Molecular Cloud Origin for the Oxygen Isotope Heterogeneity in the Solar System

See allHide authors and affiliations

Science  17 Sep 2004:
Vol. 305, Issue 5691, pp. 1763-1766
DOI: 10.1126/science.1100989


Meteorites and their components have anomalous oxygen isotopic compositions characterized by large variations in 18O/16O and 17O/16O ratios. On the basis of recent observations of star-forming regions and models of accreting protoplanetary disks, we suggest that these variations may originate in a parent molecular cloud by ultraviolet photodissociation processes. Materials with anomalous isotopic compositions were then transported into the solar nebula by icy dust grains during the collapse of the cloud. The icy dust grains drifted toward the Sun in the disk, and their subsequent evaporation resulted in the 17O- and 18O-enrichment of the inner disk gas.

Oxygen is the most abundant element in the solid phases that formed early in the solar system, and it has three stable isotopes of mass numbers 16, 17, and 18. On a three-oxygen isotope diagram, 18O/16O and 17O/16O abundance ratios of most terrestrial material constitute a line with slope of ∼0.5, called the terrestrial fractionation (TF) line. This slope is due to isotope fractionation processes that depend on the mass difference between each pair of isotopes. In contrast, most meteorites have oxygen isotopic compositions that diverge from the TF line (1). Refractory inclusions and some chondrules in primitive meteorites have the most 16O-enriched isotope compositions, shifted from the TF line with magnitudes of several percent in 17O/16O and 18O/16O ratios (1, 2). Nonradiogenic effects in the other major elements (e.g., Mg and Si) in these meteorite constituents have isotope compositions close to the terrestrial compositions, and their small deviations can be explained by isotope fractionation due to thermal processes, e.g., evaporation, condensation, aqueous alteration, and low-temperature chemical reaction (3).

The origin of mass-independent fractionation of oxygen isotopes and the lack of such fractionation in other major elements in meteorites remains poorly understood. It cannot be due to nucleosynthetic processes or nuclear reactions that involve energetic particles from the Sun or from Galactic cosmic rays, because these processes would also change the isotopic compositions of the other elements (1). In addition, presolar grains enriched in 16O are rare in meteorites (4). Although some types of molecular reactions in gaseous phases have been found to induce such mass-independent isotope fractionation in oxygen (5), they are observed among gas species (e.g., O3, O2, and CO2) that are minor in the solar nebula (6). Furthermore, even if such fractionation occurs, no plausible mechanism has been proposed for trapping the fractionated products into chondrite constituents. Oxygen isotope changes due to selective ultraviolet (UV) dissociation of molecules in the solar nebula gas have been proposed (5, 7, 8); however, a mechanism for transferring these effects to the chondritic constituents has not been identified.

Recently, variations in C16O/C18O ratio have been observed in diffuse molecular clouds (9, 10). These variations are explained by selective predissociation (11) of C18O by UV radiation. In the environment of molecular clouds, predissociation due to line spectrum absorption of UV photons is the dominant mechanism for photodissociation of CO (1217). UV intensity at the wavelengths of dissociation lines for abundant C16O rapidly attenuates in the surface layer of a molecular cloud, because of its UV self-shielding. For less abundant C17O and C18O, which have shifted absorption lines because of differences in vibrational-rotational energy levels, the attenuation is much slower. As a result, C17O and C18O are dissociated by UV photons even in a deep molecular cloud interior. This process results in selective enrichment of CO in 16O and enrichment of atomic oxygen in 17O and 18O.

Because CO and atomic oxygen are the dominant oxygen-bearing gas species in molecular clouds (18), their isotopic fractionation may propagate to other oxygen-bearing species. Water ice is the dominant oxygen-bearing species among ices in molecular clouds (19), where it nucleates and grows on silicate dust grains by surface hydrogenation reactions between atomic oxygen and hydrogen (20, 21). Therefore, oxygen isotopic compositions of H2O ice should be close to those of gaseous atomic oxygen enriched in 17O and 18O (22). Water ice is observed in molecular clouds with total visual extinction (AV) greater than 3.2 (23); abundance of water ice increases with increasing AV (24). As a molecular cloud becomes dense, most of the atomic oxygen reacts to form H2O ice, and CO becomes the most dominant gas species within 105 years (25). Thus, the oxygen isotopic composition of the gas in a dense molecular cloud becomes enriched in 16O with time.

Low-mass (less than two solar masses) stars form by collapse of individual cores or clumps in a cold, dark molecular cloud with molecular densities of hydrogen (nH2) of 104 to 105 cm–3, AV of 5 to 25, and temperatures as low as ∼10 K (26). According to a model simulating photochemical isotope fractionation in a molecular cloud (17), under these typical cloud parameters, the isotopic compositions of ice and gas are expected to be in the ranges δ18OMC =+100 to +250 per mil (‰) and δ18OMC = –60 to –400‰, respectively, where δ18OMC Ξ {[(18O/16O)/(18O/16O)MC] – 1} × 1000; (18O/16O) and (18O/16O)MC are the isotopic ratios of corresponding chemical species and the bulk molecular cloud (MC), respectively. The degrees of fractionation for calculated 18O/16O ratios are consistent with astronomical observations (9, 10). Although the lack of experimental data for C17O predissociation prevents us from a detailed analysis of 17O/16O fractionation, its degree is likely near that for 18O/16O, because the absorption lines of these minor isotope species are unsaturated at least over several tens of AV (15, 16). Such expected similarity has been recently observed for diffuse interstellar gas (27).

In denser and more evolved cold molecular cloud cores, most CO may become frozen onto dust grains. Because of the low temperature, oxygen isotope exchange between CO and H2O ices is inefficient, and the original isotope fractionation of oxygen is preserved in each phase. Transient external heating by shock waves or by other mechanisms would cause vaporization of both H2O and CO ices and local homogenization in such a cloud. However, as long as H2O and CO molecules do not decompose into radicals and atoms, the oxygen isotope fractionation in each molecule is probably preserved.

Here we examine how such isotopic heterogeneity in a molecular cloud may cause the variations of oxygen isotopic compositions observed in our solar system. Taking relative oxygen abundances of silicates, ice, and gas to be 1: 2:3 in molecular clouds (20), we assumed that both δ17OMC and δ18OMC17 and 18OMC) are 0‰ for silicates, +120‰ for ice, and –80‰ for gas (Fig. 1A). The δ17 and 18OMC values for H2O ice and CO gas were chosen to be within the simulated ranges and to conserve the mean isotopic composition of the bulk molecular cloud. The 16O-depleted nature of ice relative to silicates is consistent with evidence from a primitive meteorite. The most 16O-depleted known component formed in the solar system is the product of aqueous alteration of Fe,Ni-metal by H2O in the most primitive ordinary chondrite, Semarkona (28).

Fig. 1.

Schematic diagram of oxygen isotope evolution from a molecular cloud to a protoplanetary disk with dust sedimentation. (A) Oxygen isotopic compositions in a molecular cloud. CO (open circle) is the most abundant species next to H2 and He in a molecular cloud. UV radiation selectively destroys C17 and 18O, leaving behind CO enriched in 16O and producing atomic oxygen enriched in 17O and 18O. This heavy oxygen later becomes incorporated into water ice (snowflake). δ17,18OMC values of 0‰, +120‰, and –80‰ for silicate (solid circles), ice, and gas, respectively, are assumed. (B) Oxygen isotopic compositions in the outer disk. The oxygen isotopic composition produced in the molecular cloud is preserved in the individual phases in the outer disk after accretion because of low temperature. (C) In the inner disk, oxygen isotopic composition of gas (⊚) shifts to a 16O-poor one. (D) In the disk, solid materials settle down to the mid-plane and spiral into the proto-sun. Water ice evaporates inside the snow line (29), leading to the shift in oxygen isotopic composition shown in (C). Degrees of the shift depend on the H2O enrichment factor (Fig. 2) and on oxygen isotopic compositions of the individual phases in the molecular cloud. The relationship between δ notation relative to SMOW and that to the molecular cloud possibly corresponds to δ17 and 18OSMOW ≅ δ17 and 18OMC – 50‰, assuming the traditional 16O-rich reservoir in the solar system. The setting of lighter silicates and heavy disk gas with respect to oxygen isotopic composition in the inner disk is consistent with meteoritic observations.

A protoplanetary disk is formed by collapse of a molecular cloud core. In the outer region of the disk, because of low temperatures (29), the primordial oxygen isotopic compositions of the molecular cloud components are preserved (Fig. 1B). CO sublimes while preserving its own oxygen isotope composition outside the orbits of outer planets, even in the case of frozen CO. In the inner region of the disk, H2O ice evaporates. During an early stage of disk evolution accompanied by vigorous gas accretion, gasdust fractionation is probably minor, and the mean oxygen isotopic composition of the inner disk gas is reset to the value of the bulk molecular cloud, δ17 and 18OMC = 0‰. Because transient heating events, such as the formation of refractory inclusions and chondrules, were common in the inner solar nebula (30), silicate grains would equilibrate with such gas and have similar oxygen isotopic compositions.

As the gas accretion rate decreases, dustgas fractionation processes begin to proceed in the disk. One such fractionation process is the dust sedimentation toward the disk midplane (31) (Fig. 1D). In addition, dust particles may preferentially migrate toward the central star (32), and ice in the dust evaporates after passing the snow line, releasing 16O-depleted water vapor into the inner disk gas (33) (Fig. 1D). This increases the mean δ17 and 18OMC of disk gas along the mixing line between the oxygen isotopic compositions of CO and of H2O ice (Fig. 1C), correlating with the degree of H2O enrichment relative to the H2O/CO ratio in the parent molecular cloud (22). Although enrichment of H2O by a factor of 10 is justified in the solar nebula (34), even moderate enrichment can produce extreme 17O- and 18O-enrichment of the disk gas (Fig. 2). For example, if three times the H2O enrichment occurs (i.e., if relative oxygen abundances of ice:gas are 2:1), the mean δ17 and 18OMC of the inner disk gas will be about +50‰. Therefore, oxygen isotopic compositions of the disk gas are altered easily by dust-gas fractionation processes (35). Silicate grains equilibrated with such H2O-enriched gas during transient heating events would acquire isotope compositions with high δ17 and 18OMC values (36).

Fig. 2.

δ17 and 18OMC of inner disk gas as a function of the H2O enrichment factor relative to molecular cloudabundance. The H2O enrichment factor times the H2O/CO ratio in a molecular cloud represents the H2O/CO ratio in the disk gas, excluding chemical equilibria. The solid curve represents a case that neglects refractory organics as an oxygen carrier. Here we assume the δ17 and 18OMC values of silicate, H2O ice, and CO gas to be 0‰, +120‰, and–80‰, respectively, and their relative oxygen abundances in a molecular cloud to be 1:2:3. Significant 16O-depletion of the gas is expected even for small H2O enrichments. Long and short-dashed curves incorporate the effect of organics, providing that the δ17 and 18OMC values of silicate, ice, organics, and gas are 0‰, +120‰, +20‰, and –80‰, respectively, and that the relative oxygen abundances are 1:1.5:1:2.5 in the molecular cloud. The long-dashed curve indicates the case of higher temperature than the sublimation point of organics, assuming the same enrichment factor with H2O. The short-dashed curve indicates the case of the temperature between sublimation points of H2O ice andof organics. In either case, significant 16O-depletion of the gas occurs for small H2O enrichment factors.

Organic materials may also accommodate a significant fraction of oxygen. In hot molecular cloud cores observed for high-mass star-forming regions, CO is probably depleted because of conversion to refractory organics (37). Free radical reactions in ices are dominant processes to form refractory organics in molecular clouds (38). These organics are interpreted as UV photolysis products in H2O ice contaminated with CO during the previous cold evolutionary stage of a molecular cloud (20). Because oxygen contained in such organics seems to come from H2O and CO, its isotope composition is expected to be somewhere between those of both species, possibly δ17 and 18OMC of +20‰ if a 1:1 contribution of H2O and CO is assumed.

If we accept a refractory organic abundance and composition in a comet nucleus (20), the oxygen abundance of organics will be comparable to that of silicate (i.e., silicate: ice:organics:gas = 1:1.5:1:2.5). Because refractory organics evaporate under higher temperatures than H2O, they may affect the mean oxygen isotopic composition of inner disk gases at high temperatures. Even though this diminishes the amount of change in the isotopic composition because of H2O enrichment, the disk gas probably has 16O-poor compositions (Fig. 2).

The proposed scenario can reproduce oxygen isotope heterogeneity in the inner solar nebula with an 17O- and 18O-enriched gas, i.e., 16O-depleted gas, relative to silicate dust, consistent with the conventional O isotope reservoirs inferred from meteorite studies (1). Under such an environment, the silicate dust evolves into an 16O-depleted composition through isotope exchange with the surrounding gas, because of transient heating events in the nebula. Therefore, the average oxygen isotopic composition of the solar nebula normalized to the standard mean ocean water (SMOW) may be δ17,18OSMOW ≅ –50‰ or smaller (39) (Fig. 1).

We have shown that even small mass fractionation for CO and atomic O in the molecular cloud can explain the formation of 16O-rich or -poor reservoirs observed for the solar nebula. The 16O-rion or -poor reservoirs can easily form if we use larger mass fractionation factors as expected by chemical models (16, 17) and observations (27) of molecular clouds. Thus, the 16O isotope variations may not be unique to our solar system but instead ubiquitous in any planetary system. A direct test of this scenario would be to measure the oxygen isotopic compositions of cometary ices and that of solar wind. We predict the oxygen isotopic values as δ17 and 18OSMOW ≅ +50 to +200‰, –100 to –450‰, and –50‰ for cometary H2O, cometary CO, and solar wind, respectively.

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article