2

Boddington Industry Project, Yilgarn Craton, W.A.

by

Dale Cameron a1219014 Mineral Geoscience Project

ABSTRACT
The Boddington gold mine, situated in the Saddleback greenstone belt, Yilgarn Craton, W.A., is a geologically complex and highly varied deposit. A variety of genetic models have been invoked in the past to explain the genesis of the deposit and features observed, including porphyry- and orogenic- models, as well as more recently, an intrusion-related gold system. Mineralisation occurs as veins, veinlets, shears, lenses and disseminations with host rocks of diorite, andesite and dacites. Veins and alteration are pervasive and consist of multiple stages of quartz-sericite, quartz-biotite, quartz-albite and actinolite alteration. Detailed ore mineralogical, petrographic and mineral-chemical study of representative ore samples from five of the eight domains within the deposit have given insights into the distribution of precious metals and also provided evidence for the formation of the Boddington deposit and provide evidence for its genetic evolution. Mineralisation is characterised by a reduced assemblage, with chalcopyrite and pyrrhotite as the dominant sulphides. Pyrite (often replacing pyrrhotite), sphalerite, cubanite, cobaltite, arsenopyrite and pentlandite are minor sulphides. Molybdenite is also relatively abundant and occurs as a major mineral in localised areas throughout the deposit. The study has shown that the deposit also contains an extremely diverse array of trace minerals that can provide supporting evidence for aspects of ore genesis. Native gold and electrum are the main gold minerals; LA-ICPMS analysis of pyrite and arsenopyrite revealed that these minerals are not significant Au-carriers at Boddington. In addition to maldonite, the deposit contains a suite of Bi-minerals, including native bismuth and a suite of Bi, Bi-Ag, Ag- and Pb-tellurides and selenides. These minerals are identified both in ore samples and in Cu-concentrates. Microprobe analysis of Bi-chalcogenides of the tetradymite group shows compositions from across the full range of the series, demonstrating the multiphase character of the Boddington mineralisation and, specifically, (often incomplete) overprinting by more oxidising fluids. This dataset also includes several previously unreported and non-stoichiometric compositions of tetradymite group phases; these may represent unnamed phases, but may also be disordered at the lattice-scale. Boddington is clearly not the product of a single ore-forming event but is rather a multiphase system recording successive overprinting and replacement of minerals, often very localised, and displaying strong lithological control.

TABLE OF CONTENTS

CHAPTER 1 INTRODUCTION CHAPTER 2 GEOLOGICAL SETTING REGIONAL GEOLOGY MINERALIZATION STRUCTURAL HISTORY CHAPTER 3 METHODOLOGY OPTICAL MICROSCOPY (UNIVERSITY) OPTICAL MICROSCOPY (ADELAIDE MICROSCOPY) SCANNING ELECTRON MICROSCOPY (ADELAIDE MICROSCOPY) CHAPTER 4 PETROGRAPHY RESULTS GANGUE MINERALIZATION OTHER MINERALS CHAPTER 5 - SULPHIDE MINERAL CHEMISTRY CHAPTER 6 LA-ICP-MS MAPPING OF MOLYBDENITE CHAPTER 7 TETRADYMITE GROUP MINERALS CHAPTER 8 DISCUSSION CHAPTER 9 CONCLUSION BIBLIOGRAPHY APPENDIXES

3 4 4 5 5 6 6 6 6 7 8 9 11 12 12 13 14 15 18 20

INTRODUCTION
The Boddington Gold deposit, is economically one of Australias largest gold deposits. The world-class Boddington Cu-Au-Mo deposit has a complex genetic history. The deposit was originally discovered in 1980. The relative importance of different ore-forming processes during the period 3.0 2.6 Ga is debated, particularly with respect to the role played by the 26113 Ma Wourahming granite. LA-ICP-MS analysis of trace element concentration in molybdenite represents a valuable new metallogenetic tool to track mineralising events in deposits with protracted geologic histories. The Re content and trace-element signatures in molybdenite from diorite and granite show three distinct populations, attributed to porphyry-style (hundreds of ppm) orogenic- and granite-related systems (<1 to a few ppm, respectively). The lateritic gold mineralisation at Boddington occurred as a semi-continuous blanket over a strike length of more than 5 km and a width of around 1 km. Some 30% of the gold was hosted by the same vertical zones as the bauxite, namely the upper lateritic gravels, the hardcap and the B-zone laterite. The remaining 70% was found in the underlying clay and saprolite zone, which was 10m, and locally up to 120 m in thickness. The laterite resource was exhausted in 2001 and the mine placed on care and maintenance, after having produced 105 Mt @ 1.4 g/t recovered Au for 147 t Au and 6500 t Cu. The $2.4bn project was initially a three-way joint venture between Newmont Mining, AngloGold Ashanti and Newcrest Mining. In 2006 Newmont bought Newcrest's 22.22% share, bringing its interest to 66.67% and ending any Australian ownership. AngloGold owned the remaining 33.33%. In June 2009, Newmont became the sole owner of the mine by acquiring the 33.3% interest of AngloGold. The original, mainly oxide open-pit mine was closed at the end of 2001 and now re-opened in 2009 and in September of 2009, the first 100,000t of ore was processed establishing a re-opening into the commodities market.

GEOLOGICAL SETTING
Regional Geology The regional geology of the Boddington gold mine is situated in the Yilgarn Craton. The Yilgarn Craton is composed of volcanic and sedimentary rocks that formed between 3000 and 2600 Ma and are metamorphosed to upper greenschist/amphibolites facies during an event at 2649 Ma (Nemchin et al. 1993). The Yilgarn Craton formation scripts a major chapter of tectonic activity on Earth and is interpreted to have fashioned from the accumulation of a number of crustal terrains including volcanic arcs, back arc basins and micro continents with ages from 3730 to 2550 Ma (Myers 1993). The period from 2760 to 2620 Ma in the late Archaean marks the formation of the majority of the greenstone belts in the Yilgarn Craton of W.A. (Nelson 1998). The formation of the Greenstone belts has been suggested to have formed in ensialic rift environments (Groves et al. 1987). Though, new information gathered has swayed geologists to recognize the formation of greenstone belts (within the South- western are of the Yilgarn) to be associated with subduction manners. The chemistry of the Saddleback greenstone belt (SGB) indicates that the belt formed in an island arc setting (McCuaig et al. 2001). The Saddleback greenstone belt is an isolated fault-bounded cut of volcanic and intrusive rocks surrounded by granitoids and gneisses of the Western Gneiss Terrane. The Saddleback Greenstone belt extends for approximately 50km along strike in a north-northwest orientation and has a maximum width of 8 km (Allibone et all., 1998). The Boddington gold deposit is located within the Archaean SGB, most of the deposit lies in the northern region within the Yilgarn Block (fig. 1). The deposit is within a considerably faulted, steep-dipping sequence of metamorphosed pelites, felsic and mafic extrusive and intrusive rocks. The metamorphosed rocks are greenschist to lower amphibolite facies that have been extensively faulted (Anand 2005). The primary mineralisation is hosted in 2715-2690 Ma intermediate to felsic intrusive and volcanic rocks diorite, andesite and dacite. It is a low sulphide gold deposit with an Au-Cu-Mo-W-Bi association (Symons et al. 1990; McCuaig et al. 2001). Mineralisation at Boddington occurs as shear zones, brittle ductile faulting, veins and reactivated veins, veinlets, lenses and disseminated ores.

Mineralization The primary mineralization, Au is hosted in intermediate to felsic volcanic rocks. It is a low-sulphide system with intense silification, potassic and calc-silicate alteration. It also includes an association with Au-Cu-Mo-W- (Bi) (Symmons et al., 1988). Regolith samples indicate that the Au is located in the lower part of the bauxite zone, which is the middle of the clay saprolite and the base of the saprolite. The major components of the Cu-Mo mineralisation at Boddington are chalcopyrite, pyrrhotite and molybdenite. Other sulphides identified here are cubanite, pyrite, sphalerite, stannite, pentlandite, argentopentlandite, a Fe-Ni-thiospinel, Co-mackinawite, bornite, chalcocite and covellite; native copper was also found in one of the samples with zeolite alteration (RG67). The Au mineralisation at Boddington is characterised by an Au-Ag-Bi- (Pb)-Te- (Se) mineral association which includes native gold, electrum, maldonite (Au2Bi), native bismuth, hessite (Ag2Te), altaite (PbTe), intermediate members of the galena- clausthalite series (PbS-PbSe), naummanite (Ag2Se), abundant Bi-tellurides from the tetradymite group (Cook et al. 2007a) and subordinate Bi-Pb-sulphotellurides from the aleksite series (Cook et al. 2007b). Structural History A large portion of Western Australia is composed of Archaean metavolcanic and metasedimentary rocks initially formed at 3050-2600 Ma (Myers et al. 1993, Duuring et al. 2007). These rocks were originally part of volcanic arcs, back arc basins and micro continents that amalgamated during a time of tectonic movement and formed the Yilgarn Craton (Myers 1993). Crustal fragments were aligned and became joined during a period of increased tectonic activity dated at 2780-2630 Ma (Myers 1993). A significant section of the Yilgarn Craton is comprised of granitoids and associated greenstone belts that underwent metamorphism. All greenstone belts in W.A. contain packages of igneous and sedimentary rocks that display similar lithologies and broad-scale structural controls. Greenstone belts are widely associated with metallic mineral deposits, since the latter form in volcanic arc and back arc basin environments where faulting can occur (Barley et al. 1990). The series of deformational events at Boddington is hotly debated. Allibone et al. (1998) propose seven distinct post-mineralisation deformational events. In this scheme, molybdenite and chalcopyrite veins formed between 2714-2696 Ma before the first deformational event. Widespread metamorphism led to a quartzoligoclasealbite- clinozoisite-muscovite-biotitechlorite mineralogy and well-defined rock fabric. Sericite-quartz alteration then took place within shear zones after peak metamorphism. D3 shears dated at post-2675 Ma hold quartz-albite-epidote alteration and cut through the previous two sets. Dikes were intruded along D4 faults introducing actinolite-

bearing veins and biotite-clinozoisiteactinolite alteration halos. Further movement on the D4 faults introduced superimposed mineralisation.

METHODOLOGY
In order to determine the microscopic mineralogy of the Boddington ore body, several different methods of observation and analysis were undertaken. These included optical microscopy, both in the classroom and using the technologically advanced microscopes courtesy of Adelaide Microscopy, and Scanning Electron Microscope analysis (SEM), also within Adelaide Microscopy.

OPTICAL MICROSCOPY (ADELAIDE UNIVERSITY)

Nikon reflective light microscopes were used to examine the microscopic optical properties of polished section samples. These are equipped with varying magnification options of 2.5x, 5x, 10x and 20x zoom. These microscopes were used prior to the other methods of microscopic mineral observation as they were designed to give an overview of what was within the samples rather than for close examination or analysis. Polished sections approximately 3cm in diameter and 1cm thick were observed. These were derived from samples taken from various locations around the Prominent Hill mine site, including the open pit.

OPTICAL MICROSCOPY (ADELAIDE MICROSCOPY)

Optical analysis at Adelaide Microscopy was undertaken using a similar Nikon reflective light microscope to those used at Adelaide University, however these were also equipped with a digital camera and 2.5x, 5x, 10x, 20x and 50x air objectives. These microscopes were used to enhance what was viewed under the microscopes at Adelaide University due to their higher quality imaging and elevated zooming options. The same polished sections were used to study the microscopic mineralogy and the digital camera attached to the microscope allowed photographs of features to be taken and thus examined in closer detail.

SCANNING ELECTRON MICROSCOPY (SEM)

The SEM was performed at Adelaide Microscopy using the Philips XL30 FEG-SEM equipped with an Energy Dispersive X-ray Spectrometer (EDS). The instrument was operated in back-scatter electron mode at a voltage of

20eV and ~20 na beam current, using a spot size of 3 or 4. A small sample is placed within the machine to be analysed. Through this method of examination, the minerals contained within a sample can be identified at an elemental level. Graphs showing the proportion of elements contained within a spot are produced which helps to positively identify the minerals. Still images were taken so further analysis could be undertaken.

RESULTS
The sample chosen for the microscopy sections of the practical was sample RG43, this sample was selected because of the high amount of containing molybdenite, which is a key aspect in the mineralization in the Boddington region. The reflected light microscopy results gave vital information on the hot spots and hints on where to find economic mineralization. At Boddington the mineralisation occurs as disseminations and veinlets throughout several types of magmatic rocks, including diorite, andesite, rhyodacite, dolerite and granite, all with various degrees of alteration. All these rocks are deformed and metamorphosed at lower amphibolite facies (biotite-plagioclase present). Sample RG43 showed heavily deformed molybdenite veinlets surrounded by dark gangue. Most commonly, the sulphides occur as narrow veinlets (mm-cm size) or as patches and disseminations, a great example of this is image 1. This image contains deformed molybdenite with chalcopyrite growth in the lamellae of the molybdenite, it is also backed up in image 2 at a broader scale. Chalcopyrite is often associated with pyrrhotite and in some cases with molybdenite, an example of this is shown in image 5 with the chalcopyrite and pyrrhotite intergrowing against the gangue. The latter however is often found in monomineralic patches of cm- to mm-size. The molybdenite in the granite shows strenuous deformation in the lamellae and is best explained in image 10 and image 3. The molybdenite in image 10 is completely deformed with no additional mineralogy, the differing in colour is only due to the local twinning of the sample and should not be mistaken as gangue; however in image 3 there is a tiny amount of gangue (quartz) present. The local gangue of the granite is comprised mostly from quartz, rutile, epidote (Allanite), feldspars and sphalerite, biotite. All these minerals are found in both granite and diorite samples. Also common in the two types of rocks is the presence of sphalerite, a dark grey mineral in the RLM that occurs as small grains where the latter occurs along boundaries mostly between the pyrrhotite and chalcopyrite.

The scanning electron microscope was a useful piece of equipment as it presented sub-accurate chemical compositions of the selected areas in the sample (RG 43). The results showed that what was interpreted in the reflected light microscopy but other additions in smaller parts of the sample were discovered.

GANGUE:
ALBITE: NaAlSi3O8 or Na1.00.9Ca0.0 The albite in sample RG43 interplayed with the quartz grains amongst the granite. They were commonly found in veinlets and alterations of quartz-albite. BIOTITE: K(Mg,Fe)3(AlSi3O10)(F,OH)2 The data for biotite in the granitic sample, as well as the diorite shows minute differences around an intermediate composition between the Fe-rich (annite) and Mg-rich (phlogopite) end-members. A slightly Fe-richer variety is present in a granite sample EPIDOTE GROUP MINERALS (ALLANITE): (Ce,Ca,Y,La)2(Al,Fe )3(SiO4)3(OH)
+3

Data for minerals from the epidote group show the presence of rare earth elements, elements such as Europium (Eu), Thulium (Tm), Cerium (Ce), Praseodyium (Pr) and Promethium (Pm). These rare earth elements distinguish the Epidote mineral into Allanite, more specifically an allanite-(Ce), due to the high content of Cerium present (4.15 Mol%). The epidote was dominantly composed of Silicon Dioxide (51.31 Mol%), Calcium Oxide (16.13 Mol%) and Aluminium Oxide (13.32 Mol%). K-FELDSPAR: NaAlSi3O8 CaAl2Si2O8 Plagioclase feldspars from the granite are Na-rich with a narrow compositional variation within the albite. Likewise, the potassium feldspar is nearby to end-member K-fs presenting a slight difference in all samples. The compositional spread for the plagioclase feldspars in the diorite is also in the albite-oligoclase range but unlike in the granite, this variation is observed in each of the analysed samples. The feldspars in sample RG43 grew as a

disseminated grain in a host of other gangue minerals, mostly quartz. It had no link to the mineralisation of sulphides. RUTILE: TiO2 The rutile in sample RG43 was very disseminated. The mineral was found amongst the quartz filled gangue background and look as the quartz was almost eating away at the rutile, perhaps a sharing of the Oxygen element in both groups is applicable. QUARTZ: SiO4 The Quartz created the background of the gangue network and was dominantly the main gangue mineral in distribution. As this sample (RG 43) is contained within the granite, a granitic composition is obviously noted, eg. Quartz, biotite, k-feldspar dominant.

MINERALISATION:
MOLYBDENITE: MoS2 The Molybdenite in sample RG43 contains a strange molecular ratio, the usual ratio for molybdenite is Mo=1, S=2, however in the samples taken during the scanning electron microscopy, the formula was M= 0.6553, S=0.355. The Molybdenite is the main attraction point of sample RG 43; the presence of molybdenite in the granite is quite abundant. It is largely associated with the chalcopyrite and pyrrhotite. It displays great morphological variety from bundles and knots to lamellar aggregates along the foliation or single lamella scattered in the surroundings of the larger patches. It is mostly coarse in grain texture and is quite heavily deformed in sections (see fig. 10 in RFM). Great twinning is subdue to the molybdenite and can sometimes be mistaken for gangue if care is not taken. In molybdenite from both granite and diorite Bi-tellurides from the tetradymite group are abundant, galena and hessite are also noted. Phases from the aleksite series are present in the granite whereas chalcopyrite, native gold, electrum, native bismuth and minor altaite are seen in the diorite.

The molybdenite in the granite displays regular, dense fields of telluride inclusions. In the diorite, however, chalcopyrite inclusions are regular and dense. In the latter case all other inclusions, including tellurides, have an irregular distribution and range in abundance from scattered to very dense in certain areas. In both cases the inclusions are seen preferentially along the molybdenite lamellae with a tendency to coarsening across intervals of kink deformation or in the axial plane of tight microfolds. Ductile remobilisation of inclusions is also observed along microshears oblique to the molybdenite lamellae. The thicker chalcopyrite inclusions display pointing edges indicating they are formed under stress conditions. CHALCOPYRITE: CuFeS2 Chalcopyrite occurred as both veinlets and disseminations in sample RG43. The sulphide grew within the molybdenite and formed as veinlets with in the lamellae of deformed molybdenite (image 1 RLM). The disseminated chalcopyrite occurred still within the molybdenite but as patches that were mostly associated with the pyrrhotite. PYRRHOTITE: Fe1-xS The pyrrhotite was largely associated with chalcopyrite; they grew together in patchy disseminations. In sample RG43, the pyrrhotite was not found without a deposit of chalcopyrite next to it. This gives a strong relationship indicator between the Fe bearing minerals. Also common in the two types of rocks is the presence of sphalerite where the latter is found as small grains at the boundaries between chalcopyrite and pyrrhotite. CUBANITE: CuFe2S3 Cubanite formed as exsollution lamellae within the disseminated chalcopyrite in sample RG43 and is showing the same traits mostly throughout the granites. However in the diorite sample RG27, the cubanite occurs as a main mineral and was contained in a high mafic component mineralogy with cubanite (Cub) dominant and almost fully replaced chalcopyrite, inclusions of granular cobaltite. It also played host to small Parkerite and cobalite inclusions that may or may not have further hosted Bismuth-Tellurides.

10

SPHALERITE: (Zn,Fe)S There was minimal sphalerite present in sample RG 43, however a disseminated patch of it is existent. It occurs as inclusions within the main ore minerals, indicating that Au deposition occurred at the same time as crystallisation of the main sulphides. Sphalerite is a minor sulphide as is cubanite, which forms exsollutions from chalcopyrite

OTHER IMPORTANT MINERALS NOT FOUND IN RG43:

CHLORITE: (Mg,Fe)3(Si,Al)4O10 Chlorite was not found in the granitic sample, RG43 however in the diorite sample RG27 it most certainly was. There was abundant chlorite in the diorite, in fact there were two differing iron contents, which distinguished two different styles of the mineral chlorite. One side of the chlorite is characterised by the richness in Fe (94.9 mol% chamosite, the Fe-end-member). Otherwise, the chamosite component samples ranges from 70.9 to 81.7 mol%. In the diorite, Mg-richer compositions are noted; Fe/(Fe+Mg+Mn) ranges from 0.38 to 0.78, inferring that both chamosite and clinochlore (Mg end-member) are present. This suggests two distinct generations, therefore two different forming conditions. The two populations in RG27 give temperature estimates at the minimum of the range (219 C) and towards the upper limit (367 C), further suggesting these two populations were formed under different conditions. Temperature estimates from individual analyses are also summarised in Figure 3. (Guerin 2011). GOLD: Au The granitic sample RG43 contained no gold what so ever, however gold is mostly hosted in the intermediate to felsic rocks (diorite). The Au mineralisation at Boddington is characterised by an Au-Ag-Bi-(Pb)-Te-(Se) mineral association which includes native gold, electrum, maldonite (Au2Bi), native bismuth, hessite (Ag2Te), altaite (PbTe), intermediate members of the galena-clausthalite series (PbS-PbSe), naummanite (Ag2Se), abundant Bi-tellurides from the tetradymite group (Cook et al. 2007a) and subordinate Bi-Pb-sulphotellurides from the aleksite series (Cook et al. 2007b).

11

SULPHIDE MINERAL CHEMISTRY

Compositional variation in the sulphides was measured using EPMA data. Results are presented in terms of means and standard deviation for each sample or distinct populations in a given sample. Pyrrhotite, Fe1-xS, where 0<x<0.125, is an interesting mineral group both in geological and crystallographic terms. The pyrrhotite structures have hexagonal close packed layers of S with Fe atoms occupying octahedral interstices between the S layers. All octahedral positions are occupied in troilite (FeS; hexagonal), whereas the non- stoichiometric pyrrhotites have vacancies on the Fe positions (Posfai & Buseck 1997). The most common forms in nature are monoclinic 4C pyrrhotite (ideally Fe7S8) and hexagonal 5C pyrrhotite (ideally Fe9S10) (Becker et al. 2010). The plot of M versus S shows that the single phase Po corresponds to the Fe7S8 field in the granite and spreads into the field of Fe9S10 for single phase Po in the diorite. About a third of the Po in the granite analyses plot in a field with still lower M, which is difficult to explain unless we consider that the Po is replaced by Py and the latter is not observable on the BSE images (submicroscopic grain replacement). There is a clear compositional gap between the dark (showing a wide spread across Fe7S8 to Fe10S11) and bright lamellae (towards the upper limit of stoichiometric FeS) from the 2-phase Po in the diorite (Figure 4).

LA-ICP-MS MAPPING OF MOLYBDENITE

Sample RG45 is representative of the coarse molybdenite in the granite where the abundance of Bi-(Pb)-tellurides is greatest. The selected area includes 2 sets of kink-bends on each side; replacement of silicates strongly marks the kink-bends. Slight deformation oblique to lamella orientation is also seen on the optical microscopic image (Figure 10k). Rhenium shows zonation with strong enrichment in the areas with the kink-bends relative to the middle part of the mapping area. In contrast W only shows strong enrichment along the kink bends and also along the oblique deformation. Lead, Bi and Te show high concentration with similar pattern to the W, strongly suggesting the relationship between textures and trace element concentration. Notably, Au show spot concentrations overlapping the chalcophile elements enrichment despite the fact that the measured Au values in the sample are at best a few ppm. (Guerin 2010)

12

TETRADYMITE GROUP MINERALS

The study of the tetradymite group minerals is relatively new to the geological society, tellurides have been a struggle to study over the years, as Nigel Cook and co. have quoted; Unlike the Bi-dominant sulfosalts, for which enormous progress has been achieved in recent years in understanding the structural architecture of numerous homologous and homeotypic series (e.g., Makovicky 1989, 1997a, b), tellurides and selenides have been difficult to investigate at the same scale, owing to their invariably smaller grain-size and intergrown character. (Cook et al 2007). Extensive solid-solution has been invoked by Bayliss (1991) and termed disorder. Focusing on tsumoite in particular, Bayliss argued for interchangeable occupancy of Bi and Te sites, causing extensive (isostructural) variation in Bi:Te ratio. The compositional range of tsumoite was said to encompass other binary tellurides such as hedleyite and pilsenite. Such disorder has been since used to account for a wide variety of values of the Bi:Te ratio observed in natural tellurides of Bi. Dobbe (1993) considered that his suite of compositional data from Tunaberg, Sweden, supported the idea of a single solid-solution, a tsumoite-type phase (bismuthian tsumoite) extending between values of the Bi:Te ratio of 1.35 to 2.61, this data from the Canadian Mineralogist is consistent with the Boddington Tsumoite in that Bi:Te ratios are ranging from 2.73:1 to 2.99:1 with up to 0.0448 K-ratio of Sulphur being included from mineralogical probe data. Tellurides of Au, Ag, Bi, Pb and other elements are common trace minerals in gold deposits that span the magmatic-hydrothermal spectrum, as well as those formed in metamorphic terranes. The association between tellurides and gold is long recognized, and is most evident from the fact that a small number of deposits, mostly but not exclusively of epithermal type, contain exploitable Au-(Ag)-tellurides. These are the true Au(Ag)-telluride deposits in which a significant proportion of the precious metals are carried by the tellurides themselves. Although the majority of such deposits are epithermal, Au-(Ag) tellurides are also known from many Archaean and Proterozoic orogenic gold deposits, the giant Golden Mile deposit, W.A. being a prime example. Insights into mechanisms responsible for telluride deposition have been examined in relatively few instances and no general model currently exists. A general scheme for telluride deposition via gas condensation is attractive to explain features of mineralising systems that experienced sustained boiling, but tectonically driven hydrothermal systems could be as efficient in making an Au-telluride deposit. Depending on temperatures, scavenging of Au by Bi(Te) melts, or partial melting of a pre-existing ore, may offer alternatives to generate telluride-rich gold ores. Pressure variation, e.g., by throttling, appears to impact on telluride deposition and distribution in any given deposit.

13

Relatively recently, this strong relationship between Au and Bi has been explained in terms of the generation of bismuth melts within the ore during metamorphic overprinting (Ciobanu et al. 2006a; 2010). Underlying the concept of melt generation is the rather low melting point of bismuth (271 C) and the observed tendency for Bi- droplets to scavenge gold from fluid. Textures between Au- and Bi-minerals suggestive of deposition in the form of melts have been given in a number of studies (Ciobanu et al. 2005; Ciobanu et al. 2006a; Tooth et al. 2008). Bismuth is thus considered a powerful scavenger for gold if conditions favour melt generation and Bi-melts have been proven experimentally to scavenge gold from fluids under saturated in Au (Douglas et al. 2000). Tellurides may also be a prominent component of such assemblages, indicating that Au- Bi-Te melts can be considered analogous to simple Au-Bi melts (Ciobanu et al. 2005).

DISCUSSION
The Boddington deposit contains tetradymite-group minerals from across the full compositional range of the series, as well as probable disordered, non-stoichiometric compositions that do not correspond to named minerals. This diversity, coupled with strong variation in the associations of these minerals and reaction between phases in the group, clearly show that the Boddington deposit is not the product of a single ore-forming event, but rather suggests discrete events during the evolution of the deposit, as well as showing the major role that lithology played during ore genesis. While metal abundances in molybdenite from the granite are low and the granite is relatively unmineralised it does contain similar alteration assemblages and minor chalcopyrite/pyrrhotite mineralisation + molybdenite. Such evidence does not rule out the granite as a potential metal source but could be explained by late stage crystal fractionation and metal incompatibility. Metal enrichment from fractionation of granite has been suggested by past geologists (by Mustard et al. (2006)) to explain mineralisation associated with intrusion-related systems. Bismuth levels in the molybdenite sample are relatively elevated and could have also assisted with the removal of Au from the granite as a result of Au scavenging from a fluid under saturated in the Au. Petrographic, mineralogical and geochemical evidence support a three-stage model for Boddington. An early porphyry event can account for the bulk of the Cu mineralisation, as well as some of the Au and Mo. A subsequent orogenic-Au event led to shearing and remobilisation of ore components. New constraints on metamorphic conditions are offered by chlorite and stannite-sphalerite geothermometry and the occurrence of two co-existing

14

pyrrhotite species. The granite introduced some Au, Mo and other granitic elements, notably Bi leading to substantial upgrading of Au grades by Bi-melt scavenging.

LIMITATIONS
The results observed in this report are of limited outcomes. Due to the fact that only a few hand samples were examined, only one polished section was observed which restricted information only to the granitic component of the Boddington deposit. Future research would include more polished sections from different areas of the deposit and around the deposit looking at the bordering minerals and assemblages to tie in with the deposit itself. Laser ablation data and EPMA data was scavenged from past thesis due to lack of data upload and capabilities.

CONCLUSION
This summary highlights that at Boddington there are at least 2 main ore stages; relating to host diorite and that formed during emplacement of granite. The similarity in the main sulphide and alteration assemblages, as well as the reducing character indicate that the metals and the fluids involved in both mineralising stages have either undergone a similar mechanism of ore deposition or that a strong superposition or obliteration occurred during granite emplacement. The Wandoo basement gold mineralisation is therefore interpreted to represent a structurally-controlled, intrusion- related Au-Cu deposit, paragenetically associated with both ~2700 and 2612 Ma events, with the main stage, higher grade mineralisation being apparently synchronous with the late, K-rich post-magmatic monzogranite suite. The dominant ore minerals are chalcopyrite and pyrrhotite. Such an assemblage is typical of an ore formed under slightly reducing conditions. Sphalerite, native bismuth, various Bi-Te species and electrum occur as inclusions within the main ore minerals, indicating that Au deposition occurred at the same time as crystallisation of the main sulphides

15

Ore Minerals Chalcopyrite Pyrrhotite Gold Molybdenite Cubanite Sphalerite Gangue Minerals Albite Biotite Epidote K-Feldspar Rutile Quartz

Occurs as Disseminated Disseminated Fine grains Blocky Disseminated within lamellae of Molybdenite Disseminated fine grains around Fe-sulphides

16

Mineral Actinolite Albite Allanite Biotite Bismuth tellurides Bornite Calcite Carbonate Chalcocite Chalcopyrite Chamosite Clinozoisite Covellite Cubanite Electrum Epidote Gold Ilmenite K-feldspar Maldonite Molybdenite Muscovite Native bismuth Native copper Native silver Oligoclase Pentlandite Plagiocalse Pyrite Pyrrhotite Quartz Rare Earth Minerals Rutile Sphalerite Stannite Thiospinel Tsumoite Volynskite

Abbreviation Act Ab Aln Bt Bi-tell Bn Cal Carb Cc Cp Cham Clz Co Cub El Ep Au Ilm KFs Mld Mo Mu Bi Cu Ag Ol Pn Plag Py Po Qz REE-min Ru Sph Stn Thio Ts Vol

17

BIBLIOGRAPHY

ALLIBONE A. H., WINDH J., ETHERIDGE M. A., BURTON D., ANDERSON G., EDWARDS P. W., MILLER A., GRAVES C., FANNING C. M. & WYSOCZANSKI R. 1998. Timing relationships and structural controls on the location of Au-Cu mineralization at the Boddington gold mine, Western Australia. Economic Geology 93, 245-270. ANAND R. R. 2005. Boddington gold deposit, Western Australia. CRC LEME, CSIRO Exploration and Mining COOK N. J., CIOBANU C. L., WAGNER T. & STANLEY C. J. 2007a. Minerals of the system bi-te-se-s related to the tetradymite archetype: Review of classification and compositional variation. Canadian Mineralogist 45, 665-708. GROVES D. I., PHILLIPS G. N., HO S.E., HOUSTOUN S.M. & STANDING C. A. 1987. Craton scale distribution of Archaean greenstone gold deposits: Predictive capacity of the metamorphic model. Economic Geology 82, 2045-2058 GUERIN, R. (2010). Petrography, mineralogy and trace element chemistry of Cu-Au-Mo mineralisation from Central Diorite, Boddington, W.A.. Adelaide University MYERS J. S. 1993. Precambrian history of the west Australian craton and adjacent orogens. Annual Review of Earth Planetary Sciences, 21, 453-485 NELSON D. R. 1998. Granite-greenstone crust formation on the Archaean Earth: A consequence of two superimposed processes. Earth and Planetary Science Letters 158, 109-119 NEMCHIN A. A., PIDGEON R. T. & WILDE S.A. 1993. Timing of the late Archaean granulite facies metamorphism and the evolution of the southwestern Yilgarn Craton of Western Australia: Evidence from U-Pb changes of zircons from mafic granulites. Precambrian Research 68(3-4), 307-321

18

SYMONS, P.M., Anderson, G., Beard, T.J., Hamilton, L.M., Reynolds, G.D., Robinson, J.M., Staley, R.W., 1988. The Boddington Gold Deposit. The second international Conference on Prospecting in Arid Terrain, Perth, Excursion Guidebook, pp 77-84.

19

APPENDIXES
Figure 2: The series of structural Figure 1: Location of the Boddington Au + Cu deposit in the southwestern Yilgarn craton, Western Australia

deformation Yilgarn Craton

events in the

20


5 3

Below: Images taken from Reflected Light Microscopy at Adelaide Microscopy 2012

Chalco Mo Chalco

Mo

Po

Qtz

Sph Qtz

Mo

Po Chalco Chalco

21

8 9

Qtz

Ep

10

Mo Qtz

Rut

11

12

Fldspr

Qtz Rut Qtz

22

Below: Images taken from Scanning Electron Microscopy at Adelaide Microscopy 2012

23

10

11

24

Figure 3: Binary plots of chlorite, showing the temperature ranges for distinct populations in diorite and granite.

25

26

Figure 4: Binary plots of total metal (M) versus S for pyrrhotite in granite and diorite. (a) Single-phase pyrrhotite corresponding to Fe7S8 from granite. (b) Single-phase pyrrhotite corresponding with Fe7S8-Fe9S10 from diorite. (c) Distinct compositional fields for two-phase pyrrhotite occurring as lamellar exsolutions in diorite. The phase that is bright on BSE images forms a field close to FeS (troilite) whereas composition of the dark phase straddles across Fe10S11- Fe7S8. (d) Sphalerite-stannite geothermometry plot showing formation temperatures from logXFeS/XZnS ratios in the two minerals. The cluster of points along the 350 C calibration line is the most reliable.


Figure 5: Electron Microprobe data of Tsumoite in sample RG43

TSUMOITE

27