Crystallization sequences of Ca-AI-rich inclusions from Allende: The effects of cooling rate and maximum temperature

We have studied the crystallization sequences, mineral chemistries, and textures that develop when an average Type B Ca-Al-rich inclusion composition is cooled in air from 1275–1580° C to below 1000°C at rates between 0.5 and 1000°C/hr. Crystallization sequences, the textures of all the major phases, pyroxene chemistry, and melilite zoning patterns are functions of both the cooling rate and the temperature from which cooling begins. Determination of the order of pyroxene and plagioclase crystallization has been identified as an important goal for petrographic studies of CAIs because it can be used to set constraints on the cooling rate experienced by an individual inclusion. Overall textures plus melilite zoning patterns and pyroxene chemistry can give important clues as to whether pyroxene or plagioclase began to crystallize first. Melilite texture and chemistry appear to yield the most valuable information on the maximum temperature to which an inclusion was raised prior to cooling. Comparison of our experimental results with petrographic observations of Type B CAIs suggests that most inclusions were partially melted and then cooled at rates on the order of a few tenths to tens of degrees per hour. Maximum temperatures of about 1400°C appear most likely for intermediate Type B Allende inclusions. Our results do not support the suggestion that the textures observed in these inclusions formed by crystallization of supercooled, metastable melt droplets condensed from nebular gas. The slow cooling rates we infer for CAIs are difficult to reconcile with models for their origin that imply simple radiative cooling of individual molten or partially molten droplets in a cold, low density environment. On the other hand, cooling rates of the nebular cloud are believed to have been much slower than those we have inferred for Type B CAIs. Scenarios that could be reconciled with the thermal history that we have inferred include drag heating of particles falling through nebular gas, heating by intense radiation (e.g., via flares) from the early sun, heating in nebular shock fronts, or other thermal heterogeneities in the early nebula allowing time scales for cooling (and heating) of CAIs much shorter than those for the nebular cloud as a whole. Successful models for the origin of Type B CAIs must account for the fact that most Type B CAIs cooled relatively slowly from a partially molten state