Experimental measurements of 3He and 4He mobility in olivine and clinopyroxene at magmatic temperatures
Additional Document Info
In-vacuo heating of 0.5-0.7 mm olivine and clinopyroxene grains separated from Hawaiian ultramafic xenoliths leads to complete He loss within hours to days for temperatures of 700-1400-degrees-C. Diffusivities calculated from the observed release rates assuming spherical grains and initially homogeneous He distributions define Arrhenius relations with activation energies of 420 +/- 20 and 290 +/- 40 Kj / mol and log10 D0 of +5.1 +/- .7 and +2.1 +/- 1.2 cm2/S in olivine and pyroxene, respectively. Values at 1350-degrees-C are 5.3 X 10(-9) cm2/S in olivine and 10 times faster in pyroxene (4.7 X 10(-8) CM2/S) . These values include small corrections for grain size variations and, in the case of olivine, about 15% prior diffusive He loss. However, an important factor that has not been considered in previous studies of this type, is that the xenolith He resides predominantly within CO2 rich fluid inclusions. Theoretical description of He loss in such a case demonstrates that the diffusivities calculated using the standard approach actually represent the product of the true volume diffusivity (D) and the helium solubility, as represented by the distribution coefficient K(Dv) (defined by C(crystal)/ C(fluid)). Although He solubility in crystals is not well determined, estimates based on the CO2 concentrations in these samples suggest that it is very low (K(Dv) of 3 X 10(-4) for pyroxene and 6 X 10(-6) for olivine; which also implies low crystal-melt distribution coefficients of .05 and .00 1 ). The resultant corrected diffusion rates are significantly faster than those obtained by the standard approach (approximately 10(-4) CM2 /s at 1350-degrees-C and are thus higher than basaltic melt values). The most reasonable interpretation of this result is that He release is enhanced by internal grain fractures, including the planar healed cracks along which most fluid inclusions are arrayed. This treatment illustrates the difficulties involved in extrapolating laboratory He release. measurements to nature, in particular the strong influence of mineral defects. The diffusivities reported here probably represent upper limits for mantle He transport or magma-phenocryst He exchange. As such, they imply that He transport over kilometer length scales in the mantle is dominated by convection rather than diffusion, and that phenocrysts in extrusive rocks will retain most of their pre-eruption helium contents. In combination with the observation that He-3 diffuses only marginally faster than He-4 (4 +/- 4% faster in pyroxene and 9 +/- 4% faster in olivine) this implies that significant isotopic fractionation of residual helium contents is unlikely to occur.