Research
Research Interests
I am conducting research that combines economic geology, metamorphic and igneous petrology, and structural geology to study metallogenic processes. I am investigating a range of mid- to deep-crustal processes that influence metal distribution and mobility including: (1) sulfide melt immiscibility in magmatic mixing zones at the base of arcs (MASH zones) and its role in ore genesis; (2) the P-T-t-D window(s) for metamorphic generation of sulfur-rich hydrothermal fluids; (3) the behaviour of metals during silicate anatexis; and (4) sulfide partial melting during metamorphism. An important aim is to link the results of these studies with tectonic models for the genesis of felsic-intermediate magma-related Cu-Au deposits (relates to 1 and 3) and orogenic gold deposits (relates to 2).
Sulfide Melt Immiscibility in MASH Zones
Historically, there has been considerable debate as to whether a metal-enriched magma is necessary for genesis of Porphyry Cu-Au deposits. It has been demonstrated through modelling that a small granitoid stock with 50 km3 of magma contains enough metal to form a typical porphyry Cu deposit. Intuitively, ore deposits should be more likely to form from a relatively metal-rich magma and indeed recent research suggests that the parent magmas of some porphyry Cu deposits were enriched in Cu and S. Therefore, it is important that we understand processes that lead to metal- and sulfur-enrichment, as well as processes that inhibit this enrichment or even cause depletion.
Tectonomagmatic evolution of arc magmas
Click image for a larger view
Intermediate to felsic arc magmas are thought to evolve through a protracted sequence of events intimately tied to the tectonic setting. Initially, hydrous and oxidised basaltic magmas form in the metasomatised mantle wedge above a subducting oceanic slab. Their oxidised nature allows them to contain more sulfur and metals than would be otherwise possible. These rise and pond beneath, or within, the crust, where they begin to crystallise. In continental arcs and some complex island arcs, the resultant heating causes partial melting of the crust, liberating melt of granitic composition with comparatively low sulfur and metals. Mixing of these crust- and mantle-derived melts produces volatile-rich intermediate magmas with sufficiently low density to rise buoyantly through the crust. In simple island arcs, mixing may instead occur between primitive mantle-derived magmas and more evolved (more felsic) magmas that formed through fractionation of earlier magma pulses. This zone of magma mixing at the base of arcs is termed a MASH zone (MASH = Melting, Assimilation, Storage, Homogenisation). Within these deep-crustal mixing zones, mantle-derived magmas are thought to transfer their metals and sulfur to the rising felsic-intermediate magmas, and although this is thought to be a key step in the formation of porphyry Cu-Au deposits, it is yet to be shown in field examples.
Sulfide Droplets
Click image for a larger view
The redox state of arc magmas is thought to be controlled by hydrothermal fluids and/or magmas derived from the subducted oceanic crust, which oxidise the overlying mantle wedge. For porphyry Cu-Au deposits, oxidation of the mantle wedge to conditions within the sulfate stability field is thought to be important because it allows metals and S to be retained within the magma, rather than fractionating into a residual sulfide liquid. However, it is thought by some that formation of sulfide melts at some stage of the magmatic evolution may be important for the formation of this deposit type. To form magmatic sulfide liquids, ultramafic to mafic magmas must become saturated in sulfur. This can be done by raising the sulfur content of the magma or by lowering the sulfur content at sulfide saturation (SCSS). A range of processes can lower the SCSS, but the most effective is by lowering the FeO content of the magma, which occurs during mixing with a more felsic magma. Therefore, base-of-arc MASH zones are places where intruded mafic magmas are likely to become saturated in sulfur (i.e., through mixing) and exsolve a sulfide melt (Fig. 2). Exsolution of sulfide melt would lead to variable metal enrichment and depletion within the silicate magma, magma-migration dynamics then becoming important in determining whether metal-rich magmas can migrate upwards. However, if the mafic magma is oxidised, it may instead exsolve a sulfate melt or crystallise a sulfate mineral.
The solubility of S in a mafic melt is approximately an order of magnitude greater in the sulfate stability field (oxidised) than the sulfide stability field (relatively reduced). One mechanism that may therefore be important is mixing between oxidised S-rich mafic magma and relatively reduced crust-derived felsic magma. Nearly all lower crustal metamorphic rocks are sufficiently reduced to stabilise sulfides rather than sulfates, many with ƒO2 ≈ FMQ, or ≈ CCO in graphitic rocks. Therefore, mixing between oxidised and reduced magmas may be commonplace in MASH zones. Equilibration of the redox state between these magmas to conditions within, or on the edge of, the sulfide stability field is likely to promote extensive exsolution of sulfide melt if the oxidised mafic magma is very S-rich. In addition, to being important to the metallogenic evolution of felsic to intermediate magmas, this may also be one previously unrecognised mechanism for forming magmatic Ni-Cu-PGE deposits.
The Hidaka Belt in Japan, is an ideal location to conduct field work for this project. Here, in an exposed island arc cross-section, a large lower to mid-crustal mafic intrusive complex is preserved that contains domains where extensive mixing having taken place between the mafic magmas and tonalitic magmas derived from partial melting of lower crustal metasediments. Intimately associated with these mixed magmas are some magmatic sulfide bodies. Some of these are massive, indicating metal extraction from the hybrid magma, whereas other mafic-intermediate rocks have disseminated and elevated concentrations of sulfides indicating metal enrichment of a particular phase of magma. This project will involve Roberto Weinberg and PhD student, Kyle Rebryna, and will use field studies in combination with experimental studies to elucidate the role of magma mixing in controlling metal distribution in arc magmas. Piston cylinder experiments will be conducted to further advance the theoretical model of magma mixing represented by Figure 2.
Windows of Metamorphic Sulfur Generation in the Crust
Sulfur is intimately associated with many types of ore deposits and various sulfide minerals are our dominant resource for a number of metals. This sulfur is sourced from many geological regions. For example, the sulfur in most magmatic nickel deposits comes from the upper mantle, whereas that in SEDEX deposits is dissolved in deep basinal brines. There are also various types of metamorphogenic ore deposits that derive their sulfur from regional metamorphic fluids. Orogenic gold deposits are a well-known type of metamorphogenic ore deposit, which form over a range of P-T conditions, but aredominantly found in rocks metamorphosed at conditions near the greenschist-amphibolite facies transition. Other purportedly metamorphogenic deposits are also thought to have been deposited at similar metamorphic grades.
The source of this metamorphic sulfur has been hypothesised for some time to be largely derived from breakdown of pyrite (FeS2) to pyrrhotite (Fe1-xS) during regional metamorphism. Previous field-based studies of metamorphic sulfur production have tended to focus on graphitic metasedimentary rocks because these are typically pyrite-rich and thus easier to study. In these rocks pyrite has been found to break down over a broad range of temperatures from the lower amphibolite facies up to lower granulite facies. However, in many regions there are few or no carbonaceous rocks so these studies are not necessarily applicable, especially considering that these rocks are abnormally pyrite-rich and produce their own unique fluid (CO2-rich) and REDOX (reduced – on the CCO buffer) conditions not relevant in other potential source rocks. Pyrite is also found in trace quantities in many other metamorphic lithologies that may, through their greater volumes, be more important as source rocks on a regional scale.
Click image for a larger view
The aim of this research is to constrain windows of sulfur production in the crust by using a mass balance approach and the thermodynamic computer program Thermocalc to analyse pyrite breakdown in different bulk compositions. It is thought that most pyrite present in metamorphic sequences will breakdown in the lower amphibolite facies because these conditions coincide with: (1) a dramatic increase in the proportion of H2S required by hydrothermal fluids to remain in equilibrium with the pyrite-pyrrhotite buffer, and (2) a large increase in metamorphic fluid production through chlorite breakdown. This
work will also be applied to subduction zone metamorphism to investigate input of sulfur into the mantle wedge. Characterising the constraints on metamorphic sulfur production will influence our understanding of how, when and where metamorphogenic ore deposits form.
Metal Mobility During Silicate Anatexis
Felsic magma-related ore deposits are Earth’s largest source of Cu and Mo, and a major source of Au. Yet there have been few studies of processes that affect metal distribution in magma source regions. Some suggest that in arc regions metal is derived
Au-bearing migmatite from
the Challenger deposit
Click image for a larger view
from subducted oceanic slab or overlying metasomatized mantle wedge, remnant metallogenic heterogeneities in the mantle, adjacent intruded crust, or from the lower crust during anatexis. Intrusion-related Au systems form distally to arcs and are typically associated with reduced granitic magmas thought to be largely derived from partial melting of crustal rocks, casting into question the absolute need for oceanic slab or mantle involvement. Even within arcs, evolution of ore producing magmas may involve mixing between mantle- and crust-derived magmas. It is therefore crucial to understand processes that influence the metal content of crust-derived magmas.
P-T pseudosections calculated using Thermocalc showing the difference in melt mode produced between unaltered (a) and altered (b) host slate, if the Wattle Gully deposit were metamorphosed at granulite facies conditions.
 Click image for a larger view Together with Roberto Weinberg and Chris McFarlane at University of New Brunswick in Canada, I have compared melting of alteration zones in metamorphosed gold deposits with that of unaltered rocks of the same protolith to examine the relative contributions of metal-rich and metal-poor source regions to crust-derived magmas. The highest melt fraction is found to be generated in metal-rich, K-altered rocks. Sulfides and Au dissolve, and are physically incorporated into the resulting felsic melt, which thereby becomes metal- and S-enriched. Since the presence of melt significantly weakens rocks, strain preferentially partitions into these melt-rich mineralized zones promoting melt segregation there. As strain increases, high-melt mineralized domains become shear zones, expanding and linking with other incipient shear zones, ultimately connecting with regional magma migration networks, thereby enriching migrating magmas in metal.
Sulfide Melting During Metamorphism
Over the last several years I have investigated several ore deposits that were metamorphosed at conditions ranging from the middle amphibolite facies up to granulites facies. This work was done with John Mavrogenes at ANU, David Pattison at the University of Calgary, Ron Frost at the University of Wyoming and Eva Zaleski, who was at the Geological Survey of Canada. Descriptions of what we’ve found at two of the more interesting deposits are given here.
At the Challenger deposit gold mineralisation is concentrated in a series of shallowly plunging shoots that parallel the plunge of ptygmatically folded leucosomes and quartz veins. These leucosomes formed during granulite facies metamorphism, at 850°C or so and 750MPa, which caused partial melting of the host rock. Garnet, which formed via vapour-absent melting reactions during peak metamorphism, contains isolated spherical gold-sulfide inclusions, indicating that gold mineralisation was present prior to peak metamorphism. In relict quartz veins and in migmatitic leucosomes large isolated inclusions (up to 1mm) of gold with co-existing bismuth + arsenopyrite ± pyrrhotite ±
Inclusion in a relict quartz vein from
the Challenger deposit. ~150mm across.
Click image for a larger view
maldonite (Au2Bi) ± chalcopyrite are located at grain boundaries and also hosted in quartz, perthitic K-feldspar, plagioclase, cordierite and garnet grains. Au-Bi phase relations show that it is possible for gold to coexist with bismuth as a melt at temperatures as low as 241°C indicating that at least some of these inclusions were molten during peak metamorphism. Piston cylinder experiments performed on gold-rich samples from Challenger showed that the other minerals also were involved in this peak metamorphic polymetallic melt. It is thought that where bismuth co-existed with gold, pyrrhotite and
löllingite, an initial polymetallic melt was produced, prior to peak metamorphism. This polymetallic melt initially remained immobile as the melt:rock ratio was too low for melt migration to occur. When the host rock partially melted via vapour-absent melting, Phase relations in the system Au-Bi
Click image for a larger view
the melt/rock ratio increased to such an extent that melt migration became possible. Both the polymetallic and silicate melt had similar rheological properties relative to the unmelted residue, and as such were redistributed to form stromatic leucosomes. The two were, however,
immiscible and remained as separate entities within individual melt accumulations.
The gigantic Hemlo gold deposit in Ontario, Canada is a highly enigmatic deposit that has been the subject of intense debate over more than a decade. Most now acknowledge that mineralization was introduced prior to or during peak metamorphism, but a number of key observations that have been used to support a post-metamorphic model have until now remained unexplained. We have been able to explain how the present ore mineral assemblages, some of which are unstable even at greenschist facies conditions, came to be hosted in mid-amphibolite facies rocks and related structures, and how the heterogeneous distribution of ore minerals evolved. Phase relations indicate that some ore mineral
Complex mineral assmblage that crystallised
from an ~150mm Au-bearing sulfosalt
melt inclusion at Hemlo.
Click image for a larger view
associations at Hemlo would have been molten at peak metamorphic conditions (600-650°C, 6-7 kbar). For example, coexisting realgar and stibnite start to melt at <300°C, and coexisting gold and aurostibite start to melt at ~360°C. These and many similar ore mineral associations at Hemlo typically occur in dilational structures that developed approximately concurrently with peak metamorphism. Although there are signs of low temperature silicate alteration at some of these sites, many display no such alteration indicating that hydrothermal mobilization was not responsible for much of the observed distribution.
The preservation of widespread pyrite at Hemlo indicates that a high sulfur fugacity environment prevailed during peak metamorphism. Under these conditions, we have found that the ore mineral assemblage underwent partial melting, primarily through breakdown
Stacked pseudosections at 600°C for a bulk sulfosalt melt
composition dominated by stibnite
Click image for a larger view
of stibnite and arsenopyrite. Interaction between this early-formed Sb- and As-rich melt and a range of unmelted sulfides was facilitated by concurrent deformation-driven melt segregation, which led to further melting and incorporation of other elements into the melt. This gold-bearing melt was mobilized from compressional high strain regions into dilational domains such as boudin necks and extensional fractures developed in competent lithologies. Ore minerals that did not participate significantly in melting (pyrite, molybdenite, pyrrhotite and sphalerite) were not extensively mobilized. Fractional crystallization of the sulfosalt melt, made possible by ongoing deformation during
cooling, led to a diverse suite of ore minerals, dominated by stibnite and realgar and including an array of rare sulfosalts, native elements, intermetallic compounds and tellurides. Some sulfosalt melt persisted to low temperature (<300°C), allowing continued small-scale, localized mobilization during late deformation. Although gold occurs at moderate concentrations within the compressional high strain domains, it is particularly concentrated in the dilational domains, a consequence of its mobilization within a sulfosalt melt.
Impact Geology, Meteoritics and Planetary Science.
Impact Geology
PhD student Jess Salisbury (co-supervised by myself and Bruce Schaefer) is studying the evolution of the Lawn Hill Impact Structure in north eastern Queensland. The question of its relevance to ore genesis was investigated during Jess’s honours work because the Location of Century and the Lawn Hill Impact Structure
Click image for a larger view
world class Century Pb-Zn deposit is situated at the conjunction of the 100+ km Termite Range Fault and the previously defined margin of the impact structure. Jess has found unambiguous evidence of an extraterrestrial impact, including planar deformation features in quartz, widespread shatter cone formation, impact diamonds and
impact melt breccia within the Mesoproterozoic basement. We now believe that the impact structure is ≥ 19.5 km wide, this constrained in part by the outer margin of an Folding lubricated by impact melt breccia
 Click image for a larger view
annulus of brecciated and highly contorted limestone. Jess has also found evidence which suggests that the impact target included Cambrian limestone, including: (1) “dykes” of brecciated Cambrian limestone extending hundreds of meters into the Mesoproterozoic basement; (2) highly contorted bedding in the limestone annulus compared to essentially undeformed limestone away from the impact site; as well as (3) a 1 million ton megaclast of Mesoproterozoic Century-like ore suspended in the limestone. Through aerial photograph
analysis, large-scale convoluted flow structures within the limestone have beenidentified, and these are interpreted to indicate that parts of the Cambrian sequence may have been soft or only semi-consolidated at the time of impact. This highly contorted limestone bedding is thought to represent slump-filling of an annular trough in response to impact-induced partial liquefaction of a thin sediment veneer. The age of impact is therefore considered to be concurrent with limestone formation during the Ordian to earlyTempletonian, at 520-510 Ma. Formation of Century is found to be
unrelated to
Impact diamond from Mesoproterozoic basement
Click image for a larger view
impact- generated hydrothermal activity, although some minor hydrothermal remobilisation of metals occurred. There was however macro-scale remobilisation of gigantic ore fragments driven by impact-induced lateral and vertical injection of limestone into the Proterozoic sediments. The limestone-filled annular trough surrounds a 7.8 km diameter central uplift, consistent with formation of
a complex crater morphology.
Meteoritics
On the 2006 3rd year geological mapping camp at Eldee Station near Broken Hill
Eldee-001 An ordinary L6 chondrite
Click image for a larger view
I found a large meteorite, which turned out to be an ordinary L6 chondrite. Over two subsequent trips (the last with Bruce Schaefer and honours student Robin) we found another six meteorites that all turned out to be from a single different fall. This second group of meteorites (referred to
as Eldee-002) are from an impact melt breccia. Using these meteorites I am investigating, together with Roberto Weinberg and Bruce Schaefer, the role that impact melting plays
Eldee-002 An L chondrite impact melt breccia
Click image for a larger view
in separating metal and sulfide from silicate material. In Eldee-002 metal content and the metal/sulfide ratio (Me/S) are highly variable. Sulfide loss generally mirrors metal loss but is not as complete, with normal Me/S preserved in high-metal domains and low Me/S in depleted domains, indicating preferential extraction of metal. Differences in density and liquid viscosity may be responsible. Preferential melting of metal, and
to a lesser extent FeS, occurred because impact-induced heating is maximised at
Patchy metal distribution in Eldee-002
 Click image for a larger view
interfaces between high and low density phases. Silicates were mainly melted along narrow cm-long high-strain zones, allowing linking between melt patches and formation of pressure differentials associated with local structural
dilatancies. The lower viscosity of liquid metal allowed it to be rapidly and preferentially segregated into these dilational sites. FeS melt was similarly segregated but not as efficiently due to its higher viscosity and less extensive melting. Silicate melt was not
effectively segregated due to its high viscosity. Away from linked high strain zones no liquids could escape,
Metal-rich melts in Eldee-002
Click image for a larger view
allowing typical L chondrite Me/S to be preserved. The rapidity of this process was such that metal segregation occurred under disequilibrium conditions. Although some Fe-FeS melt did develop where the two were in contact at the time of melting, generally the metal and sulfide melts dissolved into each other to a limited extent. Two types of melt
conduit developed; those containing silicate melt with numerous immiscible sulfide- and metal-melt droplets, and those filled with only metal and/or sulfide melt through which more rapid transfer may have been possible. Localized meso-scale accumulations of metal (e.g., Portales Valley) are likely to be the next stage in the progression of this process. Gravity on small planetesimals is too low, and the duration of impact events is too short-lived, for Stokes’ Law settling to have been the dominant control on the initial stages of metal melt segregation. We therefore suggest that a significant fraction of non-magmatic iron meteorites, which have chemistry suggesting disequilibrium metal fractionation, may have formed through this process.
|
