Premium
A study of the minor element concentrations of spinels from two type B calcium‐aluminum‐rich inclusions: An investigation into potential formation conditions of calcium‐aluminum‐rich inclusions
Author(s) -
CONNOLLY Harold C.,
BURNETT D. S.
Publication year - 1999
Publication title -
meteoritics and planetary science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.09
H-Index - 100
eISSN - 1945-5100
pISSN - 1086-9379
DOI - 10.1111/j.1945-5100.1999.tb01401.x
Subject(s) - melilite , allende meteorite , anorthite , electron microprobe , crystallization , geology , silicate , population , fractional crystallization (geology) , trace element , context (archaeology) , mineralogy , mantle (geology) , spinel , geochemistry , materials science , chemistry , chondrite , metallurgy , meteorite , demography , paleontology , physics , organic chemistry , astronomy , sociology
— We have conducted an electron microprobe study of minor element distributions among spinels from two type B1 calcium‐aluminum‐rich inclusions (CAIs): Allende TS‐23 and Leoville 3537–2. We show that by maintaining the petrologic context (edge, middle, and center of the inclusion plus their host silicate phase), four populations of spinels are resolvable based on their minor element contents. One population resides within the edge area (mainly mantle melilite) and is characterized by the highest V contents. Unlike Leoville 3537–2, many edge grains from Allende TS‐23 also have high‐Fe contents (up to 4.0 wt%) and low‐Cr values. Based on their V and Ti concentrations (which is positively correlated), middle and center grains define a trend that is divided into three populations: spinels enclosed by melilite, fassaite, and anorthite. The overall range in Ti concentration based on fractional crystallization should be much less than a factor of 2; however, the observed range is considerably larger. The minor element contents of these grains are interpreted as recording alteration, primary fractional crystallization, and a complex igneous history that may involve remelting and recrystallization. From our data, Allende TS‐23 has experienced more alteration than Leoville 3537–2, which is consistent with previous petrologic studies of silicates within these objects; yet both objects have likely been remelted (at least one additional melting event, possibly two, postdating the initial formation of these CAIs). By invoking a remelting history, the large range ir Ti concentrations and the different populations of spinels can be explained. Although our data suggest that more than one generation of spinels exist within these objects, we are unable to establish any population of relic spinel grains that predate the initial melting event.