REDUCTION OF BASALTIC MELT DURING NEAR-SURFACE EVOLUTION OF SO2: A POSSIBLE EXAMPLE FROM KILAUEA

As basaltic melts approach the Earth’s surface, exsolution of dissolved volatiles (CO₂, SO₂, H₂O) accelerates. Dissolved sulfur, much of which occurs as sulfide (S^-2) at depth in Fe-rich melts, reacts with iron in the melt to exsolve as SO₂. The valence of sulfur changes from -2 to +4, so this reaction reduces six Fe⁺³ cations to Fe⁺² for each sulfide ion oxidized. This reduction must occur in shallow storage because sulfur degassing becomes significant only at low pressures (<10 MPa). However, evidence for near-surface reduction of basalt (from e.g. crystal zoning) is rare.

The picritic 1959 summit eruption of Kilauea produced highly vesicular scoria, consisting of glass plus 10-30% olivine (Fo_86.5). Most exhibit evidence for very recent mixing and rapid transport, prior to quenching. Microprobe data for olivine in textural equilibrium with their host melts (microphenocrysts and euhedral rims on olivine phenocrysts), together with analyses of the host glasses, are available for 15 scoria samples. By allowing ferric iron contents of the melt to vary from 10% to 30%, one can calculate a range of (Fe/Mg)_ol/(Fe/Mg)_melt ratios (K_D) for all olivine-glass pairs. For the microphenocrysts (Fo_77.4-87.3), the equilibrium K_D (=0.30±0.03) is best matched at ~15% Fe₂O₃. For euhedral phenocryst rims (<10 microns of the edge; Fo_79.3-88.0), the best match is at 25-30% Fe₂O₃.

Ferric iron contents of the eruption glasses vary from 11-21%, with 10 of 15 samples having 13-18% Fe₂O₃. Equilibrium K_D values for the olivine microphenocrysts lie in this same range (~ 15% Fe₂O₃), so the microphenocrysts are approximately in equilibrium with their host glasses at observed Fe₂O₃ contents. The olivine phenocryst rims generally reflect more oxidizing conditions, with only 7 of 68 points in equilibrium with the present glasses. This shift in ferric/ferrous ratio must be very recent, so reduction by evolution of SO₂ in the near-surface magma chamber seems like the most probable cause. An important implication is that the observed ferric/ferrous ratio of even the freshest subaerial basalt may not be inherited from the mantle source.