EXPERIMENTAL REPRODUCTION OF CLASSIC BARRED OLIVINE CHONDRULES: OPEN-SYSTEM BEHAVIOR OF CHONDRULE FORMATION

Abstract

Barred olivine (BO) chondrule texture with olivine rim was successfully reproduced by cooling experiments under evacuated conditions for an FeO-rich, alkali-free, low liquidus (1215° C), analog chondrule composition. The rim formation requires a heating temperature about 100° C or more above the liquidus to promote heterogeneous nucleation of olivine on the surface of a melt droplet. Partial evaporation (mainly FeO) enhances the selective nucleation too because the surface should be cooler than the interior due to the latent heat of the evaporation. The evaporation also causes subsequent growth along the cooler surface. This rim formation process suggests chondrule formation as open systems. If evaporation occurred during chondrule formation, evaporation of Na and SiO2 should enhance the selective nucleation and growth to form olivine rims more easily. Textures close to classic BO, which consists of a single set of parallel olivine platelets with olivine rim, were also reproduced by using small samples (< 10 mg and < 1 mm in diameter), where only a limited number of nucleation events is possible in a small surface area. Three-dimensional structures of reproduced BO chondrules were observed by X-ray microtomography using synchrotron radiation. Olivine bars in a two-dimensional section are really platelets and the platelets grow from the surface of a melt spherule inwards. The crystallographic orientations of the olivine platelets, which are normal to the b-axis, are also consistent with those of natural BO chondrules. BO textures close to natural chondrules were formed by heating at super-liquidus temperatures of about 100- 200° C for 10 min and cooled at 1000° C/h. However, detailed morphologies of olivine rims and platelets (glass inclusions and dendritic side arms) and olivine zoning are inconsistent with those of natural olivine irrespective of the experimental conditions. These discrepancies can be explained by annealing. Diffusion calculations for cooling and annealing of chondrules suggest that annealing at low temperatures (< ~ 1000° C), such as by thermal metamorphism, cannot explain Mg-Fe and Ca zoning in BO. The annealing was possibly due to multiple heating in connection with chondrule recycling process.