HR: 09:45h
AN: V11A-08    [PDF]
TI: Crystal nucleation theory applied to hydrous magma
AU: * Hammer, J E
EM: jhammer@hawaii.edu
AF: Univ. Hawaii, Dept. Geology and Geophysics, Honolulu, HI 96822 United States
AB: Examination of natural volcanic products shows that the crystal nucleation kinetics determine crystal number density and ultimately crystal size distributions, and thus govern the texture of crystallized materials. Crystallization in hydrous magmas undergoing decompression and devolatilization is especially pertinent in application to volcanological problems. We examine nucleation rate data in the context of kinetic theory and attempt to reconcile observations with aspects of silicate melt structure. Feldspar nucleation rate data obtained by laboratory decompression of hydrous silicate melt are interpreted in view of the classical theory of nucleation (CNT) and a non-classical variation, the diffuse interface theory (DIT). Several simplifying assumptions are needed to compare data with theory, including the Stokes-Einstein approximation, Turnbull's approximation ($\Delta$G=$\Delta$H$\Delta$T/T$_{L}$) and a means of estimating $\Delta$G of solidification. The crystal-liquid interfacial free energy ($\sigma$) is a key parameter in nucleation theory that is difficult to obtain independently. Lack of $\sigma$ data precludes direct testing of the CNT's validity. Instead, $\sigma$ is calculated using nucleation rate data assuming the CNT formalism is appropriate; this assumption would be considered valid if each experiment yielded the same value of $\sigma$. The interfacial free energies computed in the present case are not constant, but vary by a factor of four (0.024-0.098 J m$^{-2}$) and decrease systematically with increasing H$_{2}$O content (over the range 0.8-4.8 wt.%). The nucleation rate data can be modeled only if $\sigma$ is allowed to vary as a function of composition, suggesting that a non-classical theory may be justified. The DIT states that a region between the bulk solid (at the core of subcritical clusters) and bulk melt has intermediate thermodynamic properties, and that the interfacial free energy $\sigma$ may be defined as the difference between the interfacial enthalpy (H$_{int}$) and interfacial entropy (TS$_{int}$). If the DIT model is correct, the nucleation rate data for feldspar suggest (1) that dissolved H$_{2}$O content controls the spatial gradients of TS$_{int}$ and H$_{int}$ around incipient crystals, and (2) these gradients diverge during devolatilization.
DE: 3939 Physical thermodynamics
DE: 3947 Surfaces and interfaces
DE: 8439 Physics and chemistry of magma bodies
DE: 8499 General or miscellaneous
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting