Report

# Remnants of the Early Solar System Water Enriched in Heavy Oxygen Isotopes

See allHide authors and affiliations

Science  13 Jul 2007:
Vol. 317, Issue 5835, pp. 231-233
DOI: 10.1126/science.1142021

## Abstract

Oxygen isotopic composition of our solar system is believed to have resulted from mixing of two isotopically distinct nebular reservoirs, 16O-rich and 17,18O-rich relative to Earth. The nature and composition of the 17,18O-rich reservoir are poorly constrained. We report an in situ discovery of a chemically and isotopically unique material distributed ubiquitously in fine-grained matrix of a primitive carbonaceous chondrite Acfer 094. This material formed by oxidation of Fe,Ni-metal and sulfides by water either in the solar nebula or on a planetesimal. Oxygen isotopic composition of this material indicates that the water was highly enriched in 17O and 18O (δ17,18OSMOW = +180‰ per mil), providing the first evidence for an extremely 17,18O-rich reservoir in the early solar system.

Oxygen isotopic variations in chondrites provide important constraints on the origin and early evolution of the solar system (1). Oxygen isotope ratios in chondrites change not only by mass-dependent isotope fractionation law (isotope fractionation depending on mass differences among isotopes) but also by large mass-independent isotope fractionation (MIF) that keeps 17O/18O ratio nearly constant. It is generally accepted that MIF recorded by meteorites resulted from mixing of two isotopically distinct nebular reservoirs, 16O-rich and 17,18O-rich (2). The composition of the 16O-rich reservoir has been recently constrained from isotopic compositions of nebular condensates (3) and of a unique chondrule (4). The nature and composition of an 17,18O-rich nebular reservoir are still poorly constrained (5, 6). According to the currently favorite self-shielding models (711), nebular water is hypothesized to have been highly enriched in 17,18O (5 to 20%) relative to Earth, which is, however, yet to be verified by isotope measurements. Here we report an in situ discovery of a chemically and isotopically unique material in the primitive carbonaceous chondrite Acfer 094. This material is mainly composed of iron, oxygen, and sulfur, and is highly enriched in 17O and 18O (up to +18%) relative to Earth's ocean. Mineralogical observations and thermodynamic analysis suggest that this material resulted from oxidation of iron metal and/or iron sulfide by water in the solar nebula or on a planetesimal. We infer that the extreme oxygen isotopic composition of this material recorded composition of this primordial water that corresponds to an 17,18O-rich nebular reservoir in the early solar system, in agreement with the self-shielding models (712).

During our ongoing in situ survey (1316) of presolar grains of primitive meteorites (17), we discovered isotopically anomalous regions of oxygen in matrix of the ungrouped carbonaceous chondrite Acfer 094 in addition to isotopically anomalous spots corresponding to presolar grains (Fig. 1). The oxygen isotopic compositions of the regions are uniform and enriched in 17O and 18O relative to 16O (Fig. 2). The data seem to be plotted along the slope-1 line (18) rather than the carbonaceous chondrite anhydrous mineral mixing (CCAM) line (2) (Fig. 3). The representative values of δ17OSMOW = δ18OSMOW are about +180 per mil (‰); SMOW is standard mean ocean water (19). These are the heaviest oxygen isotopic compositions of the solar system materials reported so far. The less 17O- and 18O-rich compositions (δ17,18OSMOW =+50‰) of unknown origin have been recently reported in the surface layers of metal grains from lunar regolith (20).

The chemical compositions of the isotopically anomalous regions determined by an energy-dispersive x-ray spectrometer (EDS) attached to a field-emission scanning electron microscope (FE-SEM) (16) show that they are homogeneous and mainly composed of Fe, Ni, O and S (representatively, in weight percent, Fe, 61.6; Ni, 5.4; O, 19.3; S, 9.6; Mg, 0.1; Si, 0.2). In addition, analytical transmission electron microscopy (ATEM) (16) reveals that the regions consist of aggregates of nanocrystals with a size range of 10 to 200 nm (fig. S1). The electron diffraction patterns from ∼100-nm-sized individual crystals show that the main spots of the crystals are similar to those of magnetite (Fe3O4; space group Fd3m); the corresponding cell parameter a is 0.83 nm. In addition, there are weak extra spots suggesting a threefold superstructure. Characteristic x-ray spectra from individual crystals show that they consist of the same elements determined by the FE-SEM-EDS study. These observations indicate that the crystals have a magnetite-like structure and may represent a new Fe-O-S–bearing mineral; more detailed characterization is necessary to identify it. Although this mineral consists of the same elements as a poorly characterized phase (PCP) commonly observed in aqueously altered CM chondrites and largely composed of tochilinite or tochilinite-cronstedtite intergrowths (21, 22), its O/S atom ratios are about 4 times as large as those in tochilinite. Hereafter, we refer to this mineral as a new-PCP.

The chemically unique and isotopically anomalous new-PCPs are ubiquitous and scattered randomly throughout the Acfer 094 matrix. Twenty-two new-PCPs (the largest is 160 μm2) were identified in about an 11-mm2 area of the matrix using elemental mapping with a 7-μm2 spatial resolution. This corresponds to 94 ± 20 (σ) parts per million (ppm) by volume. Because the number of the new-PCP grains increases exponentially with decreasing size, the grain numbers below 7 μm2 are dominant and the actual abundance of new-PCP must be larger than the estimate.

The new-PCP often coexists with troilite (FeS) (Fig. 2), which is considered to be a reaction product between Fe, Ni-metal and H2S gas (23, 24). $Math$(1) Because new-PCP has magnetite-like diffraction patterns, magnetite can be used as its proxy. Magnetite can be formed by oxidation of Fe, Ni-metal (23, 25), or troilite. $Math$(2) $Math$(3) Oxidation of troilite or metal to form new-PCP would occur below 360 K, independent of total pressure of the solar nebula (Fig. 4). If the nebular PH2O/PH2 ratio increases from a characteristic value for a gas of solar composition (25), formation of new-PCP occurs at higher temperature. Although the complete chemical equilibrium would not be expected in the cool solar nebula (23), the new-PCP would be formed inside the water sublimation front (snowline) of the solar nebula, because water vapor is the major oxidant in the solar nebula and the sublimation temperature of water ice is below 200K even in the several-fold H2O-enriched nebula (23).

Alternatively, new-PCP may have been formed by aqueous alteration of metal and troilite on the Acfer 094 parent body, like tochilinite in the aqueously altered CM chondrites (22). To test this hypothesis, we analyzed oxygen isotopic compositions of tochilinite in the CM chondrite Murchison. In contrast to the Acfer 094 new-PCP, oxygen isotopic composition of the Murchison tochilinite is plotted near the terrestrial fractionation line, along the “CM waters” line (26) (Fig. 3), that is considered to be a reaction path between aqueous solution and matrix silicates toward the isotope equilibrium (26). These observations and the lack of mineralogical and petrographical evidence of aqueous alteration of Acfer 094 (27) exclude formation of new-PCP by the aqueous alteration previously observed in chondrites. If new-PCP resulted from oxidation of troilite or metal in a planetesimal setting, a plausible oxidant would be water vapor or aqueous solution that originated from accreted nebular ice and did not experience oxygen isotope exchange with the matrix silicates. In contrast, the inferred oxygen isotopic compositions (26, 28) and the observed chondrite parent body water composition richest in 17,18O (29) may have recorded equilibration of aqueous solutions with the chondrite matrix silicates.

We conclude that oxygen isotopic compositions of new-PCP in Acfer 094 represent composition of the primordial water of the solar system and the previously hypothesized 17O- and 18O-rich reservoir in the early solar system. The wide oxygen isotopic variations of at least –80‰ <δ17,18O < +180‰ found from hot and cold origin materials must provide new guidelines for the origin of oxygen isotope anomalies in the solar system.

Supporting Online Material

Materials and Methods

Fig. S1

Tables S1 and S2

References

View Abstract