Aubrites and diogenites: Trace element clues to their origin

Abstract

We have analyzed by RNAA 25 aubrite and 9 diogenite samples for 13 to 29 siderophile, volatile, and lithophile trace elements. Both meteorite classes show a typically igneous siderophile element pattern, with Ir, Os, Re, Ge more depleted than Au, Ni, Pd, Sb. But aubrites tend to have about 10 × higher abundances (10 −3 − 10 − 4 × Cl for the first 4 and 10 −2−10 −3 × Cl for the last 4 siderophiles), apparently reflecting smaller metal/silicate distribution coefficients at lower f( O 2), or less complete segregation of metal. Se is surprisingly abundant in aubrites (up to 0.4 × Cl), but Te is less so ( Se Te ≅ 5 × Cl ), apparently due to its stronger siderophile character. Other volatiles (Ag, Zn, In, Cd, Bi, T1) show depletions intermediate between lunar dunite and the Earth's mantle. Of 7 aubrites analyzed for REE (Ce, Nd, Eu, Tb, Yb, Lu), 6 are depleted in REE (0.08−0.5 × Cl) and 5 show negative Eu anomalies (the exceptions are Bishopville and Mt. Egerton silicate). This supports an igneous origin, as already noted by Boynton and Schmitt (1972). No samples of the complementary, basaltic and feldspathic rocks have been found thus far, but one of our samples of Khor Temiki dark is a candidate for the basalt. It is 5−7 × enriched in REE and only slightly less so in Rb, Cs, and U. Though shocked and enriched in siderophiles to ~0.05 × Cl, it apparently represents a new meteorite class. Three diogenites analyzed for REE show very diverse patterns, from strongly depleted in light REE for Tatahouine ( Ce = 0.01 × Cl) to flat for Garland (~2.5 × Cl). The data confirm the trends found by Fukuoka et al. (1977) as well as their interpretations. Factor analysis shows several parallel groupings for aubrites and diogenites: siderophiles (Re, Ir, Os, Pd, Ge), chalcophiles (Se, Te), volatiles (Ag, In, Tl) and incompatibles (U, REE, and Cs or Rb). But there are some differences for elements such as Ni, Sb, Cd, Bi, Au, and Zn, most of which behave more sensibly in aubrites than in diogenites. Several element pairs that differ greatly in volatility (Cs-U, Ge-Ir) correlate closely in aubrites, in approximately Cl-chondrite proportions. These correlations, and other lines of evidence, suggest strongly that aubrites originated by igneous processes in their parent body, not by direct nebular condensation. The source material may have resembled EL chondrites in oxidation state and depletion of refractories, metal, and volatiles.