Search

Analyses of magmatic iron meteorites, plotted on LogC(sub i) vs LogC(sub Ni) diagrams, often form linear arrays. Traditionally, this linearity has been ascribed to fractional crystallization under the assumption of constant partition coefficients (i.e., Rayleigh fractionation). Paradoxically, however, partition coefficients in the Fe-Ni-S-P system are decidedly not constant. This contribution provides a rationale for understanding how trends on LogC(sub i) vs LogC(sub Ni) diagrams can be linear, even when partition coefficients are changing rapidly.

Following the investigation of Jurewicz et al. (1991) on the composition of partial melts of eucrites and angrites, this study investigates partial melts of the Allende and Murchison chondrites and details the anhydrous phase relations of both chondrites at 1 atm, under temperatures and oxygen fugacities plausigle for the formation of basaltic meteorites. It was found that, in general, ambient oxygen fugacity exerts a strong influence on the compositions of partial melts of chondrites by controlling the amount of FeO available to the silicate system. At high f(O2), angritic magmas are produced, whereas eucritic melts are produced at low levels of f(O2).

Partitioning coefficients for metal/sulfide liquid, troilite/sulfide liquid, and schreibersite/sulfide liquid were determined for Ag, Au, Mo, Ni, Pd, and Tl (using EMPA and proton-induced X-ray microprobe and ion microprobe analyses) in order to understand the chronometer systems of iron meteorites. In general, the obtained schreibersite/metal and troilite/metal partition coefficients for 'compatible' elements were quite similar to those inferred from natural assemblages, reinforcing an earlier made conclusion that there is a class of elements for which experimental troilite/metal and schreibersite/metal partition coefficients approximate those inferred from natural samples. The consistency between experimental and natural assemblages, however, was not observed for Ag, Pb, and Tl, indicating that the abundances of these elements determined in 'metal' and 'troilite' separates from iron meteorites are influenced by trace minerals that concentrate incompatible elements.

The SNC meteorites and the measurements of the Viking landers provide our only direct information about the abundance and isotopic composition of Martian volatiles. Indirect measurements include spectroscopic determinations of the D/H ratio of the Martian atmosphere. A personal view of volatile element reservoirs on Mars is presented, largely as inferred from the meteoritic evidence. This view is that the Martian mantle has had several opportunities for dehydration and is most likely dry, although not completely degassed. Consequently, the water contained in SNC meteorites was most likely incorporated during ascent through the crust. Thus, it is possible that water can be decoupled from other volatile/incompatible elements, making the SNC meteorites suspect as indicators of water inventories on Mars.

There has been much speculation about the abundance of water and other volatiles on Mars. Attempts to calculate abundances of water on Mars indicate that Mars contains approx. 10-100 m of water. Numerous models have been put forth to determine the amount of water on Mars more closely. Some researchers infer that Chassigny parent magma contained greater than 1.5 percent water by weight and that the Martian mantle contained greater than 1000 parts per million water. This is too much water for a depleted region. Perhaps some of the water in Chassigny was assimilated at shallow depths, either in a crustal magma chamber or by interaction with superficial permafrost. Either is possible and provides an alternative to the dilemma of water-rich to depleted regions.

Shergottites are members of the SNC meteorite suite, which may be samples of Mars. If so, the shergottite in our collection that most likely represents primitive liquid from the Martian mantle is EETA 79001. EETA 79001 has the Nd isotopic signature of a long-term depleted mantle, a relatively high Mg number, and a slightly olivine-normative composition. The authors have performed experiments on the composition of EETA 79001 for traces of Eg. Other topics discussed include: comparison of calculated phase equilibria; nature of the olivine-pyroxene boundary; and interstitial liquids.

Partition coefficients for Au, Ni, P, and Ge between solid Fe-Ni metal and sulfur-bearing metallic liquids have been measured at 7, 27 and 80 kbar. These are the only such data for Au, P, and Ge at high pressure. Comparison of the present partitioning results to those obtained at 1 bar indicate that only the 80-kbar Ge data differ significantly from the 1-bar experiments. Thus, many low-pressure partitioning experiments in the Fe-Ni-S-P system may have applicability to the greater portion of the earth's upper mantle or, alternatively, the entire mantle of Mars.

Eucrites and angrites are distinct types of basaltic meteorites whose origins are poorly known. Experiments in which samples of the Allende (CV3) carbonaceous chondrite were partially melted indicate that partial melts can resemble either eucrites or angrites, depending only on the oxygen fugacity. Melts are eucritic if this variable is below that of the iron-wuestite buffer or angritic if above it. With changing pressure, the graphite-oxygen redox reaction can produce oxygen fugacities that are above or below those of the iron-wuestite buffer. Therefore, a single, homogeneous, carbonaceous planetoid greater than 110 kilometers in radius could produce melts of drastically different composition, depending on the depth of melting.

A model-based argument is made that there is a plausible relationship between the shergottites and the other SNCs which implies that the former contain an igneous component that is both young and mantle-derived. Evidence is presented that the shergottites were derived from a common mantle source within a short time interval and then interacted with other isotopic reservoirs prior to their crystallization. The model implies in general that the initial isotopic signatures of the bulk shergottites may be approximated as a mixing among three described components.

Theoretical lunar mantle seismic velocity and density profiles were calculated for a series of specific bulk composition models, thermal models, and evolutionary models. It was found that most of the calculated seismic profiles are in qualitative agreement with the Nakamura (1983) observational mantle seismic velocity model. With respect to the density modeling results, the most interesting finding is that none of the considered models yield mantle density increases that are sufficient to match the adapted upper bound on the lunar moment-of-inertia factor (0.3928). The results suggest that the moon possesses a dense metallic core with a mass in the range of 1 to 4 percent of the lunar mass and that the alumina-rich models of Morgan et al. (1978) and Taylor (1982) are best able to match the middle mantle velocity increase that characterizes the Nakamura model.

Measurements of solid metal/liquid metal trace element partition coefficients, which are used to interpret the crystallization history of magmatic iron meteorite groups differ greatly between different research groups, using different experimental techniques. Specifically, partition coefficients measured utilizing 'static' experiments which approach equilibrium cannot be reconciled with the results of 'dynamic' experiments which mimic fractional crystallization. We report new tests of our 'static' experimental technique and demonstrate that our methodology yields reliable equilibrium values for Ni, P and Ge partition coefficients. Partition coefficients in the Fe-Ni-S-P system are well matched by interpolation between the Fe-Ni-S and Fe-Ni-P subsystems. In contrast, the predictions of 'dynamic' experiments do not agree with our measurements and, consequently, the ability of 'dynamic' experiments to reproduce iron meteorite Ge vs. Ni fractionation trends successfully must be regarded as fortuitous.

The chronologies of the Shergotty, Zagami, ALHA 77005, and EETA 79001 meteorites were reexamined on the basis of shergottites' petrography and mineral chemistry data. Among the various isochrons, the concordant Rb-Sr (about 180 Myr) and U-Th-Pb (about 190 Myr) internal isochrons are interpreted as representing the time of igneous crystallization, while the Rb-Sr, Sm-Nd, and Pb-Pb whole-rock isochrons are interpreted as mixing lines, and are reasonably attributed to igneous processes such as wall-rock assimilation and magma mixing. If the approximated age of less than 200 Myr is correct, the shergottites represent the youngest known extraterrestrial basalts. This conclusion supports the hypothesis that the SNC meteorites are samples of Mars.

Although the moon is depleted in volatile elements compared to the earth, these depletions are not in accord with simple volatility. For example, the Cs/Rb ratios of the earth and moon inferred from basalt are approximately one seventh and one half of the CI ratio, respectively. Volatility considerations alone predict that the lunar Cs/Rb ratio should be equal to or lower than the terrestrial ratio if the moon was derived entirely from earth mantle material. Thus hypotheses such as rotational fission which invoke derivation of lunar material entirely from the earth's mantle may be excluded. The collisional ejection hypothesis of lunar origin requires at least 18 percent of lunar material to be derived from a projectile with dehydrated CI composition to match the lunar Cs/Rb ratio, and 25-50 percent to match both the lunar Cs/Rb ratio and absolute concentrations of Cs and Rb. It remains to be demonstrated that this relatively large contribution of projectile material is consistent with other elemental abundances and element ratios in the moon.

Two basic types exogenous models were proposed to account for siderophile and chalcophile element abundances in the Earth's upper mantle. The first model requires that the Earth be depleted in volatiles and that, after a core formation event which extracted the most siderophile elements into the core, additional noble siderophile elements (Pt, Ir, Au) were added as a late veneer and mixed into the mantle. The second model postulates a reduced Earth with approximately CI elemental abundances in which a primary core forming event depleted all siderophile elements in the mantle. The plausibility of models which require fine scale mixing of chondritic material into the upper mantle is analyzed. Mixing in liquids is more efficient, but large degrees of silicate partial melting will facilitate the separation of magma from residual solids. Any external events affecting the upper mantle of the Earth should also be evident in the Moon; but siderophile and chalcophile element abundance patterns inferred for the mantles of the Earth and Moon differ. There appear to be significant physical difficulties associated with chondritic veneer models.

The shergottites are a group of basaltic meteorites which are very similar in appearance to terrestrial basalts. On the Earth basalts are formed by volcanic activity. Because the shergottites so resemble terrestrial basalts and because they are apparently very young ( 1.3 billion years), it has been inferred that they come from a large planet. Small planets and asteroids lose heat from their interiors quickly and stop producing hot basaltic liquids early in their history. It appears that gases trapped in one shergotite found in Antarctica (BETA 79001) are chemically similar to the martian atmosphere (as measured by the Viking mission). The controversial speculation that the shergotites are samples of mars is examined in detail.

Previous estimates of the composition of the EPB have tacitly assumed chondritic abundances of refractory LIL elements. A model is presented here which utilizes refractory element ratios instead. Consideration of these ratios and of known olivine-liquid partition coefficients for Sc, Mg and Si imply that the EPB mantle may be accurately modeled as a simple mixture of 25 percent eucrite + 75 percent olivine. The model also implies that eucrites such as Juvinas and Sioux County were derived by large degrees of partial melting (20-30 percent). Comparison of the model EPB mantle to known meteorite classes indicates a similarity between the element abundances of the EPB and C3 chondrites.

Experimental problems associated with metallic solid/metallic liquid partitioning investigations are evaluated. Of particular concern is the discrepancy of partition coefficient results obtained from experiments which were static (constant temperature for the duration of the experiment) and those which were dynamic (where an experimental charge is slowly drawn from the furnace).

The partitioning behavior of Pb in the three-phase system metal-troilite-sulfide liquid was investigated. Microscopic examination and electron microprobe analysis of a polished section showed three phases--Fe metal, troilite and an Fe-S-Pb-O metallic liquid. The Pb concentration of the metallic liquid was 1.4 wt.%; the Pb contents of both the troilite and the Fe metal were below the detection limit of the electron probe. To better define the Pb concentrations of the solid phases, the same sample was analyzed for Pb and Fe using the M.I.T., Harvard and Brown ion microprobe. Even at this level of precision, some conclusions and speculations are possible: (1) The D(Pb) (troilite/metallic liquid) is low (.01 to .001) and, contrary to expectation, troilite, like metal, excludes Pb very effectively during the crystallization of metallic liquids; (2) Pb in iron meteorites is conceivably concentrated in trace phases which are less susceptible to contamination than metal and troilite; and (3) taken at face value D(Pb)(metal/troilite) is 0.1, but given the Pb and Fe ion yields from simple endmember phases (Fe, FeS, Pb, PbS), this D is more likely to be 0.05 to 0.03.

Supporting arguments for the lunar fusion hypothesis include the low density of the Moon corresponding to the density of the Earth's mantle and the low volatile content of the lunar rocks vs. those of terrestrial origin. Vapor pressures of the alkali elements and their oxides increase in the following order: Na, K, Rb and Cs. The Moon should, therefore, be more depleted in Cs relative to Rb, Rb relative to K, and K relative to Na than the Earth if the fission model is correct. Analyses of lunar mare basalts and terrestrial mid-ocean ridge and other young basalts indicate that this behavior is not observed. It is possible that monovalent alkali elements might be lost from silicate materials in a different order than that inferred from elemental and oxide vapor pressures, as a result of differences in the way they are bound in silicate materials. To test this hypothesis a series of experiments was conducted to investigate alkali loss at high temperatures. Analyses indicate that the behavior of volatiles dissolved in a silicate melt is similar to that inferred from elemental and oxide vapor pressures. It is concluded that alkali element ratios in the Earth and Moon are not readily interpreted in terms of the fission hypothesis.

The partition coefficients between synthetic whitlockite (beta Ca-phosphate) and coexisting silicate melts are determined for the actinide elements Th, U and Pu. Experiments were performed at 1 bar pressure and 1250 C at oxygen fugacities from 10 to the -8.5 to 10 to the -0.7 bars, and partitioning was determined from trace element radiography combined with conventional electron microprobe analysis. Results show Pu to be more readily incorporated into crystalline phases than U or Th under reducing conditions, which is attributed to the observation that Pu exists primarily in the trivalent state, while U and Th are tetravalent. Corrected partition coefficients for whitlockite of 3.6, less than or equal to 0.6, 1.2, 0.5 and less than or equal to 0.002 are estimated for Pu(+3), Pu(+4), Th(+4), U(+4) and U(+6), respectively. Experiments performed at trace levels and percent levels of UO2 indicate that Si is involved in U substitution in whitlockite, and show a reduced partition coefficient at higher concentrations of U that can be explained by effects on melt structure or the fraction of tetravalent U.

The partitioning of Ir, Ge, Ga, W, Cr, Au, P, and Ni between solid metal and metallic liquid is investigated as a function of temperature and the S-concentration of the metallic liquid. The partition coefficients for siderophile elements, such as Ir, W, Ga, and Ge, are found to increase by factors of 10-100 as the S-concentration of the metallic liquid increases from 0-30 wt percent. Partition coefficients for other siderophile elements, such as Ni, Au, and P, increase by factors of only 2-3. In contrast, a decrease is seen in the partition coefficients for the more chalcophile element Cr. These experimentally determined coefficients are used in conjunction with a fractional crystallization model to reproduce the geochemical behavior of Ni, P, Au, and Ir during the magmatic evolution of groups IIAB, IIIAB, IVA and IVB iron meteorites. The mean S-concentration for each group increases in the order IVB, IVA, IIIAB, IIAB, which is in accordance with cosmochemical prediction. However, the geochemical behavior of Ge, Ga, W, and Cr cannot be reproduced in an internally consistent way. It is concluded that the magmatic histories of these iron meteorite groups are more complex than has been generally assumed.

Two very different sets of Pu-244/U-238 ratios have been reported for early solar system materials, one high and the other low, and this paper examines how one of these groups may yield an incorrect ratio. Recently measured partition coefficients for Pu and Sm (Jones, 1981) are used to evaluate Pu-Nd fractionation during achondrite genesis, and Pu-244/U-238 ratios of chondrites, achondrites, and Ca-Al-rich inclusions are compared and discussed. A low Pu-244/U-238 ratio is favored, and some implications of this low ratio for galactic nucleosynthesis and meteorite age dating are briefly discussed.

The ureilite meteorites are carbonaceous olivine-pyroxene achondrites. They typically contain up to 4 wt.% carbon (carbonaceous matrix) as graphite, diamond, and lonsdaleite. Shock degradation has effectively obliterated primary textures in the carbonaceous matrix of previously described ureilites, a factor that has hampered efforts to explain the origin of this material. In contrast, the Antarctic ureilite ALHA78019 displays perfectly preserved primary textures in the carbonaceous matrix characterized by euhedral graphite blades intergrown with Fe-Ni metal and sulfide (diamonds are absent). This petrographic feature suggests that most graphite in ureilites originated by crystallization from a C-rich metallic phase. Assuming that fO2 is controlled by C-CO-CO2 reactions, the compositions of silicates and metals in ureilites imply a two-stage redox history. The noble gases and rare earths of ureilites are discussed in light of this model.

Actinide and Sm partition coefficients between silicate melt and several metallic phases have been measured. Under reducing conditions Si, Th, U and Pu can be reduced to metals from silicate melts and alloyed with a platinum-gold alloy. U and Pu enter a molten Pt-Si alloy with roughly equal affinity but U strongly partitions into the solid Pt. Th behaves qualitatively the same as Pu but is much less readily reduced than U, and Sm appears to remain unreduced. Experiments with Fe metal have shown that the partition coefficients of the actinides between Fe and silicate liquid are extremely low, suggesting a very low actinide concentration in planetary cores. Experiments show that platinum metals can efficiently fractionate actinides and fractionate actinides from lanthanides and this process may be relevant to the condensation behavior of these elements from the solar nebula. Pt-metal grains in Allende Ca-Al-rich inclusions appear to be U-poor, although the sub-class of Zr-bearing Pt metals may have high U contents.