Woods Hole Oceanographic Institution

Dr. Nicole Keller

» Sulfur Cycling at Convergent Plate Margins

» Lau Basin

Enlarge image
Poas Volcano, Costa Rica

Enlarge image
El Chichon Volcano, Mexico

Sulfur cycling at convergent plate margins

Dr. Alison Shaw (WHOI)

Prof. Shuhei Ono (MIT)

Prof. Tim Grove (MIT)

Dr. Andrew Berry (Imperial College, London)

The aim of this project is to use S isotopes to gain insights into the cycling of sulfur in subduction zones. Recent global models have shown that sulfur is intimately linked to global carbon and oxygen cycles, and a number of studies have stressed the importance of these interconnected volatile cycles on the earth’s redox history (e.g. HAYES and WALDBAUER, 2006). Most of the older models of global fluxes did not take into account processes linked to the subduction of sulfur (e.g. GARRELS and LERMAN, 1981), and only recently have scientists started to include estimates of sulfur output due to volcanoes, hydrothermal circulation as well as metamorphic pathways into global mass balance models (CANFIELD, 2004; BERNER, 2006). An estimate by WALLACE (2005) suggests that only 15-30% of the sulfur subducted into the mantle is returned to the surface. However, as the mechanisms and processes involved in the recycling of sulfur through subduction zones are not well constrained, the uncertainties of the total fluxes remain large (e.g. CANFIELD, 2004). Part of the issue is due to the fact that it is unclear where the sulfur which is exiting the subduction system comes from (CANFIELD, 2004), and what the extent of the contribution from the subducting slab might be (e.g. DE HOOG et al., 2001).

In order to answer these questions, sulfur isotope systematics (32S and 34S) of volcanic arcs have been used to trace the fate of the sulfur: because mass-dependent fractionation processes are strongly temperature dependent with increased effects at low temperatures, it is usually assumed that processes at magmatic temperatures will not change the isotopic ratios significantly, as was shown in model calculations by ALT et al. (1993) and DE HOOG et al. (2001). This implies that the isotopic signature of the output may relate to the source of the sulfur, if effects such as partial melting, crystallisation of a sulfur-bearing phase and degassing can be accounted for (e.g. WOODHEAD et al., 1987; ALT et al., 1993; DE HOOG et al., 2001). If the sulfur is derived from the subducting slab, then its isotopic signature is potentially linked to the pre-subduction biogeochemical history of sulfur, since biological activity also strongly affects isotopic fractionation. However, the isotopic composition of the sulfur-bearing phases on the downgoing slab are not well known, and on the output side, it is unclear how late-stage effects such as degassing and fluid-rock interactions affect the isotope ratios. Therefore, it is still unclear how the output signatures relate to the input signatures.

To address the problem, I use several approaches:

1) 34S/32S in olivine-hosted melt inclusions from Costa Rican volcanoes (Collaboration with Dr. Alison Shaw, WHOI):

A record of the pre-eruptive melt can be found in small glass inclusions in crystals. Inclusions in early crystallising phases are thought to record the original melt composition much more faithfully than more evolved and/or erupted materials, and may carry information which would be lost during degassing or subsequent alteration of the rock. In particular, they have retained the volatile contents of their parent melt (H2O, CO2, S, Cl – e.g., LOWENSTERN, 2003), thus making them far more suitable for understanding degassing processes than erupted material. Of particular interest are inclusions found in olivine, since olivine generally crystallises from relatively primitive melts. Using the Cameca 1280 Ion Probe at WHOI (, I will determine the chemical composition of melt inclusions from five volcanoes in Costa Rica, as well as their 34S/32S ratios.

2) Quadruple sulfur isotope systematics in volcanic environments (collaboration with Prof. Shuhei Ono, MIT).

Very recent advances in high precision isotopic measurements of quadruple sulfur isotopes (32S, 33S, 34S and 36S) have shown promising results for modern terrestrial samples such as seafloor hydrothermal vent sulfides (ONO et al., 2007) and biogenic sulfides (ONO et al., 2006). At this stage, the application of quadruple sulfur isotopes has not been extended to any type of arc-related volcanic rocks and gases; yet such materials usually contain several thousand ppm of dissolved sulfur that has cycled through the subduction factory. Using a laser fluorination isotope-ratio-monitoring GC-MS, we are conducting a systematic survey of quadruple sulfur isotopes in all sulfur-bearing materials found on a volcano: Volcanic gases (separate values for H2S and SO2), native sulfur, dissolved sulfur in volcanic lakes, tephras. Several volcanoes are used to establish a database for quadruple sulfur values in volcanic environments: - Poas and Turrialba, Costa Rica - El Chichon, Mexico Furthermore, we are planning on analysing sulfur in submarine volcano sulfur ejecta (NW Rota, Guam), as well an in back-arc basin basalts from the Lau Basin.  

3) Quadruple sulfur isotope systematics of subducted material (collaboration with Prof. Shuhei Ono, MIT).

A study by ONO et al. (2007) shows that sedimentary sulfides are characterized by slightly enriched Δ33S even when δ34S values are identical; the authors conclude that the Δ33S signature can be used as a tracer for the interaction between seawater, oceanic crust and microbial activity in subseafloor hydrothermal cycles. Prof. Ono and his collaborator Dr. O. Rouxel (WHOI) have been working on the multiple-sulfur isotope systematics of sulfides in sediments and altered basalt from an ODP core from western Pacific (801c) (ROUXEL et al., in review). I will extend this to other drill cores to constrain multiple-sulfur isotope ratios of subducting oceanic crust and trace recycling of sedimentary sulfide through subduction zones. The comparison between all volcanic sulfur-bearing phases will enable us to establish to what extent late-stage frationation processes affect sulfur isotopes; the comparison between melt inclusion data and volcanic gases will allow us to estimate the order of magnitude of fractionation in the upper part of a volcanic system; Finally, in combination with the study of subducted material, I will be able to establish the quadruple sulfur isotope ratios of materials cycling through the subduction process, and establish whether the sulfur isotopic signature of the source can be preserved through the multiple processes taking place between input (pre-subduction) and output (volcanic arc).

4) Experimental determination of fractionation factors (collaboration with Prof. Tim Grove, MIT).

The goal of these experiments is to determine the fractionation factors for 34S and 32S in magmatic systems; In particular, we will investigate the effect of volatile exsolution from a cooling magma on the sulfur isotopic composition of the magma.

5)    Relationship between sulfur isotopes, sulfur speciation and oxygen fugacity of the system using synchrotron-based spectroscopy (XANES; collaboration with Dr. Andrew Berry, Imperial College, London).

X-Ray Absorption Near Edge Spectroscopy (XANES) is a technique that can be used to gain information on the oxidation state and coordination of an element in a melt or a fluid (e.g. MOSBAH et al., 1999; BERRY et al., 2003). The use of a synchrotron microbeam enables the in-situ study of melt inclusions. Fe2+/Fe3+ ratios may be used to constrain the oxygen fugacity at time of entrapment, and S2-/S6+ ratios allow the determination of the speciation of sulfur in a melt (BONNIN-MOSBAH et al., 2001; BONNIN-MOSBAH et al., 2002; METRICH et al., 2005). This novel approach, combined with sulfur isotopes, will help me understand the link between the sulfur speciation, the prevailing oxygen fugacity, and the isotopic composition of a primitive sub-arc melt.

By combining novel quadruple isotopic systematics in both input and output materials, conventional isotopic systematics in undegassed melts and volcanic gases, experimental work and synchrotron-based x-ray spectroscopy, I will obtain new and improved data which is needed to fingerprint of the source of the sulfur and gain a more profound understanding of the recycling of sulfur through the subduction factory. These results will allow me to establish better quantitative mass balances of sulfur in particular, and volatiles in general, in the context of subduction zones.

© Woods Hole Oceanographic Institution
All rights reserved