Cristobalite- and tridymite-bearing clasts in Parnallee (LL3) and Farmington (L5)

— A set of cristobalite- and tridymite-rich igneous clasts (CB1 to CB8) have been found in Parnallee (LL3.6). They consist of clinoenstatite, minor feldspathic mesostasis and cristobalite veined by endiopsideaugite. The largest clast, CB8, is 1.6 cm in diameter and contains veined tridymite and cristobalite, clinoenstatite (zoned to ferroaugite and pyroxferroite Fs75.6Wo20.0) and plagioclase. Compared to bulk ordinary chondrites (OC), the bulk clasts are depleted in Al (0.02–0.8× OC), Na and K and enriched in Si (1.6–2.0× OC) and Ca (1.3–4.5× OC). Bulk CB8 has LREE > HREE (La/Lu = 1.6) with a positive Eu anomaly (Eu/Eu* = 2.4). Textural observations suggest that the clasts cooled rapidly (24–420 °C/h) above 1200 °C. Clasts CB1—CB8 contain the isotopically heaviest O yet found in ordinary chondrites (up to δ17O = +8.7%o, δ18O = +11.6%o). Enrichment in the heavy isotopes of O is dependent on the proportion of cristobalite (or tridymite) in the clasts. A regression line CRIL (Cristobalite Line), with slope 0.77, is defined by the isotopic compositions of CB1—CB8, the Farmington clast and ordinary chondrite chondrules. An 16O-poor gas reservoir, whose composition must lie at some point along the extension of CRIL, has undergone varying degrees of isotopic exchange with most ordinary chondrite material. Silica polymorphs have undergone the greatest degree of exchange because of their open, framework structures. Silicon in CB1—CB8 has normal isotopic ratios. A model is proposed that involves differentiation of H-group material through extraction of volatile elements in a vapour phase, loss of an Fe-Ni-S melt and metastable crystallisation (60–70%) of olivine. The calculated residual liquid is silica-oversaturated and its subsequent predicted crystallisation sequence resembles that preserved in CB1—CB8. This model may require two stages of heating, the second one prior to cristobalite crystallisation (if the silica polymorph crystallises within its predicted stability field of > 1500 °C). Isotopic exchange took place either when CB1—CB8 were ejected from their parent body due to impact or near the surface of the parent body, perhaps in an ejecta blanket setting. The latter option is preferred because it is more consistent with our igneous model.