Chemical diffusion of fluorine in jadeite melt at high pressure

The chemical diffusion of fluorine in jadeite melt has been investigated from 10 to 15 kbars and 1200 to 14OO“C using diffusion couples of jadeite melt and fluorine-bearing jadeite melt (6.3 wt.% F). The diffusion profile data indicate that the diffusion process is concentration-independent, binarv. F-O interdiffision. The F-O interdiffusion coefficient ranges from 1.3 X IO-’ to 7.1 X lo-’ cm*/ set and is much huger than those obtained by KUSHIRO (1983) for Si-Ge and AI-Ga interdiffision in jadeitic melts. The Arrhenius activation energy of di~sion is in the range of 36 to 39 kcal/mole as compared with 19 kcal/mole for fluorine tracer diffusion in a lim~aluminosilicate melt. The diffisivity and activation energy of F-O interdiffusion vary slightly with pressure, but the pressure dependence of F0, Al-Ga and Si-Ge interdiffusion may be related to the relative volumes of the interdiffusing species for each pair. The magnitude of chemical diffusivity of fluorine is comparable to that of the chemical diffusivity of water in obsidian melts. The diffusivities of various cations am significantly increased by the addition of fluorine or water to a silicate melt. This fact, combined with the high diffusivity of fluorine, suggests that the F ion is the principal diffusing species in dry ~umin~ili~te melts and that dissolved fluorine will accelerate chemical equilibration in dry igneous melts. INTBGDUCTION EXPERIMENTAL METHOD A KNOWLEDGE of the transport properties of melts of geoiogic interest is required in order to model their behavior during petrogenetic processes. In particular, cationic and anionic diffiisivities in silicate melts allow us to relate time, temperature, and physical scale in processes where diffusion is the ratelimiting step. Examples of such igneous processes include both melt-vapour and melt-crystal interactions (e.g. growth of zoned minerals, vapour phase transport of dissolved metals during magma degassing, crustal assimilation) and intra-melt processes such as thermogravitational diffusion and double-diffusive convection. The starting materials were (1) fluorine-bearing jadeite glass prepared from magent-grade sodium carbonate, alumina, aluminum fluoride and purified quartz sand and (2) jadeite glass prepared from a gel, dehydrated at 800°C for two hours. These starting glasses were analysed for Na, Al and Si by energy dispersive analysis using an ARL SEMQ microprobe fitted with an EEDS-ORTEC energy dispersive spectrometer. Operating conditions were 15 kV accelerating voltage, 4 nA sample current and 240 set counting time. The fluorine content of the fluorine-bearing starting glass was determined by neutron activation anaIysis as described in DINGWELL et al. (1984). Analyses of the starting glasses are presented in Table 1. Fluorine, like water, has a considerable effect on many properties of silicate melts including viscosities (DINGWELL et ai., 1984) phase equilibria (MANNING et al., 1980), melt-vapour partitioning (HARDS, 1978), and, as discussed below, component diffusivities in the melt. The occurrence of fluorine-rich amphiboles and micas in the lower crust and upper mantle (SMITH et al., 1981; VALLEY et al., 1982) and the suggestion that many relatively dry, fluo~ne-ash, melts originate from these regions (HARRIS and MARRINER, 1980; BURT et al., 1982) indicate the need for a better understanding of the role of fluorine in melts at high pressures. Such considerations, along with the potential for insights into the structure of Fand H&-bearing melts, prompted this study. Jadeite melt was chosen for this study because it has been used as a model for polymerized silicate melts in several studies (e.g. melt viscosity, KUSH~RO, 1976; oxygen diffusivity, SHIMIZU and KUSHIRO, 1984; cationic diffisivities, KUSHIRO, 1983; Raman spectra, SHARMA et al., 1979, SEIFERT et al., 1982; fluorinebearing melt viscosity, DINGWELL et al., 1984). The diffusion couple technique of KUSHIRO (1983) was used for this study. Glasses were ground in an agate mortar and packed into (5 mm diameter by 8-10 mm length) platinum capsules using a tight-fitting, stainless steel tool. The denser, fluorine-gee powder was packed into the capsule first and occupied the lower end of the vertical diffusion couple in all experiments. Packed and crimped capsules were dried at 8OO’C for 10 minutes and immediately welded. Sealed capsules were packed with hematite powder into % inch furnace assemblies with tapered graphite heaters, which reduce the temperature gradient along the 10 mm capsule to 15°C (KUSHIRO, 1976). The hematite acts as a trap for any water diffising into the charge from the assembly. The assemblies were stored in a drying oven at 1lO’C prior to use. Temperatures were monitored with a Pt/Ptl3Rh thermocouple without any correction for pressure, and are believed accurate to better than +lO”C. Pressures were monitored continuously with a Bourdon-tube gauge and were accurate to within *OS kbars. Pressure calibrations were performed using the melting curve of NaCl (CLARK, 1959) and a pressure correction of -7% was applied to all runs. Run durations were one hour with the exception of a zero-time experiment and one % hour experiment (run numbers 12 and 10, respectively). The zero-time experiment provided confirmation of an initially flat interface. Runs were quenched by switching off the power to the heater resulting in quench rates of approximately 125”C/sec. Quenched runs (encased in coarsely recrystallized specular