2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 25
Presentation Time: 8:00 AM-12:00 PM

REDOX STATE, SULFUR FUGACITY AND GAS CHEMISTRY AT VOLCAN VILLARRICA, CHILE


KRESS, V.C., ORTIZ, M.D. and GHIORSO, M.S., Earth and Space Sciences, Univ of Washington, Box 351310, Seattle, WA 98195-1310, kress@u.washington.edu

Volcan Villarrica is a 2847m stratovolcano in central Chile. Recent eruptive products comprise remarkably uniform basaltic andesites. Villarrica has maintained an active lava lake at its summit for over 17 years. This lava lake could not have survived for so long without sustained supply of heat by magma convection in the conduit.

As part of a larger study of the Villarrica volatile budget (Witter, Ph.D., U.W., 2003) we performed measurements designed to constrain the redox state of the active lava lake. Samples were obtained from two minor extrusive events. Pele's hair samples were obtained by Kress on 2/3/00 during minor eruptive activity. Reticulite samples were retrieved by Witter on 2/29/00, within 3 days of another minor eruption.

Ground samples were hand-picked to obtain nearly pure glass separates. Iron redox ratios were determined by combining electron microprobe measurements of total iron with wet-chemical determination of FeO. The uniformity of measured redox ratios suggests that these samples did not re-equilibrate substantially during quench. Total variation in redox state from the reticulite samples is much larger than the difference between the reticulites and Pele's hair samples. This suggests that redox state remained constant during this interval. Fe redox ratios were used to derive fO2 estimates using the calibration of Kress and Carmichael (1991). Results indicate an fO2 0.2 ± 0.3 log10 units above NNO.

Melt compositions and oxygen fugacity estimates were combined with inclusion S contents from Witter (2003) to estimate the fS2 of the lava lake. This was done using the 1400°C S solubility calibration of O'Neill and Mavrogenes (2002) and applying the solubility temperature dependence from Wallace and Carmichael (1992). Results indicate log10(fS2) near -1.4 at 1040°C. This is 0.2 log10 units above FMQP.

These results were combined with SO2 and halogen measurements from Witter (2003) to constrain the composition of the gas phase. The C/H ratio of the gas is not well constrained in these calculations. It is clear, however, that one cannot simultaneously achieve such high fO2 and fS2 in a gas stream with < 40% SO2. Such high S contents are unexpected. They may indicate that the magma reservoir at Villarrica is particularly enriched in S, or more likely depleted in H2O and CO2.