TUNGSTEN AND HAFNIUM DISTRIBUTION IN CALCIUM-ALUMINUM INCLUSIONS (CAIs) FROM ALLENDE AND EFREMOVKA

Introduction: A widely accepted paradigm for the origin of chondrites maintains that chondrites represent the earliest planetesimals to form from the solar nebula and that differentiated bodies formed by the subsequent melting of chondrites. A challenge to this paradigm comes from physical models of accretion that require rapid accretion of ~10 km planetesimals that are then melted by the heat released by 26 Al, followed by impact disruption of these bodies to form chondrules [1, 2]. Support for such models has been provided by recent 182 Hf-182 W chronology of CAIs and chondrule metals that imply that iron meteorites may have formed contemporaneously with, or even earlier than, metal from CAIs and chondrules [3]. Tungsten isotope compositions represent initial 182 W/ 184 W at the time of metal segregation. A relative chronology of iron meteorites has been possible from ε 182 W measurements for iron meteorites [3, 4, and references therein]. Kleine et al. [3] have reported ε 182 W= –3.47±0.20 from an Allende CAI isochron that included a magnetic separate (ε 182 W= –3.17±0.70). These numbers are smaller than, or comparable to, the most negative ε 182 W reported from iron meteorites. Cosmic-ray induced neutron burnout of 182 W can produce negative effects that have been estimated to be no more than ε 182 W~-0.3 [3]. This led Kleine et al. [3] to support the hypothesis that at least some iron meteorites represent an earlier generation of molten planetesimals that preceded chondrites. Since 182 W is the daughter product of the decay of 182 Hf (t 1/2 ~9 Ma), a fundamental assumption in [3] is that metal (the W host) in Allende refractory inclusions last equilibrated W isotopically with its surrounding silicates (the Hf host) at the time of formation of the refractory inclusion. This assumption must be made by any study that attempts to relate the relative chronology of 182 Hf-182 W with an absolute chronology such as U-Pb so that a test of the robustness of this assumption is required. Unlike iron meteorites where the separation of Hf and W occurred on length-scales (L) of >1 metre (m), the physical separation of Hf and W in CAIs is generally L<10-4 m, within the possible length-scale of diffusive exchange. Microanalysis of CAI metal from Efremovka showed that W had been redistributed in the solid state [5]. The same study found that W had