The role of sulphur in upper mantle processes

This thesis consists of a review of experimental studies of sulphur solubility in mafic magmas and the stability of magmatic sulphides, a theoretical examination of the thermodynamics of C-O-S-H fluids to high temperature and pressure and three experimental studies examining the role of sulphur under both volatile absent and volatile saturated conditions. Data concerning the sulphur solubility in mafic melts comes from experimental studies on natural basalts and from metallurgical sources concerned with the equilibrium between silicate slags and liquid metals. All of the metallurgical studies and most of those involving natural compositions have been conducted at one atmosphere and have identified the iron activity and the ambient oxygen and sulphur fugacities as being the major factors controlling the sulphur solubility in a silicate melt. Studies at pressures greater than one atmosphere are few and have examined only a few compositions under relatively unconstrained sulphur and oxygen fugacities. Therefore there is a lack of sulphur solubility data at high pressures and the partitioning of sulphur between solid, melt and fluid phases remains an unknown of potential significance. Recent models for the genesis of MORB have emphasized the non-primary nature of the majority of MORB glass compositions. The compositions of a range of sulphide saturated potential MORB parent picritic liquids in equilibrium with either lherzolitic or harzburgitic assemblages have been determined. The liquids produced are similar to those generated from a similar but sulphur free source composition (pyrolite) indicating that the presence of sulphur does not have the same fluxing effect as either CO\(_2\) or H\(_2\)O. Whereas volatile-poor conditions is appropriate for the modelling of MORB, xenolithic material from the upper mantle frequently shows evidence of fluid activity. A thermodynamic model for super-critical C-O-S-H fluids at high temperature and pressure has been derived. Available P-V-T data for sulphur bearing species has been combined with data for C-O-H species to calculate fugacity coefficients by means of a modified Redlich-Kwong equation of state. Using fugacity coefficients for multicomponent fluids the species distribution as has been determined as a function of the intensive variables P, T, fO\(_2\) and fS\(_2\). The results show that reduced fluids (f02≤1W) may contain a significant sulphur fraction as H\(_2\)S at moderate fS\(_2\) (IT+1) whereas more oxidized fluids (fO\(_2\)≥GC0) by comparison contain little sulphur even at high fS\(_2\). SO\(_2\) is not a significant fluid component in the range of fO\(_2\)'s thought to characterize the upper mantle redox range (IW<fO\(_2\)<NNO). Two series of experiments have been carried out to examine the role of sulphur in fluid saturated systems. Both employ simultaneous buffering of both sulphur and oxygen by a tungsten based buffer assemblage (fO\(_2\)~1W+1; S\(_2\)~IT+1). This assemblage when combined with carbon saturation results in invariant fluid compositions. The first examines the system olivine-C-O-S-H. A technique is developed by which fluid inclusions are induced to form in single crystals of natural olivine (Fo\(_{90}\)) from a tungsten assemblage buffered C-0-S-H fluid (H\(_2\)0>CH\(_4\)). By analysis of the quenched fluid from the capsule, the relationship between the parent fluid and the daughter fluid inclusions may be examined. Laser Raman micro analysis of inclusions formed from a variety of pressure/temperature combinations show compositional variation between individual inclusions and between inclusions and the parent fluid showing that the assumption of isochemical entrapment may not be valid for fluid inclusions formed under mantle conditions. Inclusions formed also show development of daughter crystals indicating appreciable cation solubility in the fluid phase at the run conditions. Infra-red spectrometry of the host olivine shows development of structural O-H during the experiments revealing a possible 'sink' for inclusion hydrogen; a model for post entrapment chemical change in the fluid phase is devised which has implications for the origin of CO2 rich inclusions in mantle xenoliths. In the second series of experiments the same oxygen/sulphur fluid buffer technique has been used to explore the fluid saturated phase relations for the system pyrolite-C-0-S-H. The condition of melting (solidus) appears to coincide with the disappearance of a hydrous phase (amphibole and/or phlogopite) and occurs midway between the reduced C-0-H solidus as determined by Taylor and Green(1988) and the water saturated solidus. Fractured natural olivine crystals were incorporated into the capsule assembly and the inclusions produced show fluids dominated by CH\(_4\)-H\(_2\)O mixtures in agreement with the analysis of the capsule gasses by capsule piercing/mass spectrometry. In addition to these volatile phases substantial amounts of daughter crystals are observed in the fluid inclusions from conditions below the 'solidus' indicating significant cation solubility in supercritical CH\(_4\)-H\(_2\)O mixtures. At higher temperatures this solid component increases continuously in abundance until true melt inclusions are obtained at conditions above the solidus. These sulphide/silicate compositional relationships are examined and a model derived which at a given pressure and temperature relates the compositions of olivine, orthopyroxene and sulphide to the ambient sulphur and oxygen fugacities. Such a reaction may be used as an oxygen/sulphur fugacity sensor if an independent means of estimating the temperature and pressure is available