THE TYPE 1 AND TYPE 2 carbonaceous chondrites (CI 1 and CM2) are among the most primitive meteorites, particularly in regard to their bulk chemical compositions, which are very similar to solar values for nonvolatile elements (ANDERS and GREVESSE, 1989). Nevertheless, mineralogical and textural evidence show that these meteorites have undergone considerable processing on their parent bodies, including brecciation, aqueous alteration and veining, and, in some cases, mild thermal metamorphism (ZOLENSKY and MCSWEEN, 1988, and references therein). CIl and CM2 chondrites have distinctive oxygen isotopic compositions in that they are considerably enriched in the heavy isotopes relative to other chondrite groups (ordinary chondrites, enstatite chondrites,
and C3 chondrites).
This enrichment is associated with the abundant phyllosilicates in CI 1 and CM2 chondrites, which in turn result from low-temperature aqueous interactions on their parent bodies (CLAYTON, 1993). The oxygen isotopic compositions may, therefore, be useful in determining both the source of the water involved and the conditions of temperature and pressure under which the interaction occurred. Carbonaceous chondrites of type CI 1 consist primarily of hydrous layer-lattice silicates and magnetite. Carbonaceous chondrites of type CM2 are more heterogeneous, but in general, consist primarily of a mixture of anhydrous minerals (predominantly olivine and pyroxene) and phyllosilicate minerals. Usually, their magnetite contents are less than l%, in contrast to the CI 1 s, in which the magnetite abundance is approximately 10% (HYMAN and ROWE, 1983, 1986). Bells and Essebi are exceptional among the CM2 chondrites in containing >lO% magnetite, almost identical in content with that observed in the Ivuna and Orgueil CII chondrites (HYMAN and ROWE, 1986; DAVIS and OLSEN, 1984).
Ever, Essebi magnetite differs in detailed morphology from that in the CII chondrites. HYMAN et al. (1985) pointed out that Essebi contains at least two of the distinctive magnetite morphologies first described in detail by JEDWAB (1965, 1967, 1968, 197 1). Other CM2 and CV3 chondrites also contain some magnetite of these distinctive morphologies, but usually only in minor amounts compared to that in the CIl chondrites (HYMAN and ROWE, 1983, 1986). The long-suspected relationship between the CI and CM chondrites, if one exists at all, remains unclear. However, on the basis of the oxygen isotope data on magnetite presented in this paper, there does seem to be a common origin of the magnetite in Essebi and in the CIl chondrites. METZLER et al. (1992) noted major petrographic differences which distinguish Bells and Essebi from other CM chondrites, including extensive brecciation. They concluded that Essebi, and possibly Bells also, originated in a different parent body. Previous oxygen isotope measurements of whole-rock and separated components of the CI, CM, CO, CV, and CR chondrites have shown some regular patterns.
These include (1) the compositions of high-temperature minerals in CM, CO, and CV fall along a 160-mixing line nearly coincident with that determined from the Ca-Al-rich inclusions in Allende and other CV chondrites and in CM chondrites as well (CLAYTON et al., 1983; CLAYTON, 1993); (2) the matrix and whole-rock analyses of the CI chondrites fall along a slope- % line which is very near the terrestrial mass fractionation line (CLAYTON and MAYEDA, 1989); (3) the compositions of matrix minerals from the CM2 chondrites fall along a slope- % line far removed from the data for anhydrous phases (CLAYTON et al., 1977; CLAYTON and MAYEDA, 1984); (4) chondrules and matrix of CR chondrites lie on a separate mixing line of slope 0.7 which lies between the CM mixing line and the terrestrial fractionation line (WEISBERG et al., 1993). Oxygen isotopes are useful in the elucidation of possible genetic relationships between the various carbonaceous chondrites and their components. An understanding of these observed trends in the oxygen isotopic data was at least partially accomplished with a model based primarily on the silicates and calcite in the Murchison CM2 chondrite (CLAYTON and MAYEDA, 1984). This model can readily incorporate some of the additional data provided by the present study.
It (a reconstruction of Fig. 2 in CLAYTON and MAYEDA, 1984) shows its essential features. It begins with the assumption that only two oxygen reservoirs of different isotopic compositions are required to explain the data. The starting point is taken to be a solar nebula consisting of gas and solids in the range of temperatures at which silicates are solid and water is a gas, i.e at temperatures of a few hundred kelvins. Thus. 17% of the oxygen resides in the solids and 83% is in the gas phase principally as Hz0 and CO, defined by solar system elemental abundances (ANDERS and GREVESSE. 1989). Evidence from carbonaceous chondrites had earlier indicated that the gas and solids originally had very different oxygen isotopic abundances, with the solid phase being enriched in “0 as a consequence of its prior nucleosynthetic history (CLAYTON et al.. 1977; CLAYTON, 1993). The interaction between these two reservoirs, with the constraints imposed by the Murchison oxygen data, is the heart of the model
as illustrated Schematic diagram representing a model for the oxygen isotopic evolution of the major components of CM2 carbonaceous chondrites. based on data from Murchison (CLAYTON and MAYEDA, 1984). Compositions labeled S. A. M, and C has observed values for S = spinel. A = mean of anhydrous silicates (olivine and pyroxenes),
M = phyllosilicate matrix, C = carbonate. Points labeled Cl. G2. L I and L2 are hypothetical fluids: G 1 and G2 = nebular gas, L 1 and L2 = parent body liquids.
The model postulates two initial oxygen-bearing reservoirs, a solid (dust) component with composition S. and a gas component with composition G I. Isotope exchange during high-temperature interaction (perhaps chondrule formation) brings about complementary isotope shifts from S to A in the solids. and from Cl to G2 in the gas. After the accretion of solids and ices to form a parent body, low-temperature aqueous alteration reactions cause further shifts of solids from A to M and liquid water from L Ito L2. Calcite (C) is precipitated from the solution at about 0°C. The terrestrial fractionation line (TF) is shown for reference.
The model begins with gas. G I, determined from Murchison data to have an oxygen composition of 6’“O = +30.0,
6”O = +24.2%0 and initial solids, S. with 6i”O = -40, 6”O = -41%~. The exchange of these two components at high temperatures led to the evolution line for gas from G 1 to G2 (6” O = +20.6%0, 6” O = + I5.2%0). while the solids went from the initial composition S to point A (6” O = -4.2%0, 6” O = -7.4%~). Then the gas. G2. which had cooled to -0°C and was mostly in the form of HzO, began to condense to liquid water, Ll. with composition 6” O = +30.3%~. 6” O =m t20.2%0. The change in the silicates from A to M (the matrix composition) by hydration reactions must be compensated by a complementary change in the liquid composition to L2 with 6’*0 = +0.6%00 and 6” O = -1. I %O. CLAYTON and MAYEI)A (I 984) considered that the most striking feature of Murchison oxygen isotopic data is the very large fractionation between calcite (point C) and phyllosilicates (point M).
The fact that the matrix and the calcite lie on the same slopei/z line led them to conclude that the two components had a simultaneous origin in a common environment at low temperatures (-0 to 20°C). It is on the basis of this model that the new oxygen isotopic
data on the separated components from the Alais, Ivuna, and Orgueil CI I chondrites and the Bells and Essebi chondrites will be discussed in this paper. Just as the coexisting phyllosilicate and calcite in Murchison were useful in assessing the conditions for aqueous alteration in CM2 meteorites, it was anticipated that coexisting phyllosilicate and magnetite in Cl1 and Essebi would provide isotopic information on conditions for aqueous reactions in the parent bodies of those meteorites.
EXPERIMENTAL PROCEDURES
Five types of samples were analyzed: whole-rock, chondrules, and matrix. magnetite. and olivine/pyroxene concentrates. For whole-rock samples, approximately 3 mg of the bulk meteorite was ground. The preparation of separated fractions differed in detail from sample to sample but was generally as follows. Chondrules Meteorite samples (I35 to 450 mg) were gently ground with a mortar and pestle, in acetone for cooling. When hard pieces were felt. the ground-up material was viewed at -25X under a microscope. A few chondrules from Essebi were thus identified and handpicked
with fine-pointed forceps. Other visible grains were also removed at this stage and placed into a vial. This grinding/microscope cycle was repeated until no further resistance due to hard pieces was felt. Some of the samples were ultrasonicated at this stage of the separation, but no difference in resulting separation was observed.
