# The Effect of Sulfur on the Fe2+/Fe3+ Ratio of MORB and its Implications for the Redox State of the Mantle.

Abstract : Assessing the redox state of the mantle source of Mid Ocean Ridge Basalts (MORBs) is important since variations of fO2 affect a number of physical and chemical properties of Earth's materials including magmas. The iron redox ratio of glassy MORBs has been widely used to infer the redox state of the magma prior to its eruption on seafloor and by, inference, that of its source zone, the upper depleted mantle (Christie et al., 1986 ; Bezos and Humler, 2005). Such an approach relies on the use of empirical equations between iron redox ratio and fO2 calibrated over a range of melt compositions (Kilinc et al., 1983 ; Kress and Carmichael, 1991 or KC). The application of these geobaromaters to the glassy rims of oceanic lavas in mid ocean ridge settings have yielded a range of fO2 values centered at around NNO-1.8 (referenced to the NNO solid buffer). However, the empirical equations have been calibrated on S-free systems, yet MORBs have S concentrations of about 1000 ppm (Wallace and Carmichael, 1992). We have therefore experimentally investigated the role of S on the iron redox state on a model MORB composition, at 1300$\deg$C and under variable fO2 conditions. Experiments were done at 1 bar, using 1 atm gas (CO-CO2-SO2) mixing furnaces equiped with Ca stabilised zirconia cells to monitor fO2. The synthetic composition was chosen to match that used in recent work aimed at constraining S solubility in silicate melts (O'Neill and Mavrogenes, 2002). Melt compositions were determined via EPMA, while the amount of Fe2+was determined volumetrically. We first performed experiments without S to check for consistency between our experimental procedure and those of previous work. Time series experiments reveal that redox equilibrium is achieved between 3 to 6 hrs, depending on fO2. Comparison between predicted Fe3+/Fe2+as constrained by the KC model, and observed values at steady state, shows excellent agreement, experimental values being always identical to calculated ones, within analytical error. In a second step we performed experiments during which we progressively increased fS2 at constant fO2. The value of fS2 was chosen so as to impose sulfide undersaturated conditions, based on O'Neill and Mavrogenes (2002) results. Time series experiments show that an Fe3+/Fe2+ steady state value is reached after 3-6 hrs. Two fO2 were explored : at NNO-2.8, sulfur bearing melts display a smaller, yet measurable, Fe3+ content than without S. At NNO-0.9, the same effect is observed : the S bearing melt has a Fe3+/Fe2+ ratio of 0.033, ie much lower than the value of 0.069 obtained under S free conditions. The results, therefore, suggest that the redox state of the mantle as deduced from iron redox ratio of quenched oceanic lavas may have been greatly underestimated. For instance, the KC model applied to our NNO-0.9 experiments yields an fO2 of NNO-2.3.
Document type :
Conference papers

https://hal-insu.archives-ouvertes.fr/hal-00078659
Contributor : Nathalie Pothier Connect in order to contact the contributor
Submitted on : Wednesday, June 7, 2006 - 9:45:54 AM
Last modification on : Wednesday, November 24, 2021 - 1:24:02 PM

### Identifiers

• HAL Id : hal-00078659, version 1

### Citation

Yann Graz, Bruno Scaillet, Michel Pichavant, Fabrice Gaillard. The Effect of Sulfur on the Fe2+/Fe3+ Ratio of MORB and its Implications for the Redox State of the Mantle.. AGU, 2006, Baltimore, United States. ⟨hal-00078659⟩

### Metrics

Les métriques sont temporairement indisponibles