OXYGEN ISOTOPIC COMPOSITION OF THE SUN AND MEAN OXYGEN ISOTOPIC COMPOSITION OF THE PROTOSOLAR SILICATE DUST: EVIDENCE FROM REFRACTORY INCLUSIONS

Preliminary analysis of the oxygen isotopic composition of the solar wind recorded by the Genesis spacecraft suggests that the Sun is O-16-rich compared to most chondrules, fine-grained chondrite matrices, and bulk compositions of chondrites, achondrites, and terrestrial planets (Delta O-17=-26.5% +/- 5.6% and -33% +/- 8% (2 sigma) versus Delta O-17 similar to +/- 5%). The inferred O-16-rich composition of the Sun is similar or slightly lighter than the O-16-rich compositions of amoeboid olivine aggregates and typical calcium-aluminum-rich inclusions (CAIs) from primitive (unmetamorphosed) chondrites (Delta O-17=-24%+/- 2%), which are believed to have condensed from and been melted in a gas of approximately solar composition (dust/ gas ratio similar to 0.01 by weight) within the first 0.1 Myr of the solar system formation. Based on solar system abundances, 26% of the solar system oxygen must be initially contained in dust and 74% in gas. Because solar oxygen is dominated by the gas component, these observations suggest that oxygen isotopic composition of the solar nebula gas was initially 16O-rich. Due to significant thermal processing of the protosolar molecular cloud silicate dust (primordial dust) in the solar nebula and its possible isotope exchange with the isotopically evolved solar nebula gas, the mean oxygen isotopic composition of the primordial dust is not known. In CO self-shielding models, it is assumed that primordial dust and solar nebula gas had initially identical, O-16-rich compositions, similar to that of the Sun (Delta O-17 similar to-25% or-35%), and solids subsequently evolved toward the terrestrial value (Delta O-17=0). However, there is no clear evidence that the oxygen isotopic compositions of the solar system solids evolved in the direction of increasing Delta O-17 with time and no 16O-rich primordial dust have yet been discovered. Here we argue that the assumption of the CO self-shielding models that primordial dust and solar nebula gas had initially identical O-16-rich compositions is incorrect. We show that igneous CAIs with highly fractionated oxygen isotopic compositions, fractionation and unidentified nuclear effects (FUN), and fractionation (F) CAIs, have Delta O-17 ranging from -0.5% to -24.8%. Within an individual FUN or F CAI, oxygen isotopic compositions of spinel, forsterite, and pyroxene define a mass-dependent fractionation trend with a constant Delta O-17 value. The degree of mass-dependent fractionation of these minerals correlates with the sequence of their crystallization from the host CAI melt. These observations and evaporation experiments on CAI-likemelts indicate that FUN and F CAIs formed by melting of solid precursors with diverse Delta O-17 values in vacuum (total pressure < 10(-6) atm). We interpret the observed range of Delta O-17 values among FUN and F CAIs as the result of varying degrees of equilibration between O-16-poor dust and O-16-rich nebular gas and suggest the former is characteristic of the primordial dust. The distinctly different oxygen isotopic compositions of the primordial solar nebula dust and gas could have resulted from Galactic chemical evolution or from pollution of the protosolar molecular cloud by a massive star (>50 circle dot) ejecta. The O-16-depleted compositions of chondrules, fine-grained matrices, chondrites, and achondrites compared to the Sun's value eflect their formation in the protoplanetary disk regions with enhanced dust/ gas ratio (up to 10(5) x solar).