Solubilities of nitrogen and noble gases in silicate melts under various oxygen fugacities: implications for the origin and degassing history of nitrogen and noble gases in the earth
Solubility experiments for nitrogen and noble gases (Ar and Ne) in silicate melts were conducted using two experimental configurations: one was conducted at 1 atmospheric pressure, T =1300°C and oxygen fugacity ( fO 2) of IW + 0.9 (i.e., 0.9 log units higher than the iron-wüstite buffer) and the other at high pressures ( Ptotal ∼ 2 × 10 8 Pa), 1500°C and fO 2 ∼ IW + 6. For the former experiment, isotopically labeled-nitrogen ( 15N 15N-enriched) was used to distinguish dissolved nitrogen from contaminating atmospheric or organic nitrogen and to examine dissolution mechanisms of nitrogen in silicate melts. The results obtained for the two series of experiments are consistent with each other, suggesting that Henry's law is satisfied for fN 2 of up to ∼250 atm (2.5 × 10 7 Pa). The results are also consistent with our earlier results (Miyazaki et al., 1995) obtained at highly oxidizing conditions ( fO 2 ∼ IW + 10). All these results support physical dissolution of nitrogen as N 2 molecules in silicate melts for fO 2 from ∼IW + 10 down to ∼IW. The observed solubility (Henry's constant) of nitrogen (3-5 × 10 -9 mol/g/atm) is comparable to that of Ar (2-4 × 10 -9 mol/g/atm), and much lower than that of Ne (11-14 × 10 -9 mol/g/atm) at 1300°C. A preliminary experiment was also performed for partitioning of nitrogen and noble gases between clinopyroxene (cpx) and basaltic melt using a piston cylinder-type apparatus at 1.5 GPa and at 1270 to 1350°C. The obtained cpx/melt partition coefficient of nitrogen is 0.06, slightly lower than those of noble gases (∼0.1 for Ne to Xe), suggesting that nitrogen is as incompatible as or even slightly more incompatible than noble gases. The present results imply that a large nitrogen/Ar fractionation would not be produced by magmatic processes. Therefore, the two orders of magnitude difference between the N 2/ 36Ar ratios in the Earth's atmosphere (∼10 4) and that in the mantle (∼10 6) must be explained by some other processes, such as incomplete segregation of metal blobs into the core and their later oxidation.