood of the crystal-solution interface is the rate-controlling mechanism and that diffusive transport of Ca^(2+) is not a significant factor. The data are simply explained by the assumptions that: a) the immediate interface of the crystal and the solution is at equilibrium with Δ(^(44)Ca/^(40)Ca) ∼ −1.5 ± 0.25‰; and b) diffusive inflow of CO_3^(2−) causes supersaturation, thus precipitating Ca from the regions exterior to the narrow zone of equilibrium. The result is that Δ(^(44)Ca/^(40)Ca) is a monotonically increasing (from negative values to zero) function of Ω. We consider this model to be a plausible explanation of most of the available data reported in the literature. The well-resolved but small and regular isotope fractionation shifts in Ca are thus not related to the diffusion of very large hydrated Ca complexes, but rather due to the ready availability of Ca in the general neighborhood of the crystal-solution interface. The largest isotopic shift which occurs as a small equilibrium effect is then subdued by supersaturation precipitation for solutions where [Ca^(2+)]>>[CO_3^(2−)] + [HCO_3^−]. It is shown that there is a clear temperature dependence of the net isotopic shifts that is simply due to changes in Ω due to the equilibrium “constants” dependence on temperature, which changes the degree of saturation and hence the amount of isotopically unequilibrated Ca precipitated. The effects that are found in natural samples, therefore, will be dependent on the degree of diffusive inflow of carbonate species at or around the crystal-liquid interface in the particular precipitating system, thus limiting the equilibrium effect.

A study of lateritic soils and samples of ground and river waters was carried out in the Nsimi-Zoetele, a tropical watershed in the southern Cameroon. The Nd isotopic compositions and concentrations of Nd and Sm were determined. It was found that the Nd isotopic composition of the river waters was much more radiogenic than the parent rocks, and that the Nd in the waters is not homogeneous but is carried by different dissolved and complexed components that are not isotopically homogenized. The soil profile shows a regular increase in e_(Nd) going from the parent rock (e_(Nd) = −36) to e_(Nd) = −18 near the top of the profile. The Nd transported in the river is thus not representative of the parent rock but reflects the results of differential weathering of constituent minerals and the redeposition of REE in phosphates and a significant contribution of radiogenic Nd from dust. The concentration of Nd in the river water is far above that found in temperate climate rivers and thus this type of tropical river may play a dominant role in the marine Nd and REE budget. It is suggested that the correlation of REE with DOC is related to DOC fixing some dissolved REE but that the REE in solution is governed by other mechanisms. No major shifts were found in Sm/Nd; however, a regular progression from the parent rock through the lateritic profile was found. The upper laterite profile shows large, almost uniform depletions in all REE below Tb and enrichment above. Complementary behavior was found in the lower part of the section. The concentration of Nd relative to the immobile elements Zr and Ti in the laterite is depleted by a factor of ∼10. Th, Nd and Sm are enriched in the lowest zone sampled and must reflect redeposition of REE from the upper part of the weathering section and is associated with phosphate formation. It is concluded that the soil evolution involves both differential dissolution of primary phases from the parent rock, significant to major input of REE from atmospheric dust from other regions, and the formation of diagenetic phases, particularly phosphates.

The possible stellar origin of those short-lived nuclei that were shown to be alive in the Early Solar System (ESS) is briefly discussed. The applicability of a close-by AGB star polluting the protosolar nebula depends on the extent to which several nuclei can be explained and in particular the strongly debated problems of the initial amount of ^(60)Fe, as inferred from recent measurements.

We present a model for chemical evolution of damped Ly alpha systems considering production of metals by SNe II and infall associated with hierarchical structure formation. The growth of metallicity in these systems is a reflection of the competition between astration and infall. The apparent late turn-on of these systems is due to the late cut-off of infall. The wide range in [Fe/H] at a given redshift is explained by the range of the times for onset of star formation and the range of the times for infall cessation in different systems. The observed lower bound of [Fe/H] = -3 follows from the very rapid initial rise of [Fe/H] subsequent to onset of star formation. To reach [Fe/H] = -3 from a metal-free initial state requires only about 30 Myr so that the probability of observing lower [Fe/H] values is very small., 4 pages, 2 figures, to appear in ApJL

A comprehensive evaluation of the transport of U, Th, Ra, and Rn nuclides of the ^(238)U- and ^(232)Th-decay series in an unconfined sandy aquifer (Long Island, NY) was conducted. Groundwater data are compared with results of a theoretical transport model of weathering of aquifer grains and interaction with surface coatings to establish relationships between the concentrations of the radionuclide activities in the water and flow line distance. The data provide estimates for geochemical parameters including weathering rates and chemical reactivities in both the vadose zone and the aquifer. A theoretical treatment of the transport is presented that considers the reaction between the water and a reactive surface layer. It is found that a model with chemical exchange between the surface layer and the water for all species is not valid, and that the effects of saturation and “irreversible” precipitation of Th is required. The water table shows a relatively wide range in U activities, the only element in the U-Th series for which vadose zone input is significant in the aquifer. High weathering of U and recoil inputs of ^(234)U to the water occur in the upper 3 m of the vadose zone, while lower weathering and removal of U from the water occur below. The deeper aquifer has variable ^(238)U activities that can be accounted for by input from the vadose zone and is not a result of non-conservative behavior. The isotopic composition of U is shown to be directly related to the recoil rate relative to the weathering rate. The wide range of ^(238)U in the aquifer waters is a reflection of diverse vadose zone inputs, showing that dispersive mixing is not a dominant effect. The higher values of δ^(234)U in the aquifer reflect the recoil/weathering input ratios from within the aquifer where the weathering rate is lower than the vadose zone. Both high U activities and high δ^(234)U values cannot be obtained in the vadose zone or within reasonable flow distances in the aquifer. Radium isotopes are found to be in exchange equilibrium with the surface layer. ^(224)Ra, ^(228)Ra, and ^(226)Ra have comparable activities throughout the aquifer. In the vadose zone, the dominant input of Ra to groundwater is weathering and recoil. As found elsewhere, the ^(222)Rn in the water is a large fraction (∼5%) of the Rn produced in the aquifer rock. This cannot be due to Ra precipitation onto surface coatings in the aquifer as supported by present weathering with Th in exchange equilibrium with the surface layer. It is found that Th is saturated in the waters under oxidizing conditions so that the weathering input is irreversibly precipitating onto surfaces. However, it is shown that under somewhat reducing conditions, Th activities are much higher and the Th/U ratio in the solution is approximately that of the rock. We propose that under oxidizing conditions the source of Rn is a surface coating enriched in ^(232)Th and ^(230)Th. This Th was precipitated in an earlier phase during rapid dissolution of readily weathered phases that contain ∼10% of the U-Th inventory of the rock, with the associated U carried away in solution. Therefore, the previously precipitated ^(230)Th and ^(232)Th produce daughter nuclides in the surface coating which are the dominant contributors of Ra and Rn to the ground water. In particular, Rn is provided by very efficient losses (by diffusion or recoil) from the surface coating. This then does not require recent, large recoil losses from the parent rock or the presence of nanopores in the rock. The first data of both long-lived ^(232)Th and short-lived ^(234)Th and ^(228)Th in ground water is reported. The Th isotope activities indicate that desorption kinetics are slow and provide the first estimate, based on field data, of the Th desorption rate from an aquifer surface. The mean residence time of Th in the surface coating is ∼3000 y while in the water it is ∼1 h. Ra is in partition equilibrium with the aquifer surface layer. However, the strong fixation of Th on surface coatings is very susceptible to changes in oxidation state as is shown by a comparison of two adjacent aquifers. This makes it difficult to define with certainty the retentive characteristics in natural systems. In general, it is shown that the distributions of naturally occurring nuclides can be used to calculate values for transport parameters that are applicable to the transport of anthropogenic nuclides.

When asked to write a summary of my professional life, I had no notion of how difficult it would be. There is the form and the substance, the remembrances and the reality, the discipline and the hope, the frustration of not finding a way, and the satisfaction of finding a trail. Of all of the experiences, the greatest one, the biggest turn-on, is the rare feeling that one has some understanding of nature. I have had the honor and privilege of receiving the Crafoord Prize of the Royal Swedish Academy from the King of Sweden, but my greatest occasions of excitement came when I thought I had possibly understood something. The greatest satisfaction comes in feeling that the work I have done or contributed to has some beauty. At the moment, I am finishing up a paper on groundwater transport in the Ojo Alamo aquifer, having just finished a paper with C.T. Lee and Frank Kyte on the chemistry and transport of PGE in the oceans. Just weeks earlier, I returned from an astrophysics conference on “Low Z at low z and high z,” (low chemical elemental abundances at low and high cosmological red shifts) led by Y.-Z. Qian at the University of Minnesota. It has been said that the virtue of being “old” is that you can live in the past as well as the present. Thinking about the past is not my way, except to tell old stories over good wines at dinner with friends. However, the truth is that the particular problem that I am working on at any instant is, to me, the most important thing in the world. Chasing some particular idea and some observations of nature make up the real present and govern my immediate future. In the chase, I met Devendra and Aruna Lal. He was fascinated with Indian puzzles and natural puzzles, particularly those cosmic-ray induced. Our mutual interest in natural puzzles has been a continuous source of entertainment and mutual support. Friendly interaction and critical comments on science and life and what might be true comes regularly from Karl Turekian (a Columbia product displaced to Yale). It has been my privilege to have had an exciting, productive career that continues to the present day. It is my intent to continue the exploration and the chase. The reader is warned that this report is not chronological, as the actual science in different areas often occurred concurrently.

The nature of Re-platinum-group element (PGE; Pt, Pd, Ir, Os, Ru) transport in the marine environment was investigated by means of marine sediments at and across the Cretaceous-Tertiary boundary (KTB) at two hemipelagic sites in Europe and two pelagic sites in the North and South Pacific. A traverse across the KTB in the South Pacific pelagic clay core found elevated levels of Re, Pt, Ir, Os, and Ru, each of which is approximately symmetrically distributed over a distance of approx. 1.8 m across the KTB. The Re-PGE abundance patterns are fractionated from chondritic relative abundances: Ru, Pt, Pd, and Re contents are slightly subchondritic relative to Ir, and Os is depleted by approx. 95% relative to chondritic Ir proportions. A similar depletion in Os (approx. 90%) was found in a sample of the pelagic KTB in the North Pacific, but it is enriched in Ru, Pt, Pd, and Re relative to Ir. The two hemipelagic KTB clays have near-chondritic abundance patterns. The approx. 1.8-m-wide Re-PGE peak in the pelagic South Pacific section cannot be reconciled with the fallout of a single impactor, indicating that postdepositional redistribution has occurred. The elemental profiles appear to fit diffusion profiles, although bioturbation could have also played a role. If diffusion had occurred over approx. 65 Ma, the effective diffusivities are approx. 10(exp -13)sq cm/s, much smaller than that of soluble cations in pore waters (approx. 10(exp -5) sq cm/s). The coupling of Re and the PGEs during redistribution indicates that postdepositional processes did not significantly fractionate their relative abundances. If redistribution was caused by diffusion, then the effective diffusivities are the same. Fractionation of Os from Ir during the KTB interval must therefore have occurred during aqueous transport in the marine environment. Distinctly subchondritic Os/Ir ratios throughout the Cenozoic in the South Pacific core further suggest that fractionation of Os from Ir in the marine environment is a general process throughout geologic time because most of the inputs of Os and Ir into the ocean have OsAr ratios greater than or = 1. Mass balance calculations show that Os and Re burial fluxes in pelagic sediments account for only a small fraction of the riverine Os (less than 10%) and Re (less than 0.1%) inputs into the oceans. In contrast, burial of Ir in pelagic sediments is similar to the riverine Ir input, indicating that pelagic sediments are a much larger repository for Ir than for Os and Re. If all of the missing Os and Re is assumed to reside in anoxic sediments in oceanic margins, the calculated burial fluxes in anoxic sediments are similar to observed burial fluxes. However, putting all of the missing Os and Re into estuarine sediments would require high concentrations to balance the riverine input and would also fail to explain the depletion of Os at pelagic KTB sites, where at most approx. 25% of the K-T impactor's Os could have passed through estuaries. If Os is preferentially sequestered in anoxic marine environments, it follows that the OsAr ratio of pelagic sediments should be sensitive to changes in the rates of anoxic sediment deposition. There is thus a clear fractionation of Os and Re from Ir in precipitation out of sea water in pelagic sections. Accordingly, it is inferred here that Re and Os are removed from sea water in anoxic marine depositional regimes.

We have made aluminum-magnesium isotopic measurements on 4 melilite-bearing calcium-aluminum-rich inclusions (CAIs), 1 plagioclase-olivine inclusion (POI), and 2 ferromagnesian chondrules from the Ningqiang carbonaceous chondrite. All of the CAIs measured contain clear evidence for radiogenic ^(26)Mg^* from the decay of ^(26)Al (τ = 1.05 Ma). Although the low Al/Mg ratios of the melilites introduce large uncertainties, the inferred initial ^(26)Al/^(27)Al ratios for the CAIs are generally consistent with the value of 5 x 10^(−5). There is clear evidence of ^(26)Al^* in one POI and two chondrules, but with considerable uncertainties in the value of (^(26)Al/^(27)Al)0. The (^(26)Al/^(27)Al)_0 ratios for the POI and the chondrules are 0.3–0.6 x 10^(−5), roughly an order of magnitude lower than the canonical value. Ningqiang shows very little evidence of metamorphism as a bulk object and the (^(26)Al/^(27)Al)_0 ratios in its refractory inclusions and chondrules are consistent with those found in other unmetamorphosed chondrites of several different classes. Our observations and those of other workers support the view that ^(26)Al was widely and approximately homogeneously distributed throughout the condensed matter of the solar system. The difference in (^(26)Al/^(27)Al)0 between CAIs and less refractory materials seems reasonably interpreted in terms of a ∼2 million year delay between the formation of CAIs and the onset of formation of less refractory objects. The POI shows clear differences in ^(25)Mg/^(24)Mg between its constituent spinels and olivine, which confirms that they are partially reprocessed material from different sources that were rapidly quenched.

Using improved analytical techniques, which reduce the Re blanks by factors of 8 to 10, we report new Re-Os data on low Re and low PGE pallasites (PAL-anom) and IIIAB irons. The new pallasite samples nearly double the observed range in Re/Os for pallasites and allow the determination of an isochron of slope 0.0775 ± 0.0008 (T = 4.50 ± 0.04 Ga, using the adjusted λ^(187)Re = 1.66 × 10^(−11) a^(−1)) and initial (^(187)Os/^(188)Os)_0 = 0.09599 ± 0.00046. If the data on different groups of pallasites (including the “anomalous” pallasites) are considered to define a whole-rock isochron, their formation would appear to be distinctly younger than for the iron meteorites by ∼60 Ma. Five IIIAB irons (Acuna, Bella Roca, Chupaderos, Grant, and Bear Creek), with Re contents ranging from 0.9 to 2.8 ppb, show limited Re/Os fractionation and plot within errors on the IIAB iron meteorite isochron of slope 0.07848 ± 0.00018 (T = 4.56 ± 0.01 Ga) and initial (^(187)Os/^(188)Os)_0 = 0.09563 ± 0.00011. Many of the meteorites were analyzed also for Pd-Ag and show ^(107)Ag enrichments correlated with Pd/Ag, requiring early formation and fractionation of the FeNi metal, in a narrow time interval, after injection of live ^(107)Pd (t_(1/2) = 6.5 Ma) into the solar nebula. Based on Pd-Ag, the typical range in relative ages of these meteorites is ≤10 Ma. The Pd-Ag results suggest early formation and preservation of the ^(107)Pd-^(107)Ag systematics, both for IIIAB irons and for pallasites, while the younger Re-Os apparent age for pallasites suggests that the Re-Os system in pallasites was subject to re-equilibration. The low Re and low PGE pallasites show significant Re/Os fractionation (higher Re/Os) as the Re and PGE contents decrease. By contrast, the IIIAB irons show a restricted range in Re/Os, even for samples with extremely low Re and PGE contents. There is a good correlation of Re and Ir contents. The correlation of Re and Os contents for IIIAB irons shows a similar complex pattern as observed for IIAB irons (Morgan et al., 1995), and neither can be ascribed to a continuous fractional crystallization process with uniform solid-metal/liquid-metal distribution coefficients.

The Fe-Ni metal within chondrites has been postulated to have originated either through condensation or as a byproduct of chondrule formation. To test these hypotheses, we studied metal in three Renazzo-like carbonaceous (CR2) chondrites from three petrographic settings: inside chondrules, on chondrule rims, and in the matrix. Abundances of Fe, Ni, Co, Cr, and P were determined in situ by electron microprobe, and those of Os, Ir, Pt, and Au were measured by a newly developed ion microprobe technique. The refractory platinum group elements Os, Ir, and Pt behave coherently in CR2 metal. They are either all enriched, all depleted, or unfractionated with respect to Fe and cosmic ratios. Metal with approximately CI Os/Fe, Ir/Fe, and Pt/Fe occur primarily in chondrule interiors. All metal grains have essentially CI values of Ni/Fe and Co/Fe. Almost all metal grains have lower-than-CI ratios of the volatile elements, Au and P. We also estimated the bulk compositions via analyses of phases and modal recombination of a subset of the chondrules whose metal we analyzed. The bulk compositions of chondrules are generally unfractionated relative to CI chondrites for elements more refractory than ∼Cr but are depleted in more volatile elements. Abundances of siderophile elements correlate strongly with the metal abundance in the chondrules, which implies that siderophile depletions are due to expulsion of metal from the chondrule melts. The metal most likely originated during melting via reduction of oxides by C that was part of the chondrule precursor. During chondrule heating, molten metal efficiently extracted siderophile elements from the silicate melt. Rim metal consists of two types. One is like metal in chondrule interiors and was in the process of being expelled when the chondrules were quenched. The other shows systematic depletions in Os, Ir, and Pt relative to Fe and higher concentrations of Au and P than interior metal. This metal is attributed to recondensation from a vapor depleted in refractory siderophiles, a vapor most likely derived via evaporation from chondrules. Large, isolated matrix metal grains comprise the same two groups as rim grains and have the same origins. Bulk chondrule compositions and siderophile abundance patterns in metal indicate that the precursors of CR2 chondrules had CI-like abundances of refractory and moderately volatile elements but was likely depleted in the more volatile elements.

The transport of Th, Ra, and ^(210)Pb from a continental source region and through an estuarine environment was investigated. Unlike previous studies, here both short- and long-lived nuclide data were obtained for river inputs, river water, and estuarine waters. The mire-rich Kalix River drainage basin was chosen as a typical northern shield area because this river may represent typical waters flowing into the Arctic and northern seas. Groundwaters from bedrock and glacial tills have comparable Th isotope concentrations and do not exhibit significant Th isotopic shifts relative to host rocks. The extensive peat deposits of the basin receive groundwater discharges and concentrate Th and U (but not Ba and Ra), which cause high ^(230)Th/^(232)Th ratios in mire waters. However, mire outflows do not have a significant impact on Th and Ra isotopic compositions of the river. Overall weathering characteristics for the basin are obtained from the river data. The ^(230)Th/^(232)Th, ^(228)Ra/^(226)Ra, and ^(226)Ra/Ba river ratios are comparable to those of source rocks, consistent with similar release rates of these nuclides from U-, Th-, and Ba-bearing minerals. River ratios of (^(230)Th/^(238)U)_(AR) and (^(226)Ra/^(238)U)_(AR) are

In this report we will try to summarize some key observations on isotopic abundances of radioactive nuclei in meteorites and their relationship to stellar sources. We will also cover the transition from the first generation of supermassive stars to the "normal" stellar population (including supernovae) in a galaxy. Two distinctive approaches are presented. The first is to utilize the observations on meteorites to infer the sources of radioactive nuclei in the solar system. We will show that some radioactive nuclides are clearly related to production in AGB stars, while others are produced in r-processes in SNII. We will show that there must be at least two distinct types of SNII sites with very different time scales for replenishing the ISM with different r-nuclides. This model leads us to a means of identifying the first generation of supernovae that contribute to the galactic inventory of r-process nuclei and of Fe. This is then related to observations of low metallicity stars and connects/disconnects more general r-process abundance patterns with/from Fe production and points to a first generation of stars having distinct characteristics. This approach is the view backward in time. The second approach uses the phenomenological model derived from the above considerations and applies it to the chronology evolving from Big Bang toward the present epoch. The latter approach proclaims an "absolute" time scale and what the evolutionary path of condensing, star-forming, baryonic matter should be. I am obviously going beyond my own capability in this approach, but it is at least interesting and may contain some seeds of truth. The assumption of a universal r-process that produces both Fe and r-nuclei in a coupled manner is in violation of the observations that the correlation of r-elements with Fe breaks down at [Fe/H] ≾ - 3 as observed in halo stars. Furthermore, the ab initio approaches to early star formation, chemical evolution, and the early universe, have not proven very successful so far. There has been no real success in establishing the relationship between the condensation of cold dark matter early in the universe (z ≳10?) with the aggregation of baryonic matter and the formation of the first generation of stars.

The recent discovery by Cayrel et al. of U in CS 31082-001 along with Os and Ir at greatly enhanced abundances but with [Fe/H]=-2.9 strongly reinforces the argument that there are at least two kinds of SNII sources for r-nuclei. One source is the high-frequency H events responsible for heavy r-nuclei (A>135) but not Fe. The H-yields calculated from data on other ultra-metal-poor stars and the sun provide a template for quantitatively predicting the abundances of all other r-elements. In CS 31082-001 these should show a significant deficiency at A, 5 pages, 1 figure

We present a three-component mixing model for the evolution of O abundance relative to Fe, taking into account the contributions of the first very massive (> 100 solar masses) stars formed from Big Bang debris. We show that the observations of O and Fe abundances in metal-poor stars in the Galaxy by Israelian et al. and Boesgaard et al. can be well represented both qualitatively and quantitatively by this model. Under the assumption of an initial Fe ([Fe/H] = -3) and O inventory due to the prompt production by the first very massive stars, the data at -3 < [Fe/H] < -1 are interpreted to result from the addition of O and Fe only from type II supernovae (SNII) to the prompt inventory. At [Fe/H] = -1, SNII still contribute O while both SNII and type Ia supernovae contribute Fe. During this later stage, (O/Fe) sharply drops off to an asymptotic value of 0.8(O/Fe)_sun. The value of (O/Fe) for the prompt inventory at [Fe/H] = -3 is found to be (O/Fe) = 20(O/Fe)_sun. This result suggests that protogalaxies with low ``metallicities'' should exhibit high values of (O/Fe). The C/O ratio produced by the first very massive stars is expected to be much less than 1 so that all the C should be tied up as CO and that C dust and hydrocarbon compounds should be quite rare at epochs corresponding to [Fe/H] < -3., 25 pages, 8 postscript figures, to appear in ApJ

Variable excesses of ^(36)S have previously been reported in sodalite in the Allende and Ningqiang meteorites and used to infer the presence of ^(36)Cl in the early solar system. Until now no unambiguous evidence of the major decay product, ^(36)Ar (98%), has been found. Using low fluence fast neutron activation we have measured small amounts of ^(36)Ar in the Allende sodalite Pink Angel, corresponding to ^(36)Cl/^(35)Cl = (1.9 ± 0.5) × 10^(−8). This is a factor of 200 lower than the highest value inferred from ^(36)S excesses in sodalite. High resolution I–Xe analyses confirm that the sodalite formed between 4561 and 4558 Ma ago. The core of Pink Angel sodalite yielded a precise formation age of 4559.4 ± 0.6 Ma. Deposition of sodalite containing live ^(36)Cl, seven million years or so after the formation of the CAI, appears to require a local production mechanism involving intense neutron irradiation within the solar nebula. The constraint imposed by the near absence of neutron induced ^(128)Xe is most easily satisfied if the ^(36)Cl were produced in a fluid precursor of the sodalite. The low level of ^(36)Ar could be accounted for as a result of residual in-situ ^(36)Cl decay, up to 1–2 Ma after formation of the sodalite, and/or later diffusive loss, in line with the low activation energy for Ar diffusion in sodalite.

The Allende calcium-rich inclusion (CAI) 5241 has been found to contain distinct initial ^(26)Al/^(27)Al in the three consecutive igneous zones that have been identified by extensive petrogenetic studies to have formed in three distinct crystallization events. The zones in order of sequence of formation from the petrologic observations are: (1) spinel-free islands (SFI) included in (2) a pyroxene–spinel-rich core (SRC) which in turn is included in (3) a melilite mantle (MM). The initial (^(26)Al/^(27)Al) values of these zones are respectively (4.6–5.0)×10^(−5), 4.3×10^(−5), and 3.3×10^(−5). It is argued that these distinct (^(26)Al/^(27)Al)0 values are not the result of metamorphism but reflect the relative times of formation by crystallization from melts. Relative to the canonical value of (^(26)Al/^(27)Al)0=5×10^(−5), we find the following chronology: t_(SFI)≈0, t_(SRC)≈10^5 yr and t_(MM)≈4×10^5 yr. The three-layer CAI is inferred to have sampled a reservoir with an initial uniform ^(26)Al/^(27)Al ratio that decreased in value due to radioactive decay. From these observations we conclude that we have resolved time differences of a few hundred thousand years at the very early stages of formation of the solar system. The SFI, SRC, and MM zones reflect sequential addition of molten CAI material which crystallized rapidly without seriously metamorphosing the previously formed material. These additions took place over a time of about 400 000 years. We believe that these sequential events are not compatible with condensation in a hot region of the solar nebula. It is proposed that a scenario involving stages of protoplanetary accretion and of melt generation in protoplanetary sites heated by ^(26)Al might provide a possible source for CAIs and chondrules.

Late Eocene microtektites and crystal-bearing microkrystites extracted from DSDP and ODP cores from the Atlantic, Pacific, and Indian oceans have been analyzed to address their provenance. A new analysis of Nd and Sr isotopic compositions confirms previous work and the assignment of the uppermost microtektite layer to the North American tektites, which are associated with the 35.5 Ma, 85 km diameter Chesapeake impact structure of Virginia, USA. Extensive major element and Nd and Sr isotopic analyses of the microkrystites from the lowermost layer were obtained. The melanocratic microkrystites from Sites 216 and 462 in the Indian and Pacific oceans possess major element chemistries, Sr and Nd isotopic signatures and Sm–Nd, T_(CHUR), model ages similar to those of tagamite melt rocks in the Popigai impact structure. They also possess Rb–Sr, T_(UR), model ages that are younger than the tagamite T_(CHUR) ages by up to ∼1 Ga, which require a process, as yet undefined, of Rb/Sr enrichment. These melanocratic microkrystites are consistent with a provenance from the 35.7 Ma, 100 km diameter Popigai impact structure of Siberia, Russia, while ruling out other contemporaneous structures as a source. Melanocratic microkrystites from other sites and leucocratic microkrystites from all sites possess a wide range of isotopic compositions (ϵ(^(143)Nd) values of −16 to −27.7 and ϵ(^(87)Sr) values of 4.1–354.0), making the association with Popigai tagamites less clear. These microkrystites may have been derived by the melting of target rocks of mixed composition, which were ejected without homogenization. Dark glass and felsic inclusions extracted from Popigai tagamites possess ϵ(^(143)Nd) and ϵ(^(87)Sr) values of −26.7 to −27.8 and 374.7 and 432.4, respectively, and T_(CHUR) and T_(UR) model ages of 1640–1870 Ma and 240–1830 Ma, respectively, which require the preservation of initially present heterogeneity in the source materials. The leucocratic microkrystites possess diverse isotopic compositions that may reflect the melting of supra-basement sedimentary rocks from Popigai, or early basement melts that were ejected prior to homogenization of the Popigai tagamites. The ejection of melt rocks with chemistries consistent with a basement provenance, rather than the surface ∼1 km of sedimentary cover rocks, atypically indicates a non-surficial source to some of the ejecta. Microkrystites from two adjacent biozones possess statistically indistinguishable major element compositions, suggesting they have a single source. The occurrence of microkrystites derived from a single impact event, but in different biozones, can be explained by: (1) diachronous biozone boundaries; (2) post-accumulation sedimentary reworking; or (3) erroneous biozonation.

The primary mineral assemblages and initial 26 Al/ 27 Al ratios of rare calcium-aluminum–rich inclusions (CAIs) from enstatite (E) chondrites are similar to those of CAIs from other chondrite classes. CAIs from all chondrite classes formed under oxidizing conditions that are much different from the reducing conditions under which the E chondrites formed. Either CAIs formed at an earlier, more oxidizing epoch in the region where E chondrites ultimately formed, or they formed at a different place in the solar nebula and were transported into the E chondrite formation region.

We apply the phenomenological model used to explain the abundances of Fe and r-process elements in very metal-poor stars in the Galaxy to [Fe/H] of damped Ly alpha systems. It is assumed that the first stars formed after the Big Bang were very massive and promptly enriched the interstellar medium to [Fe/H] ~ -3, at which metallicity formation of normal stars took over. Subsequent Fe enrichment was provided by Type II supernovae. The range of [Fe/H] at a given redshift z for damped Ly alpha systems is explained by the time t* after the Big Bang at which normal star formation started in an individual protogalactic system. The average t* is approx 80% the age of the universe for damped Ly alpha systems at z approx 1.5 to 4.5, indicating a long delay between the Big Bang and the turn-on of protogalaxies. It is inferred that a substantial fraction of the total baryonic matter may not have been aggregated into protogalaxies where normal star formation had occurred down to z ~ 1.5. The data near z = 2.2 suggest that the rate of turn-on of protogalaxies was initially very low and slowly reached a maximum at ~ 3 Gyr after the Big Bang. This may be important in understanding the rate of formation of quasars., Comment: 13 pages, 4 figures, to appear in ApJ Letters

We present Os data for axial high-temperature and off-axial low-temperature hydrothermal solutions from the Juan de Fuca Ridge. The high-temperature, H₂S-bearing axial fluids have unradiogenic Os isotopes pointing to a nearly complete domination of osmium isotopes from the basalts during hydrothermal circulation. The ridge axis fluids typically do not show large enrichment in osmium concentration over seawater although one high-temperature fluid has an Os concentration enhanced by a factor of 4 above seawater. It appears that the Os concentration of high-temperature hydrothermal fluids is typically buffered at roughly the seawater concentration. We suggest that subseafloor precipitation of pyrite from the high-temperature hydrothermal fluids controls the osmium transportation. The axial hydrothermal activity does not supply significant amounts of unradiogenic osmium to the deep oceans. In contrast, a low-temperature off-axis fluid is enriched in non-radiogenic osmium over seawater by a factor of 9, showing much less precipitation of osmium at low temperature. Because a large fraction of the cooling of oceanic lithosphere occurs on ridge flanks, Os from low-temperature, ridge-flank hydrothermal circulation may be a significant contributor to the balance of Os in the oceans. A detailed balance between Os contributions from dissolution of cosmic dust and hydrothermal fluids is still not possible. The low-temperature hydrothermal sample gives ¹⁸⁷Os/¹⁸⁸Os = 0.110 ± 0.001. This extremely unradiogenic osmium can only come from a source that underwent depletion of Re over 2.6 Ga ago. Assuming no contamination during sample collection, this result suggests that the convecting upper mantle contains ancient depleted material that imparted unradiogenic osmium to intruding basaltic melts.

We have developed a technique to measure abundances of Os, Ir, Pt, and Au in situ in Fe-Ni metal on a microscale using secondary-ion mass spectrometry. A Cs^+ beam is used to generate negative secondary ions, and interferences from molecular ions are eliminated by a combination of 40 eV of energy filtering and a mass resolution of ∼1900. Ion yields for Fe, Co, Ni, and the PGEs vary considerably from spot to spot in meteoritic metal samples. Ion yield variations for Os and Ir correlate strongly with the Fe ion yield, that for Pt is weakly correlated, and that for Au is essentially uncorrelated. From ion yields for meteorite standards and the correlations with Fe ion yield, it is possible to obtain concentrations of Os, Ir, Pt, and Au in meteorite samples. Using this technique, a Cs^+ beam current of 10 nA gives a spatial resolution of 10–20 μm and detection limits of less than 1 ppm for Os, less than 0.1 ppm for Ir, and 10–20 ppb for Pt and Au. With refinement, it should be possible to measure Rh, Pd, and Ag and to improve the spatial resolution. However, Re, Ru, W, and Hf can only be measured as positive secondary ions. We have measured the abundances of Os, Ir, Pt, and Au in kamacite, taenite, and/or plessite in five iron meteorites: Canyon Diablo (IA), Cape York (Agpalilik) (IIIA), Colomera (IIE), Cruz del Aire (Anom), and Wallapai (IID). Our measurements show that Os, Ir, Pt, and Au partition preferentially into taenite relative to kamacite during slow cooling. Measured abundance ratios (taenite/kamacite) range from ∼1.3 to ∼2.1 for Os, Ir, and Pt, and from ∼2 to ∼6 for Au. These ratios are consistent with those determined recently by laser-ablation ICPMS, but differ significantly from those determined by earlier workers. Low-temperature taenite/kamacite distribution coefficients inferred from our data are ∼2.1 for Os, ∼1.6 for Ir, ∼1.9 for Pt, and ∼6 for Au. PGEs are not enriched in taenite as much as Ni. Partitioning is controlled by the interplay of the size and electron configuration of Os, Ir, Pt, and Au and the crystal structures of kamacite and taenite. The ion probe can be used to further investigate the distribution and subsolidus redistribution of PGEs and Au in iron meteorites and in the metal in chondrites and stony irons.

We present a model to explain the wide range of abundances for heavy r-process elements (mass number A > 130) at low [Fe/H]. This model requires rapid star formation and/or an initial population of supermassive stars in the earliest condensed clots of matter to provide a prompt or initial Fe inventory. Subsequent Fe and r-process enrichment was provided by two types of supernovae: one producing heavy r-elements with no Fe on a rather short timescale and the other producing light r-elements (A < or = 130) with Fe on a much longer timescale.

Previous studies have shown that ^(187)Os/^(188)Os in seawater has become increasingly radiogenic over the last 40 Ma in a manner analogous to strontium. This rapid rise in the marine ^(187)Os/^(188)Os over the last 17 Ma has been attributed to an increase in the bulk silicate weathering rates resulting from the rise of the Himalayas and/or selective weathering and erosion of highly radiogenic organic rich ancient sediments. The key test of this hypothesis is the ^(187)Os/^(188)Os and the total osmium concentration of the Himalayan rivers. We report the concentration and isotopic composition of osmium in the Ganges, the Brahmaputra, and the Indus rivers. The ^(187)Os/^(188)Os of the Ganges close to its source (at Kaudiyal, 30°05′N, 78°50′E) is 2.65 and [Os] = 45 fM/kg. A second sample of the lower reaches of the Ganges at Patna (25°30′N, 85°10′E) gives ^(187)Os/^(188)Os =1.59 and [Os] = 171 fM/kg. The ^(187)Os/^(188)Os of the Brahmaputra at Guwahati (26°10′N, 91°58′E) is 1.07 and [Os] = 52 fM/kg. A sample of the Indus (Besham, 34°55′N, 72°51′E) has a ^(187)Os/^(188)Os of 1.2 and [Os] = 59 fM/kg. We infer that the Himalayas do not provide either a high flow of osmium or a highly radiogenic osmium component to the oceans. The overall trend for osmium and strontium could be explained by a regularly increasing input of global continental weathering sources but the Himalayas themselves appear not to be the dominant source.

We estimate gamma-ray fluxes due to the decay of nuclei beyond Bi209 from a supernova or a supernova remnant assuming that the r-process occurs in supernovae. We find that a detector with a sensitivity of about 10**(-7) photons/cm**2/s at energies of 40 keV to 3 MeV may detect fluxes due to the decay of Ra226, Th229, Am241, Am243, Cf249, and Cf251 in the newly discovered supernova remnant near Vela. In addition, such a detector may detect fluxes due to the decay of Ac227 and Ra228 produced in a future supernova at a distance of about 1 kpc. As nuclei with mass numbers A > 209 are produced solely by the r-process, such detections are the best proof for a supernova r-process site. Further, they provide the most direct information on yields of progenitor nuclei with A > 209 at r-process freeze-out. Finally, detection of fluxes due to the decay of r-process nuclei over a range of masses from a supernova or a supernova remnant provides the opportunity to compare yields in a single supernova event with the solar r-process abundance pattern.

We report the discovery of two hibonite grains (CaAl_(12)O_(19)) whose isotopic compositions show that they formed in the winds of red giant and asymptotic giant branch (AGB) stars. While hibonite is the second major phase (after corundum, Al_2O_3) expected to condense from stellar ejecta with C/O < 1, it has not previously been found. One circumstellar hibonite grain is highly enriched in ^(17)O and slightly depleted in ^(18)O relative to the solar composition and has large excesses in ^(26)Mg and ^(41)K, decay products of ^(26)Al and ^(41)Ca. The inferred initial values (^(26)Al/^(27)Al)0 ≈ 5 × 10^(-3) and (^(41)Ca/^(40)Ca)0 ≈ 1.5 × 10^(-4) of this grain are consistent with models of nucleosynthesis in an AGB star. The other hibonite is enriched in ^(17)O, strongly depleted in ^(18)O, shows no evidence of ^(41)Ca and formed with (^(26)Al/^(27)Al)0 ≈ 2 × 10^(-2). The low ^(18)O/^(16)O and very high (^(26)Al/^(27)Al)_0 may indicate substantial proton exposure during cool bottom processing in a low-mass parent star. The low upper limit on ^(41)Ca/^(40)Ca (≤ 3.2 × 10^(-5)) implies that little or no He-shell material had been dredged into the envelope when this grain formed. We also report isotopic compositions for 12 new circumstellar corundum grains. The compositions of 11 of these grains are consistent with current models for red giant and AGB stars. One corundum grain has extremely high ^(17)O/^(16)O and near-solar ^(18)O/^(16)O and may have formed in a star that was initially enriched in ^(17)O and ^(18)O.

▪ Abstract We present a review of nucleosynthesis in AGB stars outlining the development of theoretical models and their relationship to observations. We focus on the new high resolution codes with improved opacities, which recently succeeded in accounting for the third dredge-up. This opens the possibility of understanding low luminosity C stars (enriched in s-elements) as the normal outcome of AGB evolution, characterized by production of12C and neutron-rich nuclei in the He intershell and by mass loss from strong stellar winds. Neutron captures in AGB stars are driven by two reactions:13C(α,n)16O, which provides the bulk of the neutron flux at low neutron densities (Nn≤ 107n/cm3), and22Ne(α,n)25Mg, which is mildly activated at higher temperatures and mainly affects the production of s-nuclei depending on reaction branchings. The first reaction is now known to occur in the radiative interpulse phase, immediately below the region previously homogenized by third dredge-up. The second reaction occurs during the convective thermal pulses. The resulting nucleosynthesis phenomena are rather complex and rule out any analytical approximation (exponential distribution of neutron fluences). Nucleosynthesis in AGB stars, modeled at different metallicities, account for several observational constraints, coming from a wide spectrum of sources: evolved red giants rich in s-elements, unevolved stars at different metallicities, presolar grains recovered from meteorites, and the abundances of s-process isotopes in the solar system. In particular, a good reproduction of the solar system main component is obtained as a result of Galactic chemical evolution that mixes the outputs of AGB stars of different stellar generations, born with different metallicities and producing different patterns of s-process nuclei. The main solar s-process pattern is thus not considered to be the result of a standard archetypal s-process occurring in all stars. Concerning the13C neutron source, its synthesis requires penetration of small amounts of protons below the convective envelope, where they are captured by the abundant12C forming a13C-rich pocket. This penetration cannot be modeled in current evolutionary codes, but is treated as a free parameter. Future hydrodynamical studies of time dependent mixing will be required to attack this problem. Evidence of other insufficiencies in the current mixing algorithms is common throughout the evolution of low and intermediate mass stars, as is shown by the inadequacy of stellar models in reproducing the observations of CNO isotopes in red giants and in circumstellar dust grains. These observations require some circulation of matter between the bottom of convective envelopes and regions close to the H-burning shell (cool bottom processing). AGB stars are also discussed in the light of their possible contribution to the inventory of short-lived radioactivities that were found to be alive in the early solar system. We show that the pollution of the protosolar nebula by a close-by AGB star may account for concordant abundances of26Al,41Ca,60Fe, and107Pd. The AGB star must have undergone a very small neutron exposure, and be of small initial mass ([Formula: see text]). There is a shortage of26Al in such models, that however remains within the large uncertainties of crucial reaction rates. The net26Al production problem requires further investigation.

The climate of the Eastern Mediterranean region of the last 60 ky was determined by a high resolution study of the oxygen and carbon isotopic composition (1500 measurement pairs) of speleothems from the Soreq cave, Israel, with chronology provided by 53 precise ^(230)Th–^(234)U (TIMS) ages. The high precision of the speleothem TIMS ages permits us to determine the timing of regional climatic events in the Eastern Mediterranean region and to see if they correlate with global events. During the period from 60 to 17 ky, the δ^(18)O and δ^(13)C values were generally 2–2.5‰ higher than during the period from 17 ky to present. This is consistent with the climatic transition from glacial to interglacial. Within the 60 to 17 ky period, the Soreq cave stable isotope profile includes four cold peaks (at 46, 35, 25 and 19 ky) and 2 warm peaks (at 54 and 36 ky). In addition, the period

We discuss how detection of gamma-ray emission from the decay of r-process nuclei can improve our understanding of r-process nucleosynthesis. We find that a gamma-ray detector with a sensitivity of 10**(-7)/cm**2/s at 100-700 keV may detect the emission from the decay of Sb125, Cs137, Ce144, Eu155, and Os194 produced in a future Galactic supernova. In addition, such a detector may detect the emission from the decay of Sn126 in the Vela supernova remnant and the diffuse emission from the decay of Sn126 produced by past supernovae in our Galaxy. The required detector sensitivity is similar to what is projected for the proposed Advanced Telescope for High Energy Nuclear Astrophysics (ATHENA). Both the detection of gamma-ray emission from the decay of several r-process nuclei (e.g., Sb125 and Os194) produced in future Galactic supernovae and the detection of emission from the decay of Sn126 in the Vela supernova remnant would prove that supernovae are a site of the r-process. Furthermore, the former detection would allow us to determine whether or not the r-process nuclei are produced in relative proportions specified by the solar r-process abundance pattern in supernova r-process events. Finally, detection of diffuse emission from the decay of Sn126 in our Galaxy would eliminate neutron star/neutron star mergers as the main source for the r-process nuclei near mass number A=126., Comment: 14 pages, AASTeX, submitted to the Astrophysical Journal

We have investigated the Re-Os system for samples of whole rock, metal, and sulfide from ordinary chondrites. Using closed-system analytical techniques, we found complete exchange between sample and tracer isotopes for silicate-containing samples and obtained precise and reliable Re-Os concentration measurements. Results on two Group IVA iron meteorites and on a silicate-rich iron (Steinbach, IVA-AN) are consistent with the IVA-IVB isochron and support the previous observation that IVA-IVB irons may be slightly older than IIAB irons. Data on whole-rock fragments and metal-rich separates from the St. Severin chondrite (LL6) show a large range in ^(1870Re/^(188)Os and in ^(187)Os/^(188)Os, which makes possible, in principle, the determination of a Re-Os internal isochron on a chondrite, for the first time. This Re-Os fractionation may be due to partial melting of FeNiS and macroscopic redistribution of metal and sulfide. The St. Severin data show a good correlation line on a ^(187)Re-^(187)Os evolution diagram. If this is considered to represent an internal isochron, it gives an age T = 4.68 ± 0.15 AE [λ(^(187)Re) = 1.64 × 10^(−11) a^(−1)] and an initial (^(187)Os/^(188)Os)0 = 0.0953 ± 0.0013. This age is in agreement with but slightly older than the more precise ^(187)Re-^(187)Os age for the IIAB irons as well as for irons from other groups (T = 4.61 ± 0.01 AE). A St. Severin sulfide nodule has very low Re and Os concentrations and shows a young Re-Os model age (2.3 AE), indicating recent element remobilization. Whole rock and metal-rich separates of H-Group chondrites (H3 to H6) yield restricted ranges in ^(187)Re/^(188)Os (0.42–0.47) and ^(187)Os/^(188)Os (0.128–0.133). There is a systematic difference between Re/Os in the metal extracted from a chondrite and the bulk chondrite. This shows that there is a small but significant Re-Os fractionation within subsystems contained in the chondrites. From whole rock samples of H Group chondrites we calculate a mean ^(187)Re/^(188)Os = 0.423 ± 0.007 and ^(187)Os/^(188)Os = 0.12863 ± 0.00046, which may characterize the evolution of an average chondritic reservoir for Re-Os. The ordinary chondrite data plot close to the IIAB isochron, although the deviations found are larger than found for the irons. The Re-Os chronometer in iron meteorites is apparently controlled by the Re-Os fractionation due to fractional crystallization of liquid metal. Re-Os ages of iron meteorites give the time of crystallization of metal segregations and cores of early planetary bodies. In contrast, the behavior in ordinary chondrites, while also dominated by the metal phases, must reflect fractionation and transport on a local macroscopic scale within the chondrites between the metal phases after aggregation, due to partial melting of FeNiS or represent variable Re-Os fractionation of the metal phases prior to the accretion of the chondrites. However, for St. Severin, we attribute the major Re-Os fractionation to early heating of the meteorite, above the Fe-FeS eutectic. We do not consider that the Re-Os fractionation observed in other chondrites is due to the redistribution of Re and Os during chondrite metamorphism (including shock) but it may plausibly represent earlier stages of Re-Os fractionation for the different FeNi metal constituents prior to accretion.

Re-Os data on pallasite metal yield a sufficient range in Re/Os to permit the determination of a whole rock isochron, if pallasites from different groups are considered together. The results are indistinguishable from the data obtained on Groups IAB, IIAB, IID, IIIAB, IVA, and IVB iron meteorites. Pallasite data were obtained on samples of the Main Group, which have been considered related to Group IIIAB irons, and on samples of the distinctive pallasites Eagle Station, Finmarken, and Marjalahti. It appears that iron meteorites and pallasites crystallized within a time interval of 20 Ma. This result is consistent with the evidence from the short-lived chronometers ^(107)Pd-^(107)Ag and ^(53)Mn-^(53)Cr. It is also consistent with the evidence from ^(182)Hf-^(182)W that the time interval over which molten FeNi metal from different iron meteorites equilibrated with and became segregated from silicates was ∼10 Ma. The well-defined ^(187)Re-^(187)Os whole-rock isochron for pallasites and iron meteorites requires that they were formed in parent bodies that had been partially molten and segregated FeNi metal and FeS and then all crystallized within a 20 Ma time interval. This conclusion appears to apply to members of distinct groups of iron meteorites and pallasites which are viewed as not being cogenetic. The short time-scale requires that the irons and pallasites were formed in small parent planets of ≲10 km radius or near the surface of larger bodies. In contrast to the pallasites, we find that FeNi samples from mesosiderites have a narrow range in Re/Os so that it is not possible to determine a whole-rock isochron for these samples. The mesosiderite data also lie somewhat displaced from the iron meteorite and pallasite isochron indicating a more complex multi-stage evolution. Based on Sm-Nd data on mesosiderite silicate clasts, the Re-Os results on mesosiderite metal are compatible with the model of melted or partially molten FeNi cores or pods still preserved in protoplanetary bodies when these bodies were disrupted up to 150 Ma after early formation and the FeNi was splashed onto the surfaces of other small, differentiated planetesimals.

It is shown that if the ^(26)Al inventory of the early solar system, taken as (^(26)Al/^(27)Al)_☉ = 5×10^(−5), is a result of injection of fresh debris from a Type II supernova (SNII), then ^(60)Fe/^(56)Fe would have to be between 3×10^(−7) and 1×10^(−5). This inferred correlation of ^(26)Al and ^(60)Fe is based on the observation that both nuclei are produced dominantly in the O/Ne zone and that for SNII ejecta ^(26)Al/^(60)Fe is between 0.6 and 23. A similar correlation applies to ^(41)Ca,^(36)Cl,^(16)O, and ^(18)O, which are also produced in the same zone or in nearby regions. The supernova trigger hypothesis may be tested by determination of ^(60)Ni excesses correlated with Fe in samples where ^(26)Al was demonstrated to be present. From available experimental data, it appears that the observed abundance of ^(60)Fe is too low to be compatible with a supernova trigger that injected the ^(26)Al into the protosolar nebula. The same is true for ^(53)Mn, a short-lived nucleus produced in the outer edge of the Ni core.

We have studied the mineralogy and petrology of 229 calcium-aluminum-rich inclusions (CAIs) from ten CO3 meteorites of petrologic types 3.0–3.7. Subsets of these inclusions were measured by ion probe for magnesium, calcium, and titanium isotopes and REE abundances. Most CAIs from CO3 meteorites fall into three major types: (1) melilite-rich inclusions, which also contain spinel/hercynite, perovskite, and occasionally hibonite; (2) spinel-pyroxene inclusions; and (3) hibonite-hercynite inclusions. In addition, several isolated hibonite grains, two grossite (CaAl_4O_7) bearing CAIs, two hibonite-fassaite microspherules, and one anorthite-spinel-pyroxene inclusion were found. CAIs from CO3 meteorites exhibit all of the REE patterns commonly seen in inclusions from CV3 and CM2 chondrites. Most exhibit evidence of ^(26)AI, and many have inferred (^(26)AI/^(27)AI)_o ≈ 5 × 10^(−5). The relative abundances of different types of CAIs in CO3 chondrites differ from those in CV3 and CM2 chondrites. CAIs in CO3 chondrites have experienced considerable secondary alteration, both before and after accretion. Signatures of nebular alteration include Wark-Lovering rims and the high Fe contents in spinels from all hibonite/hercynite inclusions. Occasionally, melilite and anorthite show evidence of nebular alteration to feldspathoids and pyroxene. The magnesium-aluminum systematics of some melilite-rich inclusions were apparently disturbed prior to final accretion of the parent body. Parent body alteration is indicated by correlations between CAI characteristics and the petrologic type of the host meteorite. Spinel in melilite-rich and coarse-grained spinel-pyroxene inclusions becomes more Fe rich, with the development of relatively homogeneous hercynitic spinel (∼50–60 mol%) in CAIs from metamorphic grades >3.4. Perovskite has been converted to ilmenite in types >3.4. Melilite-rich inclusions are abundant in CO3.0–3.3 meteorites, rare in 3.4 meteorites, and absent meteorites of types 3.5–3.7; melilite-rich CAIs are probably replaced by inclusions rich in feldspathoids, pyroxene, and Fe-rich spinel. Isotopic disturbance of the magnesium-aluminum systematics may be more severe in higher petrologic types. Hibonite seems to be unaffected by this level of metamorphism. Three isotopically unusual inclusions were found. One single-crystal hibonite, Isna SP16, has a REE pattern strongly depleted in Ce and Y, (^(26)AI/^(27)AI)_o = (2.4 ± 0.3) × 10^(−5), and mass fractionated calcium (F_(Ca) = +12 ± 2‰/amu), but no resolvable nuclear anomalies in neutron-rich calcium isotopes. The REE pattern, which is thought to reflect nebular conditions, and mass-fractionated calcium, indicative of evaporation, are similar to those of the FUN inclusion, HAL, and related hibonites, indicating similar formation conditions. The absence in Isna SP16 of the nuclear anomalies observed in HAL and the difference in (^(26)AI/^(27)AI)_o between HAL and Isna SP16 indicate that the processes that produced HAL-type hibonites operated on diverse materials. Two hibonite-bearing microspherules, Colony SP1 and ALH82101 SP15, exhibit nearly flat REE patterns with negative europium anomalies and slightly negative δ^(26)Mg. ALH82101 SP15 has resolved excesses of ^(48)Ca and ^(50)Ti. These characteristics are similar to those of previously described microspherules from Murchison and Lance, implying that the microspherules formed via a single process from related, but not identical source materials.

Sulfide globules present in basaltic glass were analyzed for Re and Os concentrations ([Re], [Os]) and ^(187)Os/^(186)Os from a selection of MORBs and Loihi lavas. The high Re and Os concentrations measured in sulfides confirm the chalcophile character of these elements. The partition coefficients of Re and Os between sulfide globules and silicate melt are estimated to be D_(Re)=43 and D_(Os)=4.8×10^4. The ^(187)Os/^(186)Os of the samples range from 1.054 to 1.387 for MORBs and from 1.091 to 1.146 for the Loihi samples. The low Re/Os ratios of sulfides make correction for in situ ^(187)Re decay negligible. Several sulfide-rich (Os=500–1000 ppb) and sulfide-containing fractions from a Loihi sample gave essentially the same ^(187)Os/^(186)Os=1.095, demonstrating the possibility of obtaining very reliable data with our procedure. For some MORBs, shifts in ^(187)Os/^(186)Os for concentrates with low [Os] can be explained by the presence of Fe-oxyhydroxides enriched in Os by co-precipitation from seawater. The radiogenic ^(187)Os/^(186)Os ratios found in several sulfides with high [Os] cannot be explained by post-eruption contamination by Fe-oxyhydroxides or isotopic exchange between sulfides and seawater. We infer that some of the radiogenic ^(187)Os/^(186)Os ratios measured in MORB sulfides reflect the isotopic composition of the lava prior to eruption. Correlations were found between Os and B isotopic compositions and suggest that the radiogenic ^(187)Os/^(186)Os measured in MORBs are mainly due to assimilation of seawater-derived Os by the lavas during their ascent through the altered oceanic crust. MORBs with δ^(11)B≤−10‰ seem to be suitable to study the ^(187)Os/^(186)Os ratio of the mantle. Considering only such `uncontaminated' samples, we still find a significant range of ^(187)Os/^(186)Os ratios (1.054–1.093) between two MORBs from the FAMOUS area. This result suggests that the mantle underlying the FAMOUS area is heterogeneous with regard to the Re–Os system and that it may have an Os signature distinct from that of abyssal peridotites collected on the Mid-Atlantic Ridge. Osmium isotopic ratios in Loihi lavas seem less sensitive to contamination than in MORBs. Two Os-rich Loihi basalts have similar isotopic ratios that apparently provide a good estimate of the ^(187)Os/^(186)Os ratio of the Loihi source (^(187)Os/^(186)Os=1.095±0.005).

There is a large uncertainty in our understanding of the behavior of osmium during weathering and transport into deep oceans and the osmium budget of the oceans. The problem stems chiefly from the lack of osmium data on the dissolved load in the rivers and in the estuaries. In this study, the concentration and isotopic composition of osmium have been determined in three North American rivers (the Mississippi, the Columbia, and the Connecticut) and one river draining central Europe and flowing into the Baltic Sea (the Vistula). Osmium concentration in the Mississippi and the Vistula is about 45 femto mol kg^(−1); it is about 14 and 15 femto mol kg^(−1) for the Connecticut and the Columbia, respectively. The ^(187)Os/^(186)Os ratios estimated for the Mississippi and the Vistula are 10.4 and 10.7, respectively. For the Connecticut and the Columbia ^(187)Os/^(186)Os = 8.8 and 14.4, respectively. Of all the rivers examined, the Mississippi is by far the largest, supplying ∼1.6% of the total annual world river flow. Its osmium isotopic composition is identical to the upper Mississippi valley loesses (Esser and Turekian, 1993a) indicating (1) congruent dissolution of the bedrock and (2) little or no impact of anthropogenic sources on the osmium isotopic composition of the dissolved load. The latter observation indicates that the upper limit of the anthropogenic input in the dissolved osmium load of the Mississippi outflow is about 250 g yr^(−1). While the osmium concentration of the Vistula is high the isotopic composition does not appear to have been affected by substantial pollution. The river data can be used to put limits on the mean residence time of osmium in the oceans (T_(Os) and on the osmium budget of the oceans. If the bulk river influx of dissolved osmium into the oceans is similar to that of the Mississippi, we get a value for the net riverine inflow of osmium of 1680 mol yr^(−1). If there were no sequestering of osmium in the estuaries, this would give a value of T_(Os) of = 1.3 × 10^4 years. As the estuaries may trap a significant amount of river osmium, this is a lower limit so that T_(Os) » 1.3 x 10^4 years. Using the iridium data by Anbar et al. (1996), the Os/Ir ratio for the Vistula is 2.8. We infer that there is no large fractionation between osmium and iridium during the transport of these elements from the continents into the oceans. It follows that the high Os/Ir ratio (∼22) observed in seawater must be related to the different rain out mechanisms of these elements in the estuaries or in the oceans.

One hundred and forty individual SiC grains (1-9.6 µm) and twenty-two grain aggregates (2.3-7.8 µm) from the Orgueil (CI) chondrite have been measured by ion microprobe. Silicon and carbon isotopic data were obtained for all individual grains and aggregates, and nitrogen, magnesium, and aluminum abundances and isotopic compositions were measured for most grains and aggregates. Abundances of lithium, beryllium, boron, and sodium were measured for some individual grains. Orgueil SiC is remarkably similar to Murchison K-series SiC in the ranges and distributions of silicon and carbon isotopic compositions, the initial abundances of ^(26)AI, the abundances of minor and trace elements, and the proportions of isotopically unusual grains. Higher ^(15)N/^(14)N ratios in 1-4/µm Murchison K-series SiC grains relative to similar-sized Orgueil grains are inferred to be due to higher amounts of terrestrial nitrogen in the Murchison samples. Higher ^(15)N/^(14)N ratios in 3-6 µm Murchison KJH SiC grains cannot be explained by terrestrial nitrogen and imply that larger SiC grains sampled a different population of parent stars. SiC aggregates have different average silicon and carbon compositions than individual grains, indicating different source stars for the 0.1-1 µm constituent grains. However, the aggregates probably formed by clumping of small grains during laboratory procedures, not at the stellar source. Differences between Murchison L-series SiC and SiC from Murchison K-series and Orgueil are due to the presence of terrestrial SiC among L-series grains and to the larger average grain size of L-series SiC. When terrestrial contamination, sample size, and grain size are taken into account, there is no evidence of an intrinsic difference between Orgueil and Murchison SiC. The Orgueil data provide new information about stellar nucleosynthesis and the SiC parent stars. Carbon and nitrogen isotopic compositions indicate that nuclear processing in addition to that described by most stellar models occurs below the convective stellar envelope during the Red Giant and AGB phases (Cool Bottom Processing). Grains with high ^(15)N/^(14)N but with other characteristics consistent with AGB source stars indicate that ^(15)N is produced in AGB stars, contrary to the predictions of the standard models. A significantly higher rate for the ^(18)O(ɑ,n)^(15)N reaction than is typically used might reconcile the models with the observations. No data currently rule out a higher reaction rate. Addition of ^(18)O produced via ^(14)N(ɑ, y)^(18)O below the hydrogen shell and burned to ^(15)N in the envelope may also play a role in producing the high ^(15)N/^(14)N ratios observed in some SiC grains. Following previous workers, we take the slope ~2.2 silicon isotope array defined by mainstream SiC grains to reflect the initial compositions of the parent stars due to galactic evolution. A general correlation between ^(12)C/^(13)C and ^(29'30)Si/^(28)Si and more scatter around the silicon array at low ^(29'30)Si/^(28)Si indicate that SiC grains with higher ^(29'30)Si/^(28)Si ratios come from higher-metallicity parent stars. A lack of resolved isotopic effects in ^(25)Mg/^(24)Mg suggests that variations in initial magnesium compositions of the parent stars are at least a factor of three smaller than those in silicon and that the ^(22)Ne neutron source was not significantly activated in the AGB stars that produced SiC grains, in accord with theory. These observations indicate that the parent stars of most mainstream SiC grains were ≤2.3 M_☉ and experienced enough Third Dredge-up thermal pulses to supply their envelopes with 2-3% helium.shell material. A few grains have characteristics suggesting that they came from more-massive stars. Most parent stars of > 1 µm SiC grains apparently had metallicities higher than that of the sun.

Osmium is one of the rarer elements in seawater. Analytical difficulties have previously prevented the direct measurement of the osmium concentration and isotopic composition in seawater. We report a chemical separation procedure that yields quantitative extraction of osmium standard and of osmium tracer by iron hydroxide precipitation from seawater doped with osmium standard, osmium tracer, and FeCI_3. The iron hydroxide precipitate is processed to extract osmium, using techniques developed for iron meteorites. Utilizing this procedure, water samples from the Pacific and Atlantic oceans were analyzed for osmium concentration and isotopic composition. Direct determination of the osmium concentration of seawater gives between 15 and 19 fM kg^(−1). Detailed experiments on different aliquots of one seawater sample from the North Atlantic Ocean, keeping the amounts of reagents constant, yield concentrations from 16 to 19 fM kg^(−1). The variability in concentration is outside the uncertainty introduced because of blanks and indicates a lack of full equilibration between the osmium tracer and seawater osmium. The most reliable osmium concentration of the North Atlantic deep ocean water is 19 fM kg^(−1) with the ^(187)Os^(186)Os ratio being 8.7 ± 0.2 (2σ). Detailed experiments on one seawater sample from the Central Pacific Ocean indicate that the most reliable osmium concentration of the deep ocean water from the Central Pacific is 19 fM kg^(−1) with the ^(187)Os^(186)Os ratio being 8.7 ± 0.3 (2σ). The directly measured osmium isotopic composition of the oceans is in good agreement with that obtained from the analysis of some rapidly accumulating organic rich sediments (Ravizza and Turekian, 1992). A sample of ambient seawater around the Juan de Fuca Ridge gave ^(187)Os^(186)Os= 6.9 ± 0.4. This is distinctly lower than the deep-sea water value and may reflect local hydrothermal activity or some analytical difficulty with this sample. The osmium isotopic composition of the deep oceans indicates that ∼80% of the osmium is derived from the continents and the rest from extraterrestrial and hydrothermal sources. Using the iridium data of Anbar et al. (1996, 1997) we find that the Os/Ir ratio is ∼22 for seawater. The mean residence time of osmium (τ_(Os)) cannot be derived directly from these data. However, using the Os/Ir ratio of seawater and the Os/Ir ratio of continental rocks, a lower limit can be established of τ_(Os) ≥ 4.4 × 10^4 years. The Os/Ir in modern deep sea sediments should be relatively constant and is distinctly less radiogenic than seawater osmium due to limited evaporation (26%) of osmium from infalling cosmic dust. Relative to the cosmic dust infall the hydrothermal sources may, however, play a smaller but significant role in the bulk economy of dissolved osmium and other Platinum Group Elements in the oceans.

A novel technique for the precise measurement of oxygen isotopes by negative thermal ion mass spectrometry (NTIMS) is presented. The technique is ideally suited to the analysis of oxygen isotopes in phosphates which form intense PO_3^- ion beams. Since P is monoisotopic, the mass spectrum for PO_3^− at 79, 80, and 81 corresponds to ^(16)O, ^(17)O, and ^(18)O. Natural and synthetic phosphates are converted and loaded on the mass spectrometer filament as Ag_3PO_4 precipitated directly from ammoniacal solution. To lower the work function of the filament, BaCI_2, is added in a 1:1 molar ratio of PO_4:Ba. Using these procedures, Br^− mass interference (at 79 and 81 amu) is eliminated for typical analyses. Experiments with ^(18)O-enriched water show less than 1% O-exchange between sample PO_4 and adsorbed water, and there is no O-exchange with trace OZ present in the mass spectrometer source chamber. The ionization efficiency of PO_4, as PO_3^− is >10% compared to 0.01% for both conventional dual inlet Gas Isotope Ratio Mass Spectrometry (GIRMS) and secondary ion mass spectrometry (SIMS). Therefore, NTIMS offers exceptional sensitivity enabling routine and precise oxygen isotope analysis of sub-microgram samples of PO_4, (