Diffusion of helium isotopes in silicate glasses and minerals : implications for petrogenesis and geochronology

Abstract

Helium mobility in geologic materials is a fundamental constraint on the petrogenetic origins of helium isotopic variability and on the application of radiogenic and cosmogenic helium geochronology. 3He and 4He volume diffusivities determined at 25-600°C in basaltic glasses by incremental-heating and powder storage experiments (using a diffusion model incorporating grain size and shape information to obtain high precision) are three to four orders of magnitude greater than for common cations. Diffusion in tholeiitic glass can be described by an Arrhenius relation with activation energy = 16.85±.13 Kcal/mole and log Do = -2.37±.06, although low temperature data are better described by a distribution of activation energies model . The best estimate for D at 0°C in tholeiitic glass is 5±2 x 10-16 cm2/s, an order of magnitude higher than the results of Kurz and Jenkins (1981) but lower than suggested by Jambon, Weber and Begemann (1985). Measurements in an alkali basalt show that helium diffusion is composition dependent (Ea = 14.4±.5 Kcal/mole; log Do = 3.24±.2), and roughly five times faster than in tholeiites at seafloor temperatures. The corresponding timescales for 50% helium loss or exchange with seawater (1 cm spheres) are about one million years for mid-ocean-ridge- basalts, and about 100,000 years in seamount alkali basalts. Radiogenic 4He diffusion has a higher activation energy (27±2 Kcal/mole; log Do = +2.4±1.0) than inherited (magmatic) helium, suggesting very low mobility (D = 3xl0-19 cm2/s at 0°C; factor of 5 uncertainty) and that U+Th/4He geochronology of fresh seafloor basalt glasses is unlikely to be hampered by helium loss. Measured isotopic diffusivity ratios, D3He/D4He, are not composition dependent, average 1.08±.02, and vary slightly with temperature, consistent with an activation energy difference of 60±20 cal/mole. This result differs from the inverse-square-root of mass prediction of 1.15, and may be explained by quantization of helium vibrational energies. These results suggest preferential loss of 3He will be minimal at low temperature (D3He/D4He = 1.02± .03 at 0°C). Therefore, alteration of magmatic 3He/4He ratios in basaltic glasses on the seafloor will occur only by helium exchange with seawater, and be important only for samples with low helium contents (<10-8 ccSTP/g), such as those found in island arc environments. Extrapolating the glass results to magmatic temperatures yields diffusivities similar to melt values, and suggests D3He/D4He approaches 1.15 at these and higher temperatures. Helium diffusivities in olivine and pyroxene at magmatic and mantle temperatures (900-1400°C) are higher than for cations, (E = 100±5 Kcal/ mole, log Do = +5.1±.7; and 70±10 Kcal/mole, log Do = +2.1±1.2, respectively), but are still too low to transport or homogenize helium in the mantle or even in magma chambers. However, diffusion equilibrates melts and mantle minerals within decades, and interaction with wall-rocks may be enhanced for helium in comparison to other isotopic tracers because of its greater mobility. Rapid exchange of helium within xenoliths and with their host magmas set limits on origin depths and transport times for xenoliths which exhibit helium isotopic disequilibrium between minerals, or between the magma and the xenolith. Phenocrysts equilibrate helium too rapidly to exhibit zoned isotopic compositions, and are likely to retain magmatic helium quantitatively in rapidly cooled volcanic extrusives. The 100-fold higher He diffusivity in pyroxene than olivine at 1000°C allows diffusive loss effects to be evaluated in more slowly cooled rocks, when cogenetic minerals can be measured. Diffusivities of cosmic- ray produced 3He in surface exposed rocks are several orders of magnitude higher than for inherited helium. However, activation energies for olivine and quartz, 25±4 Kcal/mole (log Do = 3.7±.8) and 25.2±.9 Kcal/mole (log Do = +.2±.4) respectively, still suggest low diffusivities at surface temperatures of approximately 10-22 and 10-20 cm2/s. Equations for simultaneous helium production and diffusive loss allow model ages for surface exposure to be corrected for helium loss, and demonstrate that cosmogenic 3He geochronology will not be limited by helium loss for timescales of approximately 1 million years in quartz and 10 million years or more in olivine. The measurements also suggest that radiogenic 4He produced by U and Th decay may be a useful dating method in quartz. Application of the diffusion measurements demonstrates that part of the wide range of 3He/4He ratios (.01 to 9 Ra) of a suite of dredged basalts and andesites from the Woodlark Basin, (western Pacific) reflects post-eruptive helium addition, from seawater in glasses with low He contents and from U and Th decay in mafic mineral separates. In unaltered samples, 3He/4He ratios for tholeiites from the Woodlark Spreading Center are 8-9 Ra, similar to mid-ocean-ridges, but distinctly different than the ratio of 6.9±.2 Ra observed in Kavachi submarine volcano basaltic andesites. Helium isotopic systematics in cogenetic pyroxenes and olivines from these samples demonstrate that this is a magmatic signature, and not the result of preferential 3He loss by diffusion. Coupled Sr and He isotopic systematics in these and other samples from the region suggest the sub-arc mantle has been enriched in radiogenic helium supplied by subducted Pacific lithosphere.